T H E W O R L D B A N K 36280 Global v1 Development Finance The Development Potential of Surging Capital Flows I: R E V I E W, A N A L Y S I S, A N D O U T L O O K 2006 Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank Global Development Finance The Development Potential of Surging Capital Flows I: Review, Analysis, and Outlook Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank Global Development Finance The Development Potential of Surging Capital Flows I : R E V I E W , A N A L Y S I S , A N D O U T L O O K 2006 Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 T H E W O R L D B A N K (c) The International Bank for Reconstruction and Development / The World Bank © 2006 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved. 1 2 3 4 09 08 07 06 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and / or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Cover photo: Dennis Degnan/Corbis. Cover design: Drew Fasick. ISBN-10: 0-8213-5990-8 ISBN-13: 978-0-8213-5990-7 eISBN-10: 0-8213-6480-4 eISBN-13: 978-0-8213-6480-2 DOI: 10.1596/978-0-8213-5990-7 ISSN: 1020-5454 The cutoff date for data used in this report Delivered was by The May World 17,e-library Bank 2006. Dollars to: are current U.S. dollars unless otherwise specified. The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank Table of Contents Foreword xi Acknowledgments xiii Selected Abbreviations xv Overview and Policy Messages: The Development Potential of Surging Capital Flows 1 The broad surge in private capital flows continues 2 Global growth has propelled the surge in capital flows, but serious risks remain 4 Capital flows are being transformed 5 Net official flows continue to decline 7 To ensure economic stability, developing countries must manage capital flows effectively 8 Multilateral cooperation is key to resolving global financial imbalances 10 Chapter 1 Prospects for the Global Economy 13 Summary of the outlook 13 Global growth 16 Regional outlooks 18 Commodity markets 22 Inflation, interest rates, and global imbalances 25 World trade 30 Risks 32 Avian influenza 36 Notes 39 References 41 Chapter 2 The Growth and Transformation of Private Capital Flows 43 Private debt market developments in 2005 45 Structural changes in emerging market debt 59 Prospects for private capital flows 71 Annex: Commercial Debt Restructuring 73 Developments in 2005 and the first quarter of 2006 73 Notes 75 References 75 Chapter 3 Supporting Development through Aid and Debt Relief 79 Delivered by The World Bank e-library to: Recent trends and prospects for foreign aid 80 Bank The World IP : 192.86.100.36 Debt relief: improving and maintaining debt sustainability 87 Tue, 10 Mar 2009 16:55:33 v (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 The challenge ahead: accessing external capital, while maintaining debt sustainability 99 Annex: Debt Restructuring with Official Creditors 100 Notes 102 References 103 Chapter 4 Financial Integration among Developing Countries 107 The growth of South–South capital flows 108 Foreign direct investment in the developing world 110 South–South banking 117 Developing-country stock exchanges 124 Conclusion 127 Chapter 4 Annexes 128 Annex 1: Data on South–South capital flows 128 Annex 2: Selected South–South M&A deals by southern multinationals in service sector, 2000–5 130 Annex 3: Model of determinants of bank ownership 131 Notes 131 References 133 Chapter 5 Challenges in Managing Capital Flows 139 Two booms in capital flows—what has changed? 141 The effect of the recent influx of capital flows on domestic investment and asset prices 156 Lessons and policy agenda 159 Annex: Capital Flows and Domestic Investment 164 Notes 166 References 166 Statistical Appendix 171 Tables 1 Net capital flows to developing countries, 1997–2005 3 2 Net private capital flows to developing countries by region, 1998–2005 4 1.1 The global outlook in summary 17 1.2 Estimated impact of three risk scenarios 34 1.3 Impact of a 400-basis-point increase in interest rates in selected developing countries 35 1.4 Impact of a further $30 hike in oil prices in selected developing countries 35 1.5 Impact of a 15 percent fall in non-oil commodity 36 1.6 Impact of a widening of bird-bird flu 37 1.7 Possible economic impacts of flu pandemic 37 1.8 A breakdown of economic impacts of a potential human-to-human pandemic 38 2.1 Net private debt flows to developing countries, 2002–5 45 2.2 Gross market-based debt flows to developing countries, 2002–5 45 2.3 Gross cross-border loan flows, 2005 49 2.4 Countries’ access to international capital markets by intermediaries, 2002–5 50 2.5 Asset allocation of major international pension funds, 2004 54 2.6 Net FDI flows to developing countries, 2000–5 55 2.7 Delivered by The Selected announced privatization and World M&A Bank e-library deals in to: developing countries, 2005 57 The World Bank 2.8 International bonds and notes outstanding, by currency, 1999–2005 IP : 192.86.100.36 59 Tue, 10 Mar 2009 16:55:33 vi (c) The International Bank for Reconstruction and Development / The World Bank T A B L E O F C O N T E N T S 3.1 Net ODA disbursements, 1990–2005 80 3.2 ODA and debt relief grants in 2005 81 3.3 Main components of bilateral ODA, 1990–2005 81 3.4 Net ODA disbursements to the ten largest recipient countries 83 3.5 Donors’ shares of ODA in 2005, projected 2010 84 3.6 General government financial balances in 2004, projected 2005–7 85 3.7 ODA as a percentage of fiscal expenditures and revenues in 2004, projected 2006 85 3.8 Debt relief grants provided by DAC donor countries, by income and region of beneficiary, 1990–2005 87 3.9 Debt-service reductions to be provided by the MDRI 92 3.10 Donors’ commitment to refinance IDA for debt relief provided under the MDRI, selected years 93 3.11 Net present value of external debt relative to GNI and exports, 2004 94 3.12 Credit ratings for decision-point HIPCs 97 3.13 Average annual real GDP growth, 1990–2005 97 3.14 Indicators of external position of the 29 decision-point HIPCs, 1990–2005 99 4.1 South–South FDI as a share of global FDI, 1999–2003 111 4.2 Regional FDI by multinationals from selected countries 112 4.3 Selected southern multinationals in the oil-and-gas sector, 2004 113 4.4 South–South cross-border syndicated lending, 1985–2005 119 4.5 Source of foreign bank assets, by region 120 4.6 Performance indicators for northern and southern foreign banks, selected aggregates, 2000–4 123 4.7 Stock exchanges in selected developing countries, December 2005 125 5.1 Ratio of short-term debt to total debt in major borrowing countries, 1996–2004 144 5.2 Profile of external financial policy for developing countries considered relatively open to capital movements 149 5.3 Current account aggregated by region, 1997–2005 150 5.4 Sources of reserve accumulation, 1997–2005 151 5.5 Changes in central bank balance sheets, 2001–5 153 5.6 Foreign currency reserves and foreign assets as shares of total central bank assets in countries with high reserve accumulations, 2005 154 5.7 Investment performance during the surge in capital flows, 2002–4 158 5.8 Indicators of overheating in selected developing countries, 2002–4 160 5.9 Stock market performance in emerging markets, 2002–5 161 5A.1 Domestic investment and private capital flows 165 Figures 1 Financial flows to developing countries, 1997–2005 2 2 Benchmark spreads for emerging markets, 2001–6 4 3 Capital outflows by private entities in the developing world, 1981–2005 5 4 Net official lending and foreign aid grants to developing countries, 1980–2005 7 5 Net official lending, 1997–2005 7 1.1 Industrial production remains robust 16 1.2 Inflation in high-income countries 16 1.3 Developing-country growth remainsDelivered 18 Bank e-library to: by The World robustThe World Bank 1.4 Regional growth trends 19 IP : 192.86.100.36 1.5 10 Mar 2009 16:55:33 An end to the trend rise in oil prices? Tue,22 vii (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 1.6 Higher prices slow oil demand 23 1.7 A disappointing supply response 23 1.8 Spare production capacity remains low 24 1.9 Commodity prices 24 1.10 Moderate increases in inflation 25 1.11 Flattening yield curve 26 1.12 Changes in real effective exchange rate 26 1.13 Developing countries’ current-account balances 27 1.14 Increased aid helped finance oil costs in 2004 27 1.15 Reserves in some countries are falling rapidly or worrisomely low 28 1.16 Tensions associated with fast growth, the case of Turkey 28 1.17 Global imbalances 29 1.18 Interest rate spreads support the dollar 29 1.19 Funding the U.S. current account deficit 30 1.20 Healthy growth in world trade 30 1.21 Regional increases in market share 31 1.22 Increased product range explains most of Chinese export growth 32 1.23 Exports of developing countries have diversified 32 2.1 Net private debt flows to developing countries, 1991–2005 43 2.2 Emerging market bond issuers by type, 2002–5 48 2.3 Average spreads on new bond issues, 2001–5 48 2.4 Bond market financing by risk category, 2002–5 48 2.5 Concentration in bond and bank financing, 1993–2003 51 2.6 Comparative cost of bond and bank financing, June 2004–December 2005 51 2.7 Bank financing raised for core activities, 2002–5 52 2.8 Bank credit for high-risk borrowers: rising rates but longer maturities, 2001–5 52 2.9 IPO activities in emerging market countries 2001–5 54 2.10 Investment climate and FDI 55 2.11 The concentration of FDI, 1995–2005 56 2.12 Euro-denominated international bond issues, by region, 1999–2005 60 2.13 Yields on U.S. and German 10-year government bonds, 1999–2005 60 2.14 Comparison of euro-denominated and U.S. dollar-denominated emerging market sovereign bond issues 61 2.15 The global credit derivatives market in notional terms, 2001–5 62 2.16 Credit derivative participants, 2004 64 2.17 Five-year CDS and ASW spreads for selected countries, 2002–5 65 2.18 Trends in domestic debt securities in emerging markets, by region, 1997–2005 66 2.19 The size of the domestic bond market in selected countries 66 2.20 Bond market profile in selected countries, September 2005 67 2.21 Performance of local-currency bonds (ELMI+) against major indexes, 2002–5 69 2.22 Returns vs. volatility of selected bond indexes, 2000–5 69 3.1 Net ODA to developing countries, 1990–2005 81 3.2 Bilateral ODA loans and grants, 1990–2005 82 3.3 Share of total ODA allocated to LDCs and other low-income countries, 1990–2004 82 3.4 Net ODA as a percentage of GNI in DAC donor countries, 1990–2005 and projected 2006–10 85 3.5 Debt-service payments Delivered and HIPC by The e-library to: for 28 “decision point” HIPCs World Bankreduction debt-service The World Bank 89 3.6 Debt-service reduction provided IPby the HIPC Initiative to 25 decision-point countries : 192.86.100.36 89 3.7 Tue, 10 Mar 2009 Debt service paid by 25 decision-point 16:55:33 HIPCs, 2000 versus 2005 89 viii (c) The International Bank for Reconstruction and Development / The World Bank T A B L E O F C O N T E N T S 3.8 Debt-service reduction to be provided by the MDRI, 2006–45 92 3.9 Debt-service reduction to be provided to 18 completion-point HIPCs under the MDRI, 2006–45 92 3.10 Donors’ commitment to refinance IDA for debt relief provided under the MDRI, 2006–45 93 3.11 Debt burdens in 18 completion-point HIPCs, before and after the HIPC and MDRI debt relief 94 3.12 Net official lending to 27 decision-point HIPCs as a percent of GDP, 1990–2004 96 4.1 South–South capital flows by type, 2005 108 4.2 Growing openness of developing countries to trade and capital flows, 1995–2005 110 4.3 South–South FDI in infrastructure and by region, 1998–2003 113 4.4 Cross-border lending to all countries by banks in developing countries, 2000–5 120 4.5 South–South foreign bank entry in developing countries, by country income level 121 4.6 Developing-country firms shift away from ADRs 126 5.1 Distribution of private capital flows across developing countries, 2002–4 142 5.2 Composition of financial flows to developing countries, 1992–7 and 2002–5 143 5.3 Ratio of foreign exchange reserves to short-term debt, by region 143 5.4 Changes in exchange rate flexibility, 1991–2004 145 5.5 Frequency distribution of daily percentage changes in exchange rates for selected developing countries, 1993–6 vs. 2003–5 146 5.6 Movements in real effective exchange rates in East Asia and Latin America, 1993–2005 147 5.7 Real exchange rates for selected countries that receive higher-than-average private capital inflows as a ratio to GDP, 1994–7 and 2002–5 147 5.8 Current-account balance, developing countries, 1990–2005 150 5.9 Value of oil imports, oil-importing countries, 2001–5 150 5.10 Foreign exchange reserves, by region, 1995–2005 153 5.11 Foreign exchange reserves as a share of trade, 1970–2003 153 5.12 Currency composition of developing countries’ foreign exchange reserves, 2000 and 2005 156 5.13 Market capitalization 160 5.14 Turnover on world stock exchanges, 2004 160 5.15 Ratios of debt to equity in selected countries, 1996–2004 161 Boxes 1 International migrant remittances 3 2.1 The emerging bond market enters the mainstream 46 2.2 Strong performance of emerging stock markets in 2005 53 2.3 Growing FDI in China’s banking sector 57 2.4 Accession to the European Union and FDI 58 2.5 Credit default swaps 63 2.6 The role of multilateral development banks in developing local-currency bond markets 68 3.1 The Integrated Framework for Trade-Related Technical Assistance 84 3.2 The HIPC Initiative 88 3.3 The Paris Club 90 Delivered by The World Bank e-library to: The World Bank 3.4 The MDRI 91 IP : 192.86.100.36 3.5 The DSF for low-income countries Tue,95 10 Mar 2009 16:55:33 ix (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 4.1 Developing countries as aid donors 109 4.2 South–South FDI and trade 111 4.3 The World Bank Group and South–South flows 116 4.4 Determinants of South–South foreign bank entry 122 5.1 Preconditions for capital-account liberalization 148 5.2 Capital flows are procyclical with respect to non-oil commodity markets 152 5.3 Central bank debt in China 155 5.4 Optimizing allocations in reserve portfolios 157 5.5 Investment and private capital flows 159 Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 x (c) The International Bank for Reconstruction and Development / The World Bank . Foreword R OBUST GLOBAL GROWTH AND A credit risks associated with emerging market exter- favorable financing environment provided nal debt portfolios. Financial integration among the context for a record expansion of pri- developing countries continues to deepen with vate capital flows to developing countries in 2005. capital flows between developing countries (so- These conditions now provide a unique opportu- called South–South flows) playing a prominent nity for the international policy community to role. The role of the euro has evolved, gaining im- place development finance on a firmer footing be- portance both as an international reserve currency fore the tightening of global liquidity closes the and for debt issuance by governments and the cor- window of opportunity. porate sector in developing countries. The emerg- Most of the record $491 billion in net private ing market asset class has matured far beyond the capital bound for the developing world in 2005 earlier dominance of U.S. dollar-denominated, went to a small group of middle-income countries. high-yield, sovereign-debt instruments—indeed Many of those countries took advantage of the the Brady bonds issued in the 1980s that once ex- growing inflows to improve their external debt emplified this category have all but disappeared. profiles and accumulate large holdings of official Global growth has remained surprisingly re- foreign exchange reserves. silient to the rise in world oil prices over the past By contrast, many low-income countries still few years. Developing countries led the way with have little or no access to international private GDP growth in 2005 of 6.4 percent, more than capital, and instead depend largely on official fi- twice the rate of high-income countries (2.8 per- nance from bilateral and multilateral creditors to cent). While inflation has, on the whole, remained support their development objectives. With a subdued, there are signs of a pickup in several decade remaining to attain the Millennium Devel- rapidly growing countries, which raises the possi- opment Goals (MDGs), expectations of a “big bility of overheating and the need for a tightening push” in development assistance escalated during of macroeconomic policies. More generally, cur- 2005. Donors enhanced their efforts by scaling up rent account balances in oil-importing countries aid volumes and reallocating them to the poorest have deteriorated significantly, leaving them more countries, particularly those in Sub-Saharan vulnerable to subsequent adverse shocks. Africa. In addition, the Multilateral Debt Reduc- Looking forward, while many of the external tion Initiative (MDRI) will provide additional debt factors that have supported strong developing- relief to qualifying heavily indebted poor countries country growth are projected to weaken, eco- (HIPCs), reducing debt service and freeing up nomic growth is expected to remain relatively more fiscal resources for the MDGs. strong. However, downside risks predominate. At the same time, the development finance Persistent global imbalances, elevated current ac- landscape is being transformed. A growing num- count deficits in some developing countries, and ber of countries are issuing longer-term maturities asset price over valuation are potential sources of in international capital markets, in some cases risks to growth prospects in developing countries. even denominated in local currencies. Domestic In addition, a sharp supply shock could send oil debt markets have become a major source of fi- prices even higher, with serious consequences for nance in some countries, attracting international the most energy-dependent developing economies. investors in search of higher yields and potential A fall in non-oil commodity prices could have sim- gains from currency appreciation. Structured fi- ilar consequences for some of the poorest coun- nancial instruments such as credit defaultDelivered swaps tries, Bank by The World which haveto: e-library benefited from higher metals and The World Bank allow investors to better manage exposure to mineral prices. Finally, the Doha Round stands at IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 xi (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 a critical juncture; governments need to agree on ity for ensuring that policies are pursued that per- the key elements of a deal by mid-2006, but posi- mit imbalances to unwind in an orderly and timely tions on the central issue of market access for agri- manner. This requires cooperation. The key policy cultural and nonagricultural goods remain far prescriptions are well-known—the challenge is to apart. make meaningful progress in implementing those A key priority for developing countries going policies. Policy makers in the major economies un- forward is to pursue policies that strengthen their derstand the importance of a coordinated ap- capacity to weather whatever global storms may proach and therefore have endorsed the proposal be brewing. Continued macroeconomic stability is for the International Monetary Fund to play a vital to ensure effective management of capital more prominent role in coordinating the required flows to advance long-term investment and collective action. growth. Countries must preserve sound financial Global Development Finance is the World management, with monetary and fiscal policies Bank’s annual review of global financial condi- working in tandem to maintain debt sustainability tions facing developing countries. The current vol- and price stability. They must also build a system ume provides analysis of key trends and prospects, of risk management robust enough to respond to including coverage of capital originating from de- the needs of a more flexible exchange rate and veloping countries themselves. A separate volume open capital markets. Regulators in developing contains detailed standardized external debt statis- countries need to build their capacity to monitor tics for 135 countries as well as summary data for credit default swap transactions and define a clear regions and income groups. More information on line of responsibility and necessary expertise to the analysis, including additional material, better manage the associated risks. Oil exporters sources, background papers, and a platform for face the special challenges of managing the risks interactive dialogue on the key issues can be found surrounding volatile export revenues and using at www.worldbank.org/prospects. A companion those revenues productively. online publication, Prospects for the Global Econ- All countries would be affected by a disor- omy, is available in English, French, and Spanish derly unwinding of global imbalances, which at www.worldbank.org/globaloutlook. would destabilize international financial markets and curtail global growth. But developing coun- tries would suffer disproportionately, particularly if the imbalances were to foster a backlash of trade François Bourguignon protectionism. With deepening economic and fi- Chief Economist and Senior Vice President nancial integration, all countries share responsibil- The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 xii (c) The International Bank for Reconstruction and Development / The World Bank . Acknowledgments T HIS REPORT WAS PREPARED BY THE kets, and Financial Engineering of the Bank’s International Finance Team of the World Treasury). The main macroeconomic forecasts Bank’s Development Prospects Group were prepared by the Global Trends Team of (DECPG). Substantial support was also provided DECPG, led by Hans Timmer and including John by staff from other parts of the Development Eco- Baffes, Andrew Burns, Carolina Diaz-Bonilla, nomics Vice Presidency, World Bank operational Maurizio Bussolo, Betty Dow, Annette de Kleine, regions and networks, the International Finance Fernando Martel Garcia, Don Mitchell, Mick Corporation, and the Multilateral Investment Riordan, Cristina Savescu, Shane Streifel, and Guarantee Agency. Dominique van der Mensbrugghe. Gauresh The principal author was Mansoor Dailami, Rajadhyaksha managed and maintained the mod- with direction by Uri Dadush. The report was pre- eling and data systems. Mombert Hoppe, Denis pared under the general guidance of François Medvedev, Sebnem Sahin, and Shuo Tan pro- Bourguignon, World Bank Chief Economist and vided research assistance and technical support. Senior Vice President. The principal authors of Contributors to regional outlooks included each chapter were: Milan Brahmbhatt (East Asia and Pacific); Asad Alam, Cheryl Gray, and Ali Mansoor (Europe and Overview Mansoor Dailami, with contribu- Central Asia); Ernesto May and Guillermo Perry tions from the International (Latin America and the Caribbean); Mustapha Finance Team and William Shaw Nabli (Middle East and North Africa); Ejaz Syed Chapter 1 Andrew Burns Ghani (South Asia); and Delfin Go (Sub-Saharan Chapter 2 Mansoor Dailami, Ismail Dalla, Africa). Dilek Aykut, Eung Ju Kim The online companion publication, Prospects Chapter 3 Douglas Hostland, William Shaw, for the Global Economy, was prepared by Andrew and Gholam Azarbayejani Burns, Sarah Crow, Cristina Savescu and Shuo Chapter 4 William Shaw, Dilek Aykut, Tan with the assistance of Roula Yazigi and Shu- Jacqueline Irving, and Neeltje nalini Sarkar and the Global Trends team. Techni- Van Horen cal help in the production of that Web site was Chapter 5 Mansoor Dailami, Johanna provided by Reza Farivari, Sarubh Gupta, David Francis, and Eung Ju Kim Hobbs, Shahin Outadi, Raja Reddy Komati Reddy, Malarvizhi Veerappan, Cherin Verghese, Preparation of the commercial and official and Kavita Watsa. debt restructuring appendixes was managed by The report also benefited from the comments Eung Ju Kim, with inputs from Haocong Ren and of the Bank’s Executive Directors, given at an in- Gholam Azarbayejani. The financial flow, debt formal board meeting on May 4, 2006. estimates and the statistical appendix were devel- Many others provided inputs, comments, oped in a collaborative effort between DECPG guidance, and support at various stages of the re- and the Financial Data Team of the Development port’s preparation. Charles Collyns (International Data Group (DECDG), led by Ibrahim Levent Monetary Fund), Ishrat Husain (Former Gover- and including Nevin Fahmy, Shelly Fu, and Glo- nor, State Bank of Pakistan), Mark Sundberg, ria R. Moreno. Background notes and papers Michael Klein, and Stijin Claessens were discus- were prepared by Paul Masson (University of sants at the Bankwide review. In addition, within Toronto), Michael Pomerleano (Operations Delivered by The World and Bank e-library the Bank, commentsto: and help were provided by The World Bank Policy Department of the Bank’s Financial Sec- Alan Gelb, Alan Winters, Ali Mansoor, Asli IP : 192.86.100.36 tor), and Ivan Zelenko (Banking, Capital Mar- Tue, 10 Mar 2009 16:55:33 Demirguc-Kunt, Barbara Mierau-Klein, Anderson xiii (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Caputo Silva, Angela Gentile (MIGA), Brian Development Bank), Joyce Chang (JPMorgan Pinto, Cheryl Gray, Dan Goldblum, Deepak Bhat- Chase Bank), and William Cline (Institute for Inter- tasali, Doris Herrera-Pol, Ekaterina Vostroknu- national Economics). tova, Ellis Juan, Eric Swanson, Francis Jean-Fran- Steven Kennedy edited the report. Maria cois Perrault, Frannie Leautier, Gianni Zanini, Amparo Gamboa provided assistance to the Jeffrey Lewis, Joseph Battat (IFC), Marilou Uy, team. Araceli Jimeno and Dorota Agata Nowak Muthukumaras Mani, Punam Chuhan, Sergio managed the production of the report, while Schmukler Shahrokh Fardoust, Sona Varma, Ul- communication guidance and support for the re- rich Zachau, and Vikram Nehru. port were provided by Christopher Neal and Outside the Bank, several people contributed Cynthia Carol Case McMahon. Book design, through meetings and correspondence on issues ad- editing, production, and printing were coordi- dressed in the report. These include Hiro Ito (Port- nated by Susan Graham and Andres Ménèses of land State University), Boubacar Trore (African the World Bank Office of the Publisher. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 xiv (c) The International Bank for Reconstruction and Development / The World Bank . Selected Abbreviations ABF Asian Bond Finance IABs interest arrears bonds ABMI Asian Bond Market Inititative IDA International Development Association ADB Asian Development Bank (World Bank Group) ADRs American Depositary Receipts IDB Inter-American Development Bank AfDF African Development Fund IMF International Monetary Fund ASEAN Association of Southeast Asian Nations IPO initial public offering ASW asset swap LDCs least developed countries BIS Bank for International Settlements mbpd million barrels per day CDS credit default swap MDGs Millennium Development Goals CPIA Country Policy and Institutional MDRI Multilateral Debt Reduction Initiative Assessment MERCOSUR Southern Cone Common Market DAC Development Assistance Committee (Mercado Común del Sur) (OECD) MIGA Multilateral Investment Guarantee DRC Democratic Republic of Congo Agency DSF Debt Sustainability Framework NAFTA North American Free Trade Agreement EBRD European Bank for Reconstruction and NDF nondeliverable foreign exchange Development forward market ELMI Emerging Local Markets Index ODA official development assistance EMBI Emerging Markets Bond Index OECD Organisation for Economic Co-operation EMBIG Emerging Markets Bond Index Global and Development EMCDS emerging market credit default swap OPEC Organization of Petroleum-Exporting EMEAP Executives’ Meeting of East Asia and Countries Pacific Central Banks PAIF Pan-Asian Bond Index Fund EU European Union PPP purchasing power parity FDI foreign direct investment ROSCs reports on the observance of standards FLIRBs Front-Loaded Interest Reduction Bonds and codes (IMF and World Bank) FoBF Fund of Bond Funds SAARC South Asian Association for Regional G-3 Group of Three (European Union, Cooperation Japan, United States) SADC Southern African Development G-7 Group of Seven (Canada, France, Community Germany, Italy, Japan, United Kingdom, SBI State Bank of India United States) SME small and medium enterprise G-8 Group of Eight (G-7 plus Russian SOE state-owned enterprise Federation) UAE United Arab Emirates G-90 Group of Ninety (developing countries) UNCTAD United Nations Conference on Trade GDF Global Development Finance (World and Development Bank) WDI World Development Indicators (World GDP gross domestic product Bank) GEP Global Economic Prospects (World WDR World Development Report (World Bank) Bank) GNI gross national income WHO World Health Organization Delivered by The World Bank e-library to: HIPCs heavily indebted poor countries The World Bank WTO World Trade Organization IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 xv (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank . Overview and Policy Messages: The Development Potential of Surging Capital Flows 2 005 WAS A LANDMARK YEAR IN to scale up aid significantly and to further reduce global development finance, in both the of- the debt burdens of heavily indebted poor coun- ficial and private spheres. International pri- tries (HIPCs) to provide additional financial re- vate capital flows to developing countries reached sources needed to make progress on the MDGs. In a record net level of $491 billion. The increase in keeping with those objectives, donors have en- private capital flows in 2005 was broad-based, hanced their aid effort over the past few years and with long-term bond issuance, bank lending, and taken steps to improve the allocation of aid by pro- portfolio equity showing strong gains. A wave of viding more development assistance to the poorest privatizations and cross-border mergers and ac- countries, particularly those in Sub-Saharan Africa. quisitions drew substantial foreign direct invest- Donors also have provided targeted support for ment (FDI). Governments and private entities trade facilitation and developed a framework for took advantage of favorable financial-market improving the effectiveness of aid. Overall, aid in conditions to refinance outstanding debt and fund the form of grants and concessional loans has risen, future borrowing, while local-currency bond mar- while net lending by the official sector on noncon- kets in Asia and Latin America attracted substan- cessional terms has declined significantly. tial interest from international investors in search The global economy grew at a robust pace of of higher yields and potential gains from currency 3.6 percent in 2005, with the developing world ex- appreciation. Meanwhile, financial integration ceeding 5 percent growth for the third year run- among developing countries continued to deepen. ning. Global economic and financial conditions re- Capital flows between developing countries (so- main favorable, on the whole, despite several called South–South flows) are now growing more potentially destabilizing developments, notably rapidly than North–South flows, particularly FDI. high and volatile oil prices, growing global finan- The strong gains in private capital flows have cial imbalances, and rising short-term policy inter- been supported by financial innovations, notably est rates in some of the major industrial countries. local-currency financing and structured financial International financial markets have remained re- instruments, such as credit default swaps and silient to the test of several major credit events, in- other derivatives, which have improved the ability cluding the downgrading of two major U.S. au- of investors to manage their exposure to the risks tomakers and the settlement of backlogged credit associated with emerging market assets. derivatives contracts that had come to the attention Development finance took center stage at a se- of U.S. regulatory authorities. The upward trend ries of major international forums in 2005. With a in private capital flows appears to have continued decade remaining to attain the Millennium Devel- in the early months of 2006, and the short-run opment Goals (MDGs), expectations for a big push prospects are good. But the external environment in development assistance escalated over the course could well prove less auspicious in the future than Delivered by The World Bank e-library to: of the year, with a strong focus on Sub-SaharanThe World Bank years, depending critically on the course in recent Africa, the only region not on track to meet any of and dynamics of the necessary rebalancing of IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 the goals. There was broad agreement on the need global savings and investment patterns to underpin 1 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 the orderly unwinding of large and unsustainable Figure 1 Financial flows to developing countries, global financial imbalances. 1997–2005 The surge in private capital flows offers na- $ billions tional and international policy makers a major op- 500 portunity to bolster development efforts if they can successfully meet three challenges. The first is 400 to ensure that more countries, especially poorer 300 Total net capital flows ones, enhance their access to developmentally ben- eficial international capital through improvements 200 Net private flows in their macroeconomic performance, investment 100 climate, and use of aid. The second is to avoid sud- Net official flows den capital flow reversals by redressing global im- 0 balances through policies that recognize the grow- –100 ing interdependencies between developed and 1997 1998 1999 2000 2001 2002 2003 2004 2005 developing countries’ financial and exchange rate Source: World Bank Debtor Reporting System and staff estimates. relations in the determination of global financial liquidity and asset price movements. And the third is to ensure that development finance, both official issuance has dropped for many developing coun- and private, is managed judiciously to meet the de- tries, as long-term interest rates in industrial coun- velopment goals of recipient countries while pro- tries remain low (despite increases in short-term moting greater engagement with global financial rates in the Euro Area, the United States, and else- markets. These are the themes and concerns of this where) and spreads on emerging market sovereign year’s edition of Global Development Finance. bonds continue to decline. Those spreads reached a record low of 174 basis points in May 2006 (fig- ure 2). Short-term borrowing remained at approx- imately the same level as in 2004 and about $14 The broad surge in private capital billion higher than in 2003, in sharp contrast to flows continues the negative flows of short-term debt that were N et capital inflows from official and private sources increased from $418 billion in 2004 to $472 billion in 2005. While net official lending seen from 1998 to 2001. The rise in private flows also was widespread, with all regions experiencing an increase (table 2): was negative, net flows of private capital to devel- oping countries swelled for the third consecutive • A surge in flows to the Russian Federation year, reaching $491 billion in 2005, the highest and Turkey helped to boost flows to the Eu- level on record (figure 1 and table 1). Demand for rope and Central Asia region to $192 billion emerging market debt and equities remained in 2005, up from $160 billion in 2004. The strong, spurred by improved fundamentals in region accounts for 39 percent of developing many developing countries and investors’ search countries’ private flows, almost double the for higher yields in an environment where long- share it commanded in 2001. term interest rates remain low in major industrial • Stronger bond and equity activity increased countries, despite higher short-term interest rates. private flows to Latin America and the Developing countries’ finances also received a Caribbean from $59 billion in 2004 to $94 boost from workers’ remittances, which continued billion in 2005. But the region’s share of pri- their steady increase of the past decade (box 1). vate flows to the developing world plum- The increase in private capital flows has been meted from 45 percent in 2000 to 19 percent broad-based, extending across most debt and eq- last year. uity components and across most of the develop- • Flows to East Asia and the Pacific increased to ing world. Long-term bond flows (up $19 billion $138 billion from $125 billion the year be- Delivered by The World Bank e-library to: over 2004), medium- and long-term bank lending The World Bank fore, despite lower FDI to China. A marked (up $28 billion), and portfolio equity (up IP$24 bil- : 192.86.100.36 strengthening in flows to several regional Tue, 10 Mar 2009 16:55:33 lion) showed the strongest gains. The cost of bond economies explains the increase. 2 (c) The International Bank for Reconstruction and Development / The World Bank O V E R V I E W A N D P O L I C Y M E S S A G E S Table 1 Net capital flows to developing countries, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e Current account balance –84.5 –89.4 –4.0 47.1 18.8 69.8 122.3 153.1 248.4 as % GDP –1.5 –1.6 –0.1 0.8 0.3 1.2 1.8 1.9 2.6 Financial flows: Net equity flows 199.3 179.4 195.9 182.9 183.3 166.1 186.8 248.8 298.9 Net FDI inflows 168.7 172.4 183.3 168.8 176.9 160.3 161.6 211.5 237.5 Net portfolio equity inflows 30.6 6.9 12.6 14.1 6.4 5.8 25.2 37.3 61.4 Net debt flows 107.2 54.3 16.3 –1.0 –1.5 10.7 72.8 119.1 120.1 Official creditors 13.1 34.3 13.9 –5.7 27.4 5.2 –12.3 –28.7 –71.4 World Bank 9.2 8.7 8.8 7.9 7.5 –0.2 –0.9 1.3 0.7 IMF 3.4 14.1 –2.2 –10.7 19.5 14.0 2.4 –14.7 –41.1 Others 0.5 11.5 7.3 –2.9 0.4 –8.6 –13.8 –15.4 –31.0 Private creditors 94.1 19.9 2.5 4.7 –28.9 5.5 85.1 147.8 191.6 Net medium- and long-term debt flows 85.0 85.7 22.0 11.5 –6.2 1.2 30.2 77.8 122.3 Bonds 38.4 40.6 30.6 20.5 11.0 10.8 26.4 43.0 61.7 Banks 44.0 50.3 –7.1 –5.2 –10.8 –2.8 9.8 39.4 67.4 Others 2.7 –5.2 –1.5 –3.8 –6.3 –6.8 –5.9 –4.6 –6.7 Net short-term debt flows 9.2 –65.8 –19.6 –6.8 –22.7 4.2 54.9 70.0 69.3 Balancing itema –169.5 –127.8 –175.0 –183.6 –118.8 –74.7 –90.3 –116.2 –274.5 Change in reserves (– = increase) –52.4 –16.4 –33.2 –45.4 –81.7 –171.9 –291.6 –404.8 –393.0 Memo items: Bilateral aid grants (ex technical cooperation grants) 25.3 26.7 28.5 28.7 27.9 32.5 43.7 50.3 52.6 Net private flows (debt+equity) 293.5 199.3 198.4 187.6 154.4 171.5 271.9 396.6 490.5 Net official flows (aid+debt) 38.3 61.1 42.4 23.0 55.3 37.7 31.4 21.6 –18.8 Workers’ remittances 71.2 73.1 77.0 85.2 96.4 113.2 141.2 161.1 166.8 Sources: World Bank Debtor Reporting System and staff estimates. a. Combination of errors and omissions and net acquisition of foreign assets (including FDI) by developing-country private entities. e = estimate. Box 1 International migrant remittances R emittances are the largest source of external financing in many developing countries. According to official statistics, in 2005 remittance flows—defined as the sum of • By generating a steady stream of foreign exchange, re- mittances can improve a country’s creditworthiness and enhance its access to international capital mar- workers’ remittances, compensation of employees, and mi- kets. grant transfers in the balance-of-payments statistics col- Recorded remittance flows to developing countries lected by the International Monetary Fund—are estimated have doubled over the past five years, for several reasons. to have exceeded $233 billion worldwide, of which devel- Increased scrutiny of financial transactions since the terror- oping countries received $167 billion. Unrecorded flows ist attacks of September 2001 has made remittances more moving through informal channels push the total far visible. With the growth of competition in the remittance higher, as they are conservatively estimated to amount to industry, costs have dropped in major corridors, while at least 50 percent of the recorded flows. networks have expanded. Recently, high oil prices have Remittances bring substantial benefits to developing swelled remittance flows from oil-exporting countries. countries: Bahrain, Kazakhstan, Kuwait, Oman, Saudi Arabia, and • Household survey evidence, confirmed by cross- the Russian Federation have been important sources of re- country analyses, indicates that remittances can have mittances to developing countries. The depreciation of the a significant impact on reducing poverty. U.S. dollar (which raises the value of remittances denomi- • Remittances are associated with increased household nated in other currencies) and growth in the number of mi- investment in education, entrepreneurship, and grants and their incomes have contributed further to the health—all of which have a high social return under increase. most circumstances. Delivered by The World Bank e-library to: • Remittances tend to be countercyclical and thus sup- Bank The World Source: World Bank, Global Economic Prospects 2006; World Bank staff port economic activity in the face of adverseIPshocks. : 192.86.100.36 calculations based on various data sources. Tue, 10 Mar 2009 16:55:33 3 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 2 Benchmark spreads for emerging emerging market economies boosted portfolio eq- markets, 2001–6 uity flows to a record $61 billion, up from $37 bil- Basis points lion in 2004. However, private capital flows re- main concentrated in just a few countries. In 2005 1,000 about 70 percent of bond financing and syndi- 900 cated lending went to ten countries; three coun- 800 tries (China, India, and South Africa) accounted for almost two-thirds of all portfolio equity flows. 700 Rather than fueling domestic investment, the 600 rise in net inflows of private capital in 2005 fi- 500 nanced a substantial rise in developing countries’ 400 official reserve assets (almost as large as the record increase in 2004) and a very sharp increase in the 300 accumulation of foreign assets by private entities— 200 to $258 billion, again a record level (see figure 3). 100 The opening of capital accounts in the devel- Jan-01 Jan-02 Jan-03 Jan-04 Jan-05 Jan-06 oping world has increased opportunities for capital outflows, enabling developing-country residents to Source: JPMorgan Chase. Note: As of April 7. improve their investment returns and reduce their risks through international diversification. The creditworthiness of most developing countries continued to improve in 2005, as up- grades by credit rating agencies handily outpaced Global growth has propelled downgrades. Moreover, the pace of credit up- the surge in capital flows, grades rose to 46 in 2005, up from 31 in 2004. but serious risks remain Many developing countries have taken advantage of the favorable financial conditions by issuing bonds with longer maturities in international mar- G lobal growth has remained surprisingly re- silient to the rise in world oil prices over the past few years. Despite a doubling of oil prices kets—in some cases denominated in local cur- from early 2003 to late 2005, world GDP ex- rency. Others have been able to buy back existing panded by a robust 3.6 percent in 2005. Develop- debt using the proceeds of new bonds issued at ing countries led the way, with GDP growth of 6.4 lower rates. Also, many countries have pre-funded percent, more than twice the rate of high-income future financing requirements. Syndicated bank countries (2.8 percent). lending to developing countries set records in The impact of higher oil prices on economic 2005. Gross bank lending of $198 billion, an in- growth and inflation has been more subdued than crease of 77 percent over 2004, involved 1,261 in previous episodes. Global growth was down transactions in a broad range of sectors, domi- only 0.5 percentage points, and the expansion nated by oil-and-gas projects and oil-import fi- among developing countries was 0.7 percentage nancing. Meanwhile, booming stock markets in points, slower than in 2004. The reduced impact Table 2 Net private capital flows to developing countries by region, 1998–2005 $ billions 1998 1999 2000 2001 2002 2003 2004 2005 East Asia and Pacific 6.5 28.8 28.0 39.2 58.9 81.5 125.4 137.7 Europe and Central Asia 66.7 50.9 51.5 33.1 59.7 101.1 160.2 191.7 Latin America and the Caribbean 98.9 95.8 85.2 59.5 28.2 49.9 59.3 94.4 Middle East and North Africa 8.1 2.6 3.3 4.8 8.3 7.8 8.3 14.6 South Asia Delivered 5.3 by The World 3.5 Bank 9.7 e-library to: 5.8 10.1 15.8 22.7 23.6 The World Bank Sub-Saharan Africa 13.7 16.7 9.9 12.1 6.3 15.8 20.7 28.4 IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 Sources: World Bank Debtor Reporting System and staff estimates. 4 (c) The International Bank for Reconstruction and Development / The World Bank O V E R V I E W A N D P O L I C Y M E S S A G E S Figure 3 Capital outflows by private entities in the unwind in a disruptive fashion remains a risk— developing world, 1981–2005 particular for heavily indebted countries and those $ billions with close economic ties to the United States. A sec- 260 ond risk is that a sharp supply shock might send oil prices even higher, with potentially serious conse- 210 quences for the most energy-dependent developing economies. A fall in nonoil commodity prices could 160 have similar consequences for some of the poorest countries, which have benefited from higher metals 110 and mineral prices. There is also a possibility that the current glut of liquidity in global financial mar- 60 kets may have caused investors to underprice the risk of emerging market assets (both debt and eq- 10 uity). Political risk has reemerged as a key con- cern for investors in several emerging market – 40 economies, where elections could portend major 1981 1985 1989 1993 1997 2001 2005 changes in policy direction. Finally, there is a risk Sources: International Financial Statistics, IMF; and World Bank that avian influenza (bird flu) could mutate into a staff calculations. form that is easily transmitted between humans Note: The size of the increase in private assets is hard to judge, since it is calculated as a residual and thus includes errors and omissions and for which the population has limited immu- from elsewhere in the balance of payments. nity. Depending on the severity of the eventual dis- ease, such a pandemic could kill between 14 mil- lion and 70 million people and lower global GDP reflects several factors, notably lower oil intensi- by between 2 and 5 percent (with the latter num- ties, more flexible product and labor markets, ex- ber implying a global recession). change rate flexibility, and more credible mone- tary policy. Higher nonoil commodity prices have offset the impact of higher oil prices on the terms of trade of some countries. Capital flows are being transformed Higher oil prices have had a major influence Financial integration among developing on the external and fiscal positions of most devel- countries oping countries, however. For net oil exporters, For much of its postwar history, development fi- higher oil prices have meant significant increases nance has been characterized as a one-way flow of in external and fiscal surpluses, and higher foreign capital from industrial countries to the developing exchange reserves. For net oil importers, healthy world. But as developing countries have become current-account surpluses and ample foreign ex- more integrated with the global economy, they change reserves made it possible to cover the siz- have emerged as important sources of capital able increase in oil-import bills. Considerable in- flows in their own right. In the past decade, with creases in foreign aid for some of the poorest rising incomes in developing countries and increas- countries, particularly those in Sub-Saharan ingly open policies toward trade and financial Africa, provided an additional source of foreign markets, developing countries have become a sig- currency. But fiscal deficits have risen alarmingly nificant source of FDI, bank lending, and even of- in countries that subsidize domestic energy prices. ficial development assistance (ODA) to other de- Despite high oil prices, growth in developing veloping countries. countries is expected to remain above 5 percent per Overall, growing FDI between developing year during the period 2006–8, well above the per- countries in recent years has sometimes compen- formance of the past two decades, and with infla- sated for reductions in FDI flows from high-income tionary pressures in check. The main risks to this countries. But South–South capital flows, in partic- Delivered by The World Bank e-library to: relatively benign outlook are broadly unchangedThe World have also opened opportunities for low-income ular, Bank since the last edition of Global Development Fi- countries, because developing-country investors are IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 nance. The possibility that global imbalances might often possibly better able to handle the special risks 5 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 encountered in poor countries. Banks from develop- market because of infrastructural shortcomings, a ing countries play an increasingly prominent role lack of regulatory frameworks robust enough to in cross-border lending to low-income countries— cope with the market’s dynamic nature, and the borrowers in low-income countries received 17 concentrated participation of a small number of percent of total South-South cross-border syndi- dealers in emerging markets, which carries the risk cated lending flows in 2005, up from 3 percent in that failure of a single player could have a destabi- 1985. Moreover, 27 percent of foreign bank assets lizing impact on the market. in low-income countries are held by developing- country banks, compared to just 3 percent in mid- Domestic debt markets dle-income countries. South–South FDI is signifi- Local-currency bond markets in developing coun- cant for many low-income countries, particularly tries have, since the crises of the 1990s, emerged as those located close to major investors. a major source of long-term development finance. Although South–South capital flows remain They are now the fastest-growing segment of relatively small compared to North-South flows, emerging market debt. Driven largely by domestic they have the potential to change the landscape of institutional and individual investors, these mar- development finance over the next few years, par- kets grew from $1.3 trillion at the end of 1997 to ticularly if growth in developing countries contin- $3.5 trillion in September 2005. Their rapid ues to outstrip that in advanced countries and the growth has enabled major developing countries to trend toward deeper trade and financial integra- improve debt management by reducing currency tion persists. and maturity mismatches. Robust domestic bond markets have also improved financial intermedia- Financial innovations tion and contributed to domestic growth, as both The market for debt issued by developing coun- the government and corporate sectors have readier tries is expanding beyond the dollar-denominated, access to long-term capital. However, bringing the high-yield, sovereign debt instruments that had local-currency bond markets in emerging come to define the emerging market asset class, as economies up to the standards of mature markets exemplified by Brady bonds (which will drop to will require concerted efforts akin to those of the only 6 percent of the original amount outstanding East Asian countries, which have yielded early suc- once announced buybacks are completed). Today, cesses. But local-currency debt markets also pre- the emerging market asset class includes a range of sent new challenges for policy makers. The devel- instruments in both local and foreign currency opment of domestic debt markets requires modern that offer the capacity to tap dollar and euro in- and professional debt management procedures— vestors alike and cater to the funding needs of to manage debt on an integrated basis (that is, both sovereign and corporate borrowers on both both local and international debt)—especially in the cash and derivatives sides of the market. countries with few capital controls. Credit default swaps—derivatives that pro- vide insurance against defaults—are being applied The global role of the euro in new ways in emerging markets. This has poten- Since its introduction on January 1, 1999, the euro tially important implications for the pricing and has assumed an increasingly important interna- supply of debt capital to developing countries, of- tional role. It has emerged as a principal issuing fering investors a new way to take on exposure currency in the global debt market, as a vehicle for and enhancing the markets’ ability to gauge credit foreign exchange transactions, and as an impor- risk. By transferring banks’ credit risk from lend- tant reserve currency for official holdings of for- ing and trading activities to other market partici- eign-exchange reserves. The elimination of ex- pants, credit derivatives have altered, perhaps fun- change risk within the Euro Area has created a damentally, the traditional approach to credit risk pan-European market for euro-denominated secu- management and the lending business. While the rities, attracting both sovereign and private bor- emergence of this market could improve the ability rowers, not only from Euro Area countries, but Delivered by The World Bank e-library to: of financial systems to diversify risk acrossThe World a Bankalso from other countries—among them emerging greater number of market participants, it IPremains market economies such as Brazil, Colombia, : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 a relatively immature and potentially vulnerable China, Mexico, and Turkey. Today’s euro-denomi- 6 (c) The International Bank for Reconstruction and Development / The World Bank O V E R V I E W A N D P O L I C Y M E S S A G E S nated bond market rivals the dollar-based fixed-in- Figure 5 Net official lending, 1997–2005 come markets in important respects, including $ billions size, depth, and product range. 20 Official bilateral creditors The euro is used increasingly in debt issuance, 10 because it is the home currency of a large set of in- vestors. It is less popular as a currency of denomi- 0 nation for reserves, owing to the dominance of the –10 World Bank dollar as a vehicle for foreign exchange transac- –20 tions and currency interventions—as well as the –30 greater liquidity of the market for U.S. Treasury securities. Nevertheless, if the deteriorating U.S. –40 IMF current-account deficit sufficiently undermines –50 confidence in the dollar, more official reserve hold- 1980 1985 1990 1995 2000 2005 ings could be moved into euro-denominated as- Source: World Bank Debtor Reporting System and staff estimates. sets, with the potential for a period of financial in- stability if the shift is abrupt. The dramatic decline in net official lending over the past few years reflects, for the most part, large repayments to the International Monetary Net official flows continue to decline Fund (IMF) and large prepayments to bilateral of- Official lending falling ficial creditors (figure 4). In 2005 net debt out- Net official flows of grants and loans continued to flows from developing countries to the IMF to- fall in 2005—for the fourth consecutive year—as a taled $41.1 billion, down from a net debt inflow sharp decline in net official lending more than off- of $19.5 billion in 2001, implying a –$60.6 billion set gains in bilateral aid grants (table 1 and figure swing in net lending by the IMF over the period 4). Net official lending came to –$71.4 billion in 2001–5. The sharp decline is due to large repay- 2005, the third consecutive year of net outflows ments on emergency assistance loans made to In- from developing countries. In three years, develop- donesia and the Russian Federation in 1997/8, and ing countries have repaid $112 billion in loans to to Argentina, Brazil, and Turkey in 2001/2. The creditors. This largely reflects repayments of non- sharp decline in 2005 reflects large repayments by concessional loans mostly by middle-income coun- Argentina ($2.4 billion), Brazil ($16.8 billion), tries. In contrast, aid (comprised of concessional Indonesia ($1.0 billion), the Russian Federation loans and grants) has increased significantly during ($2.3 billion), and Turkey ($4.2 billion). More- this period, particularly for low-income countries. over, gross lending by the IMF has declined from about $30 billion in 2002–3 to only $4 billion in 2005. This reflects the marked improvement in in- Figure 4 Net official lending and foreign aid grants ternational financial stability, supported by the fa- to developing countries, 1980–2005 vorable global economic and financial conditions. $ billions The IMF’s outstanding credit has declined from 50 Bilateral aid grants special drawing rights (SDR) 71 billion in 2002/3 to SDR 23.5 billion in March 2006. Despite the 25 low level of IMF credit outstanding, net lending by the IMF could continue to decline over the next 0 few years with large scheduled repayments by In- –25 donesia, Turkey, and Uruguay. Net lending by the official bilateral creditors –50 Net official lending declined by $27.0 billion in 2005 mainly due to large prepayments to the Paris Club by the Russ- –75 Delivered by The World Bank e-library to: The ian Federation World Bank ($15 billion), Poland ($5.6 billion) 1980 1985 1990 1995 2000 2005 and Peru ($2.0 billion). Russia financed a $15 bil- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 Source: World Bank Debtor Reporting System and staff estimates. lion prepayment to the Paris Club using domestic 7 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 financial resources, as its fiscal revenues increased debt burdens of poor countries that qualify. The dramatically in the wake of higher world oil 18 countries that reached the completion point prices. The prepayments by Peru and Poland were prior to May 2006, under the HIPC Initiative, will financed by borrowing in private capital markets, see their total debt stock fall from an average level effectively substituting private debt for official of 55 percent of GDP (before HIPC debt relief) to (Paris Club) debt. Large prepayments to the Paris 13 percent (after MDRI debt relief). Club are expected to continue into 2006. In May Debt relief together with other special-purpose 2006, Algeria and the Russian Federation made grants—for technical cooperation, emergency and offers to prepay all of its remaining Paris Club disaster relief, and administrative costs—has ac- debt, totaling $22 billion and $8 billion, respec- counted for a rising portion of ODA over the past tively. Paris Club creditors have indicated their few years. The increase in ODA as a share of GNI willingness to accept the proposals. Poland has an- since 2001 reflects higher special purposes, rather nounced its intention to prepay some of its €12.3 than more flexible forms of funding. Donors have billion debt to the Paris Club, which will be due reallocated aid to the poorest countries, particu- between 2005 and 2009. larly those in Africa, and have continued to shift their resources from concessional loans to grants, More aid for the poorest countries, and more with the goal of avoiding unsustainable increases in debt relief the debt burdens of aid recipients. Net disbursements of ODA by OECD DAC mem- ber countries increased dramatically in 2005, reaching $106.5 billion, up from $79.6 billion in 2004. Expressed as a share of gross national in- To ensure economic stability, come (GNI) in donor countries, ODA has risen developing countries must manage from 0.22 percent in 2001 to 0.33 percent in capital flows effectively 2005, just below the 0.34 percent peak reached in the early 1990s. However, most of the record $27 billion increase in 2005 reflects debt relief pro- The current surge in private capital flows has occurred in the midst of much-improved do- mestic policies and global financial conditions vided by Paris Club creditors to Iraq (nearly $14 compared with those that prevailed during the billion) and Nigeria (a little over $5 billion). Nev- capital flows surge of the 1990s. This time around, ertheless, even excluding debt relief, ODA rose by governments have so far generally managed to 8.7 percent in real terms, up from a 5.6 percent av- avoid excessive expansion of aggregate demand, erage annual increase over 2002–4. large current-account deficits, and sharp apprecia- ODA is likely to decline as a percentage of tions of the real exchange rate. However, the pol- GNI in 2006–7, as the debt relief component falls icy agenda for managing capital flows is broad to more normal levels, before increasing gradually and complex, and considerable challenges remain. through the end of the decade. Donors have made commitments to increase ODA by $50 billion by Effective macroeconomic policies 2010, half of which is targeted to go to Sub-Saha- The improved response to the surge in capital ran Africa. Based on those commitments, ODA flows this time around has been supported by the should reach 0.36 percent of GNI in 2010. Extrap- adoption of more flexible exchange rate regimes olating this rate of increase would mean that the and a monetary policy framework that favors UN target of 0.7 percent would not be attained price stability. Inflation has fallen dramatically in until 2030, 15 years after the 2015 deadline set for virtually all developing countries, from a median attaining the MDGs. of 11 percent in the mid-1990s to a median of 4.5 The international community made significant percent during 2002–5. At the same time, the progress in 2005 to reduce debt burdens in some greater autonomy in monetary policy afforded by of the poorest countries. Debt relief provided more flexible exchange rates has allowed authori- under the Heavily Indebted Poor Countries ties to lower local interest rates. Flexible exchange Delivered by The World Bank e-library to: (HIPC) Initiative and the Multilateral Debt World Bankrates and lower interest rates have drastically re- Reduc- The tion Initiative (MDRI) will significantly reduce the duced the incentive to resort to short-term exter- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 8 (c) The International Bank for Reconstruction and Development / The World Bank O V E R V I E W A N D P O L I C Y M E S S A G E S nal borrowing, a major vulnerability that con- Eastern European countries. The impact of indi- tributed to the financial crises of the 1990s. Gov- vidual risks could be magnified if several were to ernments also have taken steps to accelerate devel- occur simultaneously. opment of domestic capital markets (especially local bond markets) to create more diversified fi- Prudent accumulation of reserves nancial markets that would be more capable of The current account in many developing countries, handling volatile flows in portfolio capital. These particularly major oil exporters and emerging Asia, developments, along with the shift from debt fi- has moved from deficit to sizable surplus, intensify- nance to equity (particularly FDI), have con- ing the demand for reserve accumulation. That tributed to the marked improvement in developing many of these countries have accumulated foreign countries’ net external liability position. The ratio exchange reserves far in excess of the level required of external debt to GNI for developing countries for intervention and liquidity purposes partly re- as a whole fell from a peak of 44 percent in 1999 flects a desire to self-insure against global financial to about 34 percent in 2004, while since the mid- shocks. As the volume of reserves increases, how- 1990s short-term debt has declined in most devel- ever, so does the importance of balancing their use oping countries relative to long-term debt and for- for intervention, investment, and insurance pur- eign exchange reserves. poses against their domestic resource costs. For Progress has been made in simplifying the countries with large holdings of foreign exchange very complex web of capital controls and ex- reserves, allowing local institutional investors to di- change rate restrictions imposed by many coun- versify their investment portfolio globally—while tries. But the gradual opening of capital accounts ensuring their more effective regulation—could must be accompanied by a further strengthening provide a viable channel of capital outflow, as well of macroeconomic policies, the development of as an opportunity to further diversify risk. This local capital markets and the institutions needed would transfer currency risks, currently concen- to regulate them, and the establishment of a sys- trated on the books of central banks, to domestic tem of risk management robust enough to respond institutional investors with a longer investment to the needs of a more flexible exchange rate and horizon and a greater ability to manage such risks. open capital account. Liberalization of the capital Such an approach is also more desirable for many account once implemented is difficult to reverse. A developing countries than inducing adjustments return to capital controls should be seen only as a through the current account as a way of absorbing policy of last resort, to be used to dampen exces- reserves. In addition to allowing institutional in- sive exchange rate volatility or to moderate large vestors greater scope to invest overseas, considera- inflows of capital when other policies, such as in- tion should be given to enabling local residents to terest rates and intervention in foreign exchange invest in approved international assets, as the Re- markets, prove fruitless. public of Korea has done. Despite the considerable improvement in poli- cies in recent years, the surge in capital flows still Careful management of oil-export revenues presents substantial risks to developing countries. Oil-exporting countries face particular challenges Future risks to economic and financial stability in managing volatile export revenues. Although will likely take a different form and character than high oil prices are now expected to persist, consid- those encountered in the past—and may expose erable uncertainty remains, and oil exporters institutional and macroeconomic weaknesses that should save a part of the windfall—for example, cannot be anticipated at this juncture. One warn- to reduce debt and make productive physical and ing sign of potential troubles has been the surge in social investments. Some countries have put aside portfolio inflows that has been associated with a a fraction of their oil revenues in a stable portfolio dramatic escalation of stock market prices and of diversified financial assets (referred to as “funds valuations in many developing countries, particu- for the future”), thus reducing the risk of overcon- larly in Asia, raising the risk of asset price bubbles. sumption of oil revenues and the potential for Delivered by The World Bank e-library to: Other signs of possible trouble are appreciated ex-The World Dutch disease. Such funds require robust gover- Bank change rates and current account deficits in some nance and legal frameworks to effectively insulate IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 9 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 earmarked oil wealth from political decisions bilize and disrupt international financial markets, guided by short-term agendas. The government which would cause all countries to suffer. must set and adhere to clear objectives for their in- Although a coordinated policy of interven- vestment, protection, and eventual use. Countries tion in foreign currency markets—similar to the that depend heavily on oil revenues should also Plaza Agreement of September 1985—is neither consider using derivatives to reduce the volatility desirable nor feasible (given the changes in global of future income. financial market conditions and actors over the past two decades); a degree of multilateral coop- Improvements in standards for the eration is needed to address the current global im- corporate sector balances. That approach, based on the mutual in- A growing number of developing countries have terests of deficit and surplus countries, should made considerable efforts to meet international reflect the structural asymmetry between interna- standards for transparency, corporate governance, tional reserve currencies and other currencies. At and the regulation and supervision of financial its center must be consensus on a blend of ex- systems. Although this is a global trend, individual change rate and aggregate spending adjustments countries take different approaches to adapting in- adequate to rebalance global aggregate demand ternational standards to their corporate environ- toward surplus countries without causing a global ment. Some, for example, are issuing codes that set recession. Ordinarily, policy coordination among compliance targets in tandem with laws setting key players is unnecessary, because floating ex- minimum compulsory standards, while others are change rates, accompanying monetary policies using codes to raise public awareness in advance (oriented primarily toward domestic targets for of upcoming regulatory reform. The adoption of inflation and economic activity), and independent national codes of corporate governance in at least central banks do their job to facilitate adjustment 60 countries by the end of 2005—including all of to any shocks hitting the world economy. But the Asian crisis countries, plus China, Colombia, when the sustainability of the sources of finance Turkey, and Ukraine—underscores the growing for global payment imbalances is in doubt, as it is recognition of the importance of corporate gover- at present, multilateral cooperation to prevent nance in enhancing investor confidence, a recog- sudden and disorderly market reactions becomes nition that bolsters the resilience and stability of highly desirable, especially if the growing global capital markets globally. Priority must now be imbalances create pressure for protectionist trade given to effectively implementing and enforcing policies in some countries. these new domestic policy and institutional re- Developing countries, in particular, have much forms at the national level. to gain from multilateral cooperation, and much to lose from its absence, and they would suffer dispro- portionately if instability were induced and a disor- Multilateral cooperation is key to derly unwinding of global financial imbalances en- resolving global financial imbalances sued. The world economy is moving toward a D eveloping-country policies must be reinforced by renewed international efforts to promote stability and maintain a financial environment con- multipolar international monetary system in which the monetary and financial policies of the United States, Euro Area, Japan, and several key emerging ducive to a balanced expansion and deployment of market economies, including China, all exert sub- capital flows in developing countries. One major stantial influence. Policymakers in emerging mar- risk to stability is the growing imbalance in global ket economies should therefore strive to strengthen payments and the associated market anxiety about institutions and promote policies and mechanisms the possibility of a disorderly adjustment of the im- that will improve their ability to navigate in a balance through sudden changes in exchange rates world of increasingly integrated and interdepen- and global interest rates. Such changes could desta- dent financial and production systems. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 10 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank . 1 Prospects for the Global Economy Summary of the outlook higher oil prices. As a result, developing countries C onfronted with capacity constraints in the re- source sector, sharp rises in commodity prices, and a tightening of monetary policy among Or- are much more vulnerable to potential external shocks, such as a disruptive resolution of global imbalances, a decline in nonoil commodity prices, ganisation for Economic Co-operation and Devel- or a hike in oil prices following a supply shock. opment (OECD) countries, the global economy has slowed from the record pace posted in 2004. High oil prices have had only a limited impact Nevertheless growth remains robust, especially on global growth among developing countries. Their GDP increased Lower oil intensities, more flexible product and 6.4 percent in 2005 (4.3 percent for oil-importing labor markets, exchange rate flexibility, and more developing economies, excluding India and China) credible monetary policy have all reduced the real- as compared with 2.8 percent among high-income side and inflationary impacts of higher oil prices. As countries. The resilience of developing countries— a result, and in contrast to past episodes, monetary which reflects a sustained improvement in the po- policy has remained accommodative and interest tential growth rate of many developing countries— rates low. This, plus the fact that oil deliveries have has been heartening, especially given the magnitude continued to increase rapidly (as opposed to the of the oil-price shock. This brisk expansion is pro- 1970s and 1980s, when supply was cut), helps ex- jected to continue, but slow towards a more sus- plain the resilience of output to higher oil prices. An tainable pace of 5.9 percent by 2008. Such rapid additional factor for developing countries has been growth argues against a sharp decline in oil prices, the substantial rise in the share of exports in GDP, which are expected to remain above or close to $60 which has increased the foreign currency inflows a barrel through 2008. available to finance a given increase in the oil bill. This relatively benign soft-landing scenario Adjustment was facilitated by solid initial con- for developing countries faces both internal and ditions. In particular, many oil-importing develop- external risks. First, the high growth of the past ing countries entered the period of high oil prices several years is generating tensions within individ- running current-account surpluses and building up ual countries. In several East European countries foreign currency reserves. This, plus high nonoil this has taken the form of rising inflation, currency commodity prices and a rapid expansion in trade, appreciation, and high current-account deficits, meant that finding foreign currency to pay higher while in others it has expressed itself in rising asset oil bills was relatively easy. In addition, foreign cur- prices, inflationary pressure, and growing domes- rency inflows for the poorest countries were bol- tic tensions between fast and slower growing re- stered by increasing aid flows, which in many cases gions and sectors. Second, many of the buffers that rose by more than 0.5 percent of GDP in 2004 (the permitted countries to absorb higher oil prices last year for which data is available). Delivered by The World Bank e-library to: with a minimum of disruption have been ex-The World While Bank output has remained resilient, develop- hausted, and countries have yet to fully adjust to ing countries nevertheless have endured a large hit IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 13 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 on their incomes. On average, the rise in oil prices recent rise in oil prices more easily, similar policy between 2003 and 2005 reduced real incomes in steps would reduce their vulnerability both to fur- oil-importing countries by 3.6 percent and by as ther oil shocks and other shocks, including a de- much as 10 percent for some low-income oil im- cline in nonoil commodity prices. For countries porters. For developing oil importers the addi- benefiting from fixed-price contracts at what are tional expenditure, some $137 billion annually, currently below-market prices, policy should en- exceeds by a large margin official development as- courage energy conservation now before the con- sistance (ODA, $84 billion in 2005 net of addi- tracts expire or are renegotiated. tional debt relief) and is about one-half of foreign More generally, because higher prices are direct investment (FDI) inflows ($234 billion). likely to be a more or less permanent fixture, coun- Unsurprisingly, some countries are having dif- tries need to take steps to improve their interna- ficulty adjusting. Fiscal deficits have risen alarm- tional competitiveness. Policies that stimulate pro- ingly in several countries that subsidize domestic ductivity growth and investment in the domestic energy prices. In many African countries, utility economy are most likely to be successful. Countries firms, unable to pay mounting energy bills, have with flexible exchange regimes are likely to have imposed rolling blackouts. Moreover, a few coun- more success in improving their export revenues tries appear to be financing their higher oil bill and diminishing nonoil imports so as to reestablish through an unsustainably rapid reduction in inter- a comfortable margin on the current account. national reserves. Finally, rising food and trans- Trade reform—domestic, behind-the-border re- portation prices have pushed inflation to worri- forms to improve competitiveness, accompanied by some levels in several countries in Africa and, to a progress at the multilateral level—could further ex- lesser extent, South Asia. While it is not clear that pand developing-country exports and the base an inflationary spiral has begun, an eventual eco- upon which oil and other imports essential to de- nomic slowdown appears likely if policy makers velopment can be financed. are forced to use macro policy measures to bring For oil exporters the challenge will be to use inflation back under control. petroleum revenues in a way that minimizes eco- nomic distortions and maximizes development Developing countries face further adjustment gain. Even if oil prices remain high for an extended challenges over the medium term period, most countries do not have the capacity to While the resilience of output to high oil prices is absorb these huge inflows immediately. As a result, heartening, the initially comfortable current-account they should resist the temptation to use oil-related positions that allowed many developing countries to budgetary revenues for programs that are politi- weather higher oil prices have now been absorbed. cally popular but developmentally unsound. In- Moreover, many of the factors that allowed coun- stead, they should consider introducing or expand- tries to deal with higher oil prices relatively easily in ing oil funds by sequestering that part of revenues the short run imply that much real-side adjustment that cannot be productively placed in the domestic has yet to occur. market and investing it abroad, where it will gener- Adapting to more or less permanently higher ate a permanent income stream to support develop- prices poses substantial challenges, especially for ment even after current prices ease or oil supplies those countries where high oil prices are already dwindle. Recent steps by some oil-exporting coun- generating economic strain, as evidenced by exces- tries that have unwound structural reforms for sive increases in current-account or fiscal deficits short-term political gain are unlikely to be helpful. or by unsustainable financing of oil import bills through the depletion of reserves or bank borrow- Global imbalances may have been exacerbated ing. Policy makers in these countries must take ur- by high oil prices gent steps to increase energy efficiency in general The rapid rise in oil prices has contributed to and reduce oil dependency in particular. Unwind- global imbalances by increasing the U.S. current- ing energy subsidization programs would simulta- account deficit by some $125 billion since 2002. Delivered by The World Bank e-library to: neously relieve pressure on government The World BankIt also has changed the nature of those imbalances finances and also promote private sector energy conserva- by inducing a swing in the counterparts to the IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 tion. For those countries that have managed the U.S. deficit away from oil importers and toward 14 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y oil-exporting countries. Their oil-related export economies to finance current levels of consumption earnings are up some $400 billion since 2002. and investment is vulnerable to changes in investor These are being recycled—partly through in- confidence or additional external shocks. Else- creased imports, approximately 65 percent of ad- where, rapidly rising incomes may be contributing ditional export revenues are being spent as addi- to asset bubbles in regional real estate and stock tional imports, and partly via financial flows. As a markets. In other countries, tensions arising from result, there is little likelihood that an excess in oil localized labor market shortages, combined with exporters’ savings will lead to a global slowdown. significant disparity in the degree to which regions Rather, increased financial flows—either directly or segments of the population are benefiting from or through third-party intermediaries, are con- growth, could prompt a harder-than-projected tributing to low interest rates and, both directly landing. These internal risks could generate a hard and indirectly, to the financing of the U.S. current- landing on their own or they could be triggered by account deficit. and exacerbate an external shock. In particular, Despite the ease with which the U.S. deficit is growth in several countries in South Asia and a few being financed, the continued accumulation of for- in Latin America is generating significant inflation- eign liabilities is not sustainable. Unwinding these ary pressures requiring a tightening of macroeco- imbalances will almost certainly take a long time. nomic policy if an abrupt slowdown in the future is Indeed, given the magnitude of the required ad- to be avoided. justment, a gradual approach is to be preferred to The principal external risks to the global econ- an abrupt one. However, the longer significant omy have not changed much since the publication steps to resolve the issue are delayed the greater of the last edition of the World Bank’s Global Eco- will be the tensions implicit in the disequilibrium nomic Prospects (2005). These include the possibil- and the risk that they will be resolved in a disor- ity that persistent global imbalances will resolve derly manner. Of particular concern is that some themselves in a disorderly manner, either through a of the temporary factors holding down interest significant increase in interest rates or a sharp de- rates (including corporate balance-sheet restruc- preciation of the dollar; the possibility that a signif- turing and financial flows from oil revenues) will icant supply shock will send oil prices even higher; ease, increasing the servicing costs on U.S. liabili- and the possibility that nonoil commodity prices ties. That would add to the deficit and possibly will weaken. Should any of these risks be realized, raise concerns about its sustainability, driving in- they might reduce global growth by between 1 and terest rates even higher. 3 percent, depending on the shock, with much of Resolving these imbalances is a common but the slowdown borne by developing economies. differentiated responsibility requiring increased pri- Even if the impact of the shock is relatively benign vate and public savings in the United States, in- at the global level, the increased current-account creased demand outside of the United States, and deficits of many oil-importing developing countries more flexible exchange rate management. Action make them vulnerable. For heavily indebted coun- on all fronts is required, particularly because in the tries, the most serious risk stems from the possibil- absence of higher U.S. savings, increased foreign ity of higher interest rates. For small oil-importing demand or exchange rate appreciation is unlikely African countries, the largest risk is that nonoil to have a meaningful impact on imbalances. commodity prices, particularly for metals and min- erals, will decline. The outlook for developing countries carries The outturn from the Doha trade liberaliza- both internal and external risks tion round poses a balanced risk to the outlook. Prospects for a soft landing among developing The baseline scenario assumes an unambitious ac- countries are good, but a hard landing is also possi- cord. However, an ambitious conclusion to the ble. In particular, many countries, notably in the Round, including significant liberalization of trade Europe and Central Asia region, now have current- in agricultural products and on-the-ground account deficits that exceed 5 or 6 percent of GDP. progress in the aid-for-trade agenda, could yield Delivered by The World Bank e-library to: In some instances those deficits are associated withThe World Bank benefits for developing countries. More substantial high interest rates, strong capital inflows, and ap- importantly, a failure of Doha could go beyond this IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 preciating currencies. The future ability of these agreement by weakening the whole multilateral 15 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 trade liberalization process—resulting in a more Outturns and prospects fragmented path forward with fewer benefits for in high-income countries developing countries. Growth among industrialized economies in 2005 came in at 2.8 percent, substantially lower than While a remote possibility, an influenza the 3.3 percent recorded the year before. Industrial pandemic could have serious consequences production and trade flows among these countries The continued spread of avian influenza (bird flu) were particularly anemic. Industrial production among wild birds, with limited bird-to-human growth declined from more than 5 percent in mid- transmission, comprises part of the baseline fore- 2004 to less than 1 percent in late spring. Growth cast. A serious risk to the global economy is pre- has since accelerated, reaching 3 percent (year- sented by the possibility that avian influenza mu- over-year) in the first quarter of 2006 (figure 1.1). tates into a form of the flu that is easily High oil prices, rising short-term interest transmitted between humans and to which the rates, a cooling of the housing market, and an un- population has only limited immunity. usually disruptive hurricane season helped slow The potential human and economic conse- growth in the United States to 3.5 percent in 2005 quences of such a pandemic are very large. They as compared with 4.2 percent in 2004. Partly re- depend importantly on the nature of the flu that flecting a bounce-back in activity following a emerges and on the reactions of people as it weak fourth quarter, GDP expanded 4.8 percent in spreads. Even a relatively moderate flu in terms of the first quarter of 2006. Although inflation transmission and mortality could have serious consequences for the world economy if the global Figure 1.1 Industrial production remains robust population has limited immunity. Estimates sug- gest that, depending upon the severity of the even- % change in volumes year-over-year tual disease, a combination of lost output due to 10 20 Developing countries illness, additional deaths, absenteeism, and private (ex. China) 18 and public efforts to avoid infection could lower 5 16 global GDP by between 2 and 5 percent (with the 14 latter number implying a global recession). More 0 12 High-income important, between 14 and 70 million people countries China 10 (right axis) could be killed. –5 8 Policy makers need to focus simultaneously 6 on two critical tasks: (1) further strengthening ef- –10 4 forts to monitor and curtail outbreaks of avian in- 2000 2001 2002 2003 2004 2005 2006 fluenza at points (such as domestic poultry flocks) Source: World Bank. where the likelihood is highest of the disease mu- tating into a viable human-to-human form; and (2) developing and putting systems in place to Figure 1.2 Inflation in high-income countries minimize the human cost of a pandemic if one % change, year-over-year does emerge, whether by developing effective con- 5 tainment strategies or improving the world’s capac- U.S. all goods and services ity to rapidly create and distribute vaccines. 4 EU all goods and services 3 2 Global growth D espite oil prices that reached $60 a barrel in the second half of the year, the world econ- omy grew by a very robust 3.6 percent in 2005. 1 U.S. core inflation EU core inflation 0 Delivered by The World Bank e-library to: Developing countries led the way, expanding by Bank Jan. July Jan. July Jan. July The World Jan. July Jan. 2002 2002 2003 2003 2004 2004 2005 2005 2006 6.4 percent, more than twice as fast as high-income IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 countries (table 1.1). Sources: World Bank, Datastream. 16 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y Table 1.1 The global outlook in summary % change from previous year, except interest rates and oil prices 2004 2005* 2006** 2007** 2008** Global conditions World trade volume 10.6 7.1 7.6 7.7 7.8 Consumer prices G-7 countriesa,b 2.1 2.6 2.2 1.8 1.8 United States 2.7 3.4 2.9 1.9 2.0 Commodity prices (US$ terms) Non-oil commodities 17.5 13.4 5.8 –3.2 –5.8 Oil price (US$ per barrel)c 37.7 53.4 64.2 61.0 56.9 Oil price (% change) 30.6 41.5 20.2 –5.0 –6.8 Manufactures unit export valued 6.9 0.8 1.6 2.8 1.2 Interest rates $, 6-month (%) 1.6 3.6 5.1 5.2 4.9 €, 6-month (%) 2.1 2.2 2.6 3.1 3.9 Real GDP growthe World 4.1 3.6 3.7 3.5 3.5 Memo item: World (PPP weights)f 5.3 4.6 4.6 4.5 4.5 High-income countries 3.3 2.8 3.0 2.8 2.8 OECD Countries 3.2 2.7 2.9 2.7 2.8 Euro Area 2.0 1.4 2.1 2.1 2.2 Japan 2.7 2.8 2.8 2.1 1.8 United States 4.2 3.5 3.5 3.3 3.3 Non-OECD countries 6.2 5.5 5.1 4.7 4.7 Developing countries 7.1 6.4 6.3 6.0 5.9 East Asia and Pacific 9.1 8.8 8.3 8.2 8.1 Europe and Central Asia 7.2 5.7 5.5 5.4 5.1 Latin America and Caribbean 6.0 4.4 4.6 4.0 3.7 Middle East and N. Africa 4.7 4.8 5.3 5.2 5.1 South Asia 6.7 7.7 6.8 6.5 6.2 Sub-Saharan Africa 5.2 5.2 5.4 4.9 5.4 Memorandum items Developing countries excluding transition countries 7.2 6.6 6.4 6.1 6.0 excluding China and India 6.1 5.0 5.1 4.8 4.5 Source: World Bank. Note: PPP = purchasing power parity; * = estimate; ** = forecast. a. Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States. b. In local currency, aggregated using 2000 GDP Weights. c. Simple average of Dubai, Brent and West Texas Intermediate. d. Unit value index of manufactured exports from major economies, expressed in US$. e. GDP in 2000 constant dollars; 2000 prices and market exchange rates. f. GDP mesaured at 2000 PPP weights. spiked following Katrina-related increases in gaso- quarter pause in exports following a strong accel- line prices, it has since declined and remains rela- eration in the first nine months of the year. Since tively muted at 3.4 percent in March 2006. Core then economic activity has picked up with GDP in inflation (price changes of goods and services the Euro Area estimated to have increased by other than energy and food) remains low at 2.1 around 2.4 percent in the first quarter of 2006. percent, below the rate recorded in December In Japan, growth has been strong, with indus- 2004 (figure 1.2). trial production ending the year up 5 percent and The relatively low oil intensity of European unemployment declining to 4.4 percent of the economies, significant excess capacity, and a re- labor force. Overall, GDP increased by 2.8 per- laxed macroeconomic policy stance limited the cent, with both domestic and external demand Delivered by The World Bank e-library to: slowdown in Europe. For the year as a whole,The World contributing Bank about equally to the overall result. As growth was a relatively weak 1.5 percent (1.4 per- a result, both consumer and business confidence IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 cent for the Euro Area), but this reflected a fourth- have improved. 17 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 The increase in oil prices in 2005 and early Figure 1.3 Developing-country growth remains 2006 are expected to slow growth in high-income robust countries by about 0.25 of a percentage point in GDP growth 2006 compared with what it would have been had 8 prices remained stable. In the United States, im- 7 Developing countries proved net exports are projected to maintain the 6 pace of growth in 2006, despite weaker consumer demand due to higher interest rates and a cooling 5 World of the housing market. For 2007/8, the balance of 4 these forces is expected to reverse somewhat, lead- ing to a moderate easing of growth. 3 Continued accommodative macroeconomic 2 policy and pent-up investment demand following 1 several years of very weak growth should maintain 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 the recent acceleration of output in Europe during Source: World Bank. 2006. As a result, GDP is projected to expand by about 2.1 percent in 2006 and to continue grow- ing at close to its potential rate in 2007/8. is projected to exceed 5 percent through 2008, In Japan, vigorous growth in developing East with the Latin America and Caribbean region Asia, renewed consumer and business confidence, projected to expand 4.1 percent on average over and reduced drag from consolidation are all ex- the projection period. pected to keep the recovery strong in 2006. While the economy is projected to slow somewhat (partly because of less expansionary monetary and fiscal policies), GDP should expand at or above Regional outlooks the economy’s potential rate of growth. More detailed descriptions of economic developments in de- veloping regions, including regional forecast summaries, are Developing economy outturns and prospects available at http://www.worldbank.org/globaloutlook. Notwithstanding high oil prices, economies in every developing region continued to grow at East Asia and the Pacific1 above-trend rates in 2005. Overall, the GDP of The economies of the East Asia and Pacific region low- and middle-income countries expanded by continued to expand rapidly in 2005. Their GDP is an estimated 6.4 percent. The expansion was par- estimated to have increased by 8.8 percent, down ticularly robust in China and India, where output from 9.1 percent in 2004 (figure 1.4). Growth in increased by 9.9 and about 8.0 percent, respec- China was very strong (9.9 percent), despite a sub- tively. Excluding these countries, growth in other stantial slowing in both private consumption and oil-importing developing countries came in at an investment demand, because exports continued to estimated 4.3 percent, down significantly from grow rapidly, and imports slowed. 5.7 percent in 2004. At the same time, dwindling For other countries in the region, output ex- spare capacity in the petroleum sector caused the panded by a more modest 5.3 percent, as the slow- expansion of oil-exporting developing economies down in Chinese imports, weak global high-tech to ease from 6.6 to 5.7 percent, even though oil demand, and elevated oil prices translated into re- revenues continued to rise. duced export growth and rapidly rising producer High oil prices, rising interest rates, and prices. Among larger oil-importing countries in the building inflationary pressures are expected to region, GDP growth slowed relatively sharply in restrain growth in most developing regions in the Philippines and Thailand. Among oil-exporters, 2006/8 (figure 1.3). As a group, however, low- growth slowed in Malaysia, but picked up in Viet- and middle-income countries should again out- nam and Indonesia. Delivered by The World Bank e-library to: perform high-income economies by a wide mar- Bank The World Strong exports and weak import demand in gin. Growth in five of the six developing regions IP : China 192.86.100.36 meant that the region’s current-account bal- Tue, 10 Mar 2009 16:55:33 18 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y Figure 1.4 Regional growth trends % change, GDP 10 2004 2005 2006 2007 2008 9 8 7 6 5 4 3 2 East Asia Europe & Latin America Middle East South Sub-Saharan Oil Oil importers & Pacific Central Asia & Caribbean & North Africa Asia Africa exporters (ex. China) Source: World Bank. ance improved, reaching a surplus of $143 billion trade effects and some currency appreciation are (4.9 percent of GDP). Of the larger economies, only projected to result in about a $25 billion decline in Thailand and Vietnam are running current-account the region’s current-account surplus. deficits, while the surpluses of China and Malaysia exceed 6 and 15 percent of their respective GDP. Europe and Central Asia Output in the region continues to feel the ef- Economic activity in the Europe and Central Asia fects of endemic bird-to-bird avian influenza. region grew by a robust 5.7 percent in 2005. High Cambodia, China, Indonesia, Laos, Thailand, and oil prices boosted demand in the region’s oil produc- Vietnam are the countries most affected. So far ers, particularly in the Russian Federation, where some 200 million domestic birds (less than 1 per- real GDP increased 6.4 percent. That, in turn, con- cent of domestic bird production in the region but tributed to strong exports for other countries in the rising to 12 percent in Vietnam) have died or been region, notably the Baltics and the Commonwealth killed to prevent the spread of the disease. As of of Independent States. Turkey and other Central early May 2006 no new outbreaks have been European countries participated in the export boom recorded among birds in Thailand and Vietnam, to a lesser extent, as they reoriented exports away attesting to the effectiveness of preventive mea- from a still weak European Union. sures. However, new outbreaks have been The region received record capital inflows in recorded in China, East Java, Indonesia, Malaysia 2005, reflecting favorable international credit con- and Myanmar2. ditions and the advancing EU accession process While the disease has had only a limited effect for new and candidate members. These flows con- on GDP so far (depending on the country, the sec- tributed to rapid credit growth in the Baltics, Bul- tor represents between 0.6 to 2 percent of GDP), garia, Romania, Turkey, and Ukraine, and a signif- its impact on incomes has likely been more acute. icant deterioration in current-account positions. Poultry accounts for as much as 7 percent of the High oil prices, substantial increases in the price incomes of the poor. paid for imported natural gas in some countries, As higher oil prices take hold, reduced invest- and lax fiscal policy in the Czech Republic, Hun- ment growth in China and reduced global liquidity gary, the Kyrgyz Republic, and Poland also are expected to slow regional growth to around 8.1 boosted current-account deficits. percent by 2008. This reflects a modest slowdown About half of the region’s economies posted in China, as slower export growth is partially offset current-account deficits equal to or in excess of 5 by stronger domestic demand. Excluding China, percent of GDP in 2005. Current-account deficits growth in the remaining economies in the region is exceeded 6 percent of GDP in Albania, Bulgaria, Delivered by The World Bank e-library to: expected to come in at about 5.5 percent in 2006The World Bank Estonia, Georgia, Hungary, Latvia, Croatia, through 2008. Stronger domestic demand, terms of Lithuania, Romania, and Turkey. IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 19 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 At the regional level these deficits were signifi- many countries. High nonoil commodity prices cantly offset by improved external positions of oil and strong inflows of remittance prevented most exporters, including Azerbaijan, where the deficit countries in the region from experiencing a signifi- shifted from a 30 percent share of GDP in 2004 to cant deterioration in their current-account posi- 5 percent in 2005, as new oil capacity came on tions. Indeed, with a few exceptions (Honduras, stream. This also propelled Azerbaijan’s growth to Nicaragua, Panama, Paraguay, and Uruguay), the more than 25 percent. current-account balances of most countries in the GDP growth is projected to slow slightly in region have either remained constant or improved 2006, coming in at 5.5 percent, as tighter inter- since 2002. These favorable external conditions national credit conditions and monetary policy contributed to a general pressure toward exchange are expected to slow domestic growth in the rate appreciation that has been checked by accu- Commonwealth of Independent States (CIS) sub- mulation of international reserves. region. Elevated energy revenues, investment ex- Looking forward, regional growth is pro- penditure, and the projected recovery of western jected to pick up in 2006 as easier monetary policy European demand are expected to sustain growth boosts output in Mexico and Brazil. Growth in at relatively high levels in 2007/8. High fiscal and most countries in the region is expected to be current-account deficits in a number of countries, broadly stable in 2007 and 2008, slowing only including Hungary and Turkey, pose serious risks somewhat in the face of a modest weakening in to the outlook. For regional oil exporters, key commodity prices and a gradual moderation in challenges include the need to foster greater in- capital inflows. However, the expansion for the re- vestment and productive capacity in the nonoil gion as a whole is projected to slow toward 3.7 sectors so as to improve economic diversification, percent in 2008, reflecting a significant slowing in control inflation, and prevent excessive exchange Argentina and República Bolivariana de Venezuela rate appreciation. toward more sustainable growth rates. Growth trends in Central American countries Latin America and the Caribbean are projected to improve, partly because of the re- Economic activity in Latin America and the cent Central American Free Trade Agreement. The Caribbean is estimated to have increased by some agreement should boost both trade (the United 4.4 percent during 2005. Outturns were strong States is these countries’ major trading partner) throughout the region, reflecting high levels of in- and investment, thereby lifting longer-term growth ternational liquidity, strong global demand, and prospects. However, to reap the full benefits of this high prices for the region’s exports. Macroeco- reform, further steps need to be taken towards im- nomic policy has also played a role. Except in proving road quality, increasing port and customs Brazil and Mexico, where rising interest rates con- efficiency, boosting financial depth, and raising the tributed to a slowdown in 2005, monetary policy quality and coverage of education. in the region has been generally accommodative. A central risk to this forecast remains the pos- Fiscal policy, in turn, has been relatively neutral. sibility that as growth slows and commodity prices Despite windfall revenues from high international ease, government deficits will rise, potentially rais- commodity prices and reduced debt servicing ing inflation or increasing uncertainty. Either re- charges (due to reduced interest rates and lower sult could lead to higher-than-projected interest debt stocks) most countries, with the notable ex- rates and slower growth. ception of República Bolivariana de Venezuela, have avoided a significant pro-cyclical surge in Middle East and North Africa3 spending. As a result, government deficits in the High oil prices and strong oil demand continue to region have declined and “structural” balances ac- be key drivers for the developing economies of the tually improved in some countries. Nevertheless, Middle East and North Africa4, where GDP is esti- structural rigidities in public expenditures remain mated to have increased by 4.8 percent in 2005. A an issue in a number of countries. 40 percent increase in oil revenues, to some $250 Delivered by The World Bank e-library to: Increases in coffee, sugar, and metal World Bankbillion or (66 percent of their GDP), boosted pub- Theprices largely offset the effect of higher oil prices and lic spending in oil-exporting developing countries IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 lower agricultural prices (notably soybeans) in in the region, causing their GDP to expand by 5.3 20 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y percent. This had spillover effects for the region’s monetary policy stance in both India and Pakistan oil importers in the form of strong exports, also played a role. tourism revenues, and inflows of investment and Because higher oil prices have not been passed remittances. All of these factors helped to sustain through fully, there remains significant latent in- robust growth among regional oil importers (4.2 flationary pressure from this source. In addition, percent), despite higher oil-import bills and rela- implicit energy subsidies have raised fiscal deficits tively weak demand in Europe. by as much as 0.7 percent of GDP between 2002 Looking forward, high oil prices are ex- and 2005, apparently crowding out spending on pected to continue feeding domestic demand in education and health care in India (Devarajan and oil-producing countries—outstripping domestic Ghani 2006).5 Moreover, by impeding the price supply and causing imports to continue rising mechanism from restraining energy demand, the rapidly, even as growth of export revenues slows. pass-through policy (along with robust domestic As a result, GDP in developing oil-exporting coun- demand) has contributed to a deterioration equal tries should expand by 5.2 percent in 2006 before to 4.0 percent of GDP in the region’s current- slowing to around 4.8 percent in 2008. Their cur- account balance since 2003. rent-account surpluses should decline from around Growth is projected to weaken to about 6.8 20 percent of GDP in 2005 to about 8 percent of percent in 2006, reflecting continued above trend GDP in 2008. In the oil-importing economies, growth in Pakistan and India. However, domestic growth is expected to accelerate to about 5.3 per- capacity constraints and rising inflation are pro- cent, supported by stronger European growth, con- jected to cause growth to decline to a more sus- tinued exports of goods and services to regional oil tainable 6.2 percent by 2008. exporters, and a weaker negative effect from the re- Notwithstanding this cyclical slowdown, duction in textile and clothing quotas. growth is projected to remain robust with invest- Prospects for the region remain clouded by ment in both India and Pakistan expected to con- geopolitical developments. For the region as a tinue to benefit from strong external and domes- whole, western investors’ risk perceptions have tic interest. This, plus a four-year infrastructure worsened. For the moment, this has been offset by project (Build India) valued at 5 percent of GDP, an intraregional recycling of oil revenues, which are projected to augment capacity and support has contributed to a sharp inflation in asset prices. demand over the projection period. The services sector in India is expected to continue expanding South Asia rapidly, as a result of strong FDI inflows and out- Strong external demand and private consumption sourcing. Export growth throughout the region growth, supported by generally accommodative should remain strong, despite slower growth in monetary policies, spurred growth in South Asia the United States, partly because of increased de- to a very robust 7.7 percent in 2005, led by India mand from Europe. and Pakistan, which both expanded by about 8 Solid domestic demand should cause the cur- percent. Excluding these two countries, regional rent-account deficit to grow further, reaching growth was still a strong 5.3 percent. Robust re- around 3.5 percent of GDP in 2006 before im- gional clothing exports following the removal of proving somewhat as demand slows. quotas helped limit the overall deterioration of the current account, the deficit of which is estimated Sub-Saharan Africa at 2.6 percent of regional GDP in 2005. GDP in Sub-Saharan Africa expanded by an esti- Despite some efforts to raise retail energy mated 5.2 percent in 2005, bolstered by robust prices, higher oil prices have not been completely growth in resource-rich countries. Indeed, oil- passed through to consumers. Nevertheless, infla- exporting economies grew an estimated 6.4 percent tionary pressures in the region have been building. in 2005, while growth in South Africa came in at Consumer prices rose 9.1 percent in 2005 as com- 4.9 percent, lifted by high metal prices, strong con- pared with 3.6 percent in 2003. To a significant sumer confidence, and low nominal interest rates. Delivered by The World Bank e-library to: degree, higher inflation reflects fluctuations inThe World Bank activity in small oil-importing economies Economic food prices. However, rapid growth, particularly expanded by a slower but still robust 4.3 percent, IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 strong domestic demand in response to a relaxed down from 4.7 percent in 2004. 21 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 This strong performance marks a sharp depar- Mauritius, and Swaziland) is expected to weaken as ture from the weak and relatively volatile growth European sugar preferences are withdrawn, while recorded by the region in the 1980s and 1990s. strong competition from low-cost textile producers 2005 was the fifth year in a row that regional in China and South Asia will continue to be a drag growth was at least 3.5 percent, and ended the first on regional exports. Continued rapid expansion in 5 year period since the 1960s that per capita growth South Africa is expected to spill over into the South- remained positive in every year. Hearteningly this ern Africa Development Community. A more peace- improved performance reflects stronger growth by ful and stable sociopolitical environment will serve many countries rather than very fast growth by a to accelerate growth in Liberia, Sierra Leone, and few. More than half of Sub-Saharan African coun- several other countries. On the other hand, should tries have grown by 4 percent or more on average low-level conflicts, in places such as Chad, Côte during the past five years, compared with fewer d’Ivoire, Nigeria, and the Sudan escalate, they could than one-quarter during the period 1980–95.6 bring down regional growth to a significant degree. Better subsistence and cash crops bolstered agricultural incomes and industrial production in many West African countries, while performance in East Africa was also good, despite drought in Commodity markets some areas. High metal prices bolstered growth in The oil market small resource-rich oil-importing economies. The sharp rise in oil prices since 2003,7 which was The current-account position of oil exporters driven by strong demand and dwindling spare ca- improved significantly because of higher oil rev- pacity, showed signs of ending toward the end of enues. However, external balances in many oil- 2005. Beginning in September 2005, the trend rise importing countries have come under pressure. in oil prices marked a pause, with barrel prices Excluding South Africa, the current-account posi- fluctuating around $63. However, the market re- tion of oil importers deteriorated by 2.8 percent- mains tight, and the pricing power of OPEC has age points in 2005, reaching 6.4 percent of GDP. increased. As a result, prices are volatile, and sen- In Ghana, for example, the current-account sitive to small changes in perceptions such as con- deficit is estimated to have more than doubled to cerns over future supply, which sent barrel prices reach 6.8 percent of GDP, while in Tanzania it toward the $73 mark in early May 2006, before surged close to 6.2 percent of GDP. In several declining once again (figure 1.5). other countries, a failure to fully pass through Oil demand slowed to 0.5 million barrels per higher prices has placed fiscal accounts under se- day (mbpd) in the second half of 2005, from 3.5 rious strain (Madagascar, Mauritius, Rwanda, and mbpd in the first half of 2004 (figure 1.6). While Uganda) or forced utilities to ration energy con- slower GDP growth played a role in this decline, sumption by imposing rolling electrical blackouts the most important factor appears to have been (Madagascar, Malawi). higher oil prices. Econometric models suggest that Looking forward, growth in established oil- exporting countries is projected to average more than 6 percent as new oil production is expected Figure 1.5 An end to the trend rise in oil prices? to come online in Angola, Republic of Congo, World Bank average, oil, $ per barrel Equatorial Guinea, and Sudan. Moreover, Mauri- 75 May 15, 2006 tania and São Tomé and Principe are expected to 65 begin exporting oil in 2006. 55 Small oil importers are also expected to do well, with growth remaining at about 4.5 percent 45 in 2008 as many countries benefit from debt write- 35 offs and increased aid flows. Madagascar, Tanza- 25 nia, and Uganda are expected to continue to profit 15 Delivered by The World Bank e-library to: from prudent macroeconomic policies andThe reforms World Bank Jan. July Jan. July Jan. July Jan. July Jan. May implemented in previous years. In contrast, IP : 192.86.100.362002 2002 2003 2003 2004 2004 2005 2005 2006 2006 Tue, 10 Mar 2009 16:55:33 growth in sugar and textile producers (Lesotho, Sources: Datastream, World Bank. 22 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y Figure 1.6 Higher prices slow oil demand Figure 1.7 A disappointing supply response Change in apparent oil demand, millions of barrels of oil per day Change in global oil deliveries, millions of barrels of oil per day 4.5 4 Other OPEC 4.0 Other Asia Other 3.5 China 3 FSU 3.0 OECD Total Total 2.5 2 2.0 1.5 1 1.0 0.5 0 0.0 – 0.5 –1 2002 2003 2004 2005 2005 –2 Source: International Energy Agency. –3 2001 2002 2003 2004 2005 had prices remain unchanged, oil demand would Source: International Energy Agency. have increased by some 2–2.5 mbpd.8 Incremental oil demand declined in all regions. In addition to prices, a number of special factors clining yields and the rate of increase in pro- were at work. In the United States, higher petrol duction of fields in the former Soviet Union prices in the wake of hurricane Katrina provoked a has slowed. sharp decline in both vehicle miles and gasoline 2. A deterioration in the investment climate in consumption in the autumn, while a mild winter some developing countries has lowered pro- has also eased demand. In Asia, growth in oil con- duction levels and reduced investment, despite sumption slowed, due in part to subsidy cuts in the existence of ample reserves. countries such as Indonesia and Thailand. In China 3. Low oil prices during the 1990s limited incen- energy demand eased partly because new electrical- tives to explore for new oil. More recently, generating capacity reduced the use of relatively in- uncertainty over the durability of higher oil efficient diesel-fueled backup power generators. prices led firms to be cautious about investing Notwithstanding some three years of higher in new (relatively high-cost) capacity, espe- prices9 and the coming on stream of new fields in cially given the long lead times (between three Africa and elsewhere, there has been no discernible and six years) needed to develop new fields. acceleration in aggregate oil supply (figure 1.7).10 4. Low investment in the past has contributed to This contrasts with the 1970s and 1980s, when in- a lack of skilled labor and equipment, further creased output brought substantial new capacity delaying the supply response. online, helping to reduce prices.11 5. A large share of known reserves is located in Aggregate supply has failed to respond, despite countries to which major oil companies do not a sharp increase in investment activity among oil- have access. Major oil firms have been offered exporting developing countries. Output from those service contracts to help countries develop sources has increased just 2.7 percent, or 0.9 mbpd their resources. Thus far, however, oil compa- (4.2 percent, or 0.2 mbpd, for African producers). nies appear to have found share buybacks and A number of factors have contributed to limit increased dividends to be a more profitable use the response of aggregate oil supply: of their earnings. Recent decisions in some de- Delivered by The World Bank e-library to: The World veloping Bank countries to renounce existing con- 1. Existing fields in the United States and in the tracts are unlikely to increase firms’ willing- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 North Sea have entered into a period of de- ness to invest further. 23 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 1.8 Spare production capacity remains low Nonoil commodities OPEC spare production capacity, millions of barrels per day The rise in oil prices since 2003 has been accompa- 7 nied by increasing prices for agricultural goods, met- 6 als, and minerals (figure 1.9). Reflecting continued 5 strong growth in global output, metals and minerals 4 prices increased by some 27 percent in 2005 and up an additional 24 percent in the first four months of 3 2006. Increases in 2005 were concentrated in indus- 2 trial metals, such as iron ore (up 72 percent), zinc 1 (up 38 percent), and copper (up 21 percent). Prices 0 for other metals and minerals also rose, but by less. Jan. Jan. Jan. Jan. Jan. 2002 2003 2004 2005 2006 Tin, the price of which fell by 13 percent over the year, stands out as an exception. Sources: World Bank, International Energy Agency. At the global level, prices of agricultural prod- ucts have been relatively stable, up 9.3 percent be- tween April 2006 and the same date a year earlier. The combination of still growing demand and High prices early in 2005 reflected a poor monsoon a weak supply response has meant that although season in South Asia and drought conditions in spare production capacity has improved, it re- Sub-Saharan Africa. Improved weather conditions, mains tight (figure 1.8). Looking forward, invest- in combination with increased supply in some coun- ments in new productive capacity are increasing tries, contributed to an easing in agricultural prices (up some 15 percent in 2005). Moreover, contin- through much of 2005, followed by a modest ued high prices will increase incentives to adopt pickup in prices in the first quarter of 2006. Raw more petroleum-efficient technologies and con- materials are up 11 percent since April 2005. serve fuel. As a result, demand growth is expected The recent strength of nonoil commodity to remain relatively moderate (at about 1.5–2 mil- prices is primarily a reflection of strong world de- lion barrels per day). mand in recent years and low spare capacity Unless non-OPEC supplies rise much faster brought on by low prices during the 1990s. Prices than expected (the International Energy Agency, also have been influenced by strong energy prices, 2005, projects non-OPEC supply to increase by 3 because energy is a major input in the production mbpd over the next three years), spare capacity of many commodities (notably aluminum), and will remain limited and OPEC’s pricing power because several commodities are important substi- high. The organization has signaled its willingness tutes for petroleum-based products (such as rub- to reduce output in line with demand. ber and sugar used in the production of ethanol). Prices are expected to remain volatile but Overall, about one-third of the increase in nonoil should gradually decline, reflecting the counter- vailing influences of continued strong growth in global output and limited increases in non-OPEC Figure 1.9 Commodity prices oil on the supply side, and increasing energy effi- Index, 1990 = 100 ciency on the demand side. While the precise path 300 to be taken in these conditions is largely unknow- Energy able, the forecasts reported in this chapter assume 250 that barrel prices will begin moderating in 2006, averaging $64 for the year and decline gradually 200 Metals and towards $57 in 2008. minerals 150 However, the market remains vulnerable to disruption, whether by natural disasters or geo- 100 Agricultural political events.12 Hence, the possibility of sudden products Delivered by The World Bank e-library to: upward spikes in oil prices cannot be The ignored, World Bank 50 2001 2002 2003 2004 2005 2006 even if the general trend is one of stabilization or IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 slight decline. Source: World Bank. 24 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y commodity prices between 2002 and 2005 was due Figure 1.10 Moderate increases in inflation to higher oil prices (Baffes 2005).13 Some of the End of period, year-over-year monthly inflation rate very recent strength in the prices of precious metals 10 may also reflect investor uncertainty in the face of a 9 declining dollar and continued global imbalances. 8 Mar. 2006 Improved supply should ease the prices of 7 2005 most agricultural commodities beginning in 2006. 2004 6 However, the prices of close energy substitutes and 2003 5 2002 energy-intensive products are expected to rise fur- 4 ther. Overall, agricultural prices are projected to 3 rise by about 10 percent in 2006 before easing by 2 about 3 percent in each of 2007 and 2008. Strong 1 demand from China and other developing 0 economies, low stocks, and high energy prices are a ci ia be a ia tr i e lA & Af t & ric ib r ic un om Pa As As tra pe fic a an a es projected to push metals and mineral prices up th as Af ar me si r ic co inc st h en uro or E ut n Ea & tin A h- ra N e So E some 25 percent in 2006, before they begin easing dl ig ha & id H La C Sa M C by about 5 percent in 2007 and 12 percent in b- Su 2008. Demand-driven increases in energy prices Source: World Bank. represent an upside risk to energy-sensitive non-oil commodities including food stuffs, whose yields depend on energy-intensive fertilizers. and more prudent fiscal policies. In addition, the rapidly expanding role of Asia and, to a lesser extent, the countries of the former Soviet bloc as Inflation, interest rates, low-cost manufacturing centers have served to and global imbalances dampen price inflation in high-income countries, Inflation where many of these products are consumed. Perhaps the most critical explanation for the lim- The pickup of inflation in Sub-Saharan Africa ited impact of higher oil prices on output has been and South Asia is partly explained by food prices, the weak response of inflation to higher oil which increased substantially in both regions dur- prices—especially in high-income countries, where ing the course of 2005 and should be expected to world interest rates are determined. ease in 2006 as crops improve. However, as is the While inflation is up in virtually every region, case in a few Latin American countries, it also most of the increase appears to reflect the direct likely reflects overheating in those regions, which impact of higher oil prices. With perhaps the ex- have been growing at historically high rates. ception of South Asia and Sub-Saharan Africa (see This possibility is particularly worrisome in discussion below), there is little evidence of the the case of Africa, because the credibility of mone- rapid price pass-through or the wage–price spirals tary authorities is not yet well entrenched. Should that characterized the oil shocks of the 1970s and an inflationary spiral develop, it could have serious 1980s (figure 1.10). Despite a pickup toward the consequences for macroeconomic stability and af- end of 2005 in the United States, core inflation (the fect the ability of those economies to sustain the rate of price increase of goods and services, exclud- strong growth of the past several years. In the ing food and energy) has increased relatively little meantime, continued aid flows to finance improved (see figure 1.2). As a result, inflation expectations governance and social and physical infrastructure and interest rates have remained low, eliminating investments will be essential to raising the trend one of the principal mechanisms through which growth rate that these countries can sustain. past oil shocks have slowed growth. Many factors explain this inflationary per- Interest rates Delivered by The World Bank e-library to: formance—among them more flexible labor andThe World subdued response of inflation has allowed The Bank product markets in high-income countries, lower monetary (and fiscal) policy to remain relatively ac- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 oil intensities, more credible monetary policy, commodative. While short-term interest rates are 25 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 1.11 Flattening yield curve Figure 1.12 Changes in real effective exchange rate Yield on 10- and 2-year U.S. Treasury bills Depreciation Appreciation 5 10-year Treasury bill yield R.B. de Venezuela Ecuador Algeria Malaysia 4 Mexico Chad Gabon Vietnam Cameroon Resource Rich 3 Côte d’Ivoire Argentina Indonesia Trinidad and Tobago 2 Iran, Islamic Rep. Kazakhstan Papua New Guinea 2-year Treasury bill yield Chile 1 Colombia Equatorial Guinea Nigeria Russian Fed. 0 South Africa Angola Jan. July Jan. July Jan. July Jan. Brazil 2003 2003 2004 2004 2005 2005 2006 Zambia Lesotho Egypt, Arab Rep. of Source: World Bank. Bolivia Central African Rep. St. Kitts and Nevis Gambia, The Sierra Leone Bahamas, The Tunisia rising, they remain low in real terms, and long- Mauritius Malawi term rates have only recently begun rising in high- Congo, Dem. Rep. Vatican income countries. As a result, the yield curve has Dominica Mozambique flattened significantly, with short-term bond yields Costa Rica Jordan Belarus virtually equal to longer term yields. Uganda Grenada Indeed, on several occasions during February Latvia Netherlands Antilles and March 2006 the yield on two-year U.S. Trea- Lithuania Belize sury bonds marginally exceeded that of the 10- China Guyana Antigua year bond (figure 1.11). Such yield-curve “inver- Nicaragua St. Lucia sion” has historically been a good indicator of a Taiwan, China Solomon Islands future recession (Estrella 2005).14 As such, these Morocco Macedonia, FYR Resource Poor inversions may signal a slowing of the U.S. econ- Peru Slovenia Benin omy. However, they were very small and occurred Burundi Jamaica with both short- and long-term real interest rates Pakistan Paraguay at low levels. Moreover, while the yield curve re- India Poland mains flat, long-term rates in April and early May Uruguay Hungary were once again higher than short-term rates. In Croatia Slovenia Estonia this context, the flattening of the yield curve re- Cyprus Thailand flects a broadly positive outlook for global Fiji Malta growth, characterized by stable expectations for Burkina Faso Kiribati inflation, significant spare capacity in Europe, and Ukraine Dominican Rep. Philippines an American economy that continues to expand Togo Albania quickly even as it slows in response to a more neu- Guatemala Bulgaria tral monetary policy stance. Armenia Czech Rep. Developing economies experienced a similar Ethiopia Namibia Western Samoa flattening of the yield curve. Bond spreads contin- Tonga Moldova ued to decline, reaching a historic low of 174 basis Ghana Kenya points for sovereign borrowers in May 2006. How- Romania Slovak Rep. ever, the combination of relatively stable bank Turkey spreads (around 100 basis points) and rising rates –45 –30 –15 0 15 30 45 60 75 90 in high-income countries means that the average % change, Jan. 2006–Dec. 2002 Delivered by The World Bank e-library to: interest rate paid by developing countriesThe actually World BankSources: World Bank, IMF. rose over the past 12 months (see chapter IP : 15 2). 192.86.100.36 Tue, 10 Mar 2009 16:55:33 26 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y Exchange rates Figure 1.13 Developing countries’ current-account A further factor limiting the real-side conse- balances quences of higher oil prices is the wider adoption Current account balance (% of GDP), 2002 and 2005 of flexible-exchange-rate regimes over the past 8 2002 two decades (see chapter 5). Among oil-importing 2005 6 developing countries that have not benefited from high metals and minerals prices, there was a mod- 4 est tendency toward depreciation.16 Unsurpris- 2 ingly, among developing oil exporters the tendency toward appreciation was much more pronounced, 0 with two-thirds of these countries appreciating by –2 an average of 18 percent.17 Such exchange rate fluctuations contributed to the resilient response –4 of these economies to higher oil prices by facilitat- –6 ing adjustment to the change in relative prices im- rte Oil ci ia La Asi & be a ric t ia ric n Af as ib ric Af ara Pa s As l e plied by higher oil prices (figure 1.12). For oil im- rs fic a an a a & st A th E tra op ar e h C m h or le Sa en r ut C Eu Ea & in A N dd po So porters, the depreciation transfers the price shock b- & Mi ex Su t over a wider range of tradable goods and services. Moreover, by making exports more competitive Oil importing developing countries and imports less so, the depreciation increases net Source: World Bank. exports, reducing the impact on economic output that would otherwise be observed as a result of re- duced incomes and lower consumption. Figure 1.14 Increased aid helped finance oil costs Most developing oil importers have financed in 2004 higher oil bills successfully Increases in net ODA in 2004, excl. debt relief; % of GDP Another factor behind the resilience of growth has 2 been the relative ease with which developing coun- tries were able to finance higher oil bills. Many de- 1 veloping countries entered into this period of higher oil prices with positive or near-zero current- 0 account balances. As a result, despite deteriora- tions of 2 or more percent of GDP in many cases, –1 current-account positions for most countries re- main at levels that should not pose serious financ- –2 ing difficulties (figure 1.13). Se aso M ue oz Ke l bi a M au wi ag ius ar r iL a on ka or s Ba anz co la ia Pa esh do an Et esia a a ia ga ic ca M ura am ny Sr agu in d ng an op M ala H an T oc q In ist rk an ad rit ne N as In the poorest countries, substantial increases F d n hi d k Bu Ug in ODA during 2004 and 2005 provided some of M the foreign currency necessary to finance the in- Sources: OECD, World Bank. crease in their oil bills (figure 1.14). For many African countries, the increase in foreign currency earnings from this source amounted to more than 0.5 percent of GDP in 2004 (data for 2005 are not most instances this was not the case. To the de- yet available). Simulations suggest that for oil- gree that projects financed by this aid had low im- importing poor countries, increased ODA inflows port intensities, the foreign currency, after conver- may have reduced the first-round impact of higher sion to domestic currency, would be available to oil prices by as much as two-thirds (Diaz-Bonilla finance other imports—perhaps, but not necessar- and Savescu, 2006) (figure 1.14).18 ily, more expensive oil. Moreover, if there is a pos- Delivered by The World Bank e-library to: While some countries may have used theThe World externality associated with domestic export itive Bank money directly to finance oil consumption, in activity (Frankel and Romer 1999; Ibrahim and IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 27 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 MacPhee 2003), the negative oil shock may actu- Figure 1.16 Tensions associated with fast growth, ally have improved development prospects by the case of Turkey partially offsetting the Dutch-disease effect associ- Sources of finance in Turkey’s balance of payments, US$ millions ated with the increased aid.19 25,000 Despite these offsetting factors, several coun- 20,000 tries appear to be encountering difficulties financ- 15,000 ing their higher oil bills. In Africa, current account 10,000 deficits among oil-importers (excluding South 5,000 Africa) have soared and average more than 6 per- 0 cent of GDP. Current-account deficits have also –5,000 reached worrisome levels in many European and –10,000 Central Asian countries. Many countries are expe- 2002 2003 2004 2005 riencing fiscal difficulties because of less-than- Private investment flows Bonds Other Reserves complete pass-through. Madagascar, Malawi, and Fin req (CA Bal) Sierra Leone have been forced to ration electricity Source: IMF. consumption through rotating blackouts in an ef- fort to conserve energy, suggesting that they may have met binding current-account constraints and are unable to finance additional oil imports. Sev- eral other countries appear to be consuming inter- Of particular concern are a number of coun- national reserves at unsustainable rates (Benin, tries that combine high current-account deficits, Guinea Bissau, Mali, Tanzania) (figure 1.15). In significant capital inflows, high interest rates, and still others, reserves represent a dangerously low an appreciating currency, notably Bulgaria, Roma- share of monthly import cover (Bangladesh, nia, and Turkey (figure 1.16). These conditions Madagascar, Namibia, Swaziland). In all of these pose serious problems for policy makers, as the countries, policy makers will need to take concrete capital inflows (initially in the form of direct in- steps, including currency depreciation and energy vestments) prompt an appreciation of the cur- conservation measures, so that domestic demand rency, increase domestic money supply, and raise and the country’s net revenue positions adjust to inflationary pressures. In each of these countries recent changes in relative prices. monetary institutions have responded by raising interest rates, which reduces domestic money sup- ply growth but has also induced additional finan- Figure 1.15 Reserves in some countries are falling cial inflows, adding to domestic liquidity and in- rapidly or worrisomely low flationary pressures.20 While tighter fiscal policy has helped combat these tendencies, external Import coverage in months of imports 18.9 deficits continue to rise and currencies to appreci- 10 ate in many of these countries. Should capital in- 9 flows slow or stop, financing current levels of ex- 8 penditure could be very difficult, placing these 7 2004 2005 currencies under significant pressure. A sudden de- 6 2003 preciation could generate an inflationary push— 5 partially undoing recent achievements in stabiliz- 4 ing currencies and controlling domestic inflation. 3 More generally, the deterioration in the cur- 2 rent-account position of oil-importing developing 1 countries means that they are much more vulnera- 0 ble now than they were in 2003. An important ia sh s n ia ss a r nd i supply disruption that pushed oil prices even ca al iu ni Bi uine ib an de au M rit Be ila as am nz au la Delivered by The World Bank e-library to: az G ag The World Bankhigher, or a decline in nonoil commodity prices, N ng Ta M Sw ad Ba M would be much more difficult to finance and could IP : 192.86.100.36 Source: World Bank. Tue, 10 Mar 2009 16:55:33 precipitate painful adjustments (see risks section). 28 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y Global imbalances persist its net asset position is set to deteriorate sharply, The imbalances in global spending patterns that reaching between 65 and 48 percent of GDP by have characterized the world economy over the 2015 (Higgins, Klitgaard, and Tille 2005).22 past five years, with the United States consuming So far, financing of these deficits has not significantly more than it produces and running a posed a serious problem for the United States, in large current-account deficit, persisted in 2005 part because of low interest rates and because of a (figure 1.17). High oil prices both exacerbated im- generalized willingness of foreigners to hold Amer- balances and changed their nature, contributing to ican assets that yield lower returns than the for- about 40 percent of the additional deterioration of eign assets held by Americans.23 As a result, de- the U.S. current-account deficit in 2005.21 At the spite the deterioration of its net asset position, the same time, high oil prices caused the current- United States has continued to earn a positive net account position of almost all oil-importing return on foreign investments.24 If investor’s will- countries to deteriorate and substantially boosted ingness to continue accumulating such assets those of exporters. As a result, whereas in 2002 changed, U.S. interest rates would rise and the cur- oil-importers in virtually every region except the rent account balance would deteriorate (by about United States were running a current-account 0.5 percent of GDP for every 100-basis-point rise surplus, now almost all are running deficits— in U.S. interest rates relative to foreign rates).25 with the notable exceptions China, Japan, Korea, Over the past year, short-term interest rates in the and a few other high-income countries. United States have risen by about 100 basis points The sustainability of these imbalances and more than in Europe, bringing the overall differ- their financing is a question of growing concern ence to 220 basis points. The long-term differen- (IMF 2006; World Bank 2005a, 2005b). Persistent tial is now some 100 basis points (figure 1.18). Al- current-account deficits have transformed the though it is certainly too early to tell, this United States from being the world’s most impor- movement (and the decline in emerging-market tant creditor nation (with a net international in- risk premia against the dollar) could reflect a re- vestment position of 13 percent of GDP in 1979) assessment of the dollar as a safe haven. to being the world’s largest debtor (with a net asset Independent of the reasons for these move- position of –21 percent of GDP in 2004). Unless ments, the course of long-term interest rates contin- savings in the United States increase substantially, ues to be sensitive to the willingness of nonmarket sources of finance (formerly developing-country central banks and now, increasingly, authorities in Figure 1.17 Global imbalances Current account balance, $ billions, 2002 and 2005 400 Figure 1.18 Interest rate spreads support the dollar 2002 2005 200 3-month and 10-year bond yield, % 6 0 –200 10-year U.S. yield 5 –400 4 –600 10-year euro yield –800 3 Oil importers Oil exporters –1000 2 USA Europe Other high-income importers Japan East Asia & Pacific (ex. China) China Europe & Central Asia Other developing importers High-income exporters Low-income exporters 3-month euro yield 1 3-month U.S. yield 0 Delivered by The World Bank e-library to: Jan. Bank The World Jan. Jan. Jan. Jan. 2002 IP : 192.86.100.36 2003 2004 2005 2006 Tue, 10 Mar 2009 16:55:33 Source: World Bank. Sources: World Bank, Datastream. 29 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 1.19 Funding the U.S. current account deficit ward pressure and that U.S. interest rates will con- Net flows by asset class, US$ billions tinue to exceed those in Europe. Indeed, between 1,000 January and early May 2006, dollar cross rates have been relatively sensitive to interest rate differ- 800 entials. During this period, it has depreciated 7 600 percent against the euro (4 percent against the won and 0.7 percent against the renminbi) and 2.3 400 percent in real-effective terms. Looking forward 200 these trends are expected to continue and the dol- lar to depreciate slowly by about 5 percent per 0 year over the projection period. –200 2004 2005 U.S. corporate bonds Treasuries FDI Banking World trade Equities Official assets Source: World Bank. O verall, merchandise trade growth slowed somewhat in 2005, expanding by 8.9 per- cent, as compared with 11.8 percent in 2004 (fig- ure 1.20). Most of the slowdown occurred during oil-exporting countries) to purchase low-yield dol- the first half of the year and among high-income lar-denominated assets. Lower reserve accumula- countries. For 2005 as a whole, their export vol- tion by oil-importing developing economies trans- umes increased only 6.0 percent, down from 10.2 lated into a $130 billion decline in their purchases percent the year before. However, toward the sec- of U.S. Treasury bills and official assets (figure ond half of the year and into 2006, outturns have 1.19). This was only partly offset by a $14 billion improved, in part because of increased European increase in purchases by oil exporters. The need to exports to the Middle East. meet this (nonmarket) financing shortfall may have In contrast, China’s export volume expanded been among the factors that pushed up long-term by 27.8 percent in 2005, almost exactly as fast as U.S. interest rates. in 2004. Moreover, despite a slowing in the pace The tensions implicit in the U.S. current- of Chinese foreign sales towards the end of 2005, account deficit are building and need to be ad- export volumes have once again picked up—ex- dressed. Reducing global imbalances is a shared in- panding by more than 25 percent during the first ternational responsibility, requiring a tightening of fiscal policy in the United States, increased imports abroad and increased exchange-rate flexibility. Im- Figure 1.20 Healthy growth in world trade plementation must necessarily be gradual—to avoid excessive disruption, both within the United % change, year-over-year 25 60 States as macro policy is tightened and in devel- oped and developing Asia as currencies are allowed 20 Developing countries (ex. China) 50 to appreciate. However, to be effective and pre- 15 empt market jitters the effort must be credible. In 40 particular, in the absence of increased savings in the 10 30 United States, increased domestic demand abroad 5 and greater exchange rate flexibility are unlikely to 20 0 have a significant effect on global imbalances and would likely exacerbate global capacity con- High-income countries 10 –5 straints—reducing the likelihood of a soft landing. 0 –10 Although in the near term global imbalances China (right axis) Delivered by The World Bank e-library to: are unlikely to provoke the serious currency crisis Bank–15 The World –10 2000 2001 2002 2003 2004 2005 2006 suggested by some (Roubini and Setser 2005), they IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 do imply that the dollar will face further down- Source: World Bank. 30 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y two months of 2006. Other developing countries Figure 1.21 Regional increases in market share also continued to expand their market share. Their % increase in global market share, 2005, since1995 export volumes increased 10.3 percent, only some- 5.0 what slower than the year before. Here, too, trade 4.5 growth decelerated early in the second half of 4.0 2005 but has since picked up. 3.5 Oil revenues of developing-country oil ex- porters nearly doubled between 2002 and 2005, 3.0 increasing by some $215 billion. For all oil ex- 2.5 porters, the increase was about $400 billion. 2.0 However, oil exporters have increased their own 1.5 imports markedly, and more than three-quarters 1.0 of additional export revenues have been spent on 0.5 additional imports. 0.0 Oil exporters are also recycling petrodollars through financial markets. Between 2002 and –0.5 hi ia La Asi & be ca ia ric n ci ia ric t 2005, oil-exporting developing countries increased Af as Af ara . C As As Pa As l e ib ri na a an a fic a tra p th E ar me o h ex st h st or le en ur Sa foreign currency reserves by $255 billion (with ut Ea & tin A Ea N dd C E So b- & Mi & Su C $117 billion of the increase accounted for by the Russian Federation). In total some $245 billion has Source: World Bank. flowed into the United States as securities, bonds or bank deposits, while about $50 billion has been placed directly into the European banking sector. The export boom of China is similar to past Unfortunately, because of the use of third-party in- booms in a number of countries that are now clas- termediaries and reduced reliance on the banking sified as high income (Israel, Japan, the Republic sector (as compared with past episodes of high oil of Korea, and Taiwan) in that it was mostly driven prices) it is particularly difficult to trace the desti- by an expansion in the range of goods exported. nation of these funds (BIS 2005). Thus, while technological progress, investment, Not all regions shared equally in the recycling and labor productivity growth contributed to a of petrodollars. In particular, the share of the 290 percent increase in Chinese sales to the United United States in the imports of oil-exporting coun- States of products already on sale in 1992, more tries fell from 25 to 20 percent during this period.26 than 60 percent of the total increase came from the In contrast, most developing countries increased sale of goods that China did not export to the their market share in the imports of oil-exporting United States in 1992.27 This contrasts with countries. However, the increase in their export Bangladesh, for example (figure 1.22). That coun- revenues paled in comparison with the increase in try’s revenues from exports of traditional products their oil bills. to the United States increased by an impressive 173 percent between 1992 and 2005, but compared Can developing countries continue to gain with China it managed only to generate one-tenth market share at recent rates? as much additional revenue from new products. The strong economic performance of low- and While not as marked as in China, there is evi- middle-income countries over the past several dence that other developing countries are diversi- years reflects both rapid growth in world exports fying the range of goods that they export and (up 90 percent since 1995) and an almost 50 per- moving up the value-added ladder. Today, the rev- cent increase in the market share of developing enues of developing countries from exports to economies, up from 20 percent in 1995 to almost high-income countries depend much less on raw 30 percent in 2005. This improvement is due, in materials (figure 1.23) and much more on higher- large part, to increases in the market share of value-added goods (and services). Delivered by The World Bank e-library to: China. Nevertheless, every developing region (ex-The World The Bank rapid increase in the market share of cept East Asia excluding China) has seen its global China and other developing countries resulted IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 market share increase (figure 1.21). from the exploitation of preexisting competitive 31 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 1.22 Increased product range explains parative advantages to new markets, the applica- most of Chinese export growth tion of lessons learned in existing sectors to new Index of the value of exports to the USA, 1992 = 100 ones, and a widening of the product base. China Long-term prospects for developing 1,000 900 New export products economies will depend importantly on their ability 800 Goods being sold in 1992 to continue increasing market share in this way. 700 For countries and regions that, like Bangladesh, 600 have yet to enjoy an export boom, trade liberaliza- 500 tion and facilitation comprise key agendas. For 400 China, the boost in exports associated with acces- Bangladesh sion to the World Trade Organization (WTO) may 300 be easing (accession is estimated to have increased 200 export growth by 12 percentage points). Neverthe- 100 less, China’s volume of exports can be expected to 0 1992 2005 1992 2005 continue growing at around 18 percent.29 More generally, developing countries must es- Source: World Bank. tablish and maintain low tariffs across the board, minimize administrative burdens associated with trade, and reduce transit times so that markets can Figure 1.23 Exports of developing countries have be served in a timely manner (Newfarmer 2005). diversified On the multilateral front, efforts need to be con- Product shares in exports of developing countries (ex. India & China) centrated on agriculture, the most heavily pro- to high-income countries, % 100 tected sector and one where many developing Manufactures countries enjoy a comparative advantage. Liberal- 90 Electronics ization here would allow these countries to reap 80 Machinery, equipment Motor vehicles, parts the same kind of benefits that have accrued to 70 Lumber, paper countries specialized in manufacturing following 60 Textiles, apparel, leather the liberalization of that sector. Second, countries 50 Beverages, tobacco Food products need to reduce rigidities in product, labor, and fi- 40 Metals, minerals nancial markets so that firms can react with agility 30 Gas to new opportunities to expand the range of prod- 20 Oil Fishing ucts they produce and sell. 10 Forestry 0 Crops 1978 2002 Source: World Bank. Risks advantages that have been exposed by market lib- T he relatively benign soft-landing scenario for developing countries that is described above is subject to a number of important downside risks. eralization and domestic policy reforms. These in- clude trade liberalization (both multilateral and, Managing fast growth importantly, autonomous liberalization [World Internal risks exist on both the upside and down- Bank 2005]), and behind-the-border reforms, such side. Following several years of very fast growth, a as regulatory reform, liberalization of foreign in- number of economies are showing signs of strain, vestment regimes, and improved labor market as capacity constraints appear in some sectors or regulations.28 as weaknesses in their infrastructure or institu- The important role that expanding the range tional frameworks are exposed. In several coun- of goods exported has played in China's success tries in the Europe and Central Asia region, strong Delivered by The World Bank e-library to: suggests that trade expansion need not be Thebound World BankFDI inflows attracted by privatizations and the by increases in productivity or lower IP :wages. prospects of accession to the European Union have 192.86.100.36 Tue, 10 Mar 2009 16:55:33 Rather, it reflects the exposure of preexisting com- prompted an appreciation of domestic currencies, 32 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y high current-account deficits, and domestic mone- behavior; and (3) the possibility that nonoil com- tary expansion. Subsequent increases in domestic modity prices will fall significantly. interest rates have attracted further financial in- The effects on output in the global economy, flows, exacerbating the current account and ex- should those risks be realized, have been pre- change rate pressures. While, these pressures are sented in past editions of Global Development projected to ease in our baseline projection, they Finance and Global Economic Prospects. Rather carry with them the potential to prompt a cur- than discuss them at length here (past results are rency crisis—possibly resulting in a hard landing— summarized briefly below), this section explores in one or more of these countries. their potential effects on the most vulnerable of The rapid expansion of investment and domes- low- and middle-income countries, particularly tic credit in some Asian economies may be overex- those that have significantly less room for ma- tending the banking sector in these countries in neuver than they did in 2002 because of the re- ways that are not yet obvious, potentially resulting cent increases in oil prices. in a sharp reversal of fortunes. The rapid rise in Table 1.2 summarizes the results from previ- stock-market valuations, housing prices, and prices ous simulations of three hypothetical shocks: (1) a of other assets in several oil-exporting countries reduction of 2 million barrels per day in oil supply, may also spur a crisis if conditions change rapidly. resulting in a rise in oil prices to $100 a barrel for Finally, the real-income shocks that developing three months and $80 for a further nine months; countries have been subjected to are large, and ad- (2) a 200-basis-point increase in long-term interest justment to them remains incomplete. While infla- rates and risk premia; and (3) a 15-percent decline tionary pressures in most countries have been con- in the price of nonoil commodities. tained so far, pressures on wages are being felt in While for analytical clarity these simulations some. Rising inflation in a few countries in Latin are presented independently, there are likely to be America, South Asia, and, perhaps, Sub-Saharan interactions between them. For example, were out- Africa are suggestive of the beginning of an infla- put to slow following a disruptive resolution of tionary spiral. Unless fiscal and monetary authori- global imbalances both oil and non-oil commodity ties succeed in slowing growth, inflationary expec- prices would likely decline. This kind of interac- tations may become engrained requiring a sharper tion is accounted for in table 1.2, but not in the slowdown later on as authorities intervene to con- more detailed impact analyses presented in tables tain them. 1.3–5. Similarly the probabilities of these external shocks differ. The probability of a disruptive reso- External risks lution of global imbalances is low (but grows the The external environment of the past few years longer corrective steps are not taken), while ex- has been especially propitious for growth, charac- perts argue that there is a 70 percent chance of a 2 terized by ample liquidity, rapidly expanding de- mbpd supply disruption sometime in the next 10 mand for the exports developing countries. Look- years (Beccue & Huntington, 2005). ing forward, conditions will be less benign. In the first scenario, a substantial disruption in Interest rates are rising, while very high current ac- global oil supply pushes oil prices to $100 for one count deficits in a number of developing countries quarter and to $80 for a further nine months. As a suggests that many have yet to adjust fully to result, global growth slows by about 0.75 percent a higher oil prices and that they have become more year over two years. The impact is more severe in vulnerable to additional shocks. large low-income and middle-income countries, The principal external risks facing the global both because of higher energy intensities and a economy have changed little over the past several greater inflationary impact, which requires a larger years. They include: (1) the possibility that a sup- contraction to eliminate. On average, the current- ply shock will cause the price of oil to rise even account position of oil importing countries would further; (2) the possibility that interest rates de- deteriorate by about 1.1 percent of GDP. manded by foreign investors to finance the large In the second scenario, concerns over the U.S. Delivered by The World Bank e-library to: U.S. current-account will rise, either gradually, inThe World current-account Bank deficit push long-term interest response to depreciation of the dollar, or more pre- rates up by 200 basis points. Heightened insecu- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 cipitously, because of a change in perceptions or rity, especially because the dollar—the traditional 33 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table 1.2 Estimated impact of three risk scenarios GDP (% change from baseline) First-round impact, Year 1 Year 2 Year 3 % of GDP Scenario 1: a 2-million-barrels-per-day negative supply shocka Scenario 3: a 15 percent decline in non-oil commodity prices World –1.0 –1.5 –1.1 High-income countries 0.0 High-income countries –0.7 –1.5 –1.3 Low- and middle-income countries –0.1 Middle-income countries –1.6 –1.6 –0.1 Low- and middle-income oil exporters 0.0 Large low-income countries –1.7 –2.8 –1.8 Low- and middle-income other –0.1 Current-account-constrained Low-income countries 0.0 low-income countries –0.3 0.1 0.0 Low-income oil exporters 0.2 Low-income other –0.1 Scenario 2: a 200-basis-point increase in interest ratesb East Asia & Pacific 0.1 World –1.7 –2.9 –1.9 Latin America & Caribbean –0.4 High-income countries –1.5 –2.7 –2.5 Europe & Central Asia –0.2 Low- and middle-income countries –2.4 –3.5 –3.0 Middle East & North Africa 0.5 South Asia 0.2 Sub-Saharan Africa –0.7 Oil exporters 0.3 Oil importers –1.1 Oil importers less South Africa –1.1 HIPC –0.7 Oil exporters 0.0 Other –1.2 Source: World Bank. a. For more details see (World Bank 2005b, Table 1.5) b. For more details see (World Bank 2005b, Table 1.6) safe haven currency—is the source of disruption, credibility of monetary policy, and more flexible causes developing country risk premia to increase labor and product markets) that allowed them to by an additional 200 basis points. World growth absorb the recent hike in oil prices with limited ef- slows by about one-half for a period of two years, fects on output should also permit them to deal as higher interest rates cut into investment and with the kinds of shocks modeled above without consumption demand, both through classic trans- too much difficulty. mission mechanisms and via the impact of interest For other countries, however, the recent oil rates on housing prices and consumer wealth. price hike caused a substantial deterioration in Slower growth eases inflationary pressure and their current-account position. In addition to the global tensions, allowing monetary policy to real-side consequences of higher interest rates or a loosen. Growth starts to pick up again. further increase in oil prices, the macroeconomic In the third scenario, a 15 percent fall in nonoil position of these countries could be placed under commodity prices affects global growth only mar- serious strain by the shocks assumed here—result- ginally. The bulk of the impact is felt by Sub-Saha- ing in significant disruption. In the case of an in- ran African oil-importing countries, which sustain a terest rate shock, heavily indebted countries and terms-of-trade loss equal to 1 percent of GDP. In the middle-income countries would be most vulnera- context of already elevated current-account deficits, ble, while a further increase in oil prices would this translates into a substantial reduction in domes- strike the most oil-intensive economies hardest. A tic demand but only a limited fall in output, because decline in nonoil commodity prices could also net exports increase as a lack of access to foreign have important consequences for countries that currency forces non-oil import volumes to decline are currently benefiting from strong nonoil com- in line with the increased oil bill. modity prices, notably metals and minerals. Tables 1.3 through 1.5 summarize these sensi- Potential impacts in the most tivities by highlighting the expected first-round im- vulnerable countries pacts of the three shocks outlined above on the Delivered by The World Bank e-library to: For the majority of developing countries,The fun- Bankcurrent accounts of developing economies. These theWorld damental improvements (increased globalization simulations are meant to be illustrative—not pred- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 in both product and financial markets, improved icative. Both the likelihood of a shock and its 34 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y eventual magnitude are very uncertain. As the re- interest rates and a 200-basis-point increase in sults presented in these tables are estimates of the risk premia for the most vulnerable developing first-round impact for a given size shock, they can countries30 (the most heavily indebted and those be scaled up or down to estimate the impact of a with high concentrations of short-term and other smaller or larger shock. interest sensitive debt). Such a shock could repre- Table 1.3 shows an estimate of the the cumu- sent as much as 3.5 percent of these countries’ lative impact of a 200-basis-point increase in U.S. GDP and could send their current-account deficits Table 1.3 Impact of a 400-basis-point increase in interest rates in selected developing countries % of GDP Increase in debt Interest payments on Increase in debt Interest payments on servicing costs external debt, 2004 servicing costs external debt, 2004 Estonia 3.8 3.4 Lithuania 1.4 1.6 Latvia 3.4 2.3 Jordan 1.4 1.6 Kazakhstan 3.0 2.2 São Tomé and Principe 1.3 5.0 Croatia 2.9 3.5 Poland 1.3 1.4 Moldova 2.0 2.4 Romania 1.3 1.5 Argentina 2.0 1.2 Zimbabwe 1.2 0.5 Hungary 1.9 1.8 Mauritius 1.2 1.4 Sudan 1.8 0.4 Turkey 1.2 2.3 Slovak Republic 1.8 2.2 Malaysia 1.1 1.8 Bulgaria 1.8 2.1 Paraguay 1.1 1.9 Chile 1.7 1.5 Nicaragua 1.1 1.0 Uruguay 1.6 3.6 Lebanon 1.1 6.6 Philippines 1.6 4.4 Peru 1.1 2.1 Côte d’Ivoire 1.5 0.7 Panama 1.0 4.8 Czech Republic 1.4 1.3 Colombia 1.0 2.5 Indonesia 1.4 1.8 Jamaica 1.0 3.8 Source: World Bank. Table 1.4 Impact of a further $30 hike in oil prices in selected developing countries % of GDP Change in current account Current account Change in current account Current account due to $30 hike in oil price balance in 2005 due to $30 hike in oil price balance in 2005 Guyana –8.2 –25.1 Vanuatu –3.0 –44.4 Mongolia –6.4 –2.8 Antigua and Barbuda –3.0 –5.5 Tajikistan –6.3 –4.2 Ukraine –2.9 1.3 Lesotho –5.8 –2.9 Paraguay –2.9 –0.8 Togo –5.4 –10.3 Lebanon –2.8 –16.3 Kiribati –5.3 –13.6 Mali –2.7 –8.5 Solomon Islands –5.2 –14.2 Jordan –2.7 –5.8 Swaziland –5.1 1.9 Mozambique –2.6 –5.1 Tonga –4.4 –0.5 Malawi –2.6 –7.1 Cambodia –4.4 –5.2 Bahamas, The –2.5 –11.6 Ghana –4.3 –6.9 Grenada –2.4 –71.7 Belize –4.3 –14.2 Gambia, The –2.3 –12.4 Honduras –4.0 –4.4 Dominica –2.2 –26.3 Moldova –3.8 –25.1 St. Lucia –2.2 –10.1 Nicaragua –3.5 –18.6 Nepal –2.2 –1.3 Samoa –3.5 –0.3 Pakistan –2.2 –1.9 Jamaica –3.3 –11.3 Mauritius –2.2 –4.8 São Tomé and Principe –3.3 –32.1 Madagascar –2.1 –9.1 Macedonia, FYR –3.2 –6.0 New Caledonia –2.1 — Maldives –3.2 –25.1 Kyrgyz Republic –2.1 –5.0 Micronesia, Fed. States of –3.1 — Lao PDR –2.1 –7.9 Palau –3.0 Delivered — by The World Armenia Bank e-library to: –2.1 –2.3 The World Bank IP : 192.86.100.36 Sources: World Bank; IMF. Tue, 10 Mar 2009 16:55:33 Note: — = not available. 35 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table 1.5 Impact of a 15 percent fall in non-oil pected impact with already large current-account commodity deficits. Such countries are unlikely to be able to % of GDP find additional financing for their oil bills and, as Change in current Current account a result, could be expected to undergo significant account balance balance (2005) real-side adjustments as the volume of domestic Guyana –8.3 –25.1 demand, as well as oil (and nonoil) imports, would Tajikistan –7.3 –4.2 Suriname –7.0 –12.3 have to be cut in order to finance the higher cost Solomon Islands –4.5 –14.2 of imported oil. Belize –3.8 –14.2 Table 1.5 reports the expected terms-of-trade Mauritania –3.8 –29.6 Mongolia –3.7 –2.8 impact from a 15 percent reduction in nonoil com- Paraguay –3.5 –0.8 modity prices, as well as estimates for the current- Papua New Guinea –3.5 10.0 account deficit in 2005, for those countries where Kyrgyz Republic –3.1 –5.0 Mali –3.0 –8.5 the impact would be greater than 2 percent of Côte d’Ivoire –2.9 2.2 GDP. While countries throughout the developing Ghana –2.9 –6.9 world would be hard hit, large impacts are con- Malawi –2.8 –7.1 Chile –2.5 –0.9 centrated in developing Africa. Indeed, for the re- Zimbabwe –2.3 46.6 gion as a whole, the negative impact would be 0.7 Zambia –2.2 –10.3 percent of GDP, or 1.2 percent of the GDP of Ukraine –2.0 1.3 Jamaica –2.0 –11.3 heavily indebted poor countries. While many of these countries currently have healthy current- Source: World Bank; IMF. account balances (for example, Côte d’Ivoire, Papua New Guinea, Paraguay, and Ukraine) and can be expected to absorb even such a large shock to unsustainable levels. Depending on the avail- relatively easily, many others are already in a vul- ability of additional financing, this would require nerable state. For these countries, taking steps substantial retrenchment in these countries, likely now to improve the competitiveness of their ex- implying large cuts in government spending and port industries and reduce reliance on imports is reductions in domestic demand that would likely even more critical. translate into a period of sustained lower growth or a sharp recession. Encouragingly, a number of heavily indebted countries have taken advantage of favorable financing conditions to restructure Avian influenza their debt, reducing their sensitivity to changes in interest rates. As a result, countries that have expe- rienced financial crises in the past, such as Brazil, T he continued spread of the bird-to-bird ver- sion of avian influenza (or bird flu, also known by its scientific identifier H5N1), with lim- Mexico and Thailand, appear to be much less vul- ited bird-to-human transmission comprises part of nerable to a rapid rise in interest rates and do not the baseline forecast. A serious risk to the global appear in table 1.3. economy stems from the possibility that avian in- On average, for oil-importing low-income fluenza might mutate into a form of flu that is eas- countries, the initial terms-of-trade shock of a fur- ily transmitted between humans and for which the ther $30 hike in oil prices is estimated at 4.1 per- population has limited immunity.31 The human cent of their GDP. This would translate into a 2.7 and economic consequences of such a pandemic percent decline in domestic demand, with poten- are potentially very large and depend importantly tially serious impacts on poverty. For the most oil- on the nature of the flu that emerges and on the re- intensive economies, this could amount to as much actions of people as it spreads. as 8 percent of GDP (table 1.4). While many coun- tries throughout the developing world would be Economic consequences of a further spread hard hit, most countries could be expected again of bird-to-bird flu Delivered by The World Bank e-library to: to manifest the same resilience they showed The World BankThe principal economic impact of the H5N1 virus during the previous oil hike. Problems are mostIP likely to so far has come in the rural sectors of several : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 crop up in those countries that combine a large ex- Asian economies in which the disease is endemic. 36 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y Table 1.6 Impact of a widening of bird-bird flu additional 1.4 million people worldwide. A more % change in GDP, relative to the baseline virulent form, such as the 1918-9 flu, which was Bird-birda more deadly for healthy adults than a normal flu, World total –0.1 could have much more serious consequences, killing High-income countries –0.1 as many as 1 in 40 infected individuals (Barry 2005) Low- & middle-income countries –0.4 or some 71 million, with some authors suggesting East Asia & Pacific –0.4 that as many as 180–260 million could die in a Europe & Central Asia –0.4 worst-case scenario (Osterholm 2005). Latin America & Caribbean –0.7 Table 1.7 reports the results of three separate Middle East & North Africa –0.4 South Asia –0.4 simulations of the economic consequences of a Sub-Saharan Africa –0.3 pandemic (McKibbin and Sidorenko 2006). The Source: World Bank. first (mild) scenario is modeled on the Hong Kong a. Assumes that 12 percent of domestic birds in each region die flu of 1968-9; the moderate flu has the characteris- from the disease or are killed in efforts to prevent its spread. tics of the 1957 Asian flu; and the severe simula- tion is benchmarked on the 1918-9 Spanish flu.35 Each of these scenarios assumes that efforts by in- Its appearance in a number of European and dividuals and official agencies to limit the spread African countries suggests that the disease may be- of the disease are no more effectual than those ob- come as prevalent among the wild birds of these served during previous epidemics and reflects dif- continents as it is currently in Asia. ferences in population density, poverty, and the Table 1.6 reports an effort to estimate the eco- quality of health care available. For the world as a nomic impact of such a spreading of the current whole, a mild pandemic would reduce output by bird-to-bird flu. The reported results are based on less than 1 percent of GDP, a moderate outbreak a scenario where bird-to bird flu becomes endemic by more than 2 percent, and a severe pandemic by throughout the world to the degree observed in almost 5 percent, constituting a major global re- Vietnam in 2004 (approximately 12 percent of all cession. Generally speaking, developing countries domestic birds died from the disease or were would be hardest hit, because of higher population culled to prevent spread). While direct costs are densities, poverty and weaker health infrastruc- small (only 0.1 percent of world GDP),32 differing ture.36 In addition, as modeled, less flexible mar- degrees of international specialization and cost ket mechanisms accentuate the economic impacts structures suggest that, allowing for interactions in some countries. with other sectors, regional impacts could be as Table 1.8 shows an alternative modeling of a high as 0.7 percent of GDP.33 Because the sector is pandemic. It is based on a pandemic similar in more important in developing countries and rela- terms of mortality to the Asian flu epidemic of tively labor intensive, job losses could represent 1958. This scenario is presented with a view to bet- about 0.2 percent of the global work force, or ter understanding the factors driving the aggregate some 5 million jobs during the time it takes the global economy to adjust. Table 1.7 Possible economic impacts of flu Possible economic consequences pandemic of a human pandemic % change in GDP, first-year Even a flu with “normal” characteristics in terms of Mild Moderate Severe transmissibility and deadliness could have serious World –0.7 –2.0 –4.8 consequences for the global economy if the world’s High-income countries –0.7 –2.0 –4.7 population has limited immunity. Estimates suggest Developing countries –0.6 –2.1 –5.3 East Asia & Pacific –0.8 –3.5 –8.7 that such a flu could infect as much as 35 percent of Europe & Central Asia –2.1 –4.8 –9.9 the world’s population (WHO 2005), spreading Middle-East & North Africa –0.7 –2.8 –7.0 throughout the world in as few as 180 days (RTI, South Asia –0.6 –2.1 –4.9 Delivered by The World Bank e-library to: 2006). As compared with a normal flu season,The World Deaths (millions) Bank 1.4 14.2 71.1 34 where some 0.2–1.5 million die (WHO 2003), IP : 192.86.100.36 Source: Tue, 10 Mar 2009 World Bank calculations based on McKibbin & Sidorenko 16:55:33 deaths from even a mild new flu might include an (2006). 37 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table 1.8 A breakdown of economic impacts of a potential human-to-human pandemic % of GDP Impact of illness Impact of efforts Total Mortalitya and absenteeismb to avoid infectionc Total ($ billions) World total –0.4 –0.9 –1.9 –3.1 –965.4 High-income countries –0.3 –0.9 –1.8 –3.0 –744.9 Low- and middle-income countries –0.6 –0.9 –2.1 –3.6 –220.4 East Asia & Pacific –0.7 –0.7 –1.2 –2.6 –44.8 Europe & Central Asia –0.4 –0.7 –2.3 –3.4 –21.7 Latin America & Caribbean –0.5 –0.9 –2.9 –4.4 –87.3 Middle East & North Africa –0.7 –1.2 –1.8 –3.7 –32.2 South Asia –0.6 –0.8 –2.2 –3.6 –22.7 Sub-Saharan Africa –0.6 –0.9 –2.2 –3.7 –11.8 Source: World Bank. a. Assumes a human flu pandemic similar to the 1958 Asian flu. Globally 1.08 percent of the world population dies, with regional mortallity rates varying from 0.3 percent in the U.S. to more than 2 percent in some developing countries. b. Assumes that for every person that dies 3 are seriously ill, requiring hospitalization for a week and absence from work for two weeks, 4 require medical treatment and are absent from work for a week and approximately 27 percent of the population has a mild bout of flu requiring two days absence from work. It assumes that in addition for every sick day another absentee day is registered either because people stay at home to care for a sick person or to avoid illness. c. Efforts to avoid infection are modelled as a demand shock, reflecting reduced travel, restaurant dining, hotels, tourism and theatre as individuals seek to avoid contact with others. numbers in such simulations. The first column period—implying about 15 percent decline from shows the impact in terms of GDP lost in the first trend (Siu and Wong, 2004). Higher declines on an year of the pandemic purely from additional deaths annualized basis are assumed in these simulations (here roughly equal to McKibbin and Sidorenko’s because a flu pandemic would likely last more severe scenario). The second column builds in the than a year (pandemics are typically experienced impact on aggregate productivity resulting from in at least two waves with a peak period of infec- the infection of some 35 percent of the population. tion during the winter). Even though individuals are only temporarily un- The total impact of a shock combining all available from work, the impact on output here is these elements is 3.1 percent for the global econ- more than twice as large as from the loss of life, be- omy and ranges from 4.4 percent in Latin America cause the affected population is so much larger. and the Caribbean to 2.6 percent in the East Asia The third column shows the largest impact. and Pacific region, mainly reflecting the relative Here individuals are assumed to change their be- importance and labor intensity of tourism and havior in the face of the pandemic by (a) reducing other services in each region. air travel in order to avoid infection in the en- The modeling attempted to take into account closed space of a plane, (b) avoiding travel to in- the possibility that the economic effects of an out- fected destinations, and (c) reducing consumption break would be greatest in the country where the of services such as restaurant dining, tourism, human-to-human strain originates, the main fac- mass transport, and nonessential retail shopping. tor here being private and public efforts to isolate The degree to which such reactions would occur is and contain the disease by avoiding travel and im- necessarily uncertain. In this scenario it was as- posing quarantines. However, simulations of an sumed that for the year as a whole air travel would outbreak beginning in Thailand suggest that what- decline by 20 percent and that tourism, restaurant ever additional costs the originating country may meals, and consumption of mass transportation endure, these would be dominated by secondary services would also decline by 20 percent. effects as the disease spreads to other countries This compares with a peak decline of 75 per- and global economic activity declines. cent in air travel to Hong Kong during the SARS Given the tremendous uncertainties surround- epidemic and an average decline of 50–60 percent ing the possibility and eventual nature of a pan- Delivered by The World Bank e-library to: during the four-month period the outbreak was Bankdemic, these simulations must be viewed as purely The World active. Retail sales declined by 15 percent at the illustrative. They provide a sense of the overall IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 peak and by about 9 percent over the four month magnitude of potential costs. Actual costs, both in 38 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y terms of human lives and economic losses, are 7. While large in percentage terms, the increase in oil prices from around $10 to $20 a barrel between 1999 and likely to be very different. 2000 is not considered as part of the oil shock, because it That said, these simulations serve to underline merely reflected the reversal of a similar fall in prices the the importance of mobilizing global efforts to year before. meet this potential crisis. Monitoring outbreaks of 8. The short-term price elasticity of oil demand is esti- bird-to-bird and bird-to-human infections and mated at between –0.01 and –0.2 percent (Burger 2005), implying that immediately following a 100-percent increase culling infected flocks appear to be effective strate- in oil prices, such as observed between 2002 and 2005, oil gies to reduce bird-to-human transmission and re- demand would be expected to decelerate by between 1 and duce the likelihood that the disease will mutate 20 percent. Long-term elasticities are larger (between –0.2 into a form that is easily transmissible between hu- and –0.6 percent), implying that the negative effect of higher mans. The fact that there have been no reported prices over the past few years will continue to be felt. 9. The current rise in oil prices began in early 2003. cases of bird flu in Vietnam in the 2005-6 flu sea- 10. OPEC did increase its deliveries during 2004 by son suggests that such preventative efforts can be drawing down its spare capacity, but so far investments to effective. increase that capacity have been limited. However, even with such efforts, an eventual 11. In the three years following both the 1973 and human pandemic at some unknown point in the 1979 oil price hikes, non-OPEC non-former Soviet Union oil producers increased their output by some 3.5 million future is virtually inevitable (WHO, 2004). Be- barrels per day. In contrast, since 2002, production from cause such a pandemic would spread very quickly, these sources has actually declined. substantial efforts need to be put into place to de- 12. Beccue and Huntington (2005) estimate the proba- velop effective strategies and contingency plans bility of a 2 mbpd supply shock occurring during the next that could be enacted at short notice. Much more 10 years as 70 percent for one lasting 6 months and 35 per- cent for one lasting 18 months. research and coordination at the global level are 13. Baffes (2005) estimates the elasticity of nonoil required. commodity prices to oil prices to be 0.15. 14. Normally, the yield curve is upward sloping, im- plying that bonds of shorter duration yield lower rates of re- Notes turn than longer term bonds. This upward slope is generally 1. In addition to the Prospects for the Global Economy thought to reflect individuals’ time preference for money, on web site (http://www.worldbank.org/outlook) the World the one hand, and the increased risk associated with longer Bank’s East Asia update provides more detailed information term lending. on recent developments and prospects for the East Asia and 15. For low- and middle-income countries as a whole, Pacific region (http://www.worldbank.org/eapupdate/). net bank lending actually exceeded bond emissions by a 2. The World Bank’s East Asia Update provides addi- small margin. tional detail on avian influenza in the region (http:// 16. About as many appreciated as depreciated. Over- www.worldbank.org/eapupdate/). all, the unweighted average impact was a real effective de- 3. In addition to the Prospects for the Global Economy preciation of just 1 percent. web site (http://www.worldbank.org/globaloutlook), which 17. The unweighted average appreciation of oil and provides more detail on the regional forecasts, the World mineral exporters was smaller, at around 9 percent. Bank’s Middle East and North Africa Region’s Economic De- 18. Simulations using the World Bank’s MAMS model velopments and Prospects (http://www.worldbank.org/mena) (a computable general equilibrium model for studying the provides country-specific analysis of economic developments, impact of aid on achieving the Millennium Development projections, and policy priorities. goals) indicate that a negative term-of-trade shock of 1 per- 4. For the purposes of this report, the developing coun- cent of GDP would reduce import volume growth in the tries of the region are Algeria, Egypt, Jordan, Iran, Mo- first year by 2 percent. When combined with a 1 percent of rocco, Oman, Syria, Tunisia, and Yemen. A lack of data pre- GDP increase in aid flows, imports fall by only 0.7 percent. vented inclusion of Djibouti, Iraq, Lebanon, and Libya from 19. The same simulations suggest that the real apprecia- the projections. Important regional players include the high- tion from a permanent increase in aid inflows equal to 1 per- income countries of Bahrain, Kuwait, Qatar, Saudi Arabia, cent of GDP would reduce exports by about 3 percent in the and the United Arab Emirates. first year and .66 percent per annum over a 10-year period. 5. Fiscal and quasi-fiscal spending increased by 0.7 When combined with a negative terms-of-trade effect equal to percent of GDP in Bangladesh, by 0.5 percent of GDP in 1 percent of GDP, the appreciation is reduced by half and the India, and by significant, though lesser, amounts in other impact on export growth rates reduced by 10 percent. countries of the region. 20. In the case of Turkey, the central bank has tight- 6. More than one-third grew faster than 5 percent on ened policy rates, while Delivered by The World Bank e-library to: in Bulgaria the rise in interest rates average between 2000 and 2005, compared with less thanThe World is an automatic response to capital inflows by the country’s Bank 10 percent during the period 1980–1995. currency board system. IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 39 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 21. The current-account deficit of the United States with other high-income countries, the lost market share has came in at $805 billion or about 6.4 percent of U.S. GDP. not been recouped. 22. These estimates are based on three scenarios. In the 27. Between 1972 and 2004, China went from exporting first, the current-account deficit is assumed to remain con- 510 separate goods to 10,199 (Borda and Weinstein 2004). stant at 6 percent of GDP; nominal GDP is projected to in- 28. In the case of China, many behind-the-border crease by 5 percent per annum; and exchange rates and rates changes were precipitated by the country’s desire to join the of return of U.S. and foreign assets are to remain constant. World Trade Organization. Similarly, many reforms in the Because net returns fall to –1.2 percent of GDP, this implies European transition countries were motivated by the desire an improvement in the U.S. current-account deficit equal to of those countries to join the European Union. 1 percent of GDP. A second scenario assumes that the cur- 29. Econometric estimates suggest that over the past rent-account deficit declines to 2.5 percent of GDP, implying three years the underlying trend growth in China was 11.7 a substantial improvement in the U.S. trade balance equal to percent. WTO accession contributed an additional 12 per- 0.5 percentage point per year. In the third scenario, the rates cent to Chinese export growth. Market growth was worth of return on U.S. and foreign assets are assumed to equalize, 6.3 percent. Relative price changes reduced the total by 4.2 increasing net debt-servicing costs to 2.1 percent of GDP. percent (Martel Garcia, forthcoming). 23. Empirically, this willingness takes three principal 30. Only countries where the estimated impact equals forms. First, foreigners hold a higher share of relatively or exceeds 1 percent of GDP are shown. low-yield dollar-denominated assets than do Americans— 31. There are a number of kinds of avian influenza that reducing the overall earnings on their assets. Secondly, as are carried by many wild bird species with no apparent harm. recorded in the balance of payments, American investments Some of these make other bird species, notably domestic poul- abroad earn a significantly higher rate of return than do try, sick. Typically, the birds are mildly sick, but the H5N1 foreign investments in the United States (6.9 percent vs. 2.5 virus that is currently circulating is relatively dangerous for do- percent over the past 10 years). Finally, foreigners hold mestic birds. Most forms of avian influenza viruses are highly large quantities of dollars in cash, which earn no return. species-specific and do not normally infect people. However, These three factors, in combination, mean that despite the H5N1 has crossed the species barrier to infect humans on negative net international asset position of the United three occasions in recent years—in Hong Kong in 1997 and States, the country continues to earn a small but positive during the current outbreak, which began in December 2003. net income from capital services. While deadly (115 human deaths among 208 confirmed cases 24. Haussman and Sturzenegger (2005), in a controver- as of May 12, 2006), the virus in its current form is not easily sial article, take this observation to an extreme. They argue transmitted to or between humans (WHO 2006). that if the United States earns a positive return on its net for- 32. Direct costs are small. Six percent of the world eign asset position, in economic terms, it must be positive. population of domestic poultry amounts to some 170 mil- They propose to measure it as the net present value of the in- lion birds. At a retail price of $2 per bird, and assuming come stream recorded in the balance of payments. They then (based on the Vietnamese experience) 0.75 cents in costs as- redefine the current-account balance as the change in that sociated with monitoring and culling infected birds, this net asset position (effectively 20 times the annual change in would amount to about $760 million worldwide, or about income flow). Finally, they define the difference between this 0.02 percent of world GDP. measure and the normal current account of the balance of 33. While the poultry sector represents less than 0.2 payments as exports of “dark matter,” or know-how services percent of the GDP of high-income countries, its share in embodied in FDI, insurance services provided by less risky developing countries is about 1.2 percent of GDP, rising to U.S. assets, and liquidity services deriving from the quality of 2.4 percent of GDP in the East Asia and Pacific region. the U.S. dollar as the world reserve currency. On this basis 34. The World Health Organization (2003) estimates they compute that the net asset position of the United States between 200,000 and 500,000 deaths each year. Osterholm was actually a small surplus in 2004. (2005) reports a higher death toll of between 1 and 1.5 mil- 25. Interestingly, such a change in the willingness of in- lion people worldwide from influenza infections or related vestors to hold U.S. assets would cause Haussman and complications, making it the third most deadly infectious Sturzenegger’s (2005) definition of the net international in- disease after AIDS and tuberculosis, but ahead of malaria. vestment position of the United States to deteriorate by 10 35. McKibbin and Sidorenko also model an “Ultra” flu, percent of GDP, and would imply an equal fall in their esti- which is not based on any known previous pandemic, but has mate of the current account—highlighting the sensitivity of the characteristics of the Spanish flu, plus higher mortality for their measures to interest rates and unmeasurable confi- older people. This simulation is not reported here. dence factors. 36. McKibbin and Sidorenko’s model has relatively 26. While economic factors certainly have played a limited country coverage: 20 economies. comprised of 10 role (the erosion of market share among high-income coun- high-income countries and 1 residual high-income region; 5 tries mirrors earlier developments), political factors also low- and middle-income countries in East Asia and one in played a role. In particular, the imports of oil importers South Asia; and three additional developing regions. Re- from the United States declined substantially in the period gional aggregates in table 1.7 are approximations based on 2001/2. While growth rates since then have been Delivered on World by The a par Bankthe to: and regions modeled. countries e-library The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 40 (c) The International Bank for Reconstruction and Development / The World Bank P R O S P E C T S F O R T H E G L O B A L E C O N O M Y References Kehoe, T. J., and K. J. Ruhl. 2003. “How Important Is the New Goods Margin in International Trade?” Research Baffes, John. 2005. “Oil Spills over to Other Commodities.” Department Staff Report 324, Federal Reserve Bank of Unpublished paper, World Bank, Washington, DC. Minneapolis. Barry, John M. 2005. The Great Influenza: The Epic Story Laster, David, and Robert N. McCauley. 1994. “Making of the Deadliest Plague in History. London: Penguin. Sense of the Profits of Foreign Firms in the United Beccue, Phillip C., and Hillard G. Huntington. 2005. “An States.” Quarterly Review (Federal Reserve Bank of Assessment of Oil Market Disruption Risks.” Energy New York) 19 (2). Modeling Forum, Stanford University. October. Martel Garcia, Fernando. 2006. “Understanding Long Ex- http://www.stanford.edu/group/EMF/publications/doc/ port Booms: The Case of China.” Unpublished paper, EMFSR8.pdf World Bank, Washington, DC. Billings, Molly. 1997. “The Influenza Pandemic of 1918.” Mataloni, Raymond J. 2001. “An Examination of the Low Stanford University. http://www.stanford.edu/group/ Rates of Return of Foreign-Owned U.S. Companies.” virus/uda/. Survey of Current Business (March): 55–73. Borda, C., and D. C. Weinstein. 2004. “Globalization and the McKibbin, Warwick, and Alexandra Sidorenko. 2006. Gains from Variety.” NBER Working Paper 10314, Na- “Global Macroeconomic Consequences of Pandemic tional Bureau of Economic Research, Cambridge, MA. Influenza.” Lowy Institute for International Policy, Buiter, Willem. 2006. “Dark Matter or Cold Fusion.” Sydney Australia. Global Economics Paper 136. Goldman Sachs, New Newfarmer, Richard, ed. 2005. Trade, Doha, and Develop- York. January. ment: A Window into the Issues. World Bank, Wash- Cline, William. 2005. The United States as a Debtor Nation. ington, DC. Washington, DC: Institute for International Economics. Osterholm, Michael T. 2005. “Preparing for the Next Pan- Devarajan, Shantayanan, and Ejaz Ghani. 2006. “Oil Price demic.” New England Journal of Medicine 352 (May): Shocks, Fiscal Adjustment, and Poverty Reduction in 1839–42. Asia.” Paper presented at the SAARC finance seminar on Roubini, Nouriel, and Brad Setser. 2005. “Will the Bretton “Current Oil Price Shock and Its Implications on South Woods 2 Regime Unravel Soon? The Risk of a Hard Asian Economies,” Colombo, Sri Lanka. January. Landing in 2005–2006.” Paper presented at conference Diaz-Bonilla, Carolina, and Cristina Savescu. 2006. “Simu- on “Revived Bretton Woods System: A New Paradigm lations of Economic Impacts of Official Development for Asian Development,” San Francisco, February 4. Assistance and Oil Price Increases: Ethiopia.” Unpub- Setser, Brad. 2006. “On the Origins of Dark Matter.” Janu- lished paper, World Bank, Washington, DC. ary http://www.rgemonitor.com/blog/setser/113810. Estrella, Arturo. 2005. “Why does the yield curve predict Siu, Alan, and Y. C. Richard Wong. 2004. “Economic Im- output and inflation?” Economic Journal. July. pact of SARS: The Case of Hong Kong.” Hong Kong Frankel, Jeffrey A., and David Romer. 1999. “Does Trade Institute of Economics and Business Strategy, Working Cause Growth?” American Economic Review 89 (3): Paper 1084. April. 379–99. World Bank. 2005a. Global Development Finance 2005: Higgins, Mathew, Thomas Klitgaard, and Cedric Tille. Mobilizing Finance and Managing Vulnerability. World 2005. “The Implications of Rising U.S. International Bank: Washington, DC. Liabilities.” Current Economic Issues (Federal Reserve ———. 2005b. Global Economic Prospects 2006: Economic Bank of New York) 11 (5). December. Implications of Remittances and Migration. Washing- Hausmann, Ricardo, and Federico Sturzenegger. 2005. ton, DC. “Global Imbalances or Bad Accounting? The Missing World Health Organization. 2003. “Influenza.” Fact Sheet Dark Matter in the Wealth of Nations.” Unpublished 211, Geneva. http://www.who.int/mediacentre/factsheets/ paper, Harvard University. fs211/en/index.html. Hummels, D., and P. J. Klenow. 2004. “The Variety and ———. 2004. “World Is Ill-Prepared for “Inevitable” Flu Quality of a Nation’s Exports.” Unpublished paper. Pandemic.” Bulletin of the World Health Organization Purdue University and Stanford University. 82 (4): 317–18. Ibrahim, Izani, and Craig MacPhee. 2003. “Export External- ———. 2006. “Avian Influenza (“bird flu”) Fact Sheet.” ities and Economic Growth.” Journal of International http://www.who.int/mediacentre/factsheets/avian_in- Trade & Economic Development 12 (3): 257–83. fluenza/en/index.html. February 2006. IMF (International Monetary Fund). 2006. World Economic Outlook. IMF: Washington, DC. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 41 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank . 2 The Growth and Transformation of Private Capital Flows I n 2005, global capital flows to developing on cross-border financial flows. Several countries, countries continued to grow at a record pace. especially in East Asia, made concerted efforts to Net private flows increased sharply by $94 accumulate precautionary reserves and build their billion, reaching $491 billion, reinforcing a trend domestic bond markets to better manage risks as- underway since 2002. The sharp rise came de- sociated with foreign portfolio flows. spite lingering uncertainty about the impact of The favorable environment for cross-border higher oil prices, rising global interest rates, and mergers and acquisitions and a new wave of priva- growing global payments imbalances. The flows tizations, particularly in the new member coun- have been broad-based, with bond issuance, tries of the European Union (EU), pushed FDI to bank lending, foreign direct investment (FDI), an all-time high of $238 billion. The increase and portfolio equity all recording substantial raised the share of developing countries in global gains (figure 2.1). During the year, governments FDI flows from 13 percent in 2000 to 24 percent and private entities took advantage of favorable in 2005. During the year, share prices quoted on financial-market conditions to refinance their emerging market stock exchanges turned in a stel- debt or prefund future borrowing. As a result, lar performance, receiving record flows of portfo- foreign currency–denominated bond issuance by lio equity. Stock issuance by emerging market governments and the private sector rose to a countries in international financial markets also record gross of $131 billion in 2005. The spread grew substantially. on emerging market debt dropped to historic lows, averaging 306 basis points for 2005, com- Figure 2.1 Net private debt flows to developing pared with the 2004 average of 423 basis points countries, 1991–2005 and the recent high of 832 basis points, recorded in September 2002. Meanwhile, local-currency $ billions 500 bond markets in Asia and Latin America at- Total net private capital flows tracted substantial interest from international in- 100 vestors in search of higher yields and potential gains from currency appreciation. 300 Accounting for the growth in recent years have been the policy responses to the financial 200 Net equity flows crises of the 1990s, a favorable environment for 100 Net debt mergers and acquisitions, a wave of privatizations, flows and innovations in the global marketplace. In the 0 aftermath of the financial crises of the 1990s, many major emerging markets adopted more flex- –100 Delivered by The World 1991 Bank e-library 1993 to: 1997 1999 2001 2003 2005 1995 ible exchange rate policies, while strengtheningThe World Bank domestic financial markets and relaxing controls IP : 192.86.100.36 Source: World Bank Debtor Reporting System. Tue, 10 Mar 2009 16:55:33 43 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Financial innovations in global financial mar- burgeoning government debt denominated in kets—notably local-currency financing and struc- local currencies. tured finance instruments—have allowed investors • Credit default swaps (CDSs)—derivatives that to assume greater exposures in emerging markets. provide some insurance to the buyer against The euro has emerged as a major international re- defaults and other adverse credit events—are serve currency and as an increasingly important is- being applied in new ways in emerging securi- suing currency for governments and the corporate ties markets—among them those of Bulgaria, sector in developing countries. the Republic of Korea, Mexico, Peru, the This chapter provides updates on all types of Philippines, and the Russian Federation. This private capital flows to developing countries, ex- development has important implications for ploring some implications of the increased im- the pricing and supply of debt capital to devel- portance of the euro, the fast-growing credit de- oping countries, because it offers investors an- rivatives markets, and the increasing reliance of other way of assuming exposure to emerging many countries on local-currency funding. The market risk and enhances the markets’ ability key messages emerging from this review are high- to gauge credit risk. Also, by transferring to lighted below. other market participants some of the credit risk that banks incur in their lending and trad- • Developing countries have benefited from ing activities, credit derivatives have altered, strong economic growth and sounder macro- perhaps fundamentally, the traditional ap- economic policies, leading to marked improve- proach to credit-risk management. Presently, ments in their external payment positions. De- only a few banks engage in CDSs in emerging spite the easing of financing conditions, markets, posing the risk that a failure of a however, developing countries’ access to inter- major player could create broader risks. Trad- national capital markets remains limited. Pri- ing takes place largely in the private over-the- vate capital flows to the developing world are counter market and thus lacks transparency. concentrated in just a few countries. Of the Regulators in developing countries need to 136 that report to the World Bank, 51 con- build their capacity to monitor CDS transac- tinue to rely primarily or entirely on official tions and to define a clear line of regulatory sources of cross-border finance. If they are to responsibility and expertise so as to better attract and absorb private capital effectively manage the associated risks. for long-term growth and development, they • The strong recovery of FDI in developing will need to, inter alia, further develop their countries over the past two years reflects domestic financial markets and institutions. healthy global economic conditions and a bet- • Local-currency bond markets in developing ter investment climate in developing countries. countries have, since the crises of the 1990s, While increased corporate profits, favorable emerged as a major source of long-term devel- financing conditions, and higher stock-market opment finance and are now the fastest grow- valuations fueled cross-border investments ing segment of emerging market debt. Driven globally, many developing countries managed largely by domestic institutional and individ- to attract high levels of FDI through privatiza- ual investors, these markets grew from $1.3 tions, mergers, and acquisitions. Almost all trillion at the end of 1997 to $3.5 trillion in developing countries experienced higher FDI September 2005. However, bringing the local- inflows, but the increase was especially no- currency bond markets in emerging economies table in new members of the European Union. up to the standards of mature markets will re- In China, liberalization of the financial sector quire concerted efforts. The East Asian coun- and accession to the World Trade Organiza- tries may provide a case worth watching in tion led to several important privatization this regard, given their early successes. Local- deals in the banking sector in 2005. Many currency debt markets also present new chal- middle-income countries received high levels Delivered by The World Bank e-library to: lenges for policy makers. ProfessionalismThe Worldin Bank of services-related FDI through privatizations, debt management will be needed to IPmanage : 192.86.100.36 while FDI to low-income countries grew prin- Tue, 10 Mar 2009 16:55:33 currency and duration risks associated with cipally because of high commodity prices. 44 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S • In the years ahead, policy makers in develop- 2005. As a result, foreign private debt flows have ing countries will have to remain alert to cer- become more soundly based and resilient to tain risks and vulnerabilities. The current glut swings in external financing conditions. of liquidity in the global financial markets may lead to a buildup of risky exposures, as Bond issuance set records in 2005 investors in search of higher yields settle for The investment community now accepts emerging borrowers of lower creditworthiness. The market debt as a bona fide asset class that is becom- locus of credit risk in developing countries is ing less volatile. The spread on such debt has shifting as private corporates, rather than sov- dropped to historic lows, with an average of just ereigns, are emerging as the main borrowers 306 basis points in 2005, compared with 423 basis in global credit markets. Political risk has points in 2004 (box 2.1). In 2005, developing coun- emerged once again as a key concern for tries raised a record $131 billion in 367 bond issues, emerging market investors. In several coun- an increase in proceeds of 28 percent from 2004. tries, populist candidates will stand for elec- Net issuance of foreign currency–denominated tion in 2006, raising the fear of policy changes bonds last year amounted to $62 billion, less than that could reverse the gains from recent fiscal half of the total raised. stabilization and liberalization measures. Meanwhile, the traditional policy discipline Table 2.1 Net private debt flows to developing countries, 2002–5 $ billions and frameworks agreed to with multilateral lenders are becoming less prominent with the 2002 2003 2004 2005 dwindling need for official financing. The cu- Total net debt flows 5.5 85.1 144.8 191.6 mulative risks are particularly pronounced in By region: oil-importing countries like Turkey and the East Asia and Pacific –2.4 9.3 43.3 45.8 Philippines, which have suffered from recent Europe and Central Asia 24.9 64.7 93.7 113.8 Latin America and the Caribbean –21.4 5.4 –1.0 20.5 oil price increases without benefiting from the Middle East and N. Africa 4.8 2.1 2.3 4.6 commodity price boom. South Asia 2.4 2.1 6.7 3.0 Sub-Saharan Africa –2.8 1.5 2.8 3.8 By component Bond financing 10.8 26.4 43.0 61.7 Bank financing –2.8 9.8 39.4 64.4 Private debt market developments Other financing –6.8 –5.9 –4.6 –6.7 in 2005 Short-term debt financing 4.2 54.9 70.8 69.3 I n 2005, net private debt flows to developing countries increased sharply to an estimated $192 billion, up from $148 billion in 2004 and $85 bil- Source: World Bank Debt Reporting System. Table 2.2 Gross market-based debt flows to developing countries, lion in 2003 (table 2.1). The net increase reflected 2002–5 an increase in gross financing through bonds and $ billions syndicated loans, which set record highs, with flows 2002 2003 2004 2005 54 percent higher in 2005 than in 2004 (table 2.2). Total gross flows 120.8 168.9 214.3 329.1 New bank lending was particularly strong, swelling Bonds 51.7 82.2 102.4 130.9 to $198 billion in 2005 from $112 billion the year East Asia and Pacific 12.5 11.6 15.7 20.3 before. Bank lending now accounts for 60 percent Europe and Central Asia 13.8 26.5 38.2 54.7 of gross debt flows and more than two-thirds of the Latin America and the Caribbean 21.1 38.8 35.9 43.0 Middle East and N. Africa 2.7 1.0 5.6 5.4 increase from 2004 (table 2.2). South Asia 0.2 0.5 5.1 5.3 Driving the strong upswing in foreign private Sub-Saharan Africa 1.5 3.9 2.0 2.3 debt flows are abundant global liquidity, steady Bank lending 69.1 86.9 111.8 198.1 improvements in developing-country credit qual- East Asia and Pacific 21.5 26.9 19.5 34.5 Europe and Central Asia 16.8 22.2 37.8 77.6 ity, lower yields in developed countries, and con- Latin America and the Caribbean 18.5 20.6 29.9 46.3 tinued broadening of the investor base for emerg- Middle East and N. Africa 5.8 4.6 9.7 15.7 Delivered by The World Bank e-library to: ing market assets. Upgrades in credit ratings haveThe World South Asia Bank 1.7 4.0 7.0 12.2 Sub-Saharan Africa 4.9 8.5 7.9 11.9 outpaced downgrades for eight consecutive quar- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 ters, with 46 upgrades and 18 downgrades in Sources: Dealogic Bondware and Loanware and World Bank staff. 45 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Box 2.1 The emerging bond market enters the mainstream E merging market debt is heading firmly into the main- stream of global bond trading. The traditionally high idiosyncratic risk associated with emerging market bonds • Fourth, the extraordinary narrowing of spreads has been accompanied by a parallel move to smaller daily fluctuations—both lower variability and fewer ex- has declined significantly since 2002, a trend reflected in treme changes (see figure at lower right). The fre- the spreads of such bonds over U.S. Treasuries. Four im- quency distribution of changes in daily spreads seems portant features of this transformation are: to be best characterized as nonnormal, having fatter and asymmetric tails (kurtosis and skewness). A mea- • First, emerging market bond spreads are moving in- sure of the nonnormality, the Jacques-Bera test,b indi- creasingly in tandem with U.S. high-yield bonds (see cates that the distribution became more normal in figure at top left). In the midst of uncertainty about 2002–2004 because of a decline in excess kurtosis, al- the fate of the Brazilian economy in 1998, emerging though non-normality was higher again in 2005 be- market spreads were 1,200 basis points. At the end of cause kurtosis and skewness were both higher. Skew- 2005 they were just over 200 basis points.a The de- ness was significantly negative in several years, cline occurred despite Argentina’s default in 2002, a including 2005, indicating that longer tails to the left period of tightening of U.S. monetary policy during were probably caused by the decline in spreads.c 2004–5, and turmoil in the U.S. corporate bond mar- ket caused by downgrades of car makers. Source: World Bank staff calculations based on various data sources. • Second, volatility in emerging market bond spreads, as a. The EMBIG is affected by the removal of defaulted bonds from the index; adjusting for these changes, however, gives the same picture of a dra- measured by the standard deviation of Emerging Mar- matic decline in spreads. ket Bond Index (EMBI) spreads, has declined signifi- b. The Jacques-Bera test statistic is (N/6)(.25K2+S2), where N is the number of observations, K is excess kurtosis, and S is skewness. It is distributed as a cantly since 1999 (see top right figure on next page). chi-square with 2 degrees of freedom, so that a value in excess of 6 indi- • Third, emerging bond indices are becoming more cates rejection of normality at the 5 percent level. strongly correlated with both global and U.S. bond in- c. The distribution of daily changes in EM bond spreads is becoming more normal, in the sense that the excess kurtosis displayed in changes in spreads dices (see figure at lower left). The strength of the cor- has been declining roughly since 2001, although it has increased slightly in relation between emerging market and global bond the 2004–5 period. The standard deviation has also declined significantly markets has been increasing for five years. from a high of 21.8 in 2001 to 5.7 in 2005. Over this period, the distribu- tion tended toward a normal distribution, since the Jacques-Bera test statis- tic has been declining, with the exception of 2005. Bond issuance was concentrated. Ten coun- Latin America and the Caribbean region ac- tries (Brazil, China, Hungary, India, Indonesia, counted for about 33 percent of total issuance, Mexico, Poland, the Russian Federation, Turkey, with Brazil’s government being the most active and República Bolivariana de Venezuela) ac- borrower. In 2005, the Brazilian government ex- counted for 69 percent of the issuance.1 Forty de- changed its outstanding C-bonds for U.S.-dollar- veloping countries accessed the international denominated global bonds having a face value of bond market, compared with 34 in 2002 and $4.5 billion and a maturity of 12 years, retiring a 2003. Countries from Europe and Central Asia third of its Brady debt. The Southern Copper Cor- accounted for 42 percent of total issuance in poration carried out a notable transaction in Mex- 2005, with Poland, the Russian Federation, and ico, issuing two U.S.-dollar-denominated bonds, Turkey leading the pack. Three of the five largest one with a maturity of 10 years ($200 million), issues in the region were by Russian firms, includ- and the other 30 years ($600 million). The average ing two U.S.-dollar-denominated bonds issued by maturity of fixed-rate issues by Latin American the financial entity Gazstream SA. In Poland, 13 firms in 2005 was 13.2 years. sovereign issues, totaling $12 billion, were issued Countries in East Asia and the Pacific issued to refinance the country’s Paris Club debt. Four of bonds to borrow $ 20.3 billion, with China being Delivered by The World Bank e-library to: these were publicly issued in the euro market, two Bankthe major issuer through government-owned The World in the global dollar market, and four in IP the Swiss banks. The Export-Import Bank of China and the : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 franc market. China Development Bank each issued $1 billion in 46 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S Convergence of emerging market bond spreads with Decline in emerging market bond volatility, 1994–2004 U.S. high-yield bonds, December 1998–December 2005 Rolling 36-month period; standard deviations Basis points Basis points 1190 345 EMBI Global 990 295 790 245 590 195 US High Yield 390 145 190 US AA grade 95 –10 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 8 99 00 00 01 1 02 03 03 4 4 05 5 99 00 00 00 00 Sources: JPMorgan Chase and World Bank staff calculations. 19 20 20 20 20 20 20 20 .1 .2 .2 .2 .2 ly b. p. r. ne n. g. ay ec ov ar ct ec Ap Ju Fe Se Ja Au O M Ju M D N D Sources: JPMorgan Chase and Merrill Lynch. Correlation of emerging market bond indices with Distribution of daily changes in emerging market bond global and U.S. indices, 1998–2005 spreads Rolling 36-month period, correlation Frequency 1.0 160 EMBI, US Broad 0.8 120 0.6 EMBI, Global Broad 0.4 80 0.2 0.0 40 – 0.2 – 0.4 0 1999 2000 2001 2002 2003 2004 2005 –55 –40 –25 –10 5 20 35 Sources: JPMorgan Chase and World Bank staff calculations. Sources: JPMorgan Chase and World Bank staff calculations. 10-year U.S.-dollar-denominated bonds. In Sep- sector issuers were able to borrow on better terms, tember 2005, the government of the Philippines thanks to the convergence of spreads for private completed its 2005 funding program by success- and sovereign issuers since 2003 (figure 2.3). fully issuing a 10-year U.S.-dollar-denominated In 2005, bond issuance covered the entire bond for $1 billion at a spread of 430 basis points credit spectrum, but almost half of the increase in over 10-year U.S. Treasuries. 2005 was accounted for by borrowers rated below Sovereign borrowers accounted for 46 percent investment grade. Investment-grade-rated borrow- of total issuance (figure 2.2) in 2005. They took ers accounted for 36 percent of 2005 issues, com- advantage of favorable market conditions to refi- pared to about 51 percent in 2002 (figure 2.4). Delivered by The World Bank e-library to: nance costlier debt and prefund future funding re-The World SinceBank late 2002, the favorable financing envi- quirements. Private sector issues increased as well, ronment has reduced the burden of arranging new IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 accounting for a third of issuance in 2005. Private financing for many borrowers—among them 47 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 2.2 Emerging market bond issuers by type, Figure 2.4 Bond market financing by risk category, 2002–5 2002–5 $ billions Share, $ billions 160 100 Others 140 Private Unrated 80 120 Public Sovereign/Govt. 60 Non-investment 100 grade 80 40 60 20 Investment grade 40 0 20 2002 2005 0 Source: Dealogic Bondware and Loanware. 2002 2003 2004 2005 Source: Dealogic Bondware; Bank of International Settlements. Syndicated bank loans showed a cyclical recovery Figure 2.3 Average spreads on new bond issues, Syndicated bank lending to developing countries set 2001–5 records in 2005. Gross bank lending of $198 bil- Primary market spreads (basis points) lion, an increase of 74 percent over 2004, involved 600 1,261 transactions in a broad range of sectors, Sovereign/Public dominated by oil and gas projects and oil import fi- 500 Private nancing. Europe and Central Asia accounted for 400 about 39 percent of the gross flows (table 2.3), fol- lowed by Latin America and the Caribbean (23 per- 300 cent) and East Asia (17 percent). Like FDI and bond issues, lending was highly concentrated, with the 200 top 10 countries (Brazil, Chile, China, India, Mex- 100 ico, Poland, the Russian Federation, South Africa, Thailand, and Turkey) receiving 70 percent of the 0 total bank lending to developing countries. Average 2001 2002 2003 2004 2005 gross flows to the top 10 grew by more than 107 Sources: Dealogic Bondware and World Bank staff calculations. percent, with lending to Thailand increasing by 455 percent in 2005. Sixty-seven countries, mostly low- income countries rated below investment grade or unrated, received no new loans at all. Brazil, Mexico, the Philippines, and Poland, all of In 2005, short-term debt to developing coun- which were able to prefund their 2005 financing tries increased by $61.9 billion to $556.7 billion, needs before mid-year. By late 2005, some sover- an increase of 12.5 percent from 2004. China ac- eigns were well advanced in financing their 2006 counted for 41 percent of the increase, with Brazil, and 2007 requirements. Infrequent and first-time Malaysia, the Russian Federation, and Turkey ac- borrowers, such as Pakistan and Vietnam, were counting for most of the balance. During 2000–5, able to tap international debt markets at attractive short-term loans grew considerably from the rates during 2005. Large institutional investors $316.4 billion recorded in 2000, with East Asia (such as public pension funds and endowment and Europe and Central Asia accounting for al- funds) as well as Asian central banks are now in- most all of the increase, while Latin America expe- Delivered by The World Bank e-library to: terested in investing in emerging market The debt be- Bankrienced a drop of 16 percent. (In 2005, short-term World cause of fundamental improvements IP :in the lending to Europe and Central Asia increased by 192.86.100.36 Tue, 10 Mar 2009 16:55:33 economies of major developing countries. 21 percent). Although global short-term debt has 48 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S risen, its size relative to developing countries’ for- Table 2.3 Gross cross-border loan flows, 2005 eign exchange reserves declined from 48 percent in Amount Amount Avg. loan size 2000 to around 28 percent at the end of 2005. No. of loans US$ millions % US$ millions The uses of financing raised by syndicated Total 1,261 198,135 100.0 158 bank loans vary considerably by region. Most lend- East Asia and Pacific 215 34,470 17.4 162 Europe and Central Asia 368 77,586 39.2 215 ing to the Russian Federation, which accounted for Latin America and the Caribbean 432 46,316 23.4 107 half of all flows to Europe and Central Asia, was Middle East and N. Africa 89 15,726 7.9 177 for oil and gas transactions, with the Gazprom ac- South Asia 101 12,151 6.1 121 Sub-Saharan Africa 56 11,887 6.0 203 quisition ($13.1 billion) accounting for almost two-thirds of the Russian total. In Latin America Source: World Bank staff calculations based on Dealogic Loanware data. and the Caribbean, the major borrowers were pe- troleum companies (Petrobras in Brazil and Pemex in Mexico) seeking to refinance existing loans or to ian Federation, and Thailand. All are rated finance trade. In East Asia, China received $18.5 investment-grade, have significantly lower billion (54 percent of the gross flows to East Asia spreads than the overall developing-country and the Pacific) for a broad range of transactions average, and exhibit low volatility in spreads. including oil and gas, property, project finance, and • Countries with access to bank lending only. purchase of aircraft. In Thailand, telecommunica- This category comprises countries that lack tion companies and utilities were the major bor- access to bond markets because of inadequate rowers. In South Asia, India received $11 billion, legal and institutional regulations or an un- or 91 percent of gross flows to the region. Pro- stable macroeconomic environment. Al- ceeds, most intermediated through Indian banks, though perceived as posing high credit risks, were used for projects such as a new airport in they can access bank credit because of well- Bangalore and trade financing. In Sub-Saharan defined revenue streams (such as exports and Africa, the major borrowers were central banks, remittances) or their ability to securitize bor- which refinanced existing borrowing at more at- rowing (often thanks to the presence of ex- tractive rates. In the Middle East and North Africa, tractive industries). Turkey was the major borrower, with almost all • Countries with limited access to capital mar- borrowing moving through Turkish banks for use kets. These are countries with no access to as trade financing. either bond markets or medium- and long- In several new EU member countries, includ- term bank lending. They may have access to ing Hungary and Slovenia, large financial and other types of private international finance, nonfinancial borrowers were able to borrow from such as short-term loans or FDI. Countries banks at spreads close to levels paid by their west- in this group rely mainly on official financ- ern European counterparts. Most loans were de- ing for their long-term capital needs. nominated in euros. Banks also invested in euro- denominated debt instruments issued by Poland Some 52 developing countries have accessed and Hungary. In Latin America, the oil and ce- the global bond markets each year since 2002. The ment sectors secured exceptionally cheap loans. number has not risen, despite the favorable financ- ing environment. Bond financing is more concen- The gap in access to credit persists trated than bank financing (figure 2.5). In 2005, Developing countries can be divided into three cat- 15 countries alone accounted for about 80 percent egories based on their degree and nature of access of bond volume. Non-investment-grade and un- to global capital markets (table 2.4): rated borrowers, who accounted for some 49 per- cent of total gross bond flows to emerging markets • Countries with access to bond markets. These in 2002, saw their share increase to about 64 per- are countries that have issued bonds regularly cent in 2005. Borrowers in bond markets from 10 since 2002. Included in this group are eight major emerging market economies, including Delivered by The World Bank e-library to: countries that are the developing-countryThe World Brazil, República Bolivariana de Venezuela, and Bank “stars” of the bond market—Chile, China, Turkey accounted for the bulk of the rise in high- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 Hungary, Malaysia, Mexico, Poland, the Russ- risk issuance in 2005. 49 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table 2.4 Countries’ access to international capital markets by intermediaries, 2002–5 Countries with access Countries with access to Countries with no access to bond markets Credit ratingsa bank lending onlyb Credit ratingsa to private debt marketsc Credit ratingsa Argentina B3 Albania NR Armenia NR Barbados Baa2 Algeria NR Benin B+ Belize Caa3 Angola NR Bhutan NR Brazil Ba3 Azerbaijan BB Burundi NR Bulgaria Ba1 Bangladesh NR Cambodia NR Chile Baa1 Belarus NR Cape Verde NR China A2 Bolivia B3 Central African Republic NR Colombia Ba2 Bosnia and Herzegovina B3 Chad NR Costa Rica Ba1 Botswana A2 Comoros NR Croatia Ba3 Burkina Faso B Congo, Dem. Rep. NR Czech Republic A1 Cameroon B- Côte d’Ivoire NR Dominican Republic B3 Congo, Rep. NR Dominica NR Ecuador Caa1 Djibouti NR Eritrea NR Egypt, Arab Rep. Ba1 Equatorial Guinea NR Fiji Ba2 El Salvador Baa3 Ethiopia NR Gambia, The NR Estonia A1 Gabon NR Georgia B+ Grenada B- Ghana B+ Guinea-Bissau NR Guatemala Ba2 Guinea NR Guyana NR Hungary A1 Honduras B2 Haiti NR India Baa3 Kenya NR Lesotho NR Indonesia B2 Kyrgyz Republic NR Madagascar B Iran, Islamic Rep. B+ Lao PDR NR Malawi NR Jamaica B1 Liberia NR Mauritania NR Jordan Baa3 Maldives NR Moldova Caa1 Kazakhstan Baa3 Mali B Mongolia B1 Latvia A2 Mauritius Baa2 Myanmar NR Lebanon B3 Mozambique B Nepal NR Lithuania A3 Nicaragua Caa1 Niger NR Macedonia, FYR BB+ Nigeria BB- Paraguay Caa1 Malaysia A3 Papua New Guinea B1 Rwanda NR Mexico Baa1 Senegal B+ Samoa NR Morocco Ba1 Seychelles NR São Tomé and Principe NR Oman Baa1 St. Lucia NR Sierra Leone NR Pakistan B2 Sudan NR Solomon Islands NR Panama Ba1 Tanzania NR Somalia NR Peru Ba3 Turkmenistan B2 St. Kitts and Nevis NR Philippines B1 Uzbekistan NR St. Vincent and the Grenadines NR Poland A2 Vanuatu NR Swaziland NR Romania Ba1 Yemen, Rep. NR Syrian Arab Republic NR Russia Baa2 Zambia NR Tajikistan NR Serbia and Montenegro BB- Togo NR Slovak Republic A2 Tonga NR South Africa Baa1 Uganda NR Sri Lanka BB- Zimbabwe NR Thailand Baa1 Trinidad and Tobago Baa2 Tunisia Baa2 Turkey Ba3 Ukraine B1 Uruguay B3 Venezuela, RB B2 Vietnam Ba3 Sources: Dealogic Bondware and Loanware, Moody’s, S&P and Fitch, and World Bank staff calculations. Note: This table classifies the 135 countries that report to the World Bank’s Debtor Reporting System (DRS) by accessibility to international capital markets across bond and bank segments (based on data cover transactions on international loan syndications and bond issues reported by capital-market sources, including Dealogic Bondware and Loanware). Countries are divided into three main categories: countries with access to bond markets, including all the countries that have issued bonds between 2002 and 2005; countries with access to bank lending only; coun- tries that have no access to either bond or bank lending, including countries that primarily rely on official financing for their financing needs. a. Long-term sovereign foreign currency debt ratings, as of February 3, 2006. Moody’s ratings were used for most of the countries. However, S&P and Fitch ratings were used for countries that are not rated by Moody’s, including Benin, Ghana, Grenada, Macedonia, FYR, Mali, Senegal, and Serbia and Montenegro. NR indicates countries that are not rated by either Moody’s or S&P. this table is b. For analytical purposes, bank lending inDelivered byonly The World to referred as medium- Bank and to:long-term lending (excluding short-term lending that e-library has less than 1 year of maturity). The World Bank c. The use of the term, “no access to capital markets,” is not IP intended to imply that all countries in this category do not have access to other : 192.86.100.36 types of international private capital, such as FDI and Tue, 10 Marequity. portfolio International capital defined here only refers to the bond and bank seg- 2009 16:55:33 ments of the market. 50 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S Figure 2.5 Concentration in bond and bank basis points from June 2004 to December 2005. financing, 1993–2003 However, the underlying pricing benchmark, usu- Share of top five countries (% of total) ally the six-month Libor rate, rose by almost 285 75 basis points, in step with the short-term U.S. inter- Bond est rates. In the end, this led to an increase of Bank 65 about 235 basis points in absolute borrowing costs over the cost in June 2004. The vast majority of developing countries 55 continue to rely on bank credit for their financing needs, despite rising costs. Information asymmetry 45 is one reason why bank lending is so much more common than bond financing. Because of their 35 close relations with clients and their ability to 1993 1995 1997 1999 2001 2003 monitor clients’ businesses, banks are better posi- Source: World Bank staff calculations based on Dealogic Bondware tioned than bond investors to gather information and Loanware data. on prospective borrowers, enabling banks to reach out to more borrowers. Higher-risk borrowers have no alternative to The difference between the cost of bond and bank financing. Between 2002 and 2005, some 80 bank financing narrowed substantially in 2005 due percent of bank loans were made to borrowers that to movements in spreads over benchmark pricing had no credit rating or were rated below invest- and changes in the benchmark rates (figure 2.6). ment grade. High-risk borrowers use such loans to For bond financing, spreads declined to an historic finance trade or specific projects, refinance debt, low in 2005, while the underlying benchmark long- and fund day-to-day operations (figure 2.7). Using term rate (10-year U.S. Treasury bonds) remained the bond markets for such core activities is not an depressed despite 10 hikes in short-term rates since option for high-risk borrowers. Since 2002, the June 2004. At the end of December 2005, the long- share of bank credit attributed to financing core ac- term rate was about 4.48 percent, compared with tivities has been rising, partly because borrowers 4.72 percent in June 2004, when short-term rates that could make the transition to bond financing began their rise. These developments caused ab- did so, thereby increasing the share of core financ- solute borrowing costs to drop from 8.8 percent in ing activities in remaining bank credit. June 2004 to 6.8 percent in December 2005. Although the average cost of bank borrowing For bank lending, the decline in spreads was has increased, the average maturity of bank loans not as stark as for bond financing, falling only 50 has grown as well—by about four years since Figure 2.6 Comparative cost of bond and bank financing, June 2004–December 2005 Spread and benchmark rates Average cost of borrowing Basis points Percent 500 9 Bond benchmark 400 Bond market Bank benchmark 7 300 Bond spread 200 5 100 Bank spread Bank lending 0 Delivered by The3World Bank e-library to: June Sep. Dec. Mar. June Sep. Dec. The June Sep. World Bank Dec. Mar. June Sep. Dec. 2004 2004 2004 2005 2005 2005 2005 IP : 192.86.100.36 2004 2004 2004 2005 2005 2005 2005 Tue, 10 Mar 2009 16:55:33 Source: World Bank staff calculations based on Bloomberg and J.P. Morgan Chase data. 51 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 2.7 Bank financing raised for core activities, debt and built up substantial liquidity with com- 2002–5 modity-driven windfall gains. Yet several net oil % of total bank financing raised importers, such as Thailand and South Africa, also 50 earned upgrades through strong growth and im- Acquisition proved economic management. Project finance 40 Refinancing debt General corp. purpose Portfolio equity showed major gains 30 Trade finance Portfolio equity flows to developing countries made major gains in 2005. At $61 billion, flows were up 20 sharply from $37 billion in 2004. The record gain was driven by a significant increase in international corporate equity placements in emerging markets 10 and foreign investment in emerging market stocks. The revival of interest in emerging market equity 0 2002 2003 2004 2005 can be traced to fundamental changes in emerging markets and to the growing popularity, among Sources: Dealogic Loanware and World Bank staff calculations. managers of large funds, of separate, actively man- aged emerging market portfolios. In 2005, as in the recent past, portfolio equity Figure 2.8 Bank credit for high-risk borrowers: investments remained concentrated in major rising rates but longer maturities, 2001–5 emerging markets. The Asia region continued to Margin over Libor (basis points) Maturity (years) account for the lion’s share (about 63 percent) of 350 12 total portfolio equity flows, with China, India, Average maturity and Thailand together making up about 94 per- 10 300 (line graph) cent of the region’s total. Notwithstanding the fact that the Chinese stock market performed poorly 8 250 over the last five years, China continues to attract 6 portfolio equity flows through initial public offer- ings (IPOs). In 2005, China accounted for about 200 4 31 percent of the total equity flows to all develop- ing countries and almost half of those to the Asia 150 2 region. Greater investor interest in Brazil and Mexico increased the shares of Latin America 100 0 slightly. Flows to Europe and Central Asia 2001 2002 2003 2004 2005 slumped to $2.3 billion from $4.2 billion the pre- Sources: Dealogic Loanware and World Bank staff calculations. vious year, due to outflows from the Czech Repub- lic and the Russian Federation. The volume of equity placements surged in 2004 (figure 2.8). Loan maturities normally shrink 2005, as stock markets in emerging markets out- as lending rates rise, suggesting that high-risk paced those elsewhere (box 2.2). Most of the port- countries may now be willing to pay higher costs folio equity investment in 2005 took place in return for longer maturities. through international equity placements, which Developing-country credit continued to im- were up about 60 percent over the same period in prove in 2005, as rating agency upgrades handily 2004. After a slow period in the first quarter, is- outpaced downgrades. Moreover, the pace of suance continued briskly throughout the year, on credit upgrades is accelerating. Some 46 upgrades the strength of an expanded investor base and at- occurred in 2005, in contrast to 31 in 2004. Some tractive valuations. Just 10 percent of the transac- countries enjoying upgrades are commodity ex- tions, including a few large IPOs, accounted for 64 Delivered by The World Bank e-library to: porters, (for example, Brazil, Mexico, theThe Russian World Bankpercent of the total volume. In 2005, IPOs ac- Federation, and Republica Bolivariana de counted for about 63 percent of all emerging mar- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 Venezuela). These economies paid down external ket equity transactions, up from 47 percent in 52 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S Box 2.2 Strong performance of emerging stock markets in 2005 E merging stock markets performed exceptionally well in 2005. With an increase of about 32 percent in the MSCI Emerging Market Index, these stock markets out- ily outpaced other asset classes, including both bonds and equities. Stellar performers during 2005 included Brazil (43.5 percent), India (40.2 percent), Mexico (38.6 per- performed most mature markets. However, stock prices cent), the Russian Federation (69.8 percent), and Turkey were volatile, because of rising concern about inflation (49.2 percent). Expectations of returns from emerging and the tightening of monetary policy in the United market equities in 2006 are subdued in the face of rela- States and Europe. In 2005, emerging market equity eas- tively high valuations. Performance of global equity markets, 2002–5 Total returns from global capital markets in 2005 Jan. 2004 = 100 Topix 200 MSCI EM equities Emerging Market MSCI Eurozone 160 GLOBAL MSCI Topix EMBIG 120 EURO gov. bond FTSE 100 S&P 500 S&P 500 80 ELMI+ US gov. bond US HY 40 Jan. July Jan. July Jan. July Jan. July Dec. 0 10 20 30 40 50 2002 2002 2003 2003 2004 2004 2005 2005 2005 % change in local currency Sources: Bloomberg and World Bank staff calculations. Sources: Bloomberg, JPMorgan Chase, and World Bank staff calculations. 2004. Asian countries accounted for a majority of ferings by companies in the oil and gas sector. In these transactions. China alone accounted for Sub-Saharan Africa, only South Africa had equity about 21 percent of global IPO activities in 2005 offerings, where shares of mining companies that and almost 61 percent of the total in emerging are world leaders in their sector were an attractive markets (figure 2.9). Many of these IPOs involved destination for foreign portfolio investment. sales of stakes in underperforming state-owned In recent years, major institutional investors banks and other financial institutions. Among the in the United States and elsewhere have gradually efforts was a jumbo IPO by China Construction increased their international stock holdings, in- Bank, which raised $9.2 billion. cluding stocks from emerging markets (table 2.5). Revival of interest in local equity placement The trend has accelerated since 2003, with inter- was evident in Latin America, where more than national markets generating higher adjusted re- $5.5 billion was raised on local equity markets in turns than the U.S. market. At the end of 2004, fi- 2005. Issuance volume, although still relatively nancial assets under institutional management low, contrasted markedly with the negligible activ- (pension, insurance, and mutual funds) totaled ity in the region’s equity markets over the past sev- $46 trillion,2 of which the United States accounted eral years. Issuance in emerging Europe was domi- for $20.7 trillion. Allocation to international eq- nated by the Russian Federation, which accounted uity ranged from a low of 13 percent in the United for about 64 percent of the regional total. Most States to 40 percent in the Netherlands. Because Delivered by The World Bank e-library to: equity issues in emerging Europe took the form ofThe World Bank States accounts for such a large share of the United depository receipts and IPOs issued by companies international financial assets, the recent increase in IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 in the communications sector, along with a few of- U.S. managers’ allocations to international markets, 53 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 2.9 IPO activities in emerging market clude robust global growth, increased corporate countries 2001–5 profits, favorable financing conditions, and higher $ billions No. of transactions stock market valuations, which have fueled cross- 35 140 border mergers and acquisitions. Factors specific to Others developing countries have also been at play: 30 China 120 No. of IPO 25 offerings 100 • Global economic growth has recently been much more favorable to the developing world, 20 80 bringing with it a commodity price boom and generally higher developing-country growth. 15 60 Rapid growth makes developing countries at- 10 40 tractive destinations for global FDI, particu- larly the market-seeking investments that have 5 20 become the largest share of global FDI flows 0 0 since the late 1990s. 2001 2002 2003 2004 2005 • Corporate profits have risen in developing Source: Dealogic Loanware and World Bank staff calculations. countries (UNCTAD 2005). In 2005, income generated from FDI in developing countries climbed to $120 billion from $80 billion in although small in percentage terms,3 represents a 2002. Approximately $45 billion of the 2005 major increase in flows into emerging market equi- total was reinvested. ties. The year also brought a large increase in retail • The investment climate in many developing investments in emerging markets through emerging countries, including low-income countries, market stock funds. The availability of exchange- has improved over the years (World Bank traded funds has made it much easier for private in- 2005). Many countries have revised their poli- dividuals to invest in emerging markets cies toward FDI to make them more favorable (UNCTAD 2004). After a slow down, privati- FDI grew through privatizations and zations and mergers and acquisitions (M&A) expansion of the European Union deals gained momentum in 2005, bringing in FDI flows to developing countries continued to large amounts of FDI. grow in 2005, reaching a record level of $237.5 billion, or about 2.8 percent of developing coun- The investment climate improved in many tries’ aggregate GDP (table 2.6). Much of the mo- developing countries mentum derives from the same factors that account A better investment climate in many developing for the strong recovery of FDI at the global level countries played a role in the recent rapid growth (which totaled $959.4 billion in 2005, up sharply of FDI. Many low- and middle-income countries from $666.5 billion in 2004).4 Those factors in- have taken steps, either unilaterally or in compli- ance with multilateral and regional agreements, to strengthen their foreign investment policies by eas- Table 2.5 Asset allocation of major international pension funds, 2004 ing sectoral restrictions and improving corporate Share of total governance (World Bank 2005; UNCTAD 2004). Domestic International Domestic International At the same time, better macroeconomic condi- Country Equity equity bonds bonds Cash Other tions, such as higher growth rates, increased open- Australia 31 22 17 5 6 19 ness to trade, lower external debt, and exchange Japan 29 16 26 11 11 7 rate stability made investments in developing coun- Netherlands 7 40 7 32 4 10 Sweden 21 16 29 26 2 6 tries less risky. Countries with a better investment Switzerland 13 14 34 10 8 21 climate managed to attract higher levels of FDI United Kingdom 39 28 23 1 2 7 flows as a percentage of their GDP (figure 2.10). United States 47 13 33 Delivered 1 by1The World5 Bank e-library to: The World Bank The key policy implications for countries at- tempting to attract FDI are to create a better in- IP : 192.86.100.36 Sources: International Financial Services, London, Fund Management, August 2005. Tue, 10 Mar 2009 16:55:33 54 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S Table 2.6 Net FDI flows to developing countries, 2000–5 $ billions 2000 2001 2002 2003 2004 2005e Total 168.8 176.9 160.3 161.6 211.5 237.5 East Asia & Pacific 44.3 48.5 57.2 59.8 64.6 65.3 Europe & Central Asia 30.2 32.7 34.9 35.9 62.4 75.6 Latin America & Caribbean 79.3 71.1 48.2 41.1 60.8 61.4 Middle East & North Africa 4.2 3.4 3.7 5.6 5.3 9.1 South Asia 4.4 6.1 6.7 5.7 7.2 8.4 Sub-Saharan Africa 6.5 15.0 9.5 13.6 11.3 17.6 Low-income countries 10.7 12.8 15.0 14.9 17.0 23 Middle-income countries 158.2 164.1 145.3 146.7 194.5 214.4 Global FDI Flows 1,388.4 807.8 721.0 623.8 666.5 959.4 Sources: World Bank, Global Development Finance, various years, and World Bank staff estimates for 2005. Note: Numbers may not add up due to rounding. e = estimate. vestment climate by (a) improving access to ade- Figure 2.10 Investment climate and FDI quate infrastructural and institutional facilities; Institutional Investor Ratings FDI to GDP ratio, 2003–4 (b) providing a stable, consistent, and transparent Less risky legal and regulatory framework and decreasing 50 3.0 red tape; and (c) engaging in international gover- 40 2.5 2.0 nance arrangements. More importantly, develop- 30 1.5 ing countries should identify and develop those 20 1.0 national competitive advantages that are likely to 10 0.5 be of particular interest to foreign investors. In this 0 0.0 Low-income Middle-income Low Average Good context, countries should promote local skills de- countries countries Investment climate, 2000–2 velopment and encourage private sector develop- 1995 2000 2005 ment in order to broaden the opportunities for en- trepreneurial activity. Countries also should Source: Institutional Investor Magazine, various years; Global Development Finance, strengthen their investment-promotion activities various years. Note: Investment climate (Institutional Investor Rating) is the average for the 2000–2 by establishing a broad-reaching agency that can period; FDI to GDP ratio is the weighted average for 2000–4 for 86 countries, excluding list and market investment opportunities as well major oil exporting countries. as provide information about doing business in the country.5 Countries should focus not only on policies to attract FDI, however, but also on the ence between FDI flows to the top 10 recipient policies that are necessary for FDI to generate a countries (2.7 percent of GDP) and other develop- positive development impact in the recipient ing countries (2.4 percent in other low-income and country (see chapter 5). 2.3 percent other middle-income countries) de- clined significantly over the years (figure 2.11). The concentration of FDI has declined in recent years Regional differences remain important Although the top 10 countries (China, the Russian Europe and Central Asia absorbed much of the Federation, Brazil, Mexico, the Czech Republic, increase in FDI in 2005. Investment in the region Poland, Chile, South Africa, India, and Malaysia) reached a record $76 billion in 2005, up from the accounted for almost 65 percent of FDI to devel- previous record of $62 billion in 2004. High com- oping countries in 2005, that concentration is con- modity prices encouraged significant increases in siderably less than the 75 percent share of the late FDI in the resource-rich countries of the region, 1990s. In addition, the share of low-income coun- notably the Russian Federation, Azerbaijan, and tries has increased steadily to almost 10 percent, Kazakhstan, while FDI flows to EU accession Delivered by The World Bank e-library to: mainly due to increases in resource-seeking FDI.The World Bank in the region also rose significantly. Sev- countries Relative to the size of the economies, the differ- eral of the countries in the first wave of the recent IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 55 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 2.11 The concentration of FDI, 1995–2005 to China showed their first-ever decline. Although FDI to GDP ratio (%) economic growth remains high and income from 4.5 Top 10 FDI increased, investors worried about declining 4.0 profit margins from increased competition 3.5 (IMF–World Bank Global Investor Survey 2005) 3.0 and overheating of the economy (A.T. Kearney Middle-income countries 2.5 2005). Reinvested earnings declined significantly 2.0 in 2004. FDI in services, particularly in the finan- 1.5 Low-income countries cial sector, is on the upswing, as China opens up to (excluding India) 1.0 meet the requirements of WTO membership (box 0.5 2.3). The country’s financial sector received more 0.0 than $13 billion in investment in 2005, as banks 1995 1997 1999 2001 2003 2005p (including banks from Chile and Brazil) positioned Sources: Global Development Finance, various years; World Develop- themselves by opening branches or representative ment Indicators, various years; World Bank staff estimates for 2005. Note: Top 10 countres include China, Russian Federation, Brazil, offices.7 In contrast to the situation in China, FDI Mexico, Czech Republic, Poland, Chile, South Africa, India and inflows to other Asian countries increased sharply, Malaysia. Low and Middle Income averages exclude these coun- tries. p = projected. with Indonesia receiving $2.3 billon, largely re- lated to the continuing privatization of state assets and acquisitions of private firms. Malaysia and EU expansion (Czech Republic, Hungary, and Thailand also received substantial flows. Poland) continued to receive high levels of invest- FDI in South Asia also grew in 2005. In India, ment due to buoyant corporate profits and sub- investment rose in industries such as cement, stantial reinvested earnings. Romania and Bul- sugar, plastics and rubber, and hotels. In Pakistan, garia, which are expected to join the European as in the countries of the Middle East and North Union in 2007, also received large amounts of in- Africa, privatization and resource-related FDI led vestment. In Latvia, and the Slovak Republic, FDI growth in FDI. Both the Arab Republic of Egypt levels were stabilized, mainly supported by rein- and Tunisia received significant levels of FDI in en- vested earnings. Progress in privatization of the ergy and energy services. FDI in Sub-Saharan telecom and financial sectors, along with early Africa increased significantly in 2005, mainly be- talks on EU accession, brought FDI flows to cause of two large acquisitions in South Africa.8 Turkey to an all-time high. The other countries in the region that continued to FDI in Latin America stabilized at $61.4 bil- receive high levels of FDI were resource-rich coun- lion in 2005. The continuing economic recovery in tries, notably Nigeria and Angola. the United States and resource-seeking investors were the principal forces behind the high level. A new wave of privatizations and cross-border The impact of improved competitiveness was dis- mergers and acquisitions is cresting cernible in the increase in investment in manufac- An important factor in the recovery of FDI from its turing, while FDI in services stalled (except in low point in 2002–3 has been the growing number Mexico’s financial sector). In Brazil, FDI in manu- of privatizations, mergers, and acquisitions in de- facturing increased, even as overall FDI decreased veloping countries (table 2.7). In the late 1990s, slightly because of political problems. Both Brazil FDI flows to developing countries were boosted by and Mexico were among the top developing-coun- such deals, particularly in Latin America and East- try recipients of FDI, absorbing $15 billion and ern Europe; similarly, the slowdown in activity $18 billion respectively. Colombia experienced since 2000 has been reflected in lower FDI flows. strong growth in FDI because of investments in Since 2004, however, several important privatiza- coal and the sale of a major beer company.6 tions have been completed, but their full effect on FDI in East Asia and the Pacific rose only FDI was not necessarily immediate because of the slightly in 2005, in contrast to more vigorous general lag between approval of the investments Delivered by The World Bank e-library to: growth in previous years. As expected, FDI The flows World Bankand actual implementation of the projects. IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 56 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S Box 2.3 Growing FDI in China’s banking sector S ince China joined the World Trade Organization in 2001, foreign banks have been positioning them- selves in China’s market, where restrictions on local- Despite the opportunities that come with such a large and untapped market, investing in the sector is risky. There remains some uncertainty about the financial health of currency transactions are expected to be removed by some banks, including high non-performing loans, and December 2006. Foreign banks can enter the market in credit allocation culture and standards. But foreign banks one of two ways: they may either invest in a domestic seem to be striving to replicate the success of Bank of bank and hold a minority share (less than 25 percent) America, which bought shares in China Construction Bank or open up fully owned branches. To gain immediate before its very successful public offering in October 2005 accesses to a large branch network, many foreign banks in the Hong Kong stock market. are increasing their holdings in domestic banks (see table Sources: “Bankable Prospects,” Business China (October 10, 2005); “Only below). They have invested an estimated $17 billion the Bravest of Bankers Boldly Go to China,” USA Today (January 19, since 2001. 2005). Chinese banks Date Foreign investors Investment (US$ billions) Stake % Bank of Communications Aug. 2004 HSBC $2.10 20 Bank of China Aug. 2005 Merrill Lynch, others $3.10 10 Bank of China Sept. 2005 Temasek (Singapore Gov. Fund) $3.10 10 Industrial and Commercial Bank of China Sept. 2005 Goldman Sachs, American Express, Allianz $3.00 10 China Construction Bank Sept. 2005 Bank of America $3.00 9 Huaxia bank Oct. 2005 Deutsche Bank $0.33 10 Bank of China Oct. 2005 UBS $0.50 — China Pacific Life Insurance Dec. 2005 Carlyle Group $0.41 25 Sources: JPMorgan Chase Securities (Asia Pacific); China Economic Review. Note: — = not available Table 2.7 Selected announced privatization and M&A deals in developing countries, 2005 Target (location) Sector Buyer (country) Value (US$ billions) Date NBR (Ukraine) P Banking Sberbank (Russia) $0.12 Jan-06 Texakabanka (Kazakhstan) P Banking Sberbank (Russia) $0.13 Jan-06 Turk Telekom (Turkey) P Telecom Saudi Oger (Saudi Arabia) $6.50 Jul-05 Telsim (Turkey) P Telecom Vodafone (UK) $4.50 Dec-05 BCR (Romania) P Banking Erste Bank (Austria) $4.20 Dec-05 Cesky Telecom (Czech Republic) P Telecom Telefonica (Spain) $3.60 Apr-05 PTCL (Pakistan) P Telecom Etisalat (UAE) $2.60 Jul-05 Mobitel (Bulgaria) P Telecom Austria Telekom $1.97 Jul-05 Turkcell (Turkey) P Telecom Alfa Telecom (Russia) $1.60 Dec-05 Disbank (Turkey) Banking Fortis (Belgium) $1.28 May-05 Aval Bank (Ukraine) Banking Raiffeisen International (Austria) $1.03 Oct-05 Varna and Rouse Thermal Power Plant (Bulgaria) P Energy RAO UES (Russia) $0.97 Dec-05 Al Furat (Syria) Oil CNPC (China) & ONCG (India) $0.57 Dec-05 Garanti Bank (Turkey) Banking GE Consumer Finance (U.S.) $0.25 Aug-05 Jubanka (Serbia) Banking Alpha Bank (Greece) $0.19 Jan-05 Albtelecom (Albania) P Telecom A consortium led by Turk Telekom $0.17 Jun-05 Telekom Montenegro P Telecom Matav (Hungary) $0.15 Mar-05 MISR Romaina Bank P Banking Blom Bank (Lebanon) $0.09 Dec-05 Podgoricka Banka (Montenegro) P Banking Société Générale (France) $0.02 Oct-05 Financial Times; Sources: Country Reports Economist Intelligence Unit; Delivered by Theother news World media. Bank e-library to: P = privatization deals. The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 57 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Box 2.4 Accession to the European Union and FDI T he recent enlargement of the European Union (EU) has had a salutary effect on FDI flows to Eastern Eu- rope. Seven developing countries (the Czech Republic, Es- to the liberalization of the economy can be expected to continue to attract FDI. The accession countries have access to EU Structural tonia, Hungary, Latvia, Lithuania, Poland, and the Slovak Funds intended for basic infrastructure development, Republic) have joined the European Union; two others human resources development, competitiveness and enter- (Bulgaria and Romania) are expected to join in 2007. prise development, rural development, and environmental Croatia and Turkey may join in the future. protection (Kalotay 2006). Use of such funds can be ex- EU membership requires structural changes in na- pected to bring significant improvements in the investment tional laws and regulations related to FDI. All member climate of these countries. Although implementation of countries are expected to adopt a body of EU law (the ac- structural changes is at a different stage in each accession quis communautaire). Doing so improves the business en- country, all are expected to comply eventually with EU vironment in accession countries, and thus their attractive- standards as highlighted above. ness to investors, but it may also raise the cost of doing The impact of accession on FDI inflows varies with business because of higher environmental and labor stan- the degree of implementation of the new policies. FDI dards. New EU members are also expected to amend their surged in Ireland, Portugal, and Spain following their ac- bilateral and multilateral treaties to comply with EU stan- cession, thanks to trade integration, whereas FDI in dards. Arrangements such as special zones and tax incen- Greece did not increase (left figure). Despite the adoption tives must be gradually eased, which may lead some multi- of EU standards and improved investment climate, Greece nationals to decrease their investments. lagged behind the other EU members even after accession. On the plus side, full membership in the European In newly acceding countries, particularly Romania, as customs union reduces the cost of trade with the rest of well as candidates (Croatia and Turkey), progress in priva- Europe, a significant advantage in terms of attracting in- tization has been providing opportunities for foreign in- vestors wishing to produce for the EU market. Adoption vestors. An example is the sale of the Romanian state of the euro will reduce exchange rate risk, though it may bank, the largest privatization deal in the banking sector also make the accession countries less cost-competitive. in 2005. In Turkey, recent privatizations raised the coun- Finally, in some of these countries, privatizations related try’s FDI to new heights in 2005 (right figure). FDI as share of GDP (%) FDI stock as a percentage of GDP 4.0 120 2000 3.5 100 2005 80 3.0 Portugal Ireland 60 2.5 Spain 40 2.0 20 1.5 0 a ze ary ov ep. . a a ia ia nd y ia EU ep ke ni i ni tv an at ar 1.0 la to ua R R g Greece r La ro lg Tu om Po un Es ch ak th Bu C H R Li Sl C 0.5 Source: World Bank Debtor Reporting System. 0 n–5 n–4 n–3 n–2 n–1 n n+1 n+2 n+3 n+4 n+5 Source: World Bank Debtor Reporting System. Note: Accession year (n) = 1973 for Ireland, 1981 for Greece, and 1986 for Portugal and Spain. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 58 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S The impact of privatizations on FDI was par- are the euro, credit default swap markets, and ticularly evident in many eastern European coun- local-currency bond markets. tries, particularly where upcoming or possible EU accession promises better investment climates, in- The euro’s role is growing vestment-related regulations and policies, and Since its introduction on January 1, 1999, the euro trade integration (box 2.4). However, even coun- has assumed an increasingly important interna- tries in the region that are not slated to join the tional role. It has emerged as a principal issuing European Union received notable levels of privati- currency in the global debt market, as a vehicle for zation-related FDI in 2005. foreign exchange transactions, and as an important As in the 1990s, most large privatization deals reserve currency for official holdings of foreign- occurred in banking or telecommunications. The exchange reserves. The elimination of exchange sale of BCR, a Romanian bank, was the largest risk within the Euro Area has created a wide Euro- privatization deal in the banking sector in 2005 pean market for euro-denominated securities, at- and the second-largest cross-border bank merger tracting both sovereign and private borrowers not in a developing country since the Mexican Ba- only from within the Euro Area but also from namex deal of 2001. other countries—among them emerging market economies such as Brazil, Colombia, China, Mex- ico, and Turkey. Today’s euro-denominated bond market rivals the dollar-based fixed-income mar- Structural changes in emerging kets in several respects, including size, depth, and market debt product range. As of June 30, 2005, outstanding E merging market debt markets are evolving. No longer are they dominated by the sort of dol- lar-denominated, high-yield sovereign debt typi- international bonds (debt securities marketed and sold outside a borrower’s own country) and notes issued in euros amounted to $6.2 trillion, or 45 fied by the Brady bonds of the 1980s. Today, the percent of outstanding debt obligations (table 2.8). emerging asset class includes a cluster of instru- The share of international dollar-denominated ments in both local and foreign currency that offer bonds and notes, meanwhile, has steadily declined— the capacity to tap dollar and euro investors alike from 49.4 percent in 1999 to 38.3 percent at the end and cater to the funding needs of both sovereign of June 2005. The popularity of the Japanese yen as and corporate borrowers. Active trading is occur- an issuing currency has dwindled; its share was only ring on the cash and derivatives sides of the mar- 3.6 percent in June 2005. ket. In this section, we take stock of three struc- Thus far, the major beneficiaries of the rise of tural changes that are making emerging debt the euro bond market have been the new countries markets a more diversified, robust, and liquid of the European Union. But although Poland, funding source for both sovereign and corporate Hungary, and the EU accession countries have borrowers in developing countries. Those forces been especially active in the euro-denominated Table 2.8 International bonds and notes outstanding, by currency, 1999–2005 $ billions 1999 2000 2001 2002 2003 2004 2005 (June) Euro 1,500.1 1,862.1 2,429.1 3,610.5 4,930.3 6,233.3 6,166.4 U.S. dollar 2,610.6 3,243.9 3,870.9 4,202.4 4,709.4 5,020.8 5,199.1 Yen 478.3 417.4 389.3 429.1 508.1 518.7 486.0 Pound sterling 402.3 448.4 503.6 621.6 829.6 1006.3 1019.4 Others 291.5 275.4 302.2 398.2 530.4 662.6 717.3 Total Issues 5,282.8 6,247.2 7,495.1 9,261.8 11,507.8 13,441.7 13,588.2 Euro as % of total 28.4 29.8 32.4 39.0 42.8 46.4 45.4 U.S. dollar as % of total 49.4 51.9 51.6 45.4 40.9 37.4 38.3 Yen as % of total 9.1 Delivered by The 5.2 World Bank 6.7 e-library to: 4.4 4.6 3.9 3.6 The World Bank IP : 192.86.100.36 Source: Bank for International Settlements Quarterly Review, December 2005, World Bank staff calculations. Tue, 10 Mar 2009 16:55:33 59 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 2.12 Euro-denominated international bond Emerging market issuers from China, Colom- issues, by region, 1999–2005 bia, Lebanon, Mexico, Philippines, Poland, Turkey, 30 and Ukraine have issued bonds in euros because of ECA MENA LAC SSA lower interest rates on euro-denominated bonds 25 EAP than on comparable U.S.-dollar bonds.9 Most of 20 the difference is explained by the fact that 10-year 15 euro interest rates have been lower than corre- sponding dollar rates. Spreads over the bench- 10 marks are about the same for comparable issues in 5 the two currencies. 0 For eastern European countries, the extent of 1999 2000 2001 2002 2003 2004 2005 present and future trade with Euro Area countries, and the prospective adoption of the euro by the Source: Bank for International Settlements. accession countries, has undoubtedly played a part in the choice to issue debt in euros. Poland, for ex- market, other developing countries, too, have ample, is part of the Euro Area, and its future as- found it a viable funding alternative. Among the sets will be denominated in euros; it trades already emerging market entities that have issued sizable primarily with other EU countries. Decisions to euro-denominated bonds are Mexico’s PEMEX, issue in euros also depend on the terms of issuance the Korea Development Bank, and the govern- and the liquidity of the market. Underwriting fees ments of China and the Republica Bolivariana de are roughly comparable for dollar and euro issues Venezuela. In 2005, sovereign and corporate bor- and may in fact be lower for euro issues.10 Market rowers in emerging markets issued $33.7 billion in liquidity for comparably sized issues is also simi- euro-denominated bonds in the international mar- lar. These factors all help to explain the dramatic ket, up from $21.7 billion in 2004 (figure 2.12). growth in international debt denominated in euros Much of the growth came from Argentina’s is- since 1999. suance of $9.9 billion in bonds as a part of its debt workout. No euro-denominated issues came from Credit default swap markets have grown Asia in 2005. substantially Several factors account for the increase in As anticipated in the 2003 edition of Global Devel- euro issues. The decision to issue bonds in foreign- opment Finance, trading in credit default swaps currency markets is shaped chiefly by considera- (CDSs), and especially Emerging Market Credit tions of risk and cost, but also by a desire to diver- sify funding sources (for example, to match the Figure 2.13 Yields on U.S. and German 10-year issuer’s trade patterns). Most prudent borrowers government bonds, 1999–2005 wish to match the currency denomination of their Yield bonds to their assets and cash flow over the dura- 8 tion of the bonds. (The risk of a mismatch may also 7 be covered using an appropriate derivative, such as 6 a currency swap.) Borrowing costs are influenced 5 United States by regulatory requirements (related, for example, to the withholding of tax from payments to investors) 4 and market liquidity. Otherwise, the quantity of 3 Germany bond issues in a given currency is limited only by 2 the funding requirements of borrowers, the prefer- 1 ences of institutional investors, and interest rate 0 differentials or prospective exchange rate trends. ly 9 n. 9 ly 0 n. 0 ly 1 n. 1 ly 2 n. 2 ly 3 n. 3 ly 4 n. 4 ly 5 05 Ju 199 Ja 99 Ju 00 Ja 00 0 Ja 00 Ju 00 Ja 00 0 Ja 00 Ju 00 Ja 00 Ju 00 The cost of issuing bonds in euros is determined by 20 20 20 1 2 2 2 2 2 2 2 2 2 n. Delivered by The World Bank e-library to: Ju Ju Ja the cost of the benchmark (10-year Bunds) Theplus a Bank World spread (figure 2.13) over the benchmark. IP : 192.86.100.36 Source: Bloomberg. Tue, 10 Mar 2009 16:55:33 60 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S Figure 2.14 Comparison of euro-denominated and U.S. dollar-denominated emerging market sovereign bond issues Colombia Poland Yield Yield 18 7 16 6 14 5 12 10 4 8 3 6 2 4 2 1 0 0 4 02 03 4 03 04 05 5 03 04 5 5 01 1 3 02 00 00 00 00 0 00 00 20 20 20 20 20 20 20 20 20 20 2 2 .2 .2 .2 2 b. ly n. b. g. p. r. g. g. p. ay n. ne ct ar ov Ap Ju Fe Ja Fe Au Se Au Au Se Ja O M Ju M N Yield Mexico Yield China 8 6 7 5 6 4 5 4 3 3 2 2 1 1 0 0 4 04 05 5 3 4 04 5 3 3 04 4 05 5 5 4 00 00 00 00 00 0 00 0 0 00 00 20 20 20 20 20 20 20 20 .2 .2 .2 2 2 .2 .2 .2 ay ly b. p. r. ne . ly . ly ec ct ct ct ar ar ov ov Ap Ju Fe Se Ju Ju M O O O Ju M M D N N Philippines Ukraine Yield Yield 7 25 6 20 5 15 4 3 10 2 5 1 0 0 02 3 3 1 01 4 05 04 2 4 04 5 05 5 04 04 00 00 00 00 00 00 00 00 20 20 20 20 20 20 20 20 .2 .2 .2 .2 2 2 .2 .2 p. ly e ay b. ar ct n ov n. ay ly g. . n ar ct ec Ju Se Ja Fe O Ju M Ju Ja M Au N M O M D Turkey Lebanon Yield Yield 18 12 16 10 14 12 8 10 6 8 6 4 4 2 2 0 0 03 02 0 01 03 4 1 4 05 3 4 05 03 03 04 04 05 5 00 00 00 00 00 00 00 20 20 20 20 20 20 20 20 20 20 20 .2 .2 .2 .2 .2 .2 .2 n. y ay g. ay r. . y . ly g. ct ar ec ct ov ov ar ar ec l l Ap Ju Ju Ju Ja Au Au O M O M M M M D N N D US$ Euro Source: Bloomberg. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 61 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Default Swaps (EMCDSs), has grown substantially Figure 2.15 The global credit derivatives market in over the past three years, extending beyond the Re- notional terms, 2001–5 public of Korea, Mexico, and the Russian Federa- $ trillions tion to several new countries (Bulgaria, Peru, and 14 the Philippines). A CDS contract, like an insurance 12 contract, provides the buyer some protection against a specific risk, namely the risk of default. 10 As a derivative, CDSs, the most popular type of credit derivatives, make it possible to trade credit 8 risks separately from the underlying bonds or loans (box 2.5). They can help diversify risks in financial 6 markets by allowing financial institutions to hedge 4 risks embedded in their loan portfolio by transfer- ring credit risks to other market participants, such 2 as insurance companies and hedge funds. CDSs also enable institutional investors to take a position 0 2001 2002 2003 2004 2005* on a given credit without acquiring underlying as- sets in the cash market. Source: International Swaps and Derivatives Association Market Survey, 1987–present. The growth of the global credit derivatives * = as of end-June 2005. market since the early 1990s represents a major story of financial innovation, comparable, in many respects, to the development of the interest rate de- Malaysia, Mexico, the Philippines, the Russian rivatives markets developed to manage financial Federation, Taiwan (China), Turkey, and Uruguay. risk in the 1980s. At the end of June 2005, the Quotes are now available on debt issued in more market had a total notional amount outstanding than 29 countries. Dealer banks estimate that of around $12 trillion, representing an increase of trading volumes in EMCDSs now rival those in almost 48 percent from $8.42 trillion at the end of emerging cash bonds. For some countries, such as 2004 (figure 2.15). Hungary and Lithuania, the amount of outstand- The CDS market is divided into various sectors ing CDSs dwarfs the amount of outstanding cash defined by their underlying credit: corporates, bonds by a factor of 10. banks, sovereigns, and emerging market sovereigns. The growth of the EMCDS market has coin- A CDS may be based on a single credit or several. cided with the sharp increase in emerging market So-called single-name CDSs account for 60 percent financing over the 2003–5 period and has been of the market in credit derivatives. Their outstand- driven largely by the same forces. It has also been ing notional value was approximately $7.3 trillion aided by standardization of documentation and at the end of June 2005 (BIS 2005). the development of CDS indices and index-related Emerging market credit default swaps products that improve liquidity and price trans- (EMCDS) have grown with the global expansion parency (box 2.5). In 1999 and 2003, the Interna- of CDS markets, although at a slower pace. But tional Swaps and Derivatives Association pub- with a notional outstanding value of $350 billion, lished standard CDS documentation that appears the EMCDS market is now larger than the cash to provide a robust legal framework for the in- segment of the EMBI Global (estimated to be struments. Although the CDS market has begun around $250 billion). EMCDSs currently cover a to mature, it has not yet been subjected to major broad range of sovereign credits and are actively stress testing. traded. In 2003, annual trading volumes in EM- Investor demand. The market is presently dom- CDSs were estimated at almost $200 billion, ap- inated by institutional investors seeking to invest in proximately 5 percent of total trading in emerging emerging markets by selling protection in the CDS market credit (Emerging Market Traders Associa- market as an alternative to purchasing cash bonds. Delivered by The World Bank e-library to: tion 2003). In the same year, three-quarters Theof the BankCDSs are not subject to special features that may af- World volume of transactions concerned 10 countries: fect the yield of a particular bond, and the standard- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 Brazil, Hong Kong (China), Republic of Korea, ization of CDS contracts makes it easier to compare 62 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S Box 2.5 Credit default swaps A credit default swap (CDS) is a derivative contract transacted using standard documentation developed by the International Swaps and Derivatives Association. In forms include tradable indices, options, first-to-default or tranched basket products, cash collateralized debt obligations (CDOs), and synthetic CDOs. There are two a contract, one party (the protection buyer) pays a peri- families of tradable CDS indices: the Dow Jones CDX odic fee to another party (the protection seller) in return indices for North America and the emerging markets, for a promise of compensation in the event of default (or and the Dow Jones iTraxx for Europe, Japan, and Asia. other adverse credit event) by a specified firm or sover- The first comprises equally weighted CDSs on 125 refer- eign, known as the “reference entity,” which is not a party ence entities. to the CDS. The CDS transfers the credit risk of that en- A CDS transaction depends on a clearly defined tity from one party to another. Corporate bond investors credit event and on valuation methodology. The market generally buy CDSs to insure against default by the issuer generally uses three credit events (failure to pay, restruc- of the bond, but these flexible instruments can be used in turing, and bankruptcy) as triggers for contractual pro- many ways to customize exposure to corporate credit. tection payments. Market practitioners are converging CDSs now exist for more than 1,500 “reference names” in their views on modeling and valuing single-name in every bond category. Liquidity is provided by the market CDSs. Pricing techniques currently in use are derived makers, which include commercial banks, insurance compa- from reduced-form models that apply to defaultable nies, asset managers, and, more recently in a significant bonds, as presented in Jarrow and Turnbull (1995) and manner, hedge funds. Standard trading sizes range from Duffie (1999). $10–20 million (notional value) for investment-grade credits There have been disputes in the past over whether a and $2-5 million for high yield. The most liquid CDS con- debt restructuring was to be considered a default. Accord- tracts carry a maturity of five years. ing to definitions provided by ISDA in 2003, a restructur- The single-name CDS applies to a single entity and ing is deemed a default if the obligations become less fa- is the most common form of this instrument. Other vorable to the holders. credit risk across countries. EMCDSs are also more loans hedged with purchases of CDSs to be fully liquid than emerging market cash bonds. And, since backed by capital reserves, thus freeing capital for a large segment of emerging market investors tend other uses. Institutional investors like the fact that to buy and hold, investors wishing to enter the mar- CDSs enable them to take a position on an opera- ket may find it difficult to invest in a specific coun- tion without subjecting themselves to the regula- try’s bonds. In the context of buoyant demand, in- tory restrictions that would govern a cash invest- vestors can establish a position more quickly by ment in the underlying credit. The key sellers buying EMCDSs than by going through the under- include most institutional investors such as insur- lying cash markets. Furthermore, EMCDSs are not ance companies, monoline insurers (financial guar- subject to withholding or capital gains taxes in the antee companies), hedge funds, and mutual funds. United States. In sum, for actively traded issues, Liquidity. The top 10 dealers, all large invest- EMCDSs tend to enjoy a status similar to that of ment banks, account for about 70 percent of CDS emerging economies’ benchmark bonds, with a sales (Fitch Ratings 2004). Trading in the EMCDS yield curve for maturities up to 10 years. EMCDSs market is influenced by liquidity in the repo mar- also provide investors with a slightly higher yield kets for the underlying bonds. The market practice than bonds. is for dealers to intermediate in a two-way market Market participants. The chief buyers of pro- without taking a position and without the need to tection in CDS markets are major international rely on the cash market to hedge themselves. Ad- commercial banks, hedge funds, and other institu- vances in credit-risk management have enabled tional investors seeking to eliminate credit risk dealers to take selected positions and hedge their Delivered by The World Bank e-library to: from their portfolios (figure 2.16). CommercialThe World Bank on a portfolio basis, relying heavily on position banks are attracted by the fact that banking regu- correlations between classes of emerging market is- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 lators in most developed countries do not require suers. Although there is no direct relation between 63 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 2.16 Credit derivative participants, 2004 the same maturity.13 It is not clear whether the Sellers Buyers CDS price or the cash price of the underlying bonds is the leading price. This will vary depend- 11% 9% ing on the market context and the difference 15% 38% 16% among the participants in the markets. When 51% spreads follow the long-term spread-narrowing 7% trend, traditional investors in emerging bonds will 20% set the price. If new information emerges that jus- 16% 16% tifies a reappraisal by the market, CDS premiums may be adjusted much more rapidly than the bond Banks spread, resulting in a sudden, if temporary, widen- Securities firms ing of the basis. The explanation offered here is Insurance companies that hedge funds will react—and perhaps overre- Hedge funds act—more promptly to news than will traditional Other cash investors. With the broadening of the market Source: British Bankers Association 2003/2004, Credit Derivatives and the increasing presence of hedge funds and Survey. banks’ proprietary trading desks, the bias is to- ward active trading, which should improve price discovery in the CDS market. prices in cash markets and the prices of CDSs, the The growing EMCDS market, while imma- underlying bond prices provide essential references ture, has the potential to benefit emerging for the determination of EMCDS premiums. economies. EMCDSs are very liquid and more Liquidity in EMCDS markets is driven by the available than emerging market cash bonds, most large and growing number of participants—hedge of which are held until maturity. Many partici- funds in particular. Liquidity has been improving pants with strong views on emerging names, in- over the past two years for CDSs based on issues cluding hedge funds and banks’ proprietary desks, in a broad range of countries. have joined the EMCDS market so as to engage in There are now 29 names in the liquid active value trading in credit-risk premiums. Mar- EM.CDX Diversified CDS index, a good indica- ket data show that CDS spreads react more tion of the number of names that are particularly promptly to market developments than do corre- easy to trade. The bid–ask spread is typically sponding cash market spreads (and may even around 10 basis points, with a transaction size in overreact to adverse news). On balance, that alert- the range of $10 to $20 million. For liquid names ness means greater efficiency in credit pricing and such as the Republic of Korea, Mexico, and the stronger market discipline—in other words, a re- Russian Federation, the spread narrows to 5 basis duction in the asymmetry of information between points; it can reach 20 to 30 basis points for less lender and borrower, something from which traded names such as Chile, Morocco, and the emerging market finance can benefit. Philippines. EMCDS markets are highly liquid and have Price discovery. EMCDS and bond prices tend shown strong resilience to idiosyncratic shocks, to move in tandem, although they can deviate for such as Argentina’s default. However, despite con- short periods (figure 2.17).11 The default swap siderable improvement in transparency under the basis is the difference between default swap auspices of ISDA, transparency in CDSs still lags spreads and bonds’ asset swap spreads (spreads behind emerging bond markets where, similarly relative to the Libor).12 There are several funda- trading takes place only in the private, over-the- mental reasons why the default swap basis is nor- counter market. The market has expanded to in- mally positive. The most compelling is the tradi- clude new names, such as Peru, the Philippines, tional principle of “absence of arbitrage Slovakia. It is reasonable to expect corporate opportunity.” Were the basis to become negative, names to join as well, as private entities in emerg- Delivered by The World Bank e-library to: it would offer a risk-free gain to anyone investing The World Banking market economies tap increasingly global debt in a country’s bond while buying protection for markets. The great concentration of the market in IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 64 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S Figure 2.17 Five-year CDS and ASW spreads for selected countries, 2002–5 5-yr CDS & ASW Spreads for Mexico 5-yr CDS & ASW Spreads for Russia Basis points Basis points 450 700 400 CDS spreads CDS spreads 600 350 500 300 250 400 200 300 150 200 100 ASW spreads 100 ASW spreads 50 0 0 4 ly 2 2 4 r. 2 02 4 t. 2 02 4 r. 2 03 4 t. 2 3 4 r. 2 04 4 t. 2 4 4 r. 2 05 ct 05 4 r. 2 02 4 ly 2 2 4 t. 2 02 4 r. 2 03 4 t. 2 03 4 r. 2 04 4 t. 2 4 4 r. 2 05 ct 05 4 n. 2 02 4 ly 2 3 4 n. 03 4 ly 2 4 4 n. 2 04 4 ly 2 5 5 4 n. 2 02 4 ly 2 3 4 n. 03 4 ly 2 4 4 n. 2 04 4 ly 2 5 5 0 c 0 c 0 0 c 0 0 0 0 00 0 0 0 00 Ju 0 Ap 20 O 0 Ap 0 O 0 Ap 20 O 0 Ap 0 O 0 Ap 20 Ju 0 O 0 Ap 0 O 0 Ap 20 O 0 Ap 0 O 0 Ja 0 Ju 0 Ja 0 Ju 0 Ja 0 Ju 0 Ja 0 Ju 0 Ja 0 Ju 0 Ja 0 Ju 0 .2 .2 4 n. 4 n. c c c Ja Ja 4 4 5-yr CDS & ASW Spreads for Turkey 5-yr CDS & ASW Spreads for Philippines Basis points Basis points 1600 700 CDS spreads CDS spreads 1400 600 1200 500 1000 400 800 300 600 200 ASW spreads 400 ASW spreads 200 100 0 0 4 ly 2 2 4 r. 2 02 4 t. 2 02 4 r. 2 03 4 t. 2 3 4 r. 2 04 4 t. 2 4 4 r. 2 05 ct 05 4 r. 2 02 4 ly 2 2 4 t. 2 02 4 r. 2 03 4 t. 2 03 4 r. 2 04 4 t. 2 4 4 r. 2 05 ct 05 4 n. 2 02 4 ly 2 3 4 n. 03 4 ly 2 4 4 n. 2 04 4 ly 2 5 5 4 n. 2 02 4 ly 2 3 4 n. 03 4 ly 2 4 4 n. 2 04 4 ly 2 5 5 0 c 0 c 0 0 c 0 0 0 0 00 0 0 0 00 Ju 0 Ap 20 O 0 Ap 0 O 0 Ap 20 O 0 Ap 0 O 0 Ap 20 Ju 0 O 0 Ap 0 O 0 Ap 20 O 0 Ap 0 O 0 Ja 0 Ju 0 Ja 0 Ju 0 Ja 0 Ju 0 Ja 0 Ju 0 Ja 0 Ju 0 Ja 0 Ju 0 .2 .2 4 n. 4 n. c c c Ja Ja 4 4 Sources: Bloomberg and World Bank staff calculations. Note: ASW = asset swap; CDS = credit default swap. the hands of a small number of dealers poses a economies are still relatively small, accounting for risk, however, that an adverse credit event in a just 7.9 percent of global domestic debt market as major financial center would have potentially seri- of September 30, 2005. Local currency bond mar- ous repercussions on CDS market liquidity. kets are concentrated in eight countries (Brazil, China, India, the Republic of Korea, Malaysia, Local-currency bond markets provide Mexico, Turkey, and South Africa) that together important new sources of capital make up three-quarters of the entire market. The rapid development of local-currency bond Robust domestic bond markets enable mone- markets in emerging market economies signifies tary authorities to conduct monetary policy governments’ successful responses to the string of through open-market operations. It is widely un- financial crises of the 1990s. Local-currency bond derstood that well-developed capital markets en- markets, now the fastest growing segment of hance financial stability by diversifying both the emerging market debt, are in many cases helping avenues for investing savings and the sources of to correct mismatches of currencies and maturities funding for investment activities beyond the bank- in the countries affected, thereby contributing to ing sector. A vibrant bond market, supported by Delivered by The World Bank e-library to: greater financial stability. From a global perspec-The World well-functioning Bank and well-regulated derivative tive, the local-currency bond markets in emerging markets, enables market participants to better IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 65 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 manage their financial risks through swaps and fu- local-currency debt in emerging markets, followed tures and attract foreign investors. Furthermore, by Latin America (24.3 percent), Eastern Europe domestic debt instruments with long duration are (12.2 percent), South Asia (9.1 percent), and also ideally suited for infrastructure projects, espe- Africa (2.8 percent). cially those conducted by subsovereign borrowers Local currency bond markets in developing earning revenues in local currencies. countries are diverse in their size, issuers, liquidity, Driven largely by domestic institutional and supporting infrastructure, and degree of openness individual investors, local-currency debt markets to foreign investors. Ten of the fifteen largest have grown rapidly, moving from an aggregate local-currency bond markets in the world (mea- outstanding level of $1.3 trillion at the end of sured as a percentage of GDP) in 2004 were in 1997 to $3.5 trillion in September 2005 (figure emerging markets (figure 2.19). The three largest 2.18). The countries of East Asia have led the markets (China, India, and Mexico), while small way—the region accounts for 51.7 percent of total relative to GDP (below 35 percent), have substan- tial growth potential in light of recent reforms un- dertaken by these countries. Figure 2.18 Trends in domestic debt securities in Governments are the largest issuers in emerg- emerging markets, by region, 1997–2005 ing local-currency bond markets, accounting for 65 $ billions percent of local-currency bond markets in Septem- 3,500 EAP ECA ber 2005. Governments are followed by financial 3,000 SA SSA institutions (25 percent) and corporations (10 per- 2,500 LAC cent). Relative to the United States—the world’s 2,000 most diversified local bond market—bond markets in emerging economies are still highly concentrated 1,500 in government bonds (figure 2.20). The challenge 1,000 for emerging market countries is to further diver- 500 sify their markets by building up other segments. 0 The bond markets in East Asia grew rapidly 1997 1999 2001 2003 2005 (9/05) from $400 billion in 1997 to $1.6 trillion by Sep- Sources: Bank for International Settlements data and World Bank tember 30, 2005. Since 1997, governments in East staff calculations. Asia have issued large amounts of local-currency bonds to restructure the banking system and re- vive the corporate sector. This has helped establish Figure 2.19 The size of the domestic bond market in selected risk-free interest rate benchmarks that enabled the countries corporate sector, seeking to restructure its balance Outstanding value as % of GDP sheets, to issue bonds in the local market. Bond- (December 2004) 200 market development in East Asia gained further 192% momentum in December 2002 with the launching 180 163% of the Asian Bond Market Initiatives (ABMI) by 160 the ASEAN+3 group. 140 Corporate bond markets have been more dif- 120 ficult to establish than government bond markets 100 94% 77% 82% 84% in emerging markets because of the small issue 80 61% 62% size, lack of a yield curve, difficulties with proper 60 56% 41% 49% 49% disclosure of accounting information, and general 40 32% 35% 26% weakness in corporate governance. However, sev- 20 eral countries, including Chile, the Republic of 0 Korea, and Malaysia have been able to build rela- e ey a a na a a a n o K y S nd il or re ad di ric si pa an ic tively large corporate bond markets over the past az U U rk hi ap ay In la ex Ko Af an Ja m Br Tu C ai al ng M er Delivered by The World Bank e-library to: C S. S. Th M The World Bankdecade. In the Republic of Korea, the stimulus for G Si developing a functioning corporate bond market IP : 192.86.100.36 Sources: Bank for International Settlements data and World Bank staff calculations. Tue, 10 Mar 2009 16:55:33 66 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S Figure 2.20 Bond market profile in selected countries, September 2005 Turkey Malaysia Korea, Rep. of Corporate Gov't 24% Gov't Corporate 43% 33% 39% FIS FIS 18% 43% Gov't 100% China US India Corporate FIS Corporate Corporate Treasury 2% 2% 2% 20% 16% Gov't Federal Gov't 58% agency 96% FIS Money 10% 40% market 13% Municipal 9% Asset- backed 8% Mortgage- related 24% Mexico Brazil South Africa Corporate Corporate 1% Corporate FIS 10% FIS 3% 21% 14% Gov't Gov't Gov't 87% 78% FIS 74% 12% Sources: Bank for International Settlements Quarterly Review, December 2005, Bond Market Association, and World Bank staff calculations. Note: FIS = financial institutions. came in the aftermath of the 1997–8 crisis. Before The investor base widens for local-currency 1997, all corporate bonds in Korea had been guar- bonds anteed by commercial banks, which masked the Foreign investors. Until recently, the domestic differential credit risk of corporate bonds. bond markets of major emerging markets were The development of municipal bond markets largely closed to foreign investors. The obstacles is likely to become more important, given the to investment took many forms—administrative, growing role of subnational bodies in financing in- regulatory, fiscal, infrastructural, and informa- Delivered by The World Bank e-library to: frastructure projects, which have revenues and ex-The World tional. Since the East Asian crisis of 1997, how- Bank penses in local currencies. ever, these markets have become much more open, IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 67 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Box 2.6 The role of multilateral development banks in developing local-currency bond markets M ultilateral development banks (MDBs) meet part of their general funding requirements by issuing bonds denominated in the currencies of emerging mar- active issuers. The Asian Development Bank (ADB), which recently became active, has issued bonds denomi- nated in Indian rupees, Malaysian ringgits, Chinese kets. Such issues by MDBs can be standard setters in yuan, Philippines pesos, and Thai baht. Overall, most of local-currency bond markets. Although they are likely the local-currency borrowing by the MDBs occurred in to be small relative to the size of the domestic bond Hong Kong dollars, Taiwanese dollars, South African market, they can play a catalytic role by removing the rand, Turkish lira, and Polish zlotys. policy and regulatory impediments to foreign invest- Successful bond issuance by MDBs requires several ment and accelerating the development of necessary supporting policies as well as market infrastructure. market infrastructure. They can also help create a long- These include: (i) the existence of a clearly defined and term benchmark, which in turn may facilitate issuance sound regulatory framework; (ii) a disclosure-based regu- of local-currency bonds by corporations. latory system; (iii) an efficient clearing and settlement sys- During 2000–5, the total raised by MDBs in 24 tem; and (iv) the existence of an investor base, particu- markets through 534 bond issues was $21.4 billion (see larly institutional investors such as pension funds and figure and table). MDB issuance gained momentum in insurance companies. Success in local markets also re- 2005, with 121 issues totalling $5.2 billion. The largest quires a nondiscriminatory tax structure and exemption issuer was the European Investment Bank, which ac- from exchange controls. The experience of Malaysia, and counted for $10 billion, or 47 percent of total issuance, Mexico in facilitating issuance of bonds by MDBs should followed by the World Bank ($2.7 billion). The Inter- be of interest to other emerging economies. American Development Bank (IDB) and European Bank for Reconstruction and Development (EBRD) were also Source: Dealogic Bondware. Local-currency bond issuance by multilateral Bond issuance in non-G-10 currencies by supranationals, development banks, 2000–5 January 2000–October 2005 $ millions No. of issues No. Amount Issuers of Issues ($ millions) Percent 6,000 140 Amount 121 European Investment Bank—EIB 256 10,054 47.0 No. issues 120 5,000 World Bank 91 2,722 12.7 Inter-American Development Bank—IDB 49 2,257 10.5 95 97 100 4,000 European Bank for Reconstruction 85 & Development—EBRD 50 1,628 7.6 78 80 International Finance Corp.—IFC 19 1,550 7.2 3,000 Nordic Investment Bank 38 1,490 7.0 60 Asian Development Bank 9 808 3.8 58 2,000 Council of Europe Development Bank 4 348 1.6 40 African Development Bank—AfDB 4 221 1.0 1,000 Central American Bank for Economic 20 Integration—CABEI 9 164 0.8 Eurofima 5 152 0.7 0 0 Total 534 21,395 100 2000 2001 2002 2003 2004 2005 Source: Dealogic Bondware. Source: World Bank staff estimates based on Dealogic Bondware. especially in East Asia, where many impediments ceiling of $10 billion. Several multilateral devel- to foreign investment have been removed opment banks played key roles in removing ob- (Takeuchi 2005). The only major markets in Asia stacles to foreign investment in developing mar- Delivered by The World Bank e-library to: that still limit access are China and India,The where World Bankkets by issuing bonds in the currencies of China, fixed-income investments are allowed IP only by Thailand, Malaysia, Mexico, and the Philippines : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 qualified foreign institutional investors up to a (box 2.6). 68 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S Gradual but steady liberalization of capital Figure 2.21 Performance of local-currency bonds accounts in several developing countries has led (ELMI+) against major indexes, 2002–5 to a general increase in investment interest by for- Percent eign investors (see chapter five). In the past, insti- 60 tutional investors in developed countries, espe- 50 cially the United States and United Kingdom, did 40 not view emerging market equity or bonds, re- 30 gardless of their denomination, as a separate asset 20 class. Instead, they were considered as a small 10 component of broad indexes such as Morgan 0 Stanley’s All Country World Index. Most institu- –10 tions would allocate a small amount of their in- –20 vestments to emerging market equity. In the case –30 of emerging market debt, the allocation was usu- 2002 2003 2004 2004 ally made as a part of the Lehman Aggregate Plus EMBIG (US$) ELMI+ (Local currency) Index or the High Yield Index. However, returns ELMI+ (US$) JPM-HY (JULI) S&P 500 EM equities (MSCI from U.S.-dollar-denominated emerging market UST (1-yr constant EM Free, US$) debt were attractive, and some investors have maturity) been willing to assume the associated risks to ob- tain attractive risk adjusted returns. Sources: JPMorgan Chase, Datastream, Bloomberg. Local-currency bonds were rarely considered by institutional investors, since they involved high currency convertibility risk, on top of the interest Figure 2.22 Returns vs. volatility of selected bond rate and credit risks associated with fixed-income indexes, 2000–5 investments. However, efforts by several countries Return (%) to build their domestic bond markets have begun to 30 bear fruit. Their recent performance, as well as the 25 potential for currency appreciation in several mar- kets, is drawing the attention of growing numbers 20 of fund managers. Investments by foreign institu- EMBIG (US$) 15 tional investors in local-currency bond markets ELMI+ (US$) have been facilitated by the introduction of several 10 ELMI+ (local JPM-HY (JULI) local-currency bond indexes such as JPMorgan currency) 5 S&P 500 Chase’s Emerging Market Local Currency Index UST (3mo) (ELMI) and the Lehman Global Aggregate Index, 0 which includes a small percentage of emerging mar- 0 5 10 15 20 25 Risk (%) ket bonds. During 2000–5, the JPMorgan Chase ELMI+ (Local Currency) index generated an annual Sources: JPMorgan Chase, Datastream, Bloomberg. average return of 9.9 percent, well above the aver- age return of 1.91 percent on the U.S. Treasury’s one-year, constant maturity bills (figure 2.21). been relatively modest in comparison with the size Investment in U.S. dollar–denominated debt, of these markets. Flows are reported to be higher in as measured by the EMBI Global index, outper- 2005, but no segregated information is available. formed the ELMI+ (Local Currency), with an av- In contrast to the East Asian approach of erage annual return of 15.31 percent from 2000 to opening domestic bond markets to foreign in- 2005. However, the volatility of the local-currency vestors, the major countries of Latin America and bond was less than that of the EMBI Global dur- the Russian Federation have taken a different ap- ing the same period (figure 2.22). proach, issuing bonds denominated in local cur- Delivered by The World Bank e-library to: Although data are limited, it appears from theThe World rency in the international markets (Tovar 2005). Bank IMF’s 2003 consolidated portfolio survey that for- In November 2004, the Colombian government IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 eign flows to local-currency bond markets have raised the equivalent of $375 million by issuing a 69 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 six-year bond. This was followed in January 2005 vested $3 billion in the Asian bond markets by a second issue for $124 million. In September through two funds. Although they represent a rela- 2005, Brazil issued global bonds totaling 3.4 bil- tively small share of official reserves, these invest- lion reals ($1.5 billion) with a maturity of 10 years ments are expected to play a catalytic role in the and a 12.5 percent coupon. By issuing local-cur- development of domestic bond markets. The larger rency bonds in international markets, these Latin of the funds, the Asian Bond Fund (ABF2), was American countries have tried to tap international launched in December 2004 to invest $2 billion in investors while changing the currency mix of their local currency bonds. ABF2 has two components: a debt portfolio. Pan-Asian Bond Index Fund (PAIF) and a Fund of The barriers facing foreign investors seeking Bond Funds (FoBF). The PAIF is a single bond fund to enter the domestic bond markets of countries investing in sovereign and quasi-sovereign local- (such as Brazil and Colombia) that have opted to currency-denominated bonds issued in the eight issue local-currency bonds in international mar- EMEAP markets. The FoBF is a two-layered struc- kets include registration requirements and with- ture with a parent fund investing in eight subfunds, holding taxes. In February 2006, Brazil took sev- each of which will invest in sovereign and quasi- eral steps to increase the attractiveness of its sovereign local-currency-denominated bonds is- domestic bond markets to foreign investors. These sued in the EMEAP economies. The ABF2 has included exempting investors from transaction started to invest in domestic bond markets, helping taxes and withholding tax on interest income, and in the process to create eight local-currency bond permitting tax-free migration between equities and market indices. fixed-income instruments. Domestic investors. Domestic investors, both Local-currency bond markets present new institutional and individual, thus far have been the opportunities and new challenges major investors in local-currency bond markets, Bringing local currency bond markets in emerging especially government bonds. Bond investments economies up to the standards of markets in devel- have become an acceptable and preferred asset oped countries will require concerted efforts in class in the portfolios of institutional investors several areas. Countries at an early stage of bond- (pension funds, insurance, and mutual funds) in market development should focus on the infra- emerging markets because of the high volatility ex- structure of the primary market (issuance) and re- perienced in emerging equity markets after 1997. lated markets. The pertinent areas include: (1) Pension funds and insurance companies have long- risk-free interest rate benchmarks; (2) a well-func- term liabilities, best funded by high-quality debt tioning primary dealer system (a network of finan- instruments such as long term government bonds. cial intermediaries); (3) credible credit ratings; (4) Retail investors, too, look for relatively safe in- efficient trading platforms; (5) sound and safe struments that will nevertheless bring them higher clearing and settlement systems; and (vi) a diversi- yields than bank deposits. fied investor base. The funds managed by institutional investors Countries at an advanced stage of market de- in emerging markets have grown in recent years velopment will need to undertake additional re- because of several factors—among which are the forms to improve the efficiency of their bond mar- excess of national savings over national invest- kets. These reforms include: (1) strengthening ment, particularly in several East Asian countries; primary dealer systems by offering them liquidity pension reforms (in Chile, Mexico, and Thailand, supports through repurchase agreements, in return for example); rapid growth of the insurance indus- for market making; (2) creation of a securities bor- try in many countries, especially China and Thai- rowing and lending facility to enable primary deal- land; and growth of collective investment schemes ers to borrow securities from institutional investors (mutual funds and other similar arrangements) in for trading purposes; (3) establishment of a central most emerging markets covered in this chapter. information system to disseminate bond-market in- At the end of June 2005, the East Asian central formation similar to those functioning in the Re- Delivered by The World Bank e-library to: banks, through the Executives’ Meeting The East- Bankpublic of Korea, which enable implementation of of World Asia and Pacific Central Banks (EMEAP), IP had in- market-to-market valuation of fixed-income instru- : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 70 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S ments; (4) diversification of local-currency bond funding of future requirements by developing markets through promotion of corporate and mu- countries during 2005. In February 2006, Brazil, nicipal bonds; (5) expansion of an investor base for Colombia, and República Bolivariana de Venezuela bond markets; (6) development of derivatives mar- announced debt-buyback programs that together kets to facilitate risk management; (7) increased could lower their foreign currency–denominated participation of foreign investors through removal external debt by about $16 billion (see the annex of impediments such as withholding tax and capi- to this chapter for a discussion of buybacks). tal controls. Meanwhile, Mexico announced another buyback The efforts made by the East Asian countries in in March, repurchasing $2.9 billion of its less- developing their domestic bond markets have met traded global bonds. These buybacks are in line with some early success and could provide a case with the liability-management and deleveraging worth watching by other emerging economies. practices that many developing countries have pur- Foreign institutional investors provide bene- sued over the past few years to improve the terms fits to local bond markets in several ways. First, and risk profile of their external debt. Buybacks they can increase liquidity. Second, given their could be especially significant for Brady debt—only large capital base and experience in fixed-income about $9 billion of Brady bonds (about 6 percent markets, they can play the role of primary dealers of the original issue) will remain outstanding after and market makers, the absence of which is a Brazil and República Bolivariana de Venezuela major gap in most emerging markets. Third, they conclude their buybacks. The supply of foreign can improve the efficiency of the market by de- currency bonds is likely to be limited, except from ploying state-of-the-art technology and services a few countries (Bulgaria, Hungary, Turkey) with available in the international capital markets. Fi- large external financing requirements. Turkey nally, they can introduce new investors to the do- alone is expected to account for one-third of the mestic market, help broaden the investor base, and external financing demands of emerging markets. play a key role in developing capacity in domestic In 2005, several countries were successful in alter- capital markets. ing their debt profile by refinancing foreign debt However, growing local-currency debt mar- through domestic bond markets. In coming years, kets present new challenges to decision makers. some developing countries—among them Brazil, Government debt denominated in the local cur- Colombia, Malaysia, Mexico and Thailand—are rency will need to be managed with as much care likely to raise most of their funds in domestic bond as debt denominated in international currencies markets. Therefore, sovereign issues in the interna- and on an integrated basis. The establishment of tional bond market are likely to become more an independent debt-management office should be scarce. Banking flows are also likely to taper off considered to manage both domestic and interna- from their record level in 2005, as mergers and ac- tional debt within the country’s overall macroeco- quisitions in the oil industry are completed. nomic framework. In this regard, the experiences However, the supply of bonds from corpo- of Sweden and New Zealand could be of interest rate issuers is likely to increase with the revival to developing countries. Capacity building in risk of private investment in Asia and Latin America. management (currency, interest rate, and duration) Demand from international investors is likely to will also be needed to ensure that public debt is be buoyant, because yields on corporate bonds properly managed. are higher than those on sovereign issues. For- eign flows into some local-currency bond mar- kets are likely to increase because of the limited supply of external debt denominated in foreign Prospects for private capital flows currency, and because of the potential for cur- P rivate capital flows to developing countries in- creased sharply in 2005, but the outlook through 2007 is mixed. Debt flows are likely to re- rency appreciation. FDI flows are expected to grow, although at a slower rate than the last year. High commodity Delivered by The World Bank e-library to: main subdued because of accumulated foreign ex-The World prices are likely to boost investment in extractive Bank change reserves, substantial repayments, and pre- industries, while ongoing liberalization in China, IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 71 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 India, and other countries should increase FDI in kets in developing countries are better than for the services sector. Given fundamental macro- those of the developed countries. This bodes well economic improvements in several developing for future equity flows into emerging markets in countries and projected annual growth of 2006–7. However, the pace will be more mea- around 5 percent, the prospects for equity mar- sured than in 2005. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 72 (c) The International Bank for Reconstruction and Development / The World Bank Annex: Commercial Debt Restructuring Developments in 2005 and the first standing (about $1.5 billion). In January 2006, quarter of 200614 Bulgaria exercised embedded call options to fully T he period under review saw major debt- management activities in developing coun- tries, some of which were resolutions of previ- retire just under $938 billion of interest arrears bonds (IABs) that were to expire in 2011. In July 2005, Bulgaria also bought back all of its front- ously defaulted debts, such as the conclusion of loaded interest reduction bonds (FLIRBs), worth Argentina’s debt-restructuring program and Iraq’s about $608 million, in an operation generating a restructuring of debt incurred under Saddam Hus- $648 billion reduction in outstanding debt and sein’s reign. Some of these debt-management ac- about $120 million in debt-service expenses over tivities involved stressed-debt restructuring, such the next 7 years. By retiring the entire outstanding as the Dominican Republic’s $1.1 billion debt- FLIRBs and IABS, Bulgaria fully redeemed its exchange operation. Others involved another Brady bonds, issued in 1994 to restructure its debt wave of Brady buyback operations and announce- to the London Club of commercial creditors. ments, which will retire most of the remaining República Bolivariana de Venezuela. In Feb- Brady bonds outstanding. Brady retirements are ruary 2006, República Bolivariana de Venezuela in line with the liability-management practice and announced plans to buy back $3.9 billion worth of deleveraging that many developing countries have outstanding Brady bonds, leaving only $487 mil- pursued over the past few years to improve their lion outstanding in the market, in an operation to external debt terms and risk structure. be financed by the country’s large oil revenues and international reserves. According to the govern- Brady bond restructuring ment, the deal will reduce external debt to 21 per- Brazil. Two buyback operations in 2005 retired cent of GDP by end-2006, down from 23.4 per- $5.6 billion of Brazil’s Brady bonds. In July 2005, cent at end-2005. This operation will also enable Brazil used the proceeds from a new 12-year the government to realize $670 million in interest global bond (A-bond) to buy back $4.5 billion of payment savings in 2006, and a further $600 mil- its outstanding C-bonds (or capitalization bonds). lion per year in interest and principal savings The global A-bond issue was priced at a premium through 2020. The country had previously bought and carried a coupon of 8 percent. In October back $3.8 billion of Brady bonds in 2003 and an 2005, Brazil completed its second buyback opera- additional $2.2 billion in 2004. tion, retiring the remainder of its C-bonds (worth about $1.1 billion). In February 2006, Brazil an- Other bond market restructurings nounced that it will buy back all of its remaining Argentina. In June 2005, Argentina finally com- Brady bonds by exercising the embedded call op- pleted a debt-restructuring operation involving tions, effectively marking the end of the country’s more than $100 billion in defaulted bonds and in- Delivered by The World Bank e-library to: restructured debt era. terest The World arrears. Argentina swapped about $62.3 bil- Bank Bulgaria. Two buyback operations in 2005 re- lion in defaulted bonds and $680 million in interest IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 tired all of Bulgaria’s remaining Brady debt out- payments for $35.3 billion in 11 new bond issues 73 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 denominated in yen, euros, dollars, and pesos. This $14 billion in defaulted debt for a new eurobond operation resulted in a 75 percent net present value issue worth bout $2.7 billion. In accordance with reduction in principal for bondholders, with about a December 2005 agreement, the holder of each 76 percent of bondholders accepting the deal. Ac- $100 of tendered claims received a new bond with cording to government estimates, the transaction is a $20 face value, carrying a coupon of 5.8 percent expected to result in debt-payment savings of more and amortizing between 2020 and 2028. Some than $67 billion. creditors received a floating rate note paying 50 Colombia. Two buyback operations in 2005–6 basis points over Libor in lieu of the new bond. retired about $1.1 billion of Colombia’s external Further notes up to an additional $800 million debt. In September 2005, Colombia used the pro- may be issued for other eligible outstanding claims ceeds from a reopening of its 20-year global bond on the same terms. to buy back $497 million of dollar-denominated Mexico. In October 2005, the Mexican gov- bonds maturing in 2007, 2010, 2011, 2016, 2027, ernment carried out a debt-management opera- and 2033, and 136 million in bonds maturing in tion to retire about $1.4 billion of global bonds 2008 and 2011. This operation yielded savings of with 10 different maturities between 2007 and $135 million in interest payments and improved the 2033 through open-market repurchase. In No- country’s dollar yield curve. In March 2006, the vember 2005, Mexico became the first developing Colombian government bought back about $601 country to issue warrants that allow investors to million of dollar- and euro-denominated bonds ma- exchange dollar-denominated bonds for peso-de- turing between 2006 and 2011, using $365 million nominated debt at specific strike dates in 2006. of cash on hand and $306 million of the proceeds The exchange operation involved three series of from the reopening of a 2015 peso bond. warrants, which can be exercised up to a maxi- Dominican Republic. In May 2005, the Do- mum of $2.5 billion in bonds potentially ex- minican Republic restructured about $1.1 billion changed for domestic peso bonds. The transaction of its external debt through two exchange offers, was part of the government’s continuing effort to which converted $500 million of 2006 bonds into shift its financing to local-currency debt markets. new 5-year amortizing bonds, and $600 million of In March 2006, Mexico retired $2.9 billion worth 2013 bonds into new 11-year amortizing bonds. of global bonds due to mature between 2007 and The new 5-year and 11-year bonds carry coupons 2031, and issued $3 billion of new global bonds of 9.5 percent and 9.04 percent, respectively. The due in 2017. The new global issue carried a exchange deal extended the maturities of the coun- coupon of 5.63 percent, and was priced to yield try’s outstanding bonds by 5 years and resulted in 5.74 percent, or 105 basis points above compara- about $100 million of interest savings in 2005 and ble U.S. Treasuries. 2006. Approximately 94 percent of eligible bond- Panama. In November 2005, Panama ex- holders participated in the exchanges. In July changed $820 million of short-dated dollar bonds 2005, the Dominican Republic reopened the ex- for a new $980 million global bond due in 2026. change offer, which boosted participation to about The new issue was priced at a discount with a 97 percent. coupon of 7.13 percent, yielding 7.42 percent, or Iraq. In October 2005, Iraq concluded a two- 263 basis points over the U.S. Treasury rate. In a phase commercial debt restructuring with small transaction intended to improve the long end of creditors holding $35 million or less of debt in- the government’s yield curve, in January 2006 curred under Saddam Hussein’s reign. Of about Panama exchanged about $1.1 billion of global $1.6 billion in eligible claims, it is estimated that bonds due in 2020, 2023, and 2034 for a new 71 percent of creditors accepted the deal and only $1.4 billion global bond due in 2036. This ex- 8 percent of creditors elected to reject. In January change operation retired $117 million of 2020s, 2006, the government of Iraq completed a debt- $617 million of 2023s, and $327 million of 2034s. exchange operation with commercial creditors The new issue was priced at 98.4 percent of face holding more than $35 million of debt incurred value to yield 6.94 percent, or 230 basis points Delivered by The World Bank e-library to: under Saddam Hussein’s reign, swapping Theabout World Bankover the U.S. Treasury rate. IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 74 (c) The International Bank for Reconstruction and Development / The World Bank T H E G R O W T H A N D T R A N S F O R M A T I O N O F P R I V A T E C A P I T A L F L O W S Notes floating rate note paying a coupon of Libor plus a spread; 1. The concentration pattern was similar to bank lend- (ii) fund the purchase in the repo market, paying the gen- ing except for the Philippines, which attracted very little eral-collateral repo rate, which is typically close to Libor; bank lending. and (iii) buy protection on the issuer’s name in the CDS 2. International Financial Services, London, 2005. market, paying the premium. 3. InterSec reports that U.S. fund managers’ alloca- 13. However, in practice, the arbitrage cannot be im- tions to international stocks rose from 13 percent in 2004 to plemented at all times and under all market conditions. 15 percent in 2005. 14. As of April 7, 2006. 4. The growth in FDI was led by the United Kingdom, where FDI inflows almost tripled to a record high after almost $100 billion worth of asset restructuring of a large oil com- References pany. Because of the restructuring of the Shell Transport and Trading Company and Royal Dutch Petroleum Company into A.T. Kearney. 2005. “2005 Foreign Direct Investment Con- Royal Dutch Shell, Royal Dutch Shell was classified as a for- fidence Index.” www.atkearney.com. eign company for UK balance-of-payments purposes. That re- BIS (Bank for International Settlements). Various issues. BIS sulted in a sharp increase in FDI into the United Kingdom. FDI Quarterly Review. Basel. www.bis.org. flows to Canada, Germany, and United States also increased in Blanco, Roberto, Simon Brennan, and Ian W. Marsh. 2005. 2005 after significant reductions since 2000. In 2004, after a “An Empirical Analysis of the Dynamic Relation be- continuing decline, FDI flows to Germany slumped into nega- tween Investment-Grade Bonds and Credit Default tive numbers as changes in corporate tax laws led to large re- Swaps.” Journal of Finance 60 (5): 2255–81. payments of intercompany loans (OECD 2005). Bloomberg Markets. 2005. “Deutsche Bank Ousts Citi- 5. The Foreign Investment Advisory Service (FIAS), group: Demand for Euro-Denominated Issues Puts part of the International Finance Corporation, advises gov- Sales on a Record Pace for 2005.” November. http:// ernments on how to attract and retain FDI by providing in- www.bloomberg.com/media/markets/index.html vestment climate diagnostics and developing customized Breger, Ludovic, and Darren Stovel. 2005. “Global Integra- long-term FDI promotion strategies that fit each client tion of Developed Credit Markets.” Journal of Portfo- country’s needs, objectives, and capacity. lio Management (Spring). 6. In 2005, SAB-Miller bought a brewery company for Business China. 2005. “Bankable Prospects.” October 10. $7.8 billion bringing approximately US$1 billion worth of Duffie, Darrell. 1999. “Credit Swap Valuation.” Financial FDI into Columbia. In addition, Philip Morris bought a Analysts Journal (January–February). local tobacco producer for $350 million. Emerging Market Traders Association. 2003. “2003 An- 7. Itaú BBA has opened its first office in China. Brazil’s nual Debt Trading Volume Survey.” New York. second-leading commercial bank is targeting Chinese and www.emta.org. Brazilian companies doing business in both markets. Fitch Ratings. 2004. “Global Credit Derivatives Survey 8. French Vodafone increased its share in Vodacom 2004.” http://www.fitchratings.com from 35 percent to 50 percent. The deal represents the sec- IMF (International Monetary Fund). 2003. “Coordinated ond-largest inflow of foreign direct investment into South Portfolio Investment Survey 2003.” Statistics Depart- Africa after the Barclays-ABSA deal. ment, Washington, DC. 9. In a perfect international market, covered interest ———. 2005. Global Investor Survey 2005. Washington, DC. arbitrage implies that spreads on bonds issued by the same International Financial Services. 2005. “Fund Manage- issuer in different currencies are just a function of respective ment.” London. August. Jarrow, Robert A., and Stuart M. Turnbull. 1995. “Pricing  1 + re  interest rates and net exchange rates. Thus Se =  Sd Derivatives on Financial Securities Subject to Credit  1 + rd   Risk.” Journal of Finance 50 (1). where Se and Sd are spreads on the euro and the dollar, re- Kalotay, Kalman. 2006. “The Impact of EU Enlargement on spectively and re and rd are corresponding interest rates in FDI Flows.” International Finance Review 6: 473–99. euros and dollars. See Kercheval, Goldberg, and Berger Kercheval, Alec, Lisa Goldberg, and Ludovic Breger. 2003. (2003) and Berger and Stovel (2005) for more detail. “Modeling Credit Risk: Currency Dependence in 10. See “Deutsche Bank Ousts Citigroup: Demand for Global Credit Markets.” Journal of Portfolio Manage- euro-denominated issues puts sales on a record pace for ment (Spring): 90–100. 2005,” Bloomberg Markets, November 2005. OECD (Organisation for Economic Co-operation and De- 11. Some studies have provided empirical evidence of velopment). 2005. “Recent Trends in Foreign Direct the comovement of the two asset prices for investment- Investment in OECD Countries.” Investment Division, grade bonds (Blanco, Brennan, and Marsh 2005). Paris. http://www.oecd.org/investment 12. In theory, under the absence-of-arbitrage-opportu- Takeuchi, Atsushi. 2005. “Study of Impediments to Cross- nity hypothesis, a par floating rate note and a CDS on the border Bond Investment and Issuance in Asian Coun- same issuer should have the same spread. If the spread of tries.” Paper prepared for discussion at the working the latter was strictly larger, a risk-free gain would be possi- Delivered by The Worldgroup of the ASEAN+3 Bank e-library to: Asian Bond Market Initiatives. ble by entering into the following trade: (i) purchase of a parThe World December. Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 75 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Tovar, Camilo E. 2005. “International Government Debt USA Today. 2005. “Only the Bravest of Bankers Boldly Go Denominated in Local Currency: Recent Developments to China.” January 19, 2005. in Latin America.” BIS Quarterly Bulletin (December): World Bank. 2005. World Development Report 2005: A 109–18. http://www.bis.org/publ/qtrpdf/r_qt0512.pdf Better Investment Climate for Everyone. Washington, UNCTAD. (United Nations Conference on Trade and De- DC: World Bank. velopment). 2004. World Investment Report 2004. World Bank. Various years. Global Development Finance. New York: UNCTAD. Washington, DC: World Bank. ———. 2005. World Investment Report 2005. New York: UNCTAD. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 76 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank . 3 Supporting Development through Aid and Debt Relief D evelopment finance moved to center stage ing further debt relief to HIPCs; and helping to en- at a series of major international forums in sure that developing countries can maintain sustain- 2005. The High-Level Forum on Aid Ef- able debt levels over time. It highlights recent trends fectiveness held in Paris in March set out to in each of these areas and reflects on how the policy change how aid is delivered and managed. The initiatives announced over the course of 2005 are Commission for Africa issued a report in March likely to influence development finance over the bal- urging donors to scale up aid for Africa signifi- ance of the decade. The main messages are: cantly. Expectations for a big push in development assistance with a strong focus on Africa escalated • Official development assistance (ODA) in- over the course of the year, leading up to the G-8 creased sharply in 2005, reaching 0.33 percent Summit in Gleneagles, Scotland, in July, where of gross national income (GNI) in donor coun- “Africa and Development” was one of two main tries, up from a low of 0.22 percent in 2001, themes. The United Nations World Summit fol- just below the 0.34 percent level attained in the lowed in New York in September to assess early 1990s. Although most of the record $27 progress toward the Millennium Development billion increase in 2005 is accounted for by Goals (MDGs) and reinforce commitments on the debt relief grants provided to just two countries part of donor and recipient countries. Multilateral (Iraq and Nigeria), the underlying trend indi- trade liberalization also played a central role in the cates that donors have continued to enhance development agenda in 2005. Although the World their aid effort. Based on existing commit- Trade Organization (WTO) Ministerial Meeting in ments, ODA is expected to decline in 2006–7, Hong Kong (China) in December did not complete as debt relief falls to more normal levels, but the Doha Development Round as planned, “aid then to rise gradually through the end of the for trade” surfaced as a major policy initiative, decade to reach 0.36 percent of GNI in 2010. with new commitments by advanced countries to • Donors have taken steps to improve: (1) the enrich development assistance. allocation of aid, by providing more aid re- Broad agreement surfaced at the international sources to the poorest countries, particularly forums about the need to provide more aid re- those in Sub-Saharan Africa, where the sources, particularly to poor countries in Africa, amount of aid may double by the end of the and to further reduce the debt burdens of heavily in- decade; (2) the composition of aid, by provid- debted poor countries (HIPCs) in order to free up fi- ing more grants in place of concessional loans nancial resources for meeting the MDGs. There was in an effort to reduce countries’ debt service also strong emphasis on the importance of debt sus- burden and improve debt sustainability; and tainability in underpinning growth, and thereby al- (3) the effectiveness of aid, by developing a leviating poverty over time. This chapter addresses framework that includes tangible indicators Delivered by The World Bank e-library to: these broad objectives—namely, enhancing the aidThe World and Bank targets designed to gauge development effort, particularly in the context of Africa; provid- progress over time. IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 79 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 • Debt relief provided under the HIPC Initiative rey Consensus, recognizing that a substantial in- and the Multilateral Debt Reduction Initiative crease in foreign aid was required to achieve inter- (MDRI) will significantly reduce the debt bur- nationally agreed goals, including the MDGs. dens of poor countries that qualify. The debt Donors continue to deliver on their promise. Ac- of 17 countries that have already reached the cording to the Development Assistance Committee completion point under the HIPC Initiative (DAC) of the Organisation for Economic Co-oper- will fall from 55 percent of GDP (before HIPC ation and Development (OECD), net ODA dis- debt relief) to 13 percent (after MDRI debt re- bursements by DAC member countries increased lief). Other poor countries have made consid- by a record $27 billion in 2005, reaching $106.5 erable progress in reducing their debt burdens billion (table 3.1). from very high levels, but much more needs to Relative to gross national income (GNI) in be done, particularly in Sub-Saharan Africa. DAC member countries, ODA increased to 0.33 • Debt sustainability in many of the HIPCs has percent in 2005, up from a low of 0.22 percent in been enhanced by other factors, including 2001, but still remains slightly below the 0.34 per- stronger economic growth, foreign reserve ac- cent level reached in the early 1990s (figure 3.1). cumulation, improved external balances, and higher inflows of foreign direct investment The rise reflects debt relief and other (FDI) and remittances. Going forward, low- special-purpose grants income countries, HIPCs and non-HIPCs However, much of the increase in ODA was due to alike, face the challenge of financing their de- debt relief grants, which totaled $23 billion in velopment plans without compromising debt 2005, up from $4 billion in 2004 (table 3.2). This sustainability over the long term. Countries largely reflected nearly $14 billion in debt relief can enhance debt sustainability by pursuing provided to Iraq and a little over $5 billion to macroeconomic policies that maintain eco- Nigeria by their Paris Club creditors. Excluding nomic and financial stability and by making debt relief, ODA increased by 8.7 percent in real progress on structural reforms to improve terms, up from average annual rate of 5.6 percent their policy and institutional frameworks. in 2002–4. At the UN Conference on Financing for De- velopment in Monterrey in 2002, donors pledged that debt relief would not displace other compo- Recent trends and prospects nents of ODA. It is difficult to assess whether for foreign aid donors have honored their pledge in the absence of ODA continues to rise an explicit counterfactual demonstration of the At the United Nations World Summit in Septem- amount of ODA that would have been provided in ber in New York countries reaffirmed the Monter- the absence of debt relief. The share of debt relief Table 3.1 Net ODA disbursements, 1990–2005 $ billions 1990 1995 2000 2001 2002 2003 2004 2005a DAC donors 54.3 58.8 53.7 52.4 58.3 69.1 79.6 106.5 G7 countries 42.4 44.7 40.2 38.2 42.6 50.0 57.6 80.1 United States 11.4 7.4 10.0 11.4 13.3 16.3 19.7 27.5 Japan 9.1 14.5 13.5 9.8 9.3 8.9 8.9 13.1 United Kingdom 2.6 3.2 4.5 4.6 4.9 6.3 7.9 10.8 France 7.2 8.4 4.1 4.2 5.5 7.3 8.5 10.1 Germany 6.3 7.5 5.0 5.0 5.3 6.8 7.5 9.9 Canada 2.5 2.1 1.7 1.5 2.0 2.0 2.6 3.7 Italy 3.4 1.6 1.4 1.6 2.3 2.4 2.5 5.1 Memo item: EU countries 28.3 Delivered by The World 31.2 25.3 Bank e-library 26.4 to: 30.0 37.1 42.9 55.7 The World Bank IP : 192.86.100.36 (DAC). Source: OECD Development Assistance CommitteeTue, 10 Mar 2009 16:55:33 a. Preliminary. 80 (c) The International Bank for Reconstruction and Development / The World Bank S U P P O R T I N G D E V E L O P M E N T T H R O U G H A I D A N D D E B T R E L I E F Figure 3.1 Net ODA to developing countries, Table 3.2 ODA and debt relief grants in 2005 1990–2005 $ billions Percent ODA Percent change in excluding ODA excluding debt debt relief grants 0.3 ODA Debt relief grants relief grants in real termsa Total ODA DAC donors 106.5 23.0 83.5 8.7 G7 countries 80.1 20.2 59.9 8.9 0.2 ODA excluding debt relief United States 27.5 4.1 23.4 16.2 Japan 13.1 3.6 9.5 12.1 United Kingdom 10.8 3.7 7.1 –1.7 0.1 ODA excluding special purpose grants France 10.1 3.2 6.9 0.0 Germany 9.9 3.6 6.3 –9.8 Canada 3.7 0.5 3.2 17.8 0.0 Italy 5.1 1.7 3.4 40.0 1990 1992 1994 1996 1998 2000 2002 2004 Memo item: EU countries 55.7 27.9 27.8 3.8 Source: OECD Development Assistance Committee (DAC). Source: OECD Development Assistance Committee (DAC). a. Takes into account inflation and exchange-rate movements. in ODA has risen from an average of 3.7 percent in the 1990s to 6.6 percent in 2002–4, followed by a sharp increase to 22 percent in 2005. ODA, net of debt relief, has risen relative to GNI in donor ODA net of special-purpose grants totaled $45 countries, but at a more modest pace than overall billion in 2005, unchanged from 2004, but up sig- ODA (figure 3.1). Thus, some, but not all, of the nificantly from a low of $30 billion in 2001. How- scaling-up in aid can be attributed to debt relief. ever, relative to GNI in DAC member countries, Debt relief together with other special-purpose ODA net of special-purpose grants has shown little grants—for technical cooperation, emergency and increase over the past 10 years (1996–2005), aver- disaster relief, and administrative costs—accounted aging 0.14 percent, remaining well below the 0.23 for three-quarters of the bilateral portion of ODA level attained in the early 1990s (figure 3.1). Thus, in 2005, well above the 53 percent average of the the increase in the ODA as a percent of GNI over 1990s (table 3.3). Excluding the $19 billion in debt the past few years reflects higher special purpose relief provided to Iraq and Nigeria, special-purpose grants. grants still accounted for two-thirds of bilateral ODA in 2005. Emergency and distress relief grants The shift from concessional loans increased by $5 billion in 2005, $2.2 billion of to grants continues which was provided in response to the December Bilateral donors have continued to shift their re- 2004 tsunami. However, part of remaining $2.8 sources from concessional loans to grants, with billion increase reflects a modification in the defini- the goal of limiting the rise in the debt burdens of tion to include reconstruction grants.1 aid recipients and thereby prevent a recurrence of Table 3.3 Main components of bilateral ODA, 1990–2005 $ billions 1990 1995 2000 2001 2002 2003 2004 2005a Total ODA 54.3 58.8 53.7 52.4 58.3 69.1 79.6 106.5 Bilateral ODA 38.5 40.5 36.1 35.1 40.8 49.8 54.4 82.0 Debt relief 1.5 2.7 1.6 2.0 3.7 6.8 4.2 23.0 Technical co-operation 11.4 14.3 12.8 13.6 15.5 18.4 18.8 21.6 Emergency/distress relief 1.1 3.1 3.6 3.3 3.9 6.2 7.3 12.7 Administrative costs 2.0 2.9 3.1 3.0 3.0 3.5 4.0 4.0 Special purpose grants: 15.9 23.0 21.0 21.8 26.1 34.8 34.3 61.3 Multilateral ODA 15.8 18.3 17.7 17.3 17.5 19.3 25.1 24.5 Total ODA less debt relief 52.7 56.1 52.2 50.5 54.6 62.3 75.4 83.5 Total ODA less special purpose grants 38.4 Delivered 35.8 by The World 32.7 Bank e-library 30.6 to: 32.2 34.2 45.2 45.2 The World Bank IP : 192.86.100.36 Source: OECD Development Assistance Committee (DAC). Tue, 10 Mar 2009 16:55:33 Note: a. Preliminary. 81 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 3.2 Bilateral ODA loans and grants, allocated to the Middle East rose from 4.5 percent 1990–2005 in 2002 to 11.6 percent in 2004, with most of the $ billions increase going to Iraq, Afghanistan, and Jordan 85 (table 3.4). Aid to Iraq rose from an average of 75 only $90 million in 2000–2 to $3.2 billion in 65 2003–4, making it the largest recipient of ODA. 55 Bilateral ODA Aid to Iraq is likely to rise further, as its agreement 45 35 with Paris Club creditors in November 2004 in- 25 Bilateral grants cluded $30 billion in debt relief that that will re- 15 sult in a major increase on Iraq’s share of ODA be- Bilateral loans 5 ginning in 2005. Similarly, aid to Afghanistan –5 increased from $0.5 billion to $1.4 billion over the 1990 1995 2000 2005 same period. Increases in aid to Iraq, Afghanistan, Source: OECD Development Assistance Committee (DAC). and the Democratic Republic of the Congo ac- count for over two-thirds of the increase in total lending/debt forgiveness cycles that have occurred ODA in 2003–4. over the past few decades. Net concessional lending from bilateral donors has averaged close to zero More “aid for trade” is on the way over the past five years (2001–5), implying that dis- Donors also are focusing more aid resources to bursements of new concessional loans equaled re- bolster the capacity of the poorest countries to payments (interest and principle) on existing loans participate in trade and manage the adjustment on average, whereas in the early 1990s new lending costs of liberalization. This entails providing assis- exceeded repayments by about $6 billion on aver- tance for trade policy and regulations (technical age (figure 3.2). assistance for product standards, integration of trade with development plans, trade facilitation), Donors are providing more assistance to the trade development (trade promotion, market de- least developed countries and those affected velopment activities) and building infrastructure by conflict (transport, energy, and telecommunications). The Donors have been reallocating development assis- amount of aid devoted to trade-related assistance tance to the poorest countries. The amount of has risen over the past few years, increasing from ODA allocated to the least developed countries 3.6 percent of total aid commitments in 2002 to (LDCs) has increased substantially since the late 4.4 percent in 2003, with infrastructure account- 1990s, while that allocated to other low-income ing for a further 25 percent.4 countries has been relatively constant in nominal The G-8 Summit in Gleneagles gave important terms. The share of total ODA allocated to the high-level endorsement for “aid for trade” initia- LDCs grew from a low of 30 percent in 1999 to a tives that aim to build the physical, human, and high of 45 percent in 2003, while the share allo- cated to other low-income countries declined from Figure 3.3 Share of total ODA allocated to LDCs 29.5 percent to 19 percent in 2004 (figure 3.3).2 and other low-income countries, 1990–2004 From a regional perspective, donors have been reallocating development assistance to countries in Percent Sub-Saharan Africa and the Middle East. The share 45 LDCs of total ODA allocated to Sub-Saharan Africa in- 40 creased from a low of 25 percent in 1999 to 40 35 percent in 2004,3 while that allocated to Asia de- 30 clined from 44 percent to 35 percent. Donors are 25 committed to continued increases in Africa’s share Other low-income countries of ODA over the balance of the decade. 20 Delivered by The World Bank e-library to: A portion of the rise in ODA over The past Bank 15 theWorld two years reflects increased assistance for coun- IP : 192.86.100.36 1990 1995 2000 2004 Tue, 10 Mar 2009 16:55:33 tries affected by conflict. The share of total ODA Source: OECD Development Assistance Committee (DAC). 82 (c) The International Bank for Reconstruction and Development / The World Bank S U P P O R T I N G D E V E L O P M E N T T H R O U G H A I D A N D D E B T R E L I E F Table 3.4 Net ODA disbursements to the ten largest recipient countries $ billions, average over period 1990–9 2000–2 2003–4 Egypt 2.23 Indonesia 1.35 Iraq 3.24 China 1.82 China 1.18 Dem. Rep. of Congo 3.09 Indonesia 1.47 Egypt 1.12 Afghanistan 1.45 Poland 1.33 Serbia & Montenegro 1.05 China 1.36 India 1.07 Mozambique 1.00 Vietnam 1.07 Philippines 0.90 Vietnam 0.94 Ethiopia 1.03 Bangladesh 0.75 Tanzania 0.88 Tanzania 1.00 Mozambique 0.72 India 0.78 Egypt 0.98 Thailand 0.70 Pakistan 0.76 Indonesia 0.85 Tanzania 0.68 Bangladesh 0.57 Jordan 0.76 Memo items: Iraq 0.16 Iraq 0.09 Afghanistan 0.11 Afghanistan 0.47 Dem. Rep. of Congo 0.18 Dem. Rep. of Congo 0.20 Net offical assistance disbursements by largest recipientsa Russian Fed. 1.22 Russian Fed. 1.12 Russian Fed. 1.03 Israel 1.33 Israel 0.57 Israel 0.46 Source: OECD Development Assistance Committee (DAC). a. Included in official aid (OA), but not official development assistance (ODA). institutional capacity of poor countries so that over the medium term. Five of the 22 DAC mem- they can play a more prominent role in the negoti- ber countries have already increased ODA to levels ation of multilateral trade agreements and benefit that exceed the UN target (Norway, 0.87 percent more fully from the outcomes. The G-8 asked mul- of GNI; Denmark, 0.85 percent; Luxembourg, tilateral institutions to provide additional assis- 0.83 percent; Sweden, 0.73 percent; the Nether- tance to poor countries to develop their trade ca- lands, 0.73 percent). The European Union has pacity and ease the adjustment costs arising from pledged to increase ODA provided by its member trade liberalization. In response, the World Bank countries from 0.35 percent of GNI in 2004 to 0.7 and the International Monetary Fund (IMF) pro- percent by 2015, with an interim target of 0.56 posed to enhance the Integrated Framework for percent by 2010.5 Moreover, six EU member coun- Trade-related Technical Assistance for the LDCs tries announced commitments to attain the 0.7 per- (box 3.1), a move endorsed at the annual meetings cent UN target prior to 2015 (Belgium and Finland of the IMF and the World Bank in September and by 2010; France, Ireland, and Spain by 2012; and at the WTO Hong Kong Ministerial in December. the United Kingdom by 2013). Although the Doha Development Round was Other donors have made commitments that not completed as planned at the WTO Ministerial are not linked to the UN target. For example, Meeting in Hong Kong in December 2005, modest ODA provided by the United States is projected to progress was made. In particular, participants decline from $27.5 billion in 2005 ($23.4 billion agreed to phase out agricultural subsidies by 2013, excluding debt relief grants) to $24 billion in 2006 and developed countries agreed to provide market (in real terms) and remain at that level to 2010, access (free from quotas and duties) to the LDCs based on commitments announced on the margins on 97 percent of their tariff lines. of the G-8 Summit.6 At the G-8 Summit, Japan an- nounced its intention to increase ODA by $10 bil- Donors have enhanced their commitments to lion over the next five years. Projections based on scale up aid these commitments imply that the share of total At the G-8 Summit in Gleneagles, Scotland, ODA provided by the United States will decline donors announced their commitment to increase from 26 percent in 2005 to 19 percent in 2010, Delivered by The World Bank e-library to: ODA by $50 billion by 2010 (in real terms) fromThe Worldwhile that provided by the EU member countries Bank 2004 levels. Many donor countries have made ex- as a group will increase from 54 percent to 63 per- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 plicit commitments to scale up aid significantly cent (table 3.5). 83 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Box 3.1 The Integrated Framework for Trade-Related Technical Assistance T he Integrated Framework for Trade-Related Technical Assistance (IF) brings together the International Mon- etary Fund, International Trade Centre, United Nations or applying to join. As of September 2005, 30 capacity- building projects had been approved in 12 countries, amounting to $10 million, and 17 donors, including the Conference on Trade and Development, United Nations World Bank, had pledged a total of $34 million to the IF Development Programme, World Trade Organization, the Trust Fund. World Bank, and bilateral donors to: (i) integrate trade To date, the IF has completed several capacity-build- into the national development plans of LDCs; and (ii) as- ing projects; made solid progress in the difficult task of co- sist in the coordinated delivery of trade-related technical ordinating donors and international agencies; contributed assistance. The IF is built on the principles of country to increased understanding of the constraints facing poor ownership and partnership. It consists of diagnostic stud- countries; and brought IF governments to the table on ies, technical assistance projects, and capacity-building trade. Of the eight IF countries that had completed diag- projects valued at up to $1 million per country. nostics at the time of their poverty reduction strategy, By the end of 2005, diagnostics had been completed three incorporated the recommendations, and two were in 20 countries, with a further 17 countries in the process working to do so for their next poverty reduction strategy. Table 3.5 Donors’ shares of ODA in 2005, about 0.29 percent in 2006–7 and then rise gradu- projected 2010 ally over the balance of the decade as a percent of Percent their GNI, reaching 0.36 percent in 2010, just 2005 2005 (excluding debt relief) 2010 slightly above levels attained in the early 1990s. United States 25.8 28.0 18.7 The projections imply that ODA as a ratio to Japan 12.3 11.4 9.3 GNI in donor countries will increase by about United Kingdom 10.1 8.5 11.4 0.017 of a percentage point per year on average France 9.4 8.2 11.0 Germany 9.3 7.6 12.1 over the period 2005–10. Extrapolating this rate Netherlands 4.8 5.7 4.0 of increase would mean that the UN target of 0.7 Italy 4.7 4.0 7.2 percent would not be attained until 2030, 15 years Sum: 76.5 73.4 73.7 after the 2015 deadline set for attaining the Memo item: MDGs. The UN Millennium Project (2005) esti- EU Members 53.9 49.2 63.4 mates that financing the MDGs requires an in- Source: Projections by the OECD DAC Secretariat. crease in ODA (excluding debt relief) to 0.46 per- cent of GNI by 2010, suggesting that current commitments fall short. There is, however, a high degree of uncertainty surrounding such estimates.7 ODA is expected to decline as a percentage Moreover, the quality of aid, is as, or perhaps even of GNI in the short run and then increase more, important than the quantity of aid for sup- gradually over the balance of the decade porting developing countries progress on the ODA is expected to decline in 2006 as the debt re- MDGs. For example, enriching special purpose lief component falls to more normal levels (figure grants rather than direct budgetary support could 3.4). ODA will continue to be affected by further have quite different implications for the ability of debt relief to be provided to Iraq and Nigeria by its developing countries to fund programs that are Paris Club creditors over the coming few years, but deemed to be critical for accelerating progress of in smaller amounts than in 2005. This explains the the MDGs. transitory nature of the ODA surge in 2005. Based Commitments to increase ODA have been Delivered by The World Bank e-library to: on current commitments of DAC donors, the Bankmade despite the very high level of general govern- The World OECD DAC Secretariat is projecting that ODA ment deficits in many donor countries. Fiscal IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 will decline from 0.33 percent of GNI in 2005 to deficits are expected to exceed or be close to 3 per- 84 (c) The International Bank for Reconstruction and Development / The World Bank S U P P O R T I N G D E V E L O P M E N T T H R O U G H A I D A N D D E B T R E L I E F Figure 3.4 Net ODA as a percentage of GNI in DAC Table 3.6 General government financial balances donor countries, 1990–2005 and projected 2006–10 in 2004, projected 2005–7 Percent of GDP Percent 0.35 2004 2005c 2006c 2007c 0.30 United Statesa –4.7 –3.7 –4.2 –3.9 Projected Japana –6.5 –6.5 –6.0 –6.0 0.25 Total ODA 2006–10 United Kingdom –3.2 –3.1 –3.0 –3.2 0.20 France –3.6 –3.2 –3.2 –3.0 Germany –3.7 –3.9 –3.6 –2.6 0.15 Netherlands –2.1 –1.6 –1.8 –1.5 ODA allocated to Italy –3.3 –4.3 –4.2 –4.8 0.10 Sub-Saharan Africa Projected Weighted average:b –4.3 –3.9 –4.0 –3.8 0.05 2005–10 0.00 Source: OECD Economic Outlook No. 78 Annex Table 27. a. Including social security. 1990 1995 2000 2005 2010 b. Weighted using shares of ODA in 2005 listed in Table 3.5. Source: OECD Development Assistance Committee (DAC); c. Projected. projections by the OECD DAC Secretariat. Table 3.7 ODA as a percentage of fiscal cent of GDP in 2005/6 in six of the seven largest expenditures and revenues in 2004, projected 2006 Percent DAC donor countries, which together accounted for three-quarters of total ODA in 2004 (table Expendituresa Revenuesb 3.6). However, ODA makes up less than 1 percent 2004 2006c 2004 2006c of fiscal revenues and expenditures in six of the United States 0.46 0.49 0.53 0.55 seven major donor countries (the Netherlands Japan 0.51 0.55 0.51 0.55 being the exception) (table 3.7). Donors have ex- United Kingdom 0.84 0.90 0.90 0.97 France 0.79 0.87 0.85 0.93 amined several innovative financing mechanisms Germany 0.59 0.72 0.64 0.78 that could augment aid flows, including the Inter- Netherlands 1.56 1.65 1.64 1.71 national Finance Facility for Immunization, ad- Italy 0.30 0.64 0.32 0.70 vance market commitments for vaccines, and air- Source: OECD Economic Outlook No. 78 Annex tables 2 and 26. line departure taxes.8 a. General government total outlays. b. General government total tax and nontax receipts. c. Projected. Donors have agreed to provide significant increases in aid for Africa With 10 years remaining for developing countries recommendations of the Africa Commission (in- to meet the MDGs, Africa is the only continent cluding the doubling of aid to Sub-Saharan Africa not on track to meet any of the goals. The past by 2010), while underlining the importance of year was to be the year of Africa. It began with a good governance, democracy, and transparency report issued by the Commission for Africa in on the continent. Building on this momentum, March. British Prime Minister Tony Blair had the World Bank presented its Africa Action Plan launched the commission in February 2004 to take in September, setting out a program of concrete, a fresh look at Africa’s past and present, as well as results-oriented actions for the Bank and devel- the international community’s role in its develop- opment partners to assist all African countries to ment path. The report called for a doubling of aid meet as many MDGs as possible. by 2010, while recognizing the need for African Current commitments by donors imply a sig- countries to improve governance and accelerate nificant scaling-up in aid to low-income countries policy reforms so that higher amounts of aid could in Sub-Saharan Africa. Donors have committed to be absorbed effectively. Countries at the African increase total ODA by about $50 billion by 2010 Union Summit in June reaffirmed their commit- (in real terms), at least half of which is slated for ment to promoting economic growth and reducing Sub-Saharan Africa. This would double the Delivered by The World Bank e-library to: poverty. In turn, at the G-8 Summit in July,The World amountBankof aid to the region by 2010 and raise its “Africa and Development” was adopted as one of share of total ODA from 40 percent in 2004 to al- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 two main themes. The G-8 leaders supported the most 50 percent in 2010. 85 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 The new commitments have raised concerns investments (developing public infrastructure and about absorptive capacity labor market training initiatives being prime ex- The commitment by donors to double the amount amples) and through improvements in governance. of aid to Sub-Saharan Africa by 2010 raises the Current proposals under study involve scaling up question of absorptive capacity. There is a concern aid significantly with predictable flows of grant- that a substantial increase in aid flows to some financed aid to selected countries that have rela- countries could have unfavorable macroeconomic tively strong institutions and governance. The his- repercussions. Specially, there is a risk that a surge torical record provides few examples along these in aid flows could lead to an appreciation of the lines and, hence, it is difficult to estimate the re- real exchange rate (either through inflation or the sponse of key macro variables—the real exchange nominal exchange rate), which could in turn un- rate, interest rates, inflation, and output growth— dermine competitiveness and thereby curtail ex- under such circumstances. Researchers have devel- ports. This so-called Dutch disease could under- oped modeling frameworks that can provide in- mine growth, particularly in countries where the sights into the complex linkages between the export sector provides a key source of productivity sequencing and components of aid and the growth growth (because of dynamic externalities such as process, taking into account some of the con- learning by doing). straints that can hinder development. As an exam- Assessing the overall consequences of a surge ple, model simulations reported by Sundberg and in aid flows requires considering the potential ben- Lofgren (2006) indicate that a cost-minimizing efits, along with the costs. For example, invest- strategy for achieving the MDGs in the case of ments in public infrastructure could boost produc- Ethiopia entails a front-loaded expansion in infra- tivity and thereby improve competitiveness, structure spending with constantly growing social offsetting the impact of a real exchange rate appre- spending. ciation. Moreover, higher spending on programs needed to accelerate progress on the MDGs could Improving aid effectiveness plays a critical also enhance growth over the longer term (educa- role in the development agenda tion and health being prime examples). The empir- In addition to their commitments to scale up the ical evidence on the macroeconomic consequences volume of aid, donors promised to improve the of aid surges is inconclusive.9 Recent aid surges in effectiveness of aid. Ministers of developed and a number of African countries have coincided with developing countries responsible for promoting a depreciation of the real exchange rate, contrary development, along with heads of multilateral to theory.10 It is unclear, however, whether that and bilateral development institutions, together outcome reflected productivity-enhancing benefits representing 90 countries and 26 multilateral or- of higher aid, or whether the higher aid was not ganizations, participated in the OECD High- spent or “absorbed” by recipient countries.11 Level Forum in March. Participants at the Forum Donors and recipient countries need to pay careful recognized that while the volumes of aid and attention to the macroeconomic consequences of other development resources must increase to higher aid flows for inflation, domestic interest achieve the MDGs, aid effectiveness must in- rates, and fiscal balances, taking into account the crease commensurately to support partner-coun- high degree of uncertainty surrounding the effects try efforts to strengthen governance and improve on competitiveness and productivity. development performance. To this end, the “Paris Moreover, Bourguignon and Sundberg (2006a Declaration on Aid Effectiveness” committed and 2006b) stress that absorptive capacity is a dy- donor countries, partner countries, and multilat- namic concept that depends on the composition eral institutions to: and sequencing of aid, as well as characteristics of the local economy (labor markets, institutions, de- • Strengthen partner countries’ national devel- mand side constraints, etc.). And as such, a coun- opment strategies and associated operational try’s absorptive capacity can be enhanced by frameworks Delivered by The World Bank e-library to: strategic planning that aims to identify key con- Bank• Increase alignment of aid with partner coun- The World straints to growth and expand its productive ca- IP : 192.86.100.36 tries’ priorities, systems, and procedures, and Tue, 10 Mar 2009 16:55:33 pacity through targeted and carefully sequenced help to strengthen their capacities 86 (c) The International Bank for Reconstruction and Development / The World Bank S U P P O R T I N G D E V E L O P M E N T T H R O U G H A I D A N D D E B T R E L I E F • Enhance donors’ and partner countries’ re- the past few years, reaching $23 billion in 2005, spective accountability to their citizens and largely due to $19 billion in debt relief provided by parliaments for their development policies, the Paris Club to Iraq and Nigeria. In the three strategies, and performance years prior to 2005, debt relief grants averaged $6.7 • Eliminate duplication of efforts and rational- billion, well above the $3.4 billion average in ize donor activities to make them as cost- 1990–2002, with most of the additional resources effective as possible going to the poorest countries, particularly those in • Reform and simplify donor policies and pro- Sub-Saharan Africa (table 3.8). Of the total $20 bil- cedures to encourage collaborative behavior lion in debt-relief grants provided by DAC donors and progressive alignment with partner coun- over the period 2002–4, more than half was allo- tries’ priorities, systems, and procedures cated to the LDCs, up from an average share of 29 • Define measures and standards of perfor- percent over the period 1990–2001. Countries in mance and accountability of partner-country Sub-Saharan Africa received almost three-quarters systems in public financial management, pro- of debt relief provided in 2002–4, up from just over curement, fiduciary safeguards, and environ- a third during the period 1990–2001. mental assessments, in line with broadly ac- cepted good practices and their quick and The HIPC Initiative is significantly reducing widespread application. the debt service burdens of some poor countries The HIPC Initiative has substantially eased the Tangible indicators and targets were estab- debt-service burden of a small group of poor coun- lished so that progress toward the commitments tries, most of which are in Africa (box 3.2).12 The could be tracked. To this end, donor and partner 28 countries that reached the “decision point” for countries are working together to develop an in- debt relief under the initiative prior to 2006 re- ternational monitoring system that will enable ceived $2.3 billion per year in debt relief from them to measure progress toward the targets iden- 2001 to 2005, equal to 2.2 percent of their GDP tified in the Paris Declaration. and 9.2 percent of their exports.13 The HIPC Ini- tiative has provided debt relief equal to about half of the debt service due from the group. Debt-ser- vice payments for the 28 countries equaled 1.8 Debt relief: improving and percent of their collective GDP in 2005 (down maintaining debt sustainability from 3.2 percent in 2000); were it not for debt re- P rogress continues on reducing the debt burdens of the poorest countries, particularly those in Africa. Debt relief is provided under the HIPC Ini- lief under HIPC, they would have been an esti- mated 3.8 percent of GDP in 2005 (figure 3.5). The amount of debt relief provided has varied tiative, through the Paris Club, and on a bilateral considerably across countries. In 4 of the 28 coun- basis. According to the data reported by OECD tries that reached the decision point prior to 2006, DAC donors, grants provided for debt relief from HIPC debt-service reduction exceeded 5 percent of all three sources have increased significantly over GDP on average over the period 1998–2006, but Table 3.8 Debt-relief grants provided by DAC donor countries, by income and region of beneficiary, 1990–2005 $ billions 1990 1995 2000 2001 2002 2003 2004 2005 Debt relief grants 4.3 3.7 2.0 2.5 4.5 8.3 7.1 23.0 Allocation across income classifications Least-developed countries 0.9 0.9 1.2 1.1 2.2 5.6 3.4 — Other low-income countries 1.1 0.6 0.3 0.8 1.1 2.2 2.7 — Allocation across regions Sub-Saharan Africa 1.2 by The World 2.2 Delivered 1.2 Bank e-library 1.3 to: 3.0 6.5 5.0 — The World Bank Other regions 2.2 2.6 0.8 1.2 1.6 1.9 2.1 — IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 Source: OECD Development Assistance Committee (DAC). 87 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Box 3.2 The HIPC Initiative T he HIPC Initiative was launched by the World Bank and the International Monetary Fund (IMF) in 1996, amid growing concerns that excessive debt was crippling and Zambia reached the completion point in 2005, fol- lowed by Cameroon in May 2006. The debt relief ac- corded the remaining 10 decision-point countries will not efforts to reduce poverty in some of the poorest countries. become irrevocable until they pass the completion point. It was based on agreement by multilateral organizations All 10 decision-point countries are expected to reach the and governments to offer a fresh start to countries that completion point by the end of 2007. The 11 remaining were making efforts to reduce poverty by reducing their countries that are already eligible for the HIPC Initiative external debt burdens to sustainable levels. The HIPC Ini- are referred to as the “pre-decision” countries. All 11 tiative was enhanced in 1999 to provide deeper and faster countries are expected to reach the completion point by debt relief to a larger group of countries and to increase the end of 2010.* the links with poverty reduction efforts in those countries. The HIPC initiative is estimated to cost about $41 There are currently 40 countries eligible for the HIPC billion in debt relief to the 29 countries that have reached Initiative, 33 of which are in Sub-Saharan Africa. So far the decision point, measured in net present value terms at 29 countries have reached the “decision point” at which the end of 2004. Most of the debt relief will be provided donors make a commitment to provide the debt relief nec- by multilateral creditors (50 percent) and official bilateral essary to meet a specified debt ratio. The Republic of creditors (47 percent). Commercial creditors (3 percent) Congo reached the decision point in March 2006. Of have played a relatively minor role. these, 19 have reached the “completion point,” at which *See World Bank 2006b (p. 20 Annex 2.3) for a list of estimates for they receive irrevocable debt relief. Honduras, Rwanda, completion-point dates. Estimated costs of the HIPC Initiative $ billions, net present value at end-2004 Completion point (18 countries) Decision point (11 countries) Total (29 countries) Multilateral creditors 14.5 5.8 20.3 of which: World Bank 7.0 2.3 9.3 IMF 2.2 0.8 3.0 AfDF/AfDB 1.9 1.5 3.4 IDB 1.3 0.0 1.3 Other 2.1 0.9 3.0 Official bilateral creditors 12.3 7.0 19.3 of which: Paris Club 8.9 5.8 14.7 Other 3.3 0.3 3.7 Commercial creditors 0.7 0.8 1.5 Total 27.5 13.6 41.1 Sources: World Bank and IMF 2005 (table 2) and World Bank Staff estimates. less than 1 percent in 4 other countries (figure tries only if it does not displace other components 3.6).14 There are also large differences between of foreign aid. As with the more general case of countries’ debt-service burdens. In 2005, debt- debt relief mentioned above, it is difficult to assess service payments exceeded 5 percent of GDP in 4 whether HIPC debt relief has been additional in countries, but was less than 1 percent in 4 other the absence of an explicit counterfactual showing. countries (figure 3.7). This reflects the fact that The share of ODA allocated to the 29 decision- some countries had higher debt-service burdens point HIPCs has increased substantially over the prior to HIPC debt relief and that some countries past few years, rising from 19 percent in 1999 to Delivered by The World Bank e-library to: received more HIPC debt relief than others. The World Bank28.5 percent in 2004. This suggests that HIPC Debt relief provided under the HIPC IP Initiative debt relief has not displaced other components of : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 will free up additional resources in recipient coun- ODA. However, the share of ODA allocated to 88 (c) The International Bank for Reconstruction and Development / The World Bank S U P P O R T I N G D E V E L O P M E N T T H R O U G H A I D A N D D E B T R E L I E F countries in Sub-Saharan Africa and to the LDCs Figure 3.5 Debt-service payments and HIPC debt increased by even more during this period. service reduction for 28 “decision point” HIPCs % of GDP The Paris Club plays an important role in the 5 HIPC Initiative 4 The Paris Club has made an important contribu- 3 tion to the debt relief provided to HIPCs. Initially, the Paris Club provided cash-flow relief to dis- 2 tressed debtors (debt restructuring), but no debt 1 relief in the sense of reducing the net present value of the debt (box 3.3). However, in the mid-1980s 0 2000 2001 2002 2003 2004 2005 it became apparent that debt burdens in many low-income countries were unsustainable and that HIPC debt service reduction/GDP Debt service paid/GDP debt relief was needed. Beginning in 1988, the Paris Club began providing concessional debt re- Sources: World Bank and IMF 2005 (table 1A) and staff estimates. lief to poor countries, first under Toronto Terms, Figure 3.6 Debt-service reduction provided by the HIPC Initiative to 25 decision-point countries % of GDP 12 10 8 6 4 2 0 oz Ch i bi d a a rit a G ia M ond a ag ras Be r Se nin C am l er a ia R oon o Bu an da m ina nia hi i U nea G er ia Si Pri so M ne Th N da Bo ia G ue Le e al Z ga Et aw ca am a ic n au u H han am bi ra ip op an b liv e ig M & Fa N ya M rag T an o n q ne To rk za am er nc ad u as al ui ga u w G é M Sã Source: World Bank staff estimates. Figure 3.7 Debt service paid by 25 decision-point HIPCs, 2000 versus 2005 % of GDP 2000 2005 15 12 9 6 3 0 am e Bo ia M ia G awi G na au ea am ia Si Za n ra bia Se one ar al ua G ali un a M Ug s M mb a ag ue C r d Bu an er a ia Be o Et in R pia a ca ra e cip oo H han d ha s nd b liv C itan N neg rk zan n T ig M Fa a M uin ag oz an ad iq er m o du as al uy er Le N wa Th rin hi r G P in ic a & é Delivered by The World Bank e-library to: m To The World Bank o Sã IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 Source: World Bank staff estimates. 89 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Box 3.3 The Paris Club T he year 2006 will mark the fiftieth anniversary of the establishment of the Paris Club of Creditors. Histori- cally this informal body has met in Paris to: (i) review the deficiencies that have brought about the need for debt treatment. As a general rule, such a plan takes the form of an economic adjustment program officially external debt-servicing performance of debtor countries; supported by the International Monetary Fund (IMF) (ii) develop rules and mechanisms that may be used to re- although in a few cases the Paris Club did not require solve debt-payment difficulties; and (iii) negotiate debt an IMF program. rescheduling or reduction agreements with debtor coun- 3. Comparable treatment. The debtor country must se- tries. Since 1956, the club’s 19 creditor members (along cure from all other creditors debt relief terms that in- with about a dozen invited creditor countries) have volve treatment comparable to those agreed with the reached more than 400 agreements with debtor countries. Paris Club. Formerly private creditors were not af- Initially, the Paris Club provided only cash-flow relief to fected by the comparability-of-treatment clause. countries experiencing temporary balance-of-payments However, beginning in 1998 (for Pakistan) the Paris difficulties, while maintaining the present value of credi- Club has asked some debtors to obtain comparable tors’ claims. In the past 15 years, however, the club has debt relief from bondholders. engaged increasingly in debt-reduction operations cover- 4. Agreement by consensus. This principle requires that ing not only debt flows but also debt stocks. the Paris Club act only with the concurrence of all of The Paris Club took on greater importance with the its participants. onset of the 1980s debt crisis. The number of agreements 5. Case-by-case approach. Paris Club members reserve the concluded by the club since the early 1980s has been al- right to apply the principles in a flexible manner so as most three times the number reached during the first 25 to meet the particular requirements of a specific debtor. years of its existence. Since 1983, the total amount of debt covered in agreements concluded by the Paris Club or ad In October 2003, the Paris Club adopted a new ap- hoc groups of Paris Club creditors has been $504 billion. proach to treating debt in countries that were not eligible The activities of the Paris Club have been governed for the HIPC Initiative. The Evian Approach was designed by five basic principles: to ensure that debt restructuring was granted only in cases of imminent default and that the debt treatment provided 1. Creditor solidarity. The members of the Paris Club reflected countries’ financial needs and the objective of en- act as a group in their dealings with a particular suring debt sustainability. Debt sustainability therefore debtor country. For debtors this implies that any plays a central role in determining whether and to what country seeking a debt rescheduling from the club extent countries receive debt relief. The adoption of the must agree to treat all its members in the same way; Evian Approach was followed by two major agreements for creditors it implies that club members will refuse that provided record amounts of debt relief. In November to consider a request from a debtor to reschedule debt 2004, the Paris Club agreement with Iraq considered $37 on a purely “bilateral” basis, that is, outside of the billion in debt, canceling $30 billion (80 percent) and Paris Club framework. rescheduling the rest. In October 2005, the Paris Club 2. Commitment to economic reform. Debt rescheduling reached an agreement with Nigeria concerning $30 billion requires an economic policy plan aimed at correcting in debt, $18 billion (60 percent) of which was canceled. which provided for a 33 percent reduction in the try’s debt must exceed certain threshold levels. Ei- net present value of the debt. It soon became evi- ther external debt must be at least 150 percent of dent that even more relief was required to reduce exports, or public debt must be at least 250 percent debt burdens to sustainable levels. The terms of- of revenues (in net present value terms), after receiv- fered by the Paris Club were made more generous ing debt relief from the Paris Club under Naples in a series of steps. In 1991 London Terms allowed Terms.15 Under the HIPC Initiative, countries bene- for a 50-percent reduction in net present value; in fit from debt reduction from all creditors (which in- 1994, Naples Terms allowed for debt relief of as clude the Paris Club and other bilateral official much as 67 percent. creditors, multilateral creditors, and commercial Delivered by The World Bank e-library to: Since 1997, debt relief provided by the Paris Bankcreditors) in an amount that reduces their debt bur- The World Club has been an integral part of the HIPC Initia- den to the threshold levels. In principle, the burden IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 tive. To be eligible for the HIPC Initiative, a coun- of debt relief is to be shared equally among all cred- 90 (c) The International Bank for Reconstruction and Development / The World Bank S U P P O R T I N G D E V E L O P M E N T T H R O U G H A I D A N D D E B T R E L I E F itors. However, participation is voluntary. In prac- though the amounts paid are small relative to the tice, most commercial creditors have not partici- total amount of debt relief committed by the HIPC pated,16 while the Paris Club creditors have pro- Initiative ($38 billion in net present value terms), vided much more than their share of the debt relief. judgments in favor of creditors set a precedent that In most cases, Paris Club creditors have cancelled could lead to more litigation. Debt relief frees up all of the debt owed to them by countries that have financial resources, leading creditors to reassess reached the completion point.17 In contrast, other their chances of obtaining a significant judgment official bilateral creditors have committed so far to in their favor. Thus further debt relief could make less than half of their share of debt relief.18 the litigation strategy even more alluring. HIPC debt relief could lead to more litigation Further debt relief is envisioned under the by commercial creditors HIPC Initiative and the Multilateral Debt Sharing the burden of debt relief equally across all Relief Initiative creditors is complicated by the “collective action” The HIPC initiative will continue to reduce debt- problem. Some commercial creditors have an in- service burdens. In 2006/7, some $2.6 billion in centive to “hold out” of an agreement, preferring relief will be provided annually to the 29 decision- to pursue their claims through litigation in hopes point countries, up from an average of $2.3 bil- of obtaining more favorable terms. In corporate lion provided during 2001–5. Debt service by bankruptcies, the legal system prevents creditors these countries is projected to remain unchanged from engaging in such “free-riding” and imposes in 2006/7 relative to their GDP and exports, but rules for collective action. But in the case of sover- the total amount of debt relief provided under the eign debt restructuring, there is no overriding legal HIPC Initiative will increase over time as addi- system that has such jurisdiction over all creditors. tional countries reach the decision point and com- Hence, collective action cannot be imposed pletion point. through legal means. Some commercial creditors Following on the HIPC initiative, the Multi- have prevailed in litigation against HIPCs. There lateral Debt Relief Initiative (MDRI) will achieve are currently 24 litigation cases on record against further, significant reductions in the debt burden HIPCs, 4 of which were new in 2005; court of poor countries (box 3.4). The MRDI calls for awards to creditors total $586 million, of which complete cancellation of debt owed to the Interna- countries have paid only about $35 million.19 Al- tional Development Association (IDA), the IMF, Box 3.4 The MDRI T he Multilateral Debt Relief Initiative (MDRI) was pro- posed in June 2005 by the G-8 Finance Ministers as a way to free up additional resources to help poor countries Although the MDRI is a common initiative, the ap- proach to coverage and implementation varies somewhat across the three institutions.* The IMF Executive Board with high debt levels make progress toward the Millen- modified the proposal to reflect the Fund’s requirement nium Development Goals. Under the MDRI, three multi- that the use of IMF resources be consistent with unifor- lateral institutions—the International Development Associ- mity of treatment. Thus, it was agreed that all countries ation (IDA), the International Monetary Fund (IMF), and with per capita income of $380 a year or less (HIPCs the African Development Fund (AfDF)—will cancel all and non-HIPCs) would receive MDRI debt relief fi- claims on countries that reach the completion point under nanced by the IMF’s own resources. Two non-HIPCs— the HIPC initiative. The IMF and IDA have approved debt Cambodia and Tajikistan—were certified as eligible for relief under the MDRI for 17 of the 18 HIPCs that have al- MDRI debt relief from the IMF on this basis. HIPCs ready reached the completion point. The exception, Mauri- with per capita income above that threshold would re- tania, will qualify for debt relief under the MDRI after im- ceive MDRI relief from bilateral contributions adminis- plementing key public expenditure management reforms. tered by the IMF. (Approval by the AfDF is expected to come in April Delivered 2006.) by The World Bank e-library to: The World Bank *See World Bank (2006a) for a more detailed discussion of the implementation of the MDRI. IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 91 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 and the African Development Fund (AfDF) by Figure 3.8 Debt-service reduction to be provided countries that reach the HIPC completion point. by the MDRI, 2006–45 The process of reaching the HIPC completion $ billions point includes conditions relating to governance, 2.5 accountability, and transparency. 2.0 The MDRI can be interpreted as an extension and a deepening of the HIPC Initiative. Eligibility 1.5 will require meeting the HIPC completion-point criteria, which include (i) satisfactory macroeco- 1.0 nomic performance under an IMF poverty reduc- tion and growth facility program (PRGF) or equiv- 0.5 alent; (ii) satisfactory performance in implementing a poverty reduction strategy; and (iii) the existence 0.0 2006 2012 2018 2024 2030 2036 2042 2048 2054 of a public expenditure management system that 11 pre-decision-point HIPCs meets minimum standards for governance and 11 decision-point HIPCs transparency in the use of public resources. 18 completion-point HIPCs The objective of the MDRI is to provide addi- Source: World Bank staff estimates. tional support to HIPCs to reach the MDGs, while ensuring that the financing capacity of the interna- tional financial institutions is preserved. Debt stocks The two-humped shape of the debt-service-re- in the 18 countries that reached the HIPC comple- duction profile is due to the fact that the bulk of tion point prior to 2006 will be reduced by an esti- outstanding IMF loans to these countries are mated $17 billion (in net-present-value terms, val- scheduled to mature within three to six years (fig- ued at end-2004), with most of the reduction ure 3.9). Outstanding IDA and AfDB loans have a coming from cancellation of IDA credit repayments much longer duration (extending out to 40 years), of $12 billion (table 3.9).20 If all 11 decision-point so the debt-service-reduction profile is much more countries were to reach the completion point by the gradual once the IMF loans have disappeared end of 2007, the total amount of debt relief would from the picture. be almost $22.4 billion, an amount equal to 56 per- The MDRI will affect flows of assistance from cent of the debt relief provided under the HIPC ini- IDA and the AfDF to recipient countries in two tiative to the same set of countries ($40 billion). ways. First, annual gross assistance from IDA and For the 18 HIPCs that reached the completion the AfDF to a given country will be reduced by the point prior to 2006, the MDRI will reduce debt ser- amount of debt relief provided that year. Second, vice payments by $0.9 billion on average in 2007–17 and then rise to a peak of $1.5 billion on average in 2022–4 (figure 3.8). The total amount of debt relief Figure 3.9 Debt-service reduction to be provided provided by the MDRI will rise over time as addi- to 18 completion-point HIPCs under the MDRI, tional countries reach the completion point.21 The 2006–45 modest increase in 2006 reflects the fact that the $ billions MDRI will not be implemented by IDA until July 2.0 IDA 2006 (the beginning of its fiscal year). 1.5 Table 3.9 Debt-service reductions to be provided by the MDRI $ billions, net present value at end-2004 1.0 Completion-point Decision-point Total for Pre-decision point Total for countries (18) countries (11) 29 countries countries (9) 38 countries 0.5 IMF AfDF IDA 12.1 2.8 14.9 1.2 16.1 IMF 2.8 1.4 4.2 0.3 4.5 AfDF 2.3 1.1 3.4 Delivered 0.3 by The World 3.6 Bank0.0 e-library to: Total 17.2 5.3 22.4 1.8 The World 24.2 Bank 2006 2011 IP : 192.86.100.36 2016 2021 2026 2031 2036 2041 Tue, 10 Mar 2009 16:55:33 Source: World Bank staff estimates. Source: World Bank staff estimates. 92 (c) The International Bank for Reconstruction and Development / The World Bank S U P P O R T I N G D E V E L O P M E N T T H R O U G H A I D A N D D E B T R E L I E F donors will make additional contributions to com- Table 3.10 Donors’ commitment to refinance IDA for debt relief pensate IDA and the AfDF for the total reduction provided under the MDRI, selected years $ billions in gross assistance flows in each year. Donors have specified that the additional contributions are to 2006 2007 2010 2020 2030 be calculated relative to a baseline that maintains Baseline for IDA replenishments 5.0 5.1 5.4 6.6 8.0 current contribution levels in real terms (adjusted Compensation for reduction in gross assistance flows (equal to debt service reduction) 0.8 1.2 1.6 1.7 1.6 for inflation). Under the current replenishment of Total financing commitments 5.8 6.3 7.0 8.3 9.7 IDA (IDA14) donors have agreed to make contri- butions of almost $15 billion between July 2006 Source: World Bank staff estimates. and June 2008, or about $5 billion per year. Debt relief on IDA loans under the MDRI will be fi- nanced by donors over and above the $5 billion with providing debt relief to countries with the level, measured in real terms to compensate for the highest debt burdens. In other words, debt relief effect of inflation.22 If the annual inflation rate provided under the MDRI will result in an in- were constant at 2 percent, the baseline contribu- crease in aid (above countries’ initial allocation), tion level would rise to $5.4 billion in 2010 and only to the extent that the country shares in the $8.0 billion in 2030 in nominal terms (figure 3.10 performance-based allocation. Countries that do and table 3.10). Donors’ commitment to compen- not qualify for debt relief under the MDRI may sate IDA for the total reduction in gross assistance qualify for the reallocated resources and thereby flows is equal to the debt service reduction pro- benefit from the initiative. vided to the recipient countries (figure 3.8). Although the amount of debt relief provided Donors’ total financing commitment comprised of under the HIPC Initiative has been small relative compensation for the effect of inflation and for the to the total amount of foreign aid received by all reduction in gross assistance flows rises to $7.0 developing countries, it is substantial for many of billion in 2010 and $9.7 billion by 2030. Donors’ the individual countries that qualify. In 2004, commitment to preserve financing of the AfDF is HIPC debt-service reductions provided to the 27 specified in a similar manner. countries that reached the completion point prior The additional resources provided to refi- to 2005 totaled $2.3 billion, an amount equal to nance IDA and the AfDF will be reallocated to re- just 3 percent of total ODA ($79.6 billion), but cipients using each institution’s existing perfor- 12 percent of ODA received by the 27 countries mance-based allocation mechanism, thereby ($18.6 billion). Moreover, HIPC debt-service re- alleviating the risk of “moral hazard” associated ductions exceeded 20 percent of ODA received by 8 of the 27 countries. Additional debt service re- ductions provided by the MDRI are expected to Figure 3.10 Donors’ commitment to refinance IDA for debt relief provided under the MDRI, 2006–45 keep pace with the scaling up of aid to the HIPCs. In 2007, debt-service reductions provided by the $ billions HIPC Initiative and MDRI combined are projected 12 to remain at about 12 percent of the amount of 10 ODA received by countries that reach the comple- 8 tion point. 6 A gap is opening between countries that 4 qualify for debt relief and those that do not Taken together, debt relief provided by the HIPC 2 Initiative and the MDRI will substantially reduce 0 the debt burdens of qualifying countries. For the 2006 2011 2016 2021 2026 2031 2036 2041 18 countries that reached the completion point Compensation for reduction in gross assistance prior to 2006, the HIPC Initiative reduces their flows (equal to debt service reduction) Delivered by The World Bank e-library to: Baseline for IDA replenishments The debt stock from 55 percent of their GDP to totalBank World 30 percent; the MDRI then reduces it further to 13 IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 Source: World Bank staff estimates. percent (in net present value terms). In 4 of the 18 93 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 countries the debt-to-GDP ratio will decline by compared to 99.5 percent). All 11 countries are ex- more than 90 percentage points. Debt stocks will pected to reach the completion point by the end of fall below 30 percent of GDP in all countries ex- 2007, which will reduce their debt burdens signifi- cept one (Guyana), and in 10 of the 18 countries, cantly (to less than 15 percent of GDP and 50 per- debt will fall below 10 percent of GDP, well below cent of exports in most cases). For the 11 countries the average for developing countries ([32] percent that are currently eligible for the HIPC Initiative but in 2005) (figure 3.11).23 Similar reductions would have not yet reached the decision point, the median result for other countries that reached the HIPC debt burden was 67.1 percent of GDP and 150.4 completion point. percent of exports in 2004. These countries there- For the 11 HIPCs that have reached the deci- fore have a very strong incentive to reach the deci- sion point, but not the completion point, the median sion and completion points, in order to qualify for debt burden was 41.2 percent of GDP in 2004 (in debt relief under the HIPC Initiative and the MDRI. present value terms), which is below that for middle- income countries (44.8 percent) (table 3.11). Similar Debt relief raises concerns about excessive results hold for the low-income countries that are borrowing in the future not currently eligible for the HIPC Initiative (the The low debt burdens in countries receive MDRI “other low-income countries” reported in table debt relief will improve their creditworthiness sig- 3.11). Relative to exports, however, the debt burden nificantly, raising concerns that they might borrow in the 11 decision-point HIPCs is significantly higher excessively from nonconcessional sources. This than in other low-income countries (183.7 percent could offset the efforts made to improve debt sus- tainability, leading to yet another lending-forgive- Figure 3.11 Debt burdens in 18 completion-point HIPCs, before and ness cycle. But why would countries borrow “ex- after the HIPC and MDRI debt relief cessively”? And why would private creditors be willing to lend “excessively”? % of GDP Determining whether countries are borrowing 150 Before HIPC & MDRI debt relief excessively is not straightforward. Loans used to After HIPC & MDRI debt relief finance investment projects that generate revenues 100 will not erode debt sustainability if the rates of re- turn cover the cost of financing. From this per- spective, debt sustainability is determined by the 50 quality of the investments made, not by the quan- tity borrowed. 0 Accessing external private capital entails sig- ar a au ua ia H livia am as Za i Se ia l ag ia r er na U in e Bu a Ta ina ia da al ga ca an qu nd nificant risks, but it also provides potential bene- an b p n an ig M r ag ha an m ne M thio Be rk M ndu as uy N bi ga Bo rit nz w G G R E o ic fits. Financial crises have led to major setbacks in ad N oz M many emerging market economies over the past Source: World Bank staff estimates. few decades. On the other hand, external private capital can play a valuable role in the development Table 3.11 Net present value of external debt relative to GNI and process, particularly for countries in which domes- exports, 2004 tic savings are inadequate to finance productive in- Percent vestment projects with high private and social Number of Median of present Median of present rates of return. Countries therefore face the chal- countries value of debt/exports value of debt/GNI lenge of balancing the potential risks and benefits. Completion-point HIPCs 18 Part of the concern about excessive noncon- after HIPC debt relief, prior to MDRI 102.2 27.6 cessional borrowing stems from the incentive after HIPC debt relief and MDRI 41.1 8.6 Decision point HIPCs 11 183.7 41.2 problems associated with providing publicly Pre-decision-point HIPCs 9 150.4 67.1 funded debt-relief initiatives. The public funding Other low-income countries 18 99.5 46.3 introduces an element of moral hazard into bor- Middle-income countries 76 98.0Delivered by The44.8World Bank e-library to: The World Bankrowing and lending decisions. If borrowers and Sources: World Bank Debtor Reporting System and staff estimates. IP : 192.86.100.36lenders perceive publicly funded debt relief as an Tue, 10 Mar 2009 16:55:33 94 (c) The International Bank for Reconstruction and Development / The World Bank S U P P O R T I N G D E V E L O P M E N T T H R O U G H A I D A N D D E B T R E L I E F ongoing feature of the development agenda, coun- that puts more weight on debt sustainability, with tries have an incentive to increase borrowing be- the aim of preventing a recurrence of lending- yond prudent levels, under the expectation that forgiveness cycles. debt relief will be provided by donors if they en- The Debt Sustainability Framework (DSF) for counter difficulties in meeting their debt-service low-income countries developed jointly by the IMF obligations. Similarly, some investors might be- and the World Bank provides a framework for lieve that their exposure to poor countries is re- managing the risks associated with additional bor- duced by an implicit guarantee of publicly funded rowing (box 3.5). The DSF captures the distinction debt relief, which would limit their downside risk, between concessional and nonconcessional bor- making them willing to lend at a lower rate. rowing by measuring debt in net present value Another factor underlying the concern about terms. Nonconcessional borrowing raises the debt excessive nonconcessional borrowing stems from burden by more (in net present value terms) for the the inherent trade-off in scaling up the financial re- same amount of financial resources. In other sources required to accelerate progress on MDGs words, borrowing on concessional terms improves while maintaining debt sustainability. Grants pro- the trade-off between debt sustainability and re- vide countries with financial resources without source flows. More generally, it is the overall de- sacrificing debt sustainability. But the availability gree of concessionality in a country’s loan portfolio of grants is limited. Loans provide additional fi- that determines how many more resources can be nancial resources, but raise the risk of debt dis- provided without sacrificing debt sustainability. tress, particularly when loans are made on non- The DSF can be used to assess the risks associ- concessional terms. Countries may be more willing ated with additional borrowing. Assessing the risks to accept a higher risk of debt distress in order to are complicated by the high degree of uncertainty gain additional resources. Official creditors may surrounding economic projections over long time prefer a more prudent approach to borrowing; one horizons (measured in decades), especially when the Box 3.5 The DSF for low-income countries T he International Monetary Fund (IMF) and World Bank jointly assess debt sustainability in countries that receive credits and grants from the International De- The external debt burden of each country is assessed over the projection horizon with reference to threshold levels that depend on the quality of a country’s policies velopment Association (IDA) and that are eligible for re- and institutions. The World Bank’s Country Policy and In- sources under the IMF’s Poverty Reduction and Growth stitutional Assessment (CPIA) is used to classify countries Facility (PRGF). The DSF is used by IDA and the African into three performance categories (strong, medium, and Development Fund (AfDF) to allocate credits (not loans) poor). Debt thresholds for strong policy performers are to countries. It is also used by the Paris Club to help deter- highest. The risk of external debt distress is then assessed mine whether a country’s debt is sustainable and, if it is with reference to four risk classifications: low, medium, not, how much debt relief would be required to attain high, and “in debt distress.” Empirical studies indicate debt sustainability over the long term. that low-income countries with better policies and institu- The objective is to monitor the evolution of countries’ tions have a lower risk of debt distress (see IMF and debt-burden indicators and to guide future financing deci- World Bank 2004 and the references therein). sions. The DSF traces the evolution of external and public The risk classifications do not fully capture the com- debt and debt-service indicators over the long term with ref- plexity of the assessment. For example, in cases where the erence to a baseline projection based on realistic assump- various indicators give different signals, there is still a tions. Stress tests are conducted to illustrate the implications need for careful interpretation and judgment. Further- of adverse shocks to key macroeconomic variables (typi- more, vulnerabilities related to domestic public debt cally lower growth, higher interest rates, and an exchange should also be taken into account. The past record in rate depreciation), along with other selected scenarios of meeting debt-service obligations may also be a factor in specific interest to the country under study (for example, an determining the classification, especially for countries at Delivered increase in a contingent liability of the public by The World Bank sector). e-library high to: or moderate risk of debt distress. The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 95 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 additional borrowing is used to fund projects and been declining for several years, resulting in lower programs that have the potential to enhance eco- debt service costs. Since 1990, repayments on out- nomic growth significantly over the long term. standing loans have exceeded disbursements of Managing the risks by setting limits on additional new loans by 0.4 percent of GDP on average. borrowing would require agreement among all cred- Net private debt inflows to 27 HIPCs that itors on the assessment of debt sustainability and the reached the decision point before 2005 contracted degree of risk to be tolerated. In the case of the main by $0.75 billion (0.7 percent of GDP) on average multilateral creditors, there is typically some scope over the period 2000–3, before rebounding to $0.5 for agreement on the major issues, and moral sua- billion in 2004 (0.4 percent of GDP).26 The rebound sion can be used to enforce limits on additional bor- in 2004 was lower than in other low-income coun- rowing. However, reaching agreement among all tries, where net private debt inflows increased from prospective creditors is generally be problematic, an average level of $0.4 billion (0.05 percent of and, under such conditions, moral suasion is likely GDP) in 2000–3 to $7.2 billion (0.7 percent of to be ineffective in enforcing borrowing limits. For GDP) in 2004. The rebound in 2004 was mainly countries with IMF programs, the collective-action concentrated in 4 of the 29 decision-point HIPCs: problem is addressed by setting limits on additional Tanzania ($168 million, 1.5 percent of GDP), borrowing, which help ensure that additional re- Honduras ($151 million, 2.0 percent of GDP), source flows do not endanger debt sustainability. Cameroon ($133 million, 0.9 percent of GDP), and However, for countries without an IMF program, it the Democratic Republic of the Congo ($88 mil- will be difficult to monitor and set limits on non- lion, 1.3 percent of GDP). The private debt burden concessional borrowing. of the 27 countries as a group has declined signifi- What has been the experience so far for coun- cantly, falling from more than 20 percent in the tries that have already received HIPC debt relief? early 1990s to less than 9 percent in 2004, compa- Has their borrowing increased significantly? rable to the level in other low-income countries but well below that for middle-income countries (25 Net official lending to decision-point HIPCs percent in 2004). Private debt exceeded 20 percent has been stable of GDP in only 2 of the 27 countries in 2004 Net concessional lending from the official sector (Nicaragua at 25 percent and Mozambique at 23.4 to the 27 HIPCs that reached the decision point percent), while 13 countries recorded ratios of pri- prior to 2005 declined significantly in the mid- vate debt to GDP of under 5 percent. 1990s (figure 3.12).24 The transitory increase in International credit-rating agencies have re- 2002 was partly due to a resumption in conces- cently begun issuing sovereign debt ratings for sionary lending to the Democratic Republic of the some low-income countries. Credit ratings en- Congo (DRC) in 2002.25 Non-concessional lend- hance transparency and help private investors as- ing from the official sector to the 27 countries has sess the risk of holding sovereign debt. Thirteen of the 29 decision-point HIPCs are currently rated by international agencies (table 3.12). Benin, Ghana, Figure 3.12 Net official lending to 27 decision- and Senegal are rated B+ by Standard and Poor’s; point HIPCs as a percent of GDP, 1990–2004 Ghana and Mozambique are rated B+ by Fitch. Percent These ratings, the highest among low-income countries, are three notches below investment Concessional lending grade, making it difficult for countries to expand 4 their access to international bond markets. Bank Net official lending loans and short-term debt account for most of the 2 outstanding private debt (90 percent in 2004) is- sued by the 29 countries. Medium- and long-term 0 bonds account for a negligible portion, less than Non-concessional lending 0.1 percent in 2004, down from almost 4 percent Delivered by The World Bank e-library to: The World Bankin 1993. In 2004, net inflows of medium- and –2 1990 1992 1994 1996 1998 2000 2002 2004 long-term bonds to the 29 countries totaled only IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 Source: World Bank Debtor Reporting System. $345 million (0.3 percent of GDP), and were con- 96 (c) The International Bank for Reconstruction and Development / The World Bank S U P P O R T I N G D E V E L O P M E N T T H R O U G H A I D A N D D E B T R E L I E F Table 3.12 Credit ratings for decision-point HIPCs the fact that HIPCs are required to establish a Rating of foreign currency long-term debta track record of macroeconomic stability in order Moody’s Standard & Poor’s Fitch to reach the decision point. Real GDP growth in- Benin B+ B creased by only half of a percentage point during Bolivia B3 B B– this period in the 11 countries that are eligible for Burkina Faso B the HIPC Initiative but had not yet reached the de- Cameroon CCC B– Ghana B+ B+ cision point by the end of 2004 (these are the “pre- Honduras B2 decision-point HIPCs” in table 3.13). Clearly, Madagascar B growth has been influenced by many factors be- Malawi CCC Mali B B– side debt relief. Mozambique B B+ According to the “debt overhang” hypothesis, Nicaragua Caa1 excessive debt can seriously impede countries’ Senegal B+ Uganda B growth potential.27 Much of the theoretical litera- ture has focused on the adverse incentive effects of Sources: Moody’s, Standard & Poor’s, and Fitch. a. As of March 8, 2006. excessive debt. Excessive debt raises concerns that the government may resort to inflationary finance or large tax increases to meet its debt-service centrated in just three countries: Honduras ($162 obligations or that it may default on its obligations million, 2.2 percent of GDP), Senegal ($92 mil- at some point in the future. These concerns deter lion, 1.2 percent of GDP), and Ethiopia ($71 mil- private investment, which curtails growth. More- lion, 0.9 percent of GDP). The creditworthiness of over, in countries that are unable to meet their HIPCs that reach the completion point will be en- debt-service obligations, governments can be dis- hanced by further debt relief under the MDRI. couraged from carrying out structural reforms if However, other factors such as the quality of pol- most of the benefits were used to augment debt- icy and institutional frameworks, and political service payments. risk, will continue to have an important influence The theoretical literature suggests that exter- on credit ratings by international agencies. nal borrowing may foster growth up to some threshold level, beyond which adverse incentives The decline in debt service burdens is begin to dominate. But empirical research on this supported by stronger economic growth issue has been inconclusive, on the whole. There is Growth has picked up over the past few years in a high degree of uncertainty surrounding estimates most HIPCs, helping reduce their debt service bur- of threshold levels and the effect of debt relief on den, measured relative to GDP (table 3.13). Real growth. Recent empirical studies by Clements and GDP growth in the 27 HIPCs that reached the de- others (2003) and Pattillo and others (2004) sug- cision point before 2005 averaged 4.6 percent over gest that the amount of debt relief provided by the the period 2000–5, up considerably from an aver- age rate of 2.6 percent in the 1990s and just 1.8 percent in the 1980s. The pickup in growth has Table 3.13 Average annual real GDP growth, 1990–2005 been broadly based across countries—real GDP Percent growth exceeded 4 percent in 16 of 27 decision- Average real GDP growtha point HIPCs in 2000–5. No. It is important to recognize, however, that the of countries 1980–9 1990–9 2000–5 range of outcomes was broad—annual per capita Decision-point HIPCsb 27 1.8 2.6 4.6 real GDP growth declined in 9 of the 27 countries. Pre-decision-point HIPCsc 11 2.6 1.7 2.2 Moreover, the average increase in real GDP growth Other low-income countries 19 4.7 2.7 4.6 Middle-income countries 77 3.5 2.8 4.8 in 27 decision-point HIPCs over the sub-periods 1990–9 versus 2000–5 (1.9 percentage points) was Source: World Bank Debtor Reporting System and staff estimates. a. Real GDP growth rates are first averaged over indicated sub-periods for each country and the same as in “other low-income countries” then unweighted averages Delivered by The World Bank e-library to: are calculated across countries. (countries that currently are not eligible for theThe Worldb. Burundi Bank and Congo reached the decision point prior to 2005 and are therefore classifed as pre-decision-point HIPCs for the purpose of these calculations. HIPC Initiative) and in middle-income countries. IP : 192.86.100.36 Real GDP c. 2009 data is unavailable for 2 of the 11 pre-decision-point HIPCs (Myanmar and Tue, 10 Mar 16:55:33 Furthermore, the increase in growth also reflects Somalia). 97 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 HIPC Initiative should raise countries’ annual per The external position of the 29 decision-point capita real GDP growth rates by about 1 percent- HIPCs has also been strengthened by an expansion age point.28 That estimate is broadly consistent of trade. More open economies are better able to with recent trends—annual per capita real GDP adjust to external shocks. Exports by the 29 coun- growth increased by about 2 percentage points tries as a group have increased from 20 percent of on average for the HIPCs over the periods just GDP in the early 1990s to almost 30 percent in before and just after reaching their respective de- 2005, but the figure remains well below the level cision points.29 in middle-income countries (estimated at 40 per- The “debt overhang” literature stresses that cent in 2005). debt relief can strengthen incentives to promote Non-debt-creating resource flows, notably domestic investment and structural reforms. Pro- from FDI, workers’ remittances, and foreign aid, viding more aid in the form of grants in place of can help countries meet their external financing concessionary loans can provide the same incen- needs by generating a relatively stable stream of tive effects (because it reduces the net present foreign exchange earnings. FDI and remittance in- value of debt), but this may not be the case for flows to the 29 decision-point HIPCs as a group greater aid in the form of grants allocated to re- have risen considerably since the early 1990s duce debt-service payments. In the absence of a (table 3.14). FDI and remittance inflows provide credible multiyear commitment, recipient coun- important sources of external finance to most tries face uncertainty about their ability to use countries, with FDI inflows exceeding 3 percent of grants to service debt. The irrevocable nature of GDP in one-half of the countries, and remittances HIPC debt relief upon reaching the completion exceeding 3 percent of GDP in about one-third. point provides such a commitment. ODA has risen from a low of 12 percent of GDP Debt burdens are not the only indicator of in 29 decision-point HIPCs as a group to 20.5 per- sustainability; other factors are important, as well. cent in 2003–4, which is comparable to the level Episodes of debt distress often have occurred in received in the early 1990s. emerging market economies with moderate, or even low, debt.30 Moreover, adverse shocks to eco- Sizable external and fiscal imbalances remain nomic growth and the terms of trade have had a The current-account deficit for the decision-point greater influence on debt burdens in low-income HIPCs as a group narrowed from 8.9 percent of countries than has the amount of borrowing un- GDP in 1999 to 5.1 percent in 2005. But large im- dertaken (IMF 2003). Various indicators of coun- balances remain in some countries: deficits exceed tries’ external positions can provide additional in- 10 percent of GDP in one-third of the 29 coun- sights into debt sustainability. tries. Those countries still rely heavily on external financing, making them vulnerable to external Debt sustainability in some countries has shocks. In 2005, current-account deficits widened been enhanced by reserve accumulation, by more than 3 percent of GDP in 6 of the 29 higher exports, and higher inflows of FDI, countries, mainly due to higher oil-import bills. remittances, and aid The value of oil imports increased from 3.5 per- Foreign reserves enable countries to meet their cent of GDP in decision-point countries in 2002 to debt-service obligations in the event of adverse fi- 7.6 percent in 2005. nancial or economic developments, thereby reduc- The analysis to this point has focused mainly ing of the risk of a liquidity crisis. Reserves in the on external debt burdens. However, there is grow- 29 decision-point HIPCs as a group have increased ing concern about fiscal imbalances and rising do- substantially since the early 1990s, rising from 2.6 mestic debt burdens in some countries. Data limi- percent of GDP in 1990 to a high of 13.3 percent tations make this issue difficult to analyze. in 2004, before declining to 11.9 percent in 2005 Nonetheless, the available data indicate cause for (table 3.14). In 2004, reserves provided cover for concern in some HIPCs. General government bud- more than six months of imports in one-third of get balances have improved over the past few Delivered by The World Bank e-library to: the countries, whereas in 1990 none of the The coun- World Bankyears for the decision-point HIPCs as a group, tries had enough reserves to cover six months of reaching –3.2 percent of GDP in 2005, up from IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 imports. –4.3 percent in 2003. However, fiscal deficits ex- 98 (c) The International Bank for Reconstruction and Development / The World Bank S U P P O R T I N G D E V E L O P M E N T T H R O U G H A I D A N D D E B T R E L I E F ceed 5 percent of GDP in 8 of the 26 countries for Table 3.14 Indicators of external position of the 29 decision-point which data are available. Gross domestic debt is- HIPCs, 1990–2005 Percent sued by the public sector increased by more than 5 percent of GDP over the period 1998 to 2004 in 4 1990 1995 2000 2001 2002 2003 2004 2005 of the 11 HIPCs where data are available.31 In Foreign reserves/GDP 2.6 6.8 8.2 8.9 11.5 12.8 13.3 11.9 2004 gross domestic debt exceeded 20 percent of Exports/GDP 21.1 23.9 28.3 25.3 24.9 25.6 28.4 29.5 Current account/GDP –4.8 –6.0 –7.1 –7.0 –7.9 –5.9 –5.4 –5.1 GDP in 5 of the 11 HIPCs. For countries where FDI/GDP 0.5 1.8 3.6 3.7 4.0 3.3 3.1 — the public debt burden is high and rising, the gains Remittances/GDP 1.2 1.4 2.0 2.3 2.5 2.7 2.8 — in debt sustainability provided by HIPC debt relief ODA/GDP 21.2 18.1 15.3 17.2 16.0 20.3 20.7 — have been eroded by financing public debt in the Sources: World Bank Debtor Reporting System and staff estimates. domestic market. vided to the countries. Countries can enhance debt sustainability by building up foreign reserves to The challenge ahead: accessing levels that provide adequate insurance against ex- external capital, while maintaining ternal shocks, and by pursuing macroeconomic debt sustainability policies that aim to maintain a low and stable in- L ow-income countries, HIPCs and non-HIPCs alike, face the challenge of balancing the po- tential risks of external borrowing against the ben- flation environment, along with a sound fiscal framework. Debt sustainability can also be en- hanced by implementing structural reforms de- efits. The debt burden is an important factor in as- signed to improve institutional frameworks. This sessing those risks, but it is not the only factor. includes initiatives aimed to promote trade, FDI, Much of the buildup in the debt burden in the and remittance inflows; advance export diversifi- HIPCs from the mid-1980s to the mid-1990s can cation; augment capacity for debt management; be explained by their weak policy and institutional raise fiscal revenue capacity; and improve the in- frameworks, low capacity for debt management, vestment climate through better governance and lack of export diversification, and limited fiscal sound institutions. In addition to helping to main- revenue capacity (Sun 2004). To the extent that tain debt sustainability over the long term, im- these factors have not improved significantly in proving policies and institutional frameworks countries that reach the HIPC completion point, along these lines will play a critical role in improv- debt sustainability will be an ongoing concern, de- ing aid effectiveness and more generally, in helping spite the substantial amount of debt relief pro- countries attain their development objectives. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 99 (c) The International Bank for Reconstruction and Development / The World Bank Annex: Debt Restructuring with Official Creditors T his appendix lists official debt restructuring were to be rescheduled at the appropriated market agreements concluded in 2005. Restructur- rate. Paris Club creditors also agreed to review the ing of intergovernmental loans and offi- external financing needs of the Dominican Repub- cially guaranteed private export credits take place lic in December 2005 in connection with satisfying under the aegis of the Paris Club. These agree- the conditions for the third review under the IMF ments are concluded between the debtor govern- Stand-by Arrangement, with a view to providing ment and representatives of creditor countries. additional relief in 2006, if needed. The terms of Paris Club debt treatments are Honduras In May 2005, the Paris Club credi- recorded in an agreed-upon minute. To make the tors reached agreement on debt reduction for Hon- terms effective, debtor countries must sign a bilat- duras, which had reached its completion point eral implementing agreement with each creditor under the enhanced HIPC Initiative on April 5, (see box 3.3). 2005. Of the $1.474 billion due to the Paris Club Burundi On July 29, 2005, Burundi reached creditors as of March 31, 2005, $1.171 billion was its decision point under the enhanced Initiative for treated on Cologne terms (debt reduction to 90 per- Heavily Indebted Poor Countries (enhanced HIPC cent of the net present value [NPV] of eligible exter- Initiative). In accordance with the agreement nal debt), of which $206 million was cancelled as reached in March 2004,32 Paris Club creditors in- the Paris Club share of the effort in the enhanced creased debt reduction to 90 percent of the net HIPC Initiative, $110 million was rescheduled, and present value of eligible external debt (Cologne $855 million was cancelled on a bilateral basis. As a terms), from 67 percent (Naples terms, or “tradi- result of the agreement and additional bilateral as- tional relief”), for maturities falling due between sistance, Honduras’ debt to Paris Club creditors July 29, 2005, and December 31, 2006. was reduced from $1,474 million to $413 million. Dominican Republic In October 2005, the Kyrgyz Republic In March 2005, Paris Club Paris Club creditors reached agreement with the creditors agreed with the government of the Kyr- Dominican Republic to consolidate around $137 gyz Republic to a reduction of its public external million of debt service payments falling due in debt. The comprehensive debt treatment under the 2005, of which $50 million related to ODA loans. Evian Approach covered $555 million of debt due The rescheduling was conducted according to to the Paris Club creditors as of March 1, 2005, of “classic terms,” whereby claims are to be repaid which $124 million was cancelled and $431 mil- progressively over 12 years, including a 5-year lion rescheduled. According to the agreed resched- grace period, with 14 semi-annual repayments in- uling terms, non-ODA commercial credits were creasing from 5.5 percent of the amount resched- cancelled by 50 percent ($124 million) and the re- uled to 9.08 percent. ODA loans were to be maining 50 percent will be repaid over 23 years, rescheduled at interest rates at least as favorable as with a 7-year grace period at the appropriate mar- Delivered by The World Bank e-library to: the original concessional rates and no higher than Bankket rate. ODA credits ($306 million) will be repaid The World the appropriate market rate, and non-ODA loans over 40 years with a 13-year grace period at inter- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 100 (c) The International Bank for Reconstruction and Development / The World Bank S U P P O R T I N G D E V E L O P M E N T T H R O U G H A I D A N D D E B T R E L I E F est rates at least as favorable as the concessional all creditors. Participation by Paris Club members rates applying to these loans. Moratorium interest was voluntary, although a majority of the group’s due under the agreement will be capitalized at 85 creditors agreed to accept the prepayment offer. percent in 2005, 75 percent in 2006, 70 percent in Poland In January 2005, Poland announced 2007 and 65 percent in 2008. The capitalized in- its intention to prepay portions of its €12.3 billion terest amounts will be repaid over 23 years includ- debt falling due to the Paris Club between 2005 ing a 7-year grace period. and 2009. Although Poland had prepaid around Nigeria The October 2005 debt deal with €4.5 billion of its Paris Club debt by end-May Nigeria was the single largest debt relief granted to 2005, because the prepayment was financed by any African country, effectively providing debt sovereign bond issues, it did not contribute to any cancellation estimated at $18 billion (including appreciable reduction of external debt. However, moratorium interest), which represents about 60 the deal lengthened the average maturity terms of percent of its debt owed to Paris Club creditors Poland’s external debt, removing the bulge in the (an overall reduction in its debt stock by an esti- country’s debt repayments in 2005–9 and reducing mated $30 billion). This Paris Club agreement was refinancing risk. made possible following the achievement by Nige- Russian Federation In May 2005, the Paris ria of (1) progress in pursuing an ambitious eco- Club creditors agreed on the Russian Federation’s nomic reform program, which aims to accelerate offer to prepay $15 billion of its debt at par. Par- growth and reduce poverty; (2) the World Bank’s ticipation by Paris Club members was voluntary, reclassification of the country from “blend” to although an overwhelming majority of the group’s “IDA only,” paving the way for Paris Club credi- creditors agreed to participate. This prepayment tors to grant debt relief along the Naples terms; offer translates into major interest savings for Rus- and (3) negotiation of an agreement with the IMF sia and is the largest such offer by a debtor coun- for a non-lending Policy Support Instrument (PSI), try to the Paris Club creditors. which formalizes continuing IMF surveillance. Rwanda The Paris Club creditors agreed on The debt relief agreement was to be imple- 100 percent cancellation of Rwanda’s debt in May mented in two phases in consonance with the 2005, following a month after Rwanda reached its implementation of the IMF PSI approved on Oc- completion point under the enhanced HIPC Initia- tober 17, 2005. In the first phase, Paris Club tive. Around $90 million in debt due to Paris Club creditors grant a 33 percent cancellation of eligi- creditors as of March 31, 2005, was treated on ble debts after payment of arrears estimated at Cologne terms (debt reduction to 90 percent of the $6.3 billion by Nigeria. In the second phase, NPV of eligible external debt), of which $82.7 mil- after approval of the first review under the PSI lion ($61.7 million in ODA loans and $21 million by the IMF and repayment of post-cutoff-date in non-ODA commercial credits) was cancelled as debt, Paris Club creditors would grant an addi- the Paris Club share of the effort in the enhanced tional tranche of cancellation of 34 percent on HIPC Initiative. A further $7.7 million in ODA eligible debts and Nigeria will buy back the re- loans was to be cancelled as a result of additional maining eligible debt. Paris Club creditors are to debt relief granted by creditors on a bilateral basis. be paid $12.4 billion in total, with $6.3 billion São Tomé and Principe In September 2005, to clear arrears and $6.1 billion for the buyback. the Paris Club creditors reached agreement on the Full implementation of the Paris Club deal, retroactive rescheduling of São Tomé and scheduled to be completed in April 2006 and fol- Principe’s debt service payments falling due be- lowing the IMF first review of the PSI, would re- tween May 01, 2001, and December 31, 2007. duce Nigeria’s total outstanding external debt The treatment was on Cologne terms (cancellation from $35 billion to $5 billion. of 90 percent of the NPV of eligible external debt), Peru In June 2005, the Paris Club creditors with ODA credits to be repaid over 40 years with agreed on Peru’s offer to prepay up to $2 billion of a 16 year-grace period. its non-ODA debt falling due between August Zambia In May 2005, the Paris Club creditors Delivered by The World Bank e-library to: 2005 and December 2009. Under the agreement,The World agreed to reduce Zambia’s debt stock under the En- Bank prepayment would be made at par and offered to hanced HIPC Initiative, a month after Zambia had IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 101 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 reached its completion point. Of the total $1.92 2. OECD DAC data on the allocation of ODA across income classifications and regions in 2005 will not be avail- billion due to the Paris Club creditors as of March able until December 2006. The calculations refer to the por- 31, 2005, $1.763 billion was treated on Cologne tion of ODA that is allocated across income classifications terms (90 percent cancellation rate). Of this latter and regions. In 2000–4, 26 percent of ODA was not allo- amount, $1.403 billion in pre-cut-off date debt cated across income classifications and 17 percent was not ($461 million in ODA loans and $942 million in allocated across regions, on average. 3. The United Nations’ LDC income classification non-ODA commercial credits) was cancelled as the overlaps the World Bank’s Sub-Saharan Africa region, but Paris Club share of the effort in the enhanced HIPC not completely—32 of the 49 LDCs are in the Sub-Saharan Initiative. A further $360 million ($298 million in Africa region; 32 of the 43 low-income countries in the Sub- pre- and post-cutoff-date ODA loans and $62 mil- Saharan Africa region are LDCs. lion in post-cutoff-date non-ODA commercial 4. As measured by the WTO/OECD-DAC Trade Ca- pacity Building Database. credits) was to be cancelled on a bilateral basis. As 5. The collective interim target of 0.56 percent in 2010 a result of the agreement and additional bilateral entails individual targets of 0.51 percent for the 15 “origi- assistance, Zambia’s debt to Paris Club creditors nal” EU countries, along with 0.17 percent targets for the was reduced from $1.92 billion to $124 million. 10 countries that joined the European Union in 2004. Paris Club creditors also agreed to reschedule 50 6. Projections of ODA based on donor commitments are reported by OECD (2006, table 1.1). percent of the debt service payments due in 2005, 7. See World Bank (2004, chapter 11) for a discussion 2006, and 2007 on the debt remaining due after of the difficulties entailed in estimating the amount of aid additional bilateral cancellation. required to finance the MDGs. 8. See World Bank (2006, pp. 77–8) for a detailed dis- Debt treatment for countries affected cussion of innovative financing mechanisms. 9. See IMF (2005a, annex 2, and 2005b) and Isard and by the tsunami others (2006) for a survey of recent studies. Following meetings in January and March 2005, 10. Documented in IMF (2005b). Paris Club creditors reviewed the debt treatment of 11. See IMF (2005b) for an analysis of whether recent the tsunami-affected countries and agreed not to ex- large aid surges in five African countries were “spent” or pect any debt payments on eligible sovereign claims “absorbed.” 12. All but 4 of the 29 HIPCs that have reached the de- from these countries until December 31, 2005. Two cision point are in Sub-Saharan Africa. countries, Indonesia and Sri Lanka, took up the 13. The Republic of the Congo only reached the deci- offer. According to the terms of treatment set in May sion point in March 2006 and hence did not receive any 2005, these two countries were to repay the deferred debt service reduction from the HIPC initiative over the pe- debt over 5 years with 2-year grace periods. Under riod 2000–5. 14. The Democratic Republic of the Congo and treatment were 100 percent of the amounts of prin- Gineau-Bissau are both excluded from the calculations un- cipal and interest due between January 1, 2005 and derlying figures 3.6 and 3.7 because they did not service December 1, 2005 on loans from Paris Club credi- their debt payments in 2000–3. tors having an original maturity of more than one 15. To qualify for “traditional debt relief” provided by year. For Indonesia, the total amount treated was the Paris Club, countries must generally have a Poverty Re- duction and Grant Facility (PRGF) program with the IMF. $2.704 billion, including $2.056 billion of principal 16. Commercial creditors account for only 2 percent of and interest on ODA loans and $648 million of non- debt relief due under the HIPC Initiative. ODA credits. For Sri Lanka, the total amount 17. See IMF and World Bank (2005, section III). treated was around $227 million, including $213 18. See IEG (2006: 8–9) for a more detailed discussion million of principal and interest on ODA loans and of creditors’ commitments to HIPC debt relief. 19. World Bank and IMF (2005: 18–20). 15 million of non-ODA credits. 20. Mauritania has reached the completion point under the HIPC Initiative but has not yet qualified for debt relief under the MDRI, pending implementation of key pub- lic expenditure management reforms. The calculations re- Notes ported in the text assume that Mauritania will qualify by 1. The definition of emergency and distress relief the end of 2006. grants was modified in 2005 to included reconstruction 21. The calculations underlying figure 3.8 assume that grants. The modification was not applied to previous Delivered by The years. World Bank countries will reach their respective completion points on e-library to: The amount of reconstruction grants reported byThe donors in Bankthe dates listed in World Bank 2006b (annex 2.3). World 2005 will not be known until the OECD DAC IP : 192.86.100.36 reports the 22. The baseline for refinancing IDA is specified in components of ODA in December 2006. Tue, 10 Mar 2009 16:55:33 SDRs with an inflation adjustment factor based on a three- 102 (c) The International Bank for Reconstruction and Development / The World Bank S U P P O R T I N G D E V E L O P M E N T T H R O U G H A I D A N D D E B T R E L I E F year moving average, so the U.S. dollar equivalent will vary DESA Working Paper ST/ESA/2006/DWP/15, United over time. Nations Department of Economic and Social Affairs, 23. The higher debt burdens in Guyana, Nicaragua, New York. March. http://www.un.org/esa/desa/ Bolivia, and Honduras largely represent debt owed to the papers/2006/wp15_2006.pdf Inter-American Development Bank, which is not forgiven Bourguignon, François, and Mark Sundberg. 2006b. “Ab- under the MDRI. sorptive Capacity and Achieving the MDGs.” Unpub- 24. Burundi and the Republic of Congo are excluded lished paper, World Bank, Washington, DC. March. from these calculations because they reached the decision Clements, Benedict, Rina Bhattacharya and Toan Quoc point after 2004. Net official lending in figure 3.12 includes Nguyen. 2003. “External Debt, Public Investment, and concessional and non-concessional loans from official credi- Growth in Low-Income Countries” IMF Working tors, whereas all bilateral loans discussed in the context of Paper 03/249, International Monetary Fund, Washing- ODA (figure 3.2) are concessional (by definition). ton, DC. 25. Prior to 2002, the DRC was in arrears with multi- IEG (Independent Evaluation Group). 2006. “Debt Relief lateral institutions and hence did not receive any conces- for the Poorest: An Evaluation Update of the HIPC Ini- sional loans from official sources. In 2002 the DRC received tiative.” World Bank, Washington, DC. $607 million in net concessional lending from official IMF (International Monetary Fund). 2003. “Fund Assis- sources, an amount equal to 11 percent of GDP. tance for Countries Facing Exogenous Shocks.” Policy 26. Net private debt inflows are comprised of net Development and Review Department, Washington, changes in public and publicly guaranteed debt, private DC. August. http://www.imf.org/external/np/pdr/ nonguaranteed debt, commercial bank loans and other pri- sustain/2003/080803.htm. vate credit. ———. 2005a. The Macroeconomic Challenges of Scaling 27. See Clements and others (2003) and Pattillo and Up Aid to Africa—A Checklist for Practitioners. others (2004) for recent reviews of the theoretical and em- SM/05/179. Washington, DC: IMF. www.imf.org/ pirical literature on “debt overhang.” external/pubs/ft/afr/aid/2006/eng/index.htm. 28. The empirical results reported by Clements and ———. 2005b. “The Macroeconomics of Managing In- others (2003) also imply that the impact on growth could be creased Aid Flows—Experiences of Low-Income stronger if some of the debt-service reduction were allocated Countries and Policy Implications.” SM/05/306, Policy to public investment. For instance, annual per capita GDP Development and Review Department, Washington, growth would be augmented by an additional 0.5 percent- DC. August. http://www.imf.org/external/np/pp/eng/ age point if half of HIPC debt relief were allocated to public 2005/080805a.pdf investment. IMF (International Monetary Fund) and World Bank. 2004. 29. This result is strongly influenced by large increases “Debt Sustainability in Low-Income Countries: Fur- in just a few countries, notably Chad, where real GDP per ther Considerations on an Operational Framework capita increased from an average rate of –0.9 percent over and Policy Implications.” September. http://www.imf the period 1991–2000 to 13.6 percent in 2001–04. The me- .org/external/np/pdr/sustain/2004/091004.htm. dian increase is only 1.5 percentage points. ———. 2005. “Heavily Indebted Poor Countries (HIPC) 30. This is documented by Reinhart, Rogoff, and Savas- Initiative—Status of Implementation.” August. tano (2003), who coined the term “debt intolerance.” They http://siteresources.worldbank.org/INTDEBTDEPT/Re examined 33 debt-distress episodes in emerging market sources/081905.pdf economies over the period 1970–2001. Of these, four involved Isard, Peter, Leslie Lipschitz, Alexandros Mourouras, and countries with ratios of external debt to GDP of less than 40 Boriana Yontcheva, eds. 2006. The Macroeconomic percent; another seven involved ratios of less than 50 percent. Management of Foreign Aid: Opportunities and Pit- 31. Calculations are based on World Bank staff esti- falls. Washington, DC: IMF. mates of gross general government debt. OECD. 2006. 2005 Development Cooperation Report. 32. The March 2004 agreement treated $85 million in Paris: OECD. arrears in principal and interest as of December 31, 2003 Pattillo, Catherine, Hélène Poirson, and Luca Ricci. 2004. and of maturities in principal and interest falling due from “What Are the Channels through Which External January 1, 2004 to December 31, 2006. The rescheduling Debt Affects Growth?” Working Paper 04/14, Interna- was on Naples terms (67 percent NPV debt reduction of eli- tional Monetary Fund, Washington, DC. gible external debt), with non-ODA credits cancelled by 67 Reinhart, Carmen, Kenneth Rogoff, and Miguel Savastano. percent (around $4.4 billion) and the remainder rescheduled 2003. “Debt Intolerance.” Brookings Papers of Eco- over 23 years with a 6-year grace period, at market interest nomic Activity 1: 1–74. rates, and ODA credits rescheduled over 40 years with a 16- Sun, Yan. 2004. “External Debt Sustainability in HIPC year grace period. Completion Point Countries.” Working Paper 04/160, International Monetary Fund, Washington, DC. Sundberg, Mark, and Hans Lofgren. 2006. “Absorptive Ca- pacity and Achieving the MDGs: The Case of Ethiopia.” References Delivered by The WorldIn The Bank Macroeconomic e-library to: Management of Foreign Aid: The World Opportunities Bank and Pitfalls, ed. Peter Isard, Leslie Lip- Bourguignon, François, and Mark Sundberg. 2006a. “Con- schitz, Alexandros Mourouras, and Boriana Yontcheva. IP : 192.86.100.36 Aid”10 Mar 2009 straints to Achieving the MDGs with Scaled-Up Tue, 16:55:33 DC: International Monetary Fund. Washington, 103 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 UN Millennium Project. 2005. Investing in Development: A World Bank. 2006a. Global Monitoring Report 2006. Practical Plan to Achieve the Millennium Development Washington, DC: World Bank. Goals. Overview. http://www.unmillenniumproject World Bank. 2006b. “IDA’s Implementation of the Multilat- .org/reports/index.htm. eral Debt Relief Initiative.” Resource Mobilization De- World Bank. 2005. Global Economic Prospects 2006. partment, Washington, DC. March. Washington, DC: World Bank. World Bank. 2004. Global Monitoring Report 2004. Wash- ington, DC: World Bank. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 104 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank . 4 Financial Integration among Developing Countries D eveloping countries have become impor- cial integration has progressed more rapidly tant sources of lending and investment to than North–South integration, as South–South other developing countries. In years past, trade has expanded more rapidly than North- most of the capital exported from developing coun- South trade (capital flows often follow trade) tries found its way to industrial countries, usually and developing countries have eased constraints to help wealthy individuals safeguard their assets. on outward investment. During the past decade, however, developing coun- • Developing-country multinationals enjoy tries have become a significant source of foreign di- some advantages over industrial-country rect investment (FDI), bank lending, and even offi- firms when investing in developing countries cial development assistance (ODA) for other because of their greater familiarity with developing countries. This expansion in technology and business practices suitable South–South capital flows reflects developing for developing-country markets. However, countries’ increasing integration into global finan- developing-country multinationals also face cial markets. As developing countries’ incomes rise greater impediments in their home countries and their banks and firms become increasingly so- than do industrial-country multinationals. phisticated, it is natural that they should become Impediments may take the form of bureau- more important sources of foreign lending and in- cratic constraints on outward investment, vestment and that a portion of these flows should other financial constraints, and a paucity of go to other developing countries. At the same time, institutional support and business services. South–South capital flows may have implications • South–South capital has helped to sustain FDI for developing-country recipients that differ from flows in developing countries even as FDI from the implications of capital flows coming from rich industrial countries has declined. It has made countries. The purpose of this chapter is to present more capital available to low-income coun- data on this growing trend and to evaluate its im- tries, because developing-country investors are plications for development. The principal issues are often more willing to handle the special risks (i) the forces that have propelled South–South fi- encountered in poor countries. In some cases, nancial integration, and (ii) the differences between South–South investment may also confer bene- South–South interactions and financial integration fits because firms in receiving countries may between developing and high-income countries. find it easier to absorb technology from a de- The main messages are: veloping-country investor than from an indus- trial-country investor, as developing-country • Capital flows among developing countries in- investors are likely to rely on technology ap- creased rapidly over the past 10 years, driven by propriate for a developing-country setting. the technological innovations that support glob- • Most South–South capital flows occur within Delivered by The World Bank e-library to: alization generally, rising incomes in developingThe World the Bank same geographic region, both because countries, and increasingly open policies toward they follow trade (and a large share of trade IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 trade and financial markets. South–South finan- is regional) and because proximity, common 107 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 language, and cultural and ethnic ties reduce country stock markets have shown little integra- the risks of lending and investment. tion in the form of cross-border listings or estab- • Developing-country banks are more likely lishment of regional stock exchanges.3 than industrial-country banks to invest in The growing financial integration of develop- small developing countries with weak insti- ing countries is driven by the same forces that are tutions. Especially in low-income countries, increasing integration between developing and the performance of foreign banks from de- high-income countries. Technological advances veloping countries (both in terms of asset have reduced the costs of transport and communi- quality and efficiency) does not differ from cations, facilitating greater cross-border integration that of foreign banks from rich countries, and encouraging the growth of cross-border pro- suggesting that developing-country banks do duction networks that involve expanded trade and not pose an additional risk to vulnerable financial transactions. Income growth has been ac- low-income countries because of poor man- companied by increased sophistication in financial agement or weak finances. systems, facilitating outward investment. Income • Initiatives to promote the integration of de- growth also is associated with more diverse con- veloping countries’ stock exchanges have sumption choices, stimulating international trade. made little progress, and many developing- In turn, the rise in international trade has provoked country capital markets remain more inte- greater cross-border financial transactions. The grated with major international financial very large differences in wage levels and capital in- markets than with other developing-country tensity of production within the developing world markets. Nevertheless, there are recent signs also have stimulated South–South flows. of change, with fewer new issues on U.S. ex- The rise in capital flows among developing changes, in particular, and increased local is- countries also reflects the increased importance of suance. Many exchanges may benefit from developing countries in the global economy. The closer South–South cooperation, including by developing world’s share of global GDP rose encouraging cross-border listings and invest- modestly from about 18 percent in 1990 to 20 ment, and information/technology sharing. percent in 2004, but its share in international trade grew more quickly—from 15 percent in 1991 to 26 percent in 2004. The growing impor- tance of some of the larger developing countries is The growth of South–South reflected in their increasingly prominent role in capital flows global economic negotiations, particularly within F inancial transactions among developing coun- tries increased substantially in the past decade (see annex 1).1 South–South FDI, for example, in- Figure 4.1 South–South capital flows by type, 2005 creased from $14 billion in 1995 to $47 billion in % of total flows to developing countries 2003. The share of South–South flows in total FDI 100 South–South to developing countries rose from 16 percent in 90 North–South 1995 to 36 percent in 2003, a higher share than 80 that of South–South exports in developing coun- 70 tries’ total trade and of South–South remittances 60 in their total remittance receipts (figure 4.1).2 Syn- 50 dicated loans grew from $0.7 billion in 1985 to 40 $6.2 billion in 2005. The share of South–South 30 flows in total cross-border syndicated lending was 20 3 percent in 1985, during the Latin American debt 10 crisis, when syndicated loans to Latin America and 0 other major debtors plummeted. That share fell to Export Remittances Syndicated FDI Delivered by The World Bank e-library to: 1 percent in 1995, with the recovery fromThe debt Bank theWorld revenues Loans crisis, and then rose to 3.4 percent in 2005. By IP : 192.86.100.36 Sources: UN Comtrade database; World Bank staff estimates. Tue, 10 Mar 2009 16:55:33 contrast with FDI and bank lending, developing- Note: Data are for 2005, except for FDI (2003). 108 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S the World Trade Organization (WTO). WTO’s negotiations on issues of concern to them Ministerial Meeting in Cancun in 2003 showed (Narlikar and Tussie 2004). The emergence of the that coalitions of developing countries (notably G-20 has been characterized as moving the WTO the G-20, but also the G-90 group of the poorest from a group dominated by the Quad (Canada, countries4), if they maintained solidarity, could the European Union, Japan, and the United play a major role in determining the outcome of States) to a multipolar environment (Amorim Box 4.1 Developing countries as aid donors T he Millennium Development Goals call for a global partnership for development. Historically, that part- nership has been understood as a matter of North–South be dominated by disbursements from just a few countries and show large variability from year to year because of sub- stantial, one-time loans. China accounted for 58 percent of cooperation, but that interpretation fails to acknowledge concessional lending from developing countries from 1994 the growing role of developing countries as sources of offi- to 2004, and Turkey (due to one disbursement in 1996), the cial development assistance (ODA). In recent years, how- Russian Federation, and Mauritius (due to one disburse- ever, recognition of the importance of South–South coop- ment in 2004) for another 30 percent. Fifteen (mostly low- eration has come from several quarters—among them the income) countries received some 70 percent of South–South Development Assistance Committee (DAC) of the Organi- concessional loans during 1994–2004. Sub-Saharan Africa sation for Economic Co-operation and Development received the greatest amount of South-South concessional (OECD), the European Union, and the United Nations loans (47.5 percent), followed by Latin America and the Development Programme (UNDP). Caribbean (26.5 percent) and Europe and Central Asia Brazil, Chile, China, India, South Africa, and Thai- (19.1 percent). In 2004 South–South concessional loans land are among the developing countries that now provide made up just 2 percent of all concessional lending to devel- aid to others in the developing world. There is evidence oping countries. Data on grants are not available. that the resources involved in South–South aid initiatives Like other South–South flows, South–South conces- may be increasing. China recently announced an increase sional loans are, once we exclude disbursements by China, in its assistance to developing countries over the next three mostly intraregional (78 percent). Case studies confirm the years, including $10 billion dollars in concessional loans strong intraregional pattern of South–South development and preferential export credits. In February 2006, Turkey assistance. For example, 90 percent of Thailand’s ODA became a member of the OECD Development Centre, supports infrastructure projects in Cambodia, Laos, demonstrating its commitment to providing development Myanmar, and the Maldives (Ministry of Foreign Affairs assistance to developing countries. of Thailand), and 73 percent of India’s non-plan grants Developing countries often provide aid through part- and loans from 1997–2004 went to neighboring countries nerships with traditional donors and international institu- (Ministry of Finance of India). tions (so-called triangular cooperation). For example, in co- Most emerging donors appear to have a special inter- operation with Britain’s Department for International est in providing development assistance to African coun- Development (DFID) and the U.N. Aids Program, the gov- tries. Long a donor in Africa, China, since 2000, has for- ernment of Brazil launched the International Centre for Hor- malized its relationship with the continent through the izontal Technical Cooperation to fight HIV/AIDS in Latin Forum for China-African Cooperation. Brazilian coopera- American countries. The center has allowed Brazil, which tion with Africa encompasses many areas, including agri- already has the region’s best record in fighting HIV/AIDS, culture, infrastructure, trade, and public administration. to strengthen its capacity to provide AIDS-related technical The country has written off more than $1 billion in debts assistance to other Latin American countries. of African countries. The Russian Federation, too, has Data on the magnitude of South-South development written off a substantial amount of African debt, partly assistance are scarce, although initiatives to improve collec- under the HIPC initiative. It is studying the possibility of a tion are underway.a DAC, the South-South Unit of the full HIPC debt write-off for loans not falling under ODA. UNDP, and the World Bank have formed a partnership to a. DAC provides data on official development assistance for its members collect information about South-South aid and provide a and for some non-DAC donors. These include high-income donors, such as platform for developing countries to share their experiences. Saudi Arabia, where development assistance has accounted for more than Data from the World Bank Debtor Reporting System 1.3 percent of GDP over the past five years, and some developing-country Delivered by The World Bank e-library donors, to: in Eastern Europe. Since the most prominent emerging mostly indicate that concessional loans from developing countries The World Bank donors are not included in the DAC database, however, the numbers do have shown no clear trend over the past decade, but IP : tend to 192.86.100.36 not provide an accurate picture of South–South aid. Tue, 10 Mar 2009 16:55:33 109 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 2005). More than any previous round of trade between high-income and developing countries is negotiations, the Doha Round has been shaped by less stark for foreign investment than for trade, as the actions and positions of developing countries high-income countries in 1995 were only slightly (Zedillo, Messerlin, and Neilson 2005). Another more open than developing countries, and 9 of the indication of their increasing importance is that a 21 rated high-income countries adopted more open few developing countries have become sources of regimes over the past 10 years. However, the major official development assistance (box 4.1). sources of outward investment and lending from South–South financial integration has been high-income countries, such as the United States given a boost by the rapid opening of developing and Germany, already had relatively open regimes economies. About half of 77 developing countries in the early 1990s, so their outward capital flows rated on a leading index of openness to trade did not receive any further impetus from policies showed some improvement from 1995 to 2005, becoming more open. whereas only 5 showed deterioration (figure 4.2). Another spur to South–South capital flows The rest were unchanged.5 By contrast, nearly all has been the rise of regional trade agreements high-income countries showed no major change in (RTAs) among developing countries. RTAs have their trade policies over this period, because they mushroomed: since 1990, their number rose from already were relatively open economies: the aver- 50 to nearly 230.7 Activity has been particularly age trade index of high-income economies was intense in Latin America, Africa, and Asia. more than two points better (on a 1-to-5 scale) than that of developing countries.6 Moreover, • In Latin America, Mexico and Chile have con- South–South trade expanded more quickly than cluded a series of agreements since the launch North–South (box 4.2). Because capital flows of the North American Free Trade Agreement often follow trade, this has meant more rapid (NAFTA) in 1994. South–South financial integration as well. • In Africa, the countries of eastern and south- Similarly, a majority of 76 rated developing ern Africa established a common market in countries became more open to foreign investment 1993; the East African Community was over the past 10 years, while only 8 instituted more formed in the mid-1990s; and the Southern restrictive policies. In part this reflects an easing of Africa Development Community (SADC) constraints on outward investment, leading to in- signed a trade cooperation protocol in 1996. creased South–South capital flows. The difference • In Asia since 2000, India has made agreements with the Southern Cone Common Market (MERCOSUR) and Thailand; China has con- Figure 4.2 Growing openness of developing cluded bilateral trade accords with the coun- countries to trade and capital flows, 1995–2005 tries of the Association of Southeast Asian Na- No. of countries tions (ASEAN); and the countries of South 40 Asia reached a free trade agreement in 2004. More open 35 Unchanged 30 Less open It is unclear whether such agreements have made a major contribution to South–South trade 25 and capital flows—or simply reflect their increase. 20 15 10 Foreign direct investment 5 in the developing world 0 South–South FDI is increasing Trade Foreign Investment FDI flows from developing countries to other de- veloping countries increased from an estimated Source: Heritage Foundation. Delivered by The World Bank e-library to: The Note: Number of countries rated more open, unchanged, or World less open Bank $14 billion in 1995 to $47 billion in 2003 (table IP : 192.86.100.36 on Heritage Foundation index of openness for period 1995–2005. 4.1). Increased South–South flows have provided Tue, 10 Mar 2009 16:55:33 110 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S Box 4.2 South–South FDI and trade T rade and FDI flows are closely linked (Aizenman and Noy 2005; Albuquerque, Loayza, and Serven 2005; Swenson 2004). At times FDI is a substitute for trade, as The impact of increased investment on South–South trade is hard to measure. However, the surge in trade in raw materials (126 percent from 1995 to 2003) was in when the investment is designed to serve the host market line with increasing South–South FDI flows in extractive while reducing transport costs or circumventing tariff barri- sectors (see figure). Also, the growth in trade in intermedi- ers. However, as trade barriers have come down and the im- ate goods (91 percent) and capital goods (213 percent) re- portance of global production networks has risen, FDI and flects the increased integration of production networks trade have become increasingly complementary (World Bank among developing countries, which is stimulated both by 2005a). Trade flows also can facilitate FDI by increasing in- North–South and South–South investments. vestors’ access to information (Portes and Rey 2005). South–South trade grew rapidly over the past decade, Composition of South–South exports, 1995 and 2003 reaching $562 billion in 2004 compared to $222 billion in 1995. From 2000 to 2004, South–South trade grew at an $ billions annual rate of 17.6 percent, faster than South–North and 140 1995 North–South exports (12.6 percent and 9.7 percent, respec- 120 2003 tively). South–South trade made up 26 percent of develop- 100 ing countries’ exports in 2004. 80 Most South–South trade occurs within the same re- 60 gion, although cross-regional trade has also been growing 40 rapidly. In 2004, for example, China was the fourth- 20 largest export destination for Argentina and Brazil. The 0 rapid growth in South–South trade is linked to high Raw Intermediate Consumer Capital materials goods goods goods growth rates in developing countries, substantial reduc- tions in tariff barriers, and falling transport costs. Source: UN Comtrade database. partial compensation for the decline in FDI flows Table 4.1 South–South FDI as a share of global FDI, 1999–2003 from high-income countries—from $130 billion in $ billions 1999 to $82 billion in 2003. 1995 1999 2000 2001 2002 2003e More than 50 developing countries have re- Total inflows (1) 90.3 163.5 154.7 159.3 135.3 129.6 ported FDI outflows over the past decade, although from high-income OECD (2) 48.1 95.4 93.7 84.8 55.1 59.4 the data are notoriously understated (World Bank from high-income non-OECD (3) 28.2 35.0 22.7 24.8 27.2 22.8 South–South FDI (1)-(2)-(3) 14.0 33.1 38.3 49.7 53.0 47.4 2004). It is clear that most developing-country FDI South–South FDI (percent) 15.5 20.2 24.8 31.2 39.2 36.6 comes from the same middle-income countries that account for the lion’s share of developing-country Source: World Bank staff estimates. Note: The South–South estimates are based on 35 countries that account for 85 percent economic activity. The 10 countries that accounted of total FDI flows to developing countries. The estimates are based on the World Bank’s for 73 percent of FDI inflows from 2000 to 2004 classification of developing countries. e = estimate. also were the source of 87 percent of the total out- flows (both to developed and developing countries) during the same period. The expansion of FDI outflows has been dri- strategic assets. As many developing-country gov- ven by developing countries’ increasing openness ernments have eased their policies toward capital to capital and trade, and by their increasing partic- outflows, their companies, like industrial-country ipation in international production networks. Be- multinationals, have expanded their operations cause of increased globalization of economic activ- abroad. South–South FDI flows have also in- Delivered by The World Bank e-library to: ities, developing-country companies face growingThe World creasedBank in response to the significant rise in competition in sales and in access to resources and South–South trade. IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 111 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Most South–South FDI goes to countries frastructure, in particular) and the extractive indus- in the same region tries, as shown by data on mergers and acquisitions Many expanding developing-country firms tend to (M&A) and privatization transactions (annex 2). invest regionally before taking on the rest of the South–South FDI in services increased over world because of familiarity gained through trade the last decade, in tandem with the global surge in or ethnic and cultural ties. The regional agreements services sector FDI and the liberalization of the that began to proliferate in the mid-1990s (World services sector in many developing countries.8 De- Bank 2005b) also have encouraged intraregional veloping economies attracted substantial FDI trade and investments. For example, 75 percent of flows from both high-income and other develop- the outward investments of Hungarian firms were ing countries through the privatization of state- within Europe (Elteto and Katalin 2003 and table owned assets. Developing-country firms enjoy 4.2); almost 40 percent of Russian firms’ invest- some advantages in services sector FDI, because ments abroad have been in Europe and Central services often require proximity between producers Asia (Vahtra and Liuhto 2004); and the Russian and consumers, and often favor cultural and ethnic Federation accounts for one-third of Turkey’s re- familiarity.9 Moreover, developing-country firms cent FDI outflows. Encouraged by cooperation can take advantage of their experience in managing arrangements, ASEAN countries have been the top the regulatory process (De Sol 2005; Lisitsyn and destination for Thai companies (Mathews 2005). others 2005) and create regional networks. Never- South African investments in other developing theless, FDI from high-income countries is also countries are largely in the southern part of Africa highly concentrated in the services sector. (Goldstein 2003). Following trade liberalization in The significant rise of South–South FDI in the Latin America, multinationals from Argentina, infrastructure sector, which began in the late 1990s, Brazil, and Chile expanded their regional opera- often was achieved through partnerships between tions (Chudnovsky and Lopez 2000). developing- and industrial-country firms. This ex- Nevertheless, some developing-country multi- pansion by northern investors slowed following nationals are venturing beyond their region. For stock market declines in the industrial countries and example, in 2004 about half of China’s outward in response to problems of corporate governance in FDI went to natural resources projects in Latin some companies and poor regulation in many de- America; Malaysia has emerged as a significant veloping countries. But developing-country firms new source of FDI in South Africa (Padayachee and continued their expansion through buyouts of the Valodia 1999); and Brazil has considerable invest- assets of their northern partners, privatization and ments in Angola and Nigeria (Goldstein 2003). acquisitions deals, and licenses (annex 2).10 Between 1998 and 2003, developing countries received al- South–South FDI is concentrated in services most $160 billion in foreign investment in infra- and extractive industries structure, while developing-country firms invested While data on the sectoral composition of South- more than $30 billion in developing-country infra- South FDI are not available, a substantial amount structure projects. These data represent commit- of South-South FDI is known to be in services (in- ments for selected projects, and thus the totals can- not be compared to the net-flows data usually shown for FDI (see World Bank 2005a for details). Table 4.2 Regional FDI by multinationals from Nevertheless, the commitments data do show that a selected countries Share of total investment occurring within region very significant proportion of FDI flows to develop- ing countries (from both the North and the South) Regional (South–South) is devoted to infrastructure. South–South flows China 20.7 were greatest in telecommunications and, geograph- India 25.4 Hungary 75.1 ically, in Africa (figure 4.3). Thailand 58.8 Almost 30 percent of FDI in developing coun- Turkey 32.0 tries’ telecommunications during 1998–2003 came Russian Fed. 37.0 Delivered by The World Bank e-library to: The World Bankfrom southern telecommunications companies, Source: Goldstein (forthcoming). more than 85 percent of it intraregional. Financial IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 112 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S and equity investors from the South—such as in- Figure 4.3 South–South FDI in infrastructure and by region, 1998–2003 vestment banks, private equity funds, and mutual By sector By region funds—also have become direct investors in the % of total foreign investment % of total foreign investment sector, in addition to participating through 30 60 South–South cross-border lending and syndicated 25 50 20 40 loans, as discussed in detail later in the chapter 15 30 (World Bank 2006). 10 20 Developing-country multinationals also invest 5 10 0 0 in noninfrastructure services, taking advantage of Telecom Transport Energy Water ric a As ia ea n ric a sia ific brand-name recognition, physical proximity, re- Af h ibb Af alA P ac an ut r rt h nt r r So Ca No & gional distribution networks, taste similarities, ha & Ce As ia - Sa ica s t & e & st and advantages offered by bilateral arrangements. Su b er Ea ro p Ea Am dle Eu Considerable South–South investment has oc- tin i d La M curred in banking, as we shall see later in this Source: World Bank (2005c). chapter. Other examples include the growing num- ber of supermarket chains, food companies, phar- maceutical firms, hospitals, and airline carriers growth economies, such as China and India, have from developing countries.11 In some cases, north- acquired oil-and-gas assets or licenses in other de- ern investors undertake investments in developing veloping countries (annex 2). Developing-country countries through their subsidiaries in another de- companies also are investing in exploration pro- veloping country—for example, Wal-Mex, Wal- jects. For example, Petronas (Malaysia), which has Mart’s joint venture with a Mexican company. strong technical competencies in deep-water ex- Developing-country firms (mainly in Asia) ploration, has invested in exploration and produc- have made a small but increasing number of in- tion projects in more than 20 developing countries vestments in research and development (R&D) in (Goldstein forthcoming). Countries that are large other developing countries (UNCTAD 2005a). oil-and-gas producers, such as República Bolivari- China and India are among the largest recipients ana de Venezuela, invest in other developing coun- of R&D-related investments from developing tries as they integrate their downstream operations countries, with investment from one another and such as refining, distribution, and retailing. from Malaysia and Thailand.12 South–South FDI in the nonoil mining sector The extractive sector (particularly oil and gas) is also increasing. The resource-rich African re- also attracts increasingly large amounts of gion has attracted the interest of companies from South–South FDI, mostly through state-owned China, India, South Africa, and other developing companies (table 4.3). In recent years, high- countries.13 Chinese investments in nonoil mining Table 4.3 Selected southern multinationals in the oil-and-gas sector, 2004 Total assets in 2004 Corporation (home country) Ownership ($ billions) Areas of activity CNPC (China) State 110.6 Canada, Ecuador, Kazakhstan, Mauritania, Myanmar, Sudan, R. B. de Venezuela Indian Oil Corp. State 10.9 Islamic Rep. of Iran, Libya Lukoil (Russian Federation) Private 29.8 Iraq, Romania, Ukraine, Bulgaria, Canada, Uzbekistan PDVSA (R. B. de Venezuela) State 13.4 Argentina, Belgium, Brazil, Chile, Germany, Paraguay, United States (Citgo) PEMEX (Mexico) State 84.1 Argentina Petrobras (Brazil) State 19.4 Libya, Mexico, Nigeria, Tanzania Petro China (China) State 58.8 Nigeria, Sudan, R. B. de Venezuela Petronas (Malaysia) State 53.5 Cambodia, Chad, Islamic Rep. of Iran, Myanmar, Sudan, Turkmenistan Saudi Aramco (Saudi Arabia) State Delivered by The World Bank e-library Canada, to: States China, United The World Bank Sources: UNCTAD, ECLAC, and Oil & Gas Journal Special ReportIP : 192.86.100.36 2001, company annual reports, company Web sites. Tue, 10 Mar 2009 16:55:33 113 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 projects have been growing in Latin America, State-owned and small and medium and several Russian companies have investments enterprises are investing abroad in Central Asia and the Middle East (Vahtra and State-owned enterprises (SOE) in extractive indus- Liuhto 2004).14 tries and infrastructure are a considerable source Recent bilateral and regional initiatives among of South–South FDI flows.17 The role of SOEs in developing countries are centered on cooperation in overseas investments is significant in China, where resource-seeking projects, including a proposal to 43 percent of outward FDI stock in 2003 was held create a regional state-owned energy company in by SOEs (Giroud 2005). This indicates that a con- Latin America; joint-venture projects involving siderable portion of South–South FDI may be dri- India, Bangladesh, Myanmar, and Thailand; ven not only by economic but also by political and China’s agreements with Argentina and Brazil to strategic factors.18 SOEs usually have an advan- cooperate in mining, oil, and infrastructure projects tage over privately owned firms, since they enjoy (UNCTAD 2005b); and partnerships between better financing terms when funded by state- China and India for the acquisition of energy assets. owned banks. In some cases, governments negoti- Manufacturing also receives considerable ate packages of investment deals that may give ad- South–South FDI flows, although projects tend to ditional bargaining power to SOEs.19 be smaller than the large privatization and M&A Small and medium enterprises (SMEs) also deals in services and the extractive industries. provide a significant amount of investment in Developing-country multinationals have invested other developing countries.20 In India, for exam- in efficiency-seeking activities abroad following ple, SMEs accounted for 26 percent of overseas erosion in their competitiveness, at home and in ex- projects (6.7 percent of the value) in manufac- port markets, because of currency appreciation, in- turing and 41.1 percent (47.1 percent of the creased labor costs, or other causes (Mirza 2000). value) in the software industry (Pradhan 2005). In many middle-income countries, higher living Almost three-fourths of companies investing standards are reflected in increased labor costs.15 abroad in Poland and Estonia, and about one- Developing-country manufacturing firms also in- third in the Czech Republic and in Hungary, are vest abroad to sell into the target markets or to ac- SMEs (Sevtlicic and Rojec 2003). cess other markets, sometimes through special arrangements. Examples include the investments in Southern multinationals are supported India and Thailand of Chinese white goods pro- by government incentives ducer Haier, and the plants in China, Egypt, India, In addition to easing restrictions on capital out- and Ethiopia of Russian automobile manufacturer flows, some developing-country governments have UralAZ plants (Vahtra and Liuhto 2004). Special provided fiscal and other incentives for outward arrangements play an important role in attracting investment, particularly South–South FDI. China’s South–South FDI to low-income countries. Chi- Export-Import Bank, for example, provides loans nese, Indian, Malaysian, and Sri Lankan textile for investments in resource development and infra- companies have investments in Africa to export structure, as well as for projects that facilitate garments to U.S. and European markets through trade. If the investment is in an aid-receiving coun- free trade agreements. Some developing-country try, firms can receive preferential loans under Chi- firms are investing in the manufacture of generic nese aid programs or projects (UNCTAD 2005b). drugs in Africa because WTO provides that patents Malaysia supports special deals for FDI outflows may be broken in cases of national emergency. A to countries such as India, the Philippines, Tanza- few Indian and Chinese companies are introducing nia, and Vietnam (Mirza 2000). The Thai govern- anti-malarial and AIDS drugs under such arrange- ment promotes Thai firms’ involvement in infra- ments (Goldstein and others 2006).16 structure projects in selected developing countries In some cases, FDI from high-income coun- in the region (UNCTAD 2005b). tries has facilitated South–South flows in the man- Some regional arrangements, such as SADC, ufacturing sector. For example, Mexican Bimbo, a ASEAN, MERCOSUR, and the Andean Commu- Delivered by The World Bank e-library to: food producer, has invested abroad since becom- The World Banknity offer various incentives for outward invest- ing McDonalds’ exclusive supplier in Latin IP : Amer- ment within the region, including lower tax and 192.86.100.36 Tue, 10 Mar 2009 16:55:33 ica and more recently in Europe. tariff rates and easier profit repatriation. Some 114 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S members of the regions maintain bilateral invest- Geographical proximity and cultural similari- ment agreements and double-taxation treaties. ties can make coordination of foreign operations Whether these incentives encourage or direct more effective (IMF–World Bank 2005; UNCTAD FDI outflows, and at what fiscal cost, is unclear. 2005b). Developing-country firms may have a UNCTAD (1998) found that incentives had a pos- comparative advantage over companies from de- itive, but minimal, effect. On the other hand, veloped countries in doing business in challenging Hallward-Dreimeier (2003), using only OECD economic and political conditions because of their countries, and Tobin and Rose-Ackerman (2005), experience in their home economies (Claessens using a larger sample of countries, found that incen- and Van Horen 2006). This sort of advantage tives can further increase FDI flows in countries only brought higher rates of return for northern in- where the environment for FDI is already strong. vestors that partnered with Chilean companies to Banga (2003) shows that India’s fiscal incentives and invest in Latin America than for those that in- lower tariff rates attracted investors from developing vested alone (De Sol 2005). The relative success in countries only; the removal of restrictions was neces- Uganda of MTN (the South African telecommuni- sary to attract investments from developed coun- cations company), compared with its competitors tries. Interviews with Malaysian investors suggest from developed countries, was traceable to its in- that tax and fiscal incentives were not important house expertise in managing pertinent economic (UNCTAD 2005a). In some cases, incentives simply and political risks (Goldstein 2003). generate so-called round-tripping (capital outflows Developing-country firms may also be more to finance investment back in the home country). willing to assume the risks of postconflict and For example, India’s advantageous tax treaty with other politically difficult situations (Sull and Esco- Mauritius encourages many Indian investors to in- bari 2004). For example, Chinese companies (not corporate in Mauritius in order to benefit from this all of them SOEs) are the only foreigners that have tax treatment (Shah and Patnaik 2005). invested in Sierra Leone since the end of the civil war. Egypt’s Orascam is the only foreign telecom Developing-country multinationals may enjoy company operating in Iraq (EIU 2005). some advantages over industrial-country firms when investing in developing countries Institutional, financial, and operational Compared to their northern counterparts, develop- impediments constrain FDI from ing-country multinationals may enjoy some advan- developing countries tages when investing in developing countries. Com- Despite these advantages, developing-country panies with a significant regional presence often firms face institutional procedures, financial re- benefit from well-established distribution net- strictions, and operational problems in their home works. Because of their experience in their home countries that can make it difficult for them to in- markets, they are often in a position to use locally vest abroad. available inputs more efficiently. And some devel- Institutional procedures. Many developing oping-country firms are more familiar than north- countries still have various levels of capital con- ern firms with lower-cost production processes that trols, and firms may be subject to regulatory bur- are appropriate for developing-country markets. dens to obtain access to foreign exchange. For ex- For example, India’s Tata Group produces a car ample, in addition to several capital-control that is significantly less expensive than those of the procedures, China’s regulations require its multi- major automobile companies.21 While the car lacks nationals (state-owned or private) to submit a cer- some of the qualities desired by industrial-country tificate of establishment of the firm in China, con- consumers, it has found a ready market in India tracts and agreements relating to the overseas and several other developing countries. Finally, de- project, various elements of a project feasibility veloping-country firms may also use technologies study, assessments of the project made by the Chi- that are better suited to conditions in developing nese embassy in the host country, and audited fi- countries. For example, in Vietnam, TVs made by nancial reports and bank statements—all before Delivered by The World Bank e-library to: China’s TCL are the most popular brand, as theirThe World proceeding Bank with an overseas investment (FIAS powerful color TV receivers provide clear reception 2005). Such requirements have increased costs and IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 even in remote areas (Yi 2004). in some cases prevented SMEs from investing 115 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 abroad, while some larger firms have used off- that vary with the firm’s level of experience as a shore platforms for their foreign investments. For foreign investor and to some extent with the busi- example, many Chinese companies use their Hong ness environment in the firm’s home country. For Kong affiliates as a base from which to expand example, with limited experience in FDI, some overseas (UNCTAD 2003). Chinese investors find it difficult to formulate pro- Developing countries often lack the institu- jects that fit in with the culture, market character- tional infrastructure needed to provide foreign in- istics, and regulatory environment of foreign coun- vestors with the support services that their counter- tries (FIAS 2005). Some developing-country parts in developed countries take for granted. multinationals may have overbid for large assets Access to knowledgeable consulting firms, business due to lack of experience (IMF–World Bank 2005; associations, banks, and other sources of informa- Financial Times 2004). This is not an unusual phe- tion about overseas markets and practices is more nomenon: Japanese firms experienced similar chal- difficult to obtain in most parts of the developing lenges when they started to venture abroad in the world. Unlike in developed economies, services late 1980s (Goldstein forthcoming). The World that promote outward investment are nonexistent Bank Group has made efforts to assist developing- or in their infancy in most developing countries. country multinationals in overcoming the institu- These handicaps have affected the development tional, financial, and operational challenges they and operations of overseas projects, particularly face (box 4.3). for companies relatively new to outward FDI. Financial restrictions. Developing-country South–South FDI may generate important firms, particularly SMEs, face more severe finan- benefits for developing countries cial constraints than do their industrial-country The emergence of the South as a substantial source counterparts, because local financial markets are of FDI for developing countries may have signifi- less developed. And access to international finan- cant implications for economic development. First, cial markets is limited and costly for many of these South–South FDI represents an opportunity for firms, since they carry the sovereign risk of the low-income countries. Except in the extractive sec- home country in addition to their company risks tor, most northern multinationals are unlikely to (IMF–World Bank 2005). These challenges some- invest in small markets, as market size is a major times lead large and successful developing-country determinant of North–South FDI (Levy-Yeyati, multinationals to migrate to industrial countries. Ugo, and Stein 2002; Stein and Daude 2001). In For example, South African Brewery moved its contrast, southern multinationals tend to invest in headquarters to Britain in 2001 to improve its risk neighboring developing countries with a similar or rating and position itself for global expansion. lower level of development than their home coun- India’s Ispat Corporation moved to the Nether- try (World Bank 2005a). South–South FDI flows, lands for similar reasons. however small, are significant for many poor Operational challenges. Developing-country countries, particularly those that are close to firms that invest abroad face operational issues major investors. For example, India (in hotels and Box 4.3 The World Bank Group and South–South flows T he World Bank Group, particularly the International (MIGA) supports the efforts of local export-credit agen- Finance Corporation, has several programs to help cies to serve emerging South-South investors through developing-country multinationals. IFC’s Foreign Invest- coinsurance and reinsurance arrangements. In addition, ment Advisory Service (FIAS) is surveying firms and as- MIGA’s recently launched Small Investment Program— sessing the need for technical assistance to governments which offers a streamlined insurance package and un- to enhance the investment climate as it affects outward derwriting process—is designed to increase South-South Delivered by The World Bank e-library to: FDI. The Multilateral Investment Guarantee Agency investment. The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 116 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S manufacturing) and China (in manufacturing) ac- developing countries is increasing, although signifi- count for more than half of FDI in Nepal. Most cant regional and sectoral variations in compliance FDI in Mongolia comes from China and the Russ- remain (OECD 2005b). Ultimately, of course, it is ian Federation. An Indian company is securing ap- the host country’s responsibility to improve its proval for a $2.5 billion investment project in business environment and regulatory system to re- Bangladesh, which will be the largest foreign in- alize the development potential of FDI. vestment in the country. Moreover, low-income Outward investment (including to high- countries receive almost one-third of their banking income countries) may also generate benefits to sector FDI from other developing countries (see the investing economy through increased competi- the section on banking in this chapter). tiveness and exports. Surveys report that direct Second, in some cases, developing countries presence in foreign markets has enabled many may see greater positive spillovers from FDI origi- Southern firms to increase their competitiveness nating in developing countries than from invest- and to respond better to consumer demand.24 22 ments originating in industrial countries. To the Geographic risk diversification and market access extent that developing-country firms provide tech- can be crucial for some southern firms that are nologies that are more suitable for other develop- faced with volatile home markets. ing countries (compared with more sophisticated technologies used by industrial-country firms), de- veloping countries may be in a better position to absorb them. Baldwin and Winters (2004) find that South–South banking a country’s absorption capacity is greater with a smaller technological gap between the foreign firm and domestic firms. Kabelwa (2004) finds that nar- T raditionally, banks have followed their clients overseas. Thus the growing importance of de- veloping-country firms in overseas trade and in- rower technological gaps between developing- vestment has led to an expansion of cross-border country multinationals and host economies, com- activities by developing-country banks, both pared with their industrial-country counterparts, through lending and through investment carried foster positive spillovers. Schiff and others (2002) out by branches and subsidiaries. As is the case found that the extent of spillovers from participa- with other financial flows to developing countries, tion in trade (as opposed to FDI) depends on the foreign bank lending is dominated by industrial- sector: companies in low R&D-intensive industries country banks. However, developing-country benefit more from trading with other developing- banks are playing a growing and already important country firms than with firms from industrial role, especially in low-income countries. Because countries, while companies in high R&D industries they are willing to penetrate markets where banks benefit more from trading with firms from indus- from industrial countries are reluctant to go, these trial countries. However, the importance of this ad- banks may provide an important new source of ex- vantage, which is most significant in manufactur- ternal finance for low-income countries. ing, is unclear, as South-South FDI is heavily concentrated in extractive industries and infra- The rise in South–South cross-border banking structure, where such spillovers are limited. is driven by several factors South–South FDI is not always more beneficial The recent increase in banks’ cross-border activities than North–South FDI. Over the years, many has come in response to global economic trends, northern multinationals have participated in initia- liberalization of the financial sector in many devel- tives to improve the transparency of their foreign oping countries, and advances in technology. operations, as well as the environmental and labor Economic trends. The general expansion of standards observed in those operations.23 Such ini- syndicated lending to developing countries and the tiatives are less likely to have been implemented by growing importance of developing-country lenders southern companies, which also may have low en- in such lending reflect a favorable external financ- vironmental and labor standards (Goldstein forth- ing environment characterized by ample global liq- Delivered by The World Bank e-library to: coming; IMF–World Bank 2005). That said, com-The World uidity, as well as improved economic conditions Bank pliance with corporate governance standards by and greater openness to trade and capital flows in IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 117 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 many developing countries. As South–South trade ogy to support a strategy of greater focus on serv- and FDI have expanded, many banks have fol- ing SMEs and retail credit clients. lowed their clients. FDI in banking is correlated The several motives behind the expansion of with bilateral trade and FDI between source and South–South banking can be illustrated by the ex- host countries (Grosse and Goldberg 1991; Brealey perience of the State Bank of India (SBI) and ICICI and Kaplanis 1996; Williams 1998; Yamori 1998). Bank, India’s largest privately owned bank. Both Preferential trade agreements, which have bur- are undertaking overseas expansions in Asia, geoned in number and scope since the 1990s Africa, and the Middle East to tap retail credit (WTO 2003), are opening new opportunities for clients, to facilitate increasing trade and investment banks to provide trade finance. For example, flows between India and other countries, to pro- Banco de Chile, the country’s second-largest bank vide foreign currency–denominated loans to the in terms of assets, recently opened a branch in Bei- overseas affiliates of Indian companies, and to pro- jing—reportedly to position itself to benefit from a vide remittance and retail credit services for Indian new free-trade accord between the two countries expatriates (Capital Intelligence, various issues; (Latin Finance 2005). A number of Central Ameri- State Bank of India 2005; and ICICI Bank 2005). can banks (e.g., Panama’s Banistmo, El Salvador’s Banco Cuscatlan) are seeking growth opportunities South–South bank lending has grown in other Central American retail financial markets There are two sources of data on developing to capitalize on regional trade integration and the countries’ foreign bank lending (see annex 1 for recently concluded Central American Free Trade data sources and definitions). The Bank for Inter- Agreement (CAFTA). Standard Bank of South national Settlements (BIS) in Basel publishes data Africa has established a sizable presence in south- on the foreign lending of banks from a few devel- ern and eastern Africa, reflecting South Africa’s in- oping countries. Dealogic Loanware reports data creased investment in and trade with the region. on syndicated loan transactions, which are loans Migration. Banks have expanded cross-border arranged by a group of banks (referred to as a activities to serve growing numbers of expatriates. syndicate). For example, Pakistan’s Habib Bank has targeted a Syndicated lending. Most syndicated loans to well-established customer base of expatriates developing countries are made by groups of banks through its branch network in South Asia. in high-income countries. In the past 20 years, how- Financial sector liberalization. The liberaliza- ever, the volume of syndicated lending from devel- tion of developing countries’ banking sectors and oping countries and the number of banks partici- the sale of state-owned banks have increased op- pating in syndicates have grown sharply. portunities for cross-border lending and invest- South–South syndicated flows are estimated to have ment by developing-country banks. Rules govern- increased from $0.7 billion in 1985 to $6.2 billion ing cross-border lending and the establishment of in 2005, although the data have shown substantial branches and subsidiaries by foreign banks have variability across years and countries.26 The number been eased—in many cases under the impetus of of developing countries receiving such flows also WTO commitments, notably in the Asia-Pacific re- has grown, from 19 in 1985 to 41 in 2005.27 gion (Capital Intelligence and EIU, various issues). The rise in South–South syndicated lending Technology. Advances in telecommunications partly reflects the overall rise in syndicated lending and information technology are enabling banks to developing countries from all sources, which in- and other financial institutions—including those creased by almost the same amount from 1985 to based in developing countries—to better manage 2005. Indeed, the share of South–South lending in cross-border activities. Banks based in Asia-Pa- total developing-country borrowing from the syn- cific, the Middle East, and elsewhere have been in- dicated loan market equaled 3 percent in 1985 vesting heavily in electronic delivery systems and during the debt crisis. However, once lending from other technologies to enhance their ability to offer industrial countries picked up, the share of a wider array of financial services at a distance South–South lending fell to 1 percent in 1995, but Delivered by The World Bank e-library to: from headquarters.25 Sri Lanka’s Commercial The World Bankthen rose to 3.4 percent in 2005 (table 4.4). Bor- Bank of Ceylon and Hungary’s OTP Bank, IP :among rowers in Europe and Central Asia and the Middle 192.86.100.36 Tue, 10 Mar 2009 16:55:33 others, have boosted their investment in technol- East and North Africa sourced the largest portion 118 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S Table 4.4 South–South cross-border syndicated lending, 1985–2005 $ millions Borrower’s region of domicile 1985 1995 2004 2005 Eastern Europe & Central Asia 234.4 31.2 1,420.0 2,719.7 Middle East & North Africa 326.8 109.1 694.2 1,120.9 Sub-Saharan Africa 8.7 130.0 986.8 364.0 South Asia 15.0 12.9 349.7 463.8 East Asia & Pacific 56.4 431.6 470.5 872.0 Latin America & Caribbean 54.2 165.1 301.7 686.1 Total 695.5 879.9 4,222.9 6,226.5 Total syndicated lending to developing-country borrowers 22,895.6 91,943.2 112,238.2 184,034.7 South–South share in syndicated lending to developing countries 3.0 1.0 3.8 3.4 Source: World Bank staff estimates based on loan syndicate transactions reported in Dealogic Loanware dataset. of their syndicated loans from nonlocal develop- Eastern Europe and Central Asia attracted the ing-country banks (4.2 percent overall for both re- highest share from both source groupings (35 per- gions), while borrowers in Latin America contin- cent and 44 percent, respectively), while borrow- ued to source the smallest portion (about 1.8 ers in Sub-Saharan Africa attracted the lowest percent overall). share (6 percent) from both source groupings. No- The growing participation of developing- tably, East Asia and Pacific attracted a much country banks in syndicated lending also reflects smaller share in 2005 from both source groupings the increasing size and sophistication of those (14 percent and 17 percent, respectively) com- banks. As syndicates typically are unwilling to in- pared with a decade earlier, just a few years ahead clude banks that are relatively unknown or unreli- of the financial crisis. In 1995, East Asia and Pa- able, the growing role of developing-country cific received nearly half of syndicated lending banks in syndicates is one indication of their ar- flows destined for developing countries—sourced rival as major players in global finance. For many both on a cross-border South–South basis and banks, participation in recent South–South syndi- from all lending sources worldwide The share of cated loans has been one element in a strategy of Eastern Europe and Central Asia, in particular, expansion into other developing countries through was significantly smaller (at just 4 percent and 7 loans, acquisitions, and greenfield investments.28 percent, respectively). Despite the growth of South–South lending, Cross-border lending reported to the Bank for some aspects of developing countries’ participa- International Settlements (BIS). Cross-border lend- tion have changed little over the years. Participa- ing by banks located in developing countries that tion by local banks in syndicated loan transactions report to the BIS (that is, countries with significant remains strong.29 Also, banks domiciled in devel- cross-border lending) has increased significantly, oping countries tend not to be the lead arrangers reaching $94 billion in 2005 (figure 4.4).31 While or major participants in a syndicate, given their in 1999 no developing country reported to the BIS, relative capital constraints compared with major by 2005 six developing countries (Brazil, Chile, industrial-country banks. Nevertheless, nonlocal India, Mexico, Panama, and Turkey) were report- developing-country banks participated in a man- ing data; more are expected to follow soon. About dated lead arranger role in nearly one-quarter of 85 percent of the cross-border lending was to the all South–South cross-border syndicated loan banking sector (the average across all countries transactions in 2005 (49 of 206 transactions).30 was 65 percent), indicating that a substantial share South Africa’s Standard Bank was particularly ac- of this lending represents international transactions tive, as a mandated lead arranger for 28 transac- between affiliates of the same bank. tions in 2005. The above data indicate the growing impor- The regional distribution of South–South syn- tance of certain developing countries as banking Delivered by The World Bank e-library to: dicated lending flows as compared with syndicatedThe World centers Bankfrom which domestic and foreign banks lending flows to developing countries from all operate, but they capture external positions in all IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 sources was broadly similar last year. Borrowers in countries (including high-income countries). Data 119 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 4.4 Cross-border lending to all countries by South–South bank ownership is significant banks in developing countries, 2000–5 Banks from 40 developing countries (most of them $ billions middle-income) hold 5 percent of the $944 billion 100 Turkey Panama dollars in foreign bank assets in developing coun- 90 India Chile tries (based on Bankscope data; see annex 1).33 Ex- 80 Brazil Mexico cluding Panama (an important offshore center), 70 60 the biggest investors are banks in South Africa, 50 Malaysia, and Hungary. The pattern of ownership 40 differs significantly by region. In South Asia, 20 30 percent of foreign bank assets are held by banks in 20 other developing countries.34 In Europe and Cen- 10 tral Asia the same share is just 2 percent (table 0 4.5).35 While these data indicate that participation 2000 2001 2002 2003 2004 2005 by developing-country banks is significant, banks Source: Bank for International Settlements. Note: Yearly data are averages based on quarterly data. from high-income countries still account for 95 percent of total foreign bank assets in developing countries. Moreover, all foreign banks account for from countries that report the destination of their only 16 percent of total banking sector assets in foreign claims (so far only Brazil, Chile, Mexico, developing countries. South–South bank owner- and Panama) indicate that the South–South com- ship thus accounts for less than 1 percent of total ponent is growing.32 For example, foreign claims bank assets in developing countries. Northern for- on developing countries reported by Brazilian eign banks in developing countries—with median banks rose from $1 billion in the fourth quarter of assets of $361 million—tend to be larger than 2002 to $2 billion in the third quarter of 2003, southern foreign banks—with median assets of while Chilean banks’ foreign claims on developing $92 million. Southern bank participation is more countries rose from $176 million to $891 million important in terms of the number of banks. in the same period. The increase in South–South foreign claims by banks from Panama rose only by South–South banking increases opportunities 10 percent, and foreign claims on developing for low-income countries countries by Mexican banks decreased in the last Banks from industrialized countries and develop- two years. However, on average the increase of ing countries alike tend to invest in countries with South–South foreign claims reported by these four which they have strong trade linkages, that share a countries has been more significant than the 58 common language and legal system, and that are percent rise in total North–South foreign claims nearby. But because developing-country banks (from all high-income to all developing countries). have more experience doing business in a challeng- Table 4.5 Source of foreign bank assets, by region % of foreign bank assets in host region owned by banks in other regions Source region East Asia Europe & Latin America Middle East & South Sub-Saharan High-income Host region & Pacific Central Asia & Caribbean North Africa Asia Africa countries Total East Asia & Pacific 6.39 .. .. .. .. .. 93.57 100 Europe & Central Asia .. 1.84 .. 0.01 .. 0.03 98.11 100 Latin America & Caribbean .. .. 4.78 .. .. .. 95.26 100 Middle East & North Africa .. .. .. 8.91 .. .. 91.19 100 South Asia .. .. .. .. 0.74 19.51 79.83 100 Sub-Saharan Africa 0.07 0.03 0.02 0.29 1.99 14.12 83.54 100 Source: World Bank Staff estimates based on Bankscope. Delivered Note: Foreign assets are averages over the 2000–4 by The period. Worldbank A foreign Bankise-library to: defined to have at least 50 percent foreign ownership as of December 2005. The World Bank .. = Negligible. IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 120 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S ing economic environment, they have a compara- Figure 4.5 South–South foreign bank entry in developing countries, tive advantage over industrialized-country banks by country income level when entering low-income countries (box 4.4). As Share of Banks in Total Foreign Banksa Share of Assets in Total Foreign Assetsb a result, low-income countries, which have prob- Percentage Percentage lems attracting bank lending from industrial coun- 50 50 try banks, are benefiting disproportionately from 40 40 the increased supply of banking services from 30 30 other developing countries. 20 20 Cross-border investment by developing- 10 10 country banks is more significant in low-income 0 0 countries (27 percent of foreign bank assets and Low- Middle- All Low- Middle- All income income developing income income developing 47 percent of the number of foreign banks) than countries countries countries countries countries countries in middle-income countries (4 percent of foreign Source: World Bank Staff estimates based on Bankscope. assets and 22 percent of foreign banks) (figure Note: “Southern foreign banks” are those banks headquartered in a developing country. A foreign bank is one that had at least 50 percent foreign ownership as of December 2005. 4.5). The correlation between income level and a. Number of southern foreign banks as a percentage of all foreign banks (left panel). the share of banks from developing countries in b. Bank assets held by southern foreign banks as a percentage of total foreign assets, averaged over 2000–4 (right panel). foreign bank assets is –0.37, which is statistically significant. In addition, low-income countries are also important in South–South syndicated lend- ing; their share increased from 3 percent ($24 mil- was particularly important in India (where 76 per- lion) in 1985 to 17 percent ($1 billion) in 2005, cent of South–South lending was cross-regional), although the vast majority of this latter amount Kazakhstan (83 percent), and the Russian Federa- was concentrated in a few countries in East and tion (77 percent). Important motivations for cross- South Asia (notably, India). regional South–South bank lending include trade financing (which accounted for the vast majority of South–South banking takes place largely cross-regional loans in 2005) and the desire to within the region serve expatriates. In addition to these purposes, Foreign investment and lending by developing- major uses of intraregional loans were the financ- country banks is regionally concentrated. In East ing of acquisitions and other expansion plans (par- Asia and the Pacific, Europe and Central Asia, and ticularly in East Asia) and infrastructural develop- the Middle East and North Africa, practically all ment projects in power, telecommunications, and developing-country foreign banks are from the transport (in both East Asia and Latin America). same region (table 4.5). In Sub-Saharan Africa, The dominance of intraregional cross-border banks from other regions account for only 14 per- banking in part reflects the importance of intrare- cent of developing-country foreign banks. By con- gional trade and FDI flows (discussed earlier) and trast, almost all developing-country foreign banks the priority being given to regional cooperation in South Asia are from Sub-Saharan Africa. How- and integration in policy agendas. In addition, ge- ever, these data reflect ownership by branches and ographic proximity often implies a common cul- holding companies of banks from OECD countries tural heritage, language, or ethnic ties, making it based in Mauritius (an offshore banking center) easier for banks to assume more risk. that own Indian banks. Just as local banks have an advantage over Intraregional transactions are becoming less foreign banks due to their greater knowledge of dominant in South–South cross-border syndicated local conditions and their ability to screen and lending. In 2005, 52 percent of this lending was to monitor local borrowers (Nini 2004), foreign borrowers in the same region as the lenders, down banks from within the same geographic region from 66 percent in 1985.36 Intraregional lending may have an advantage over other nonlocal remained particularly important in East Asia lenders. This greater familiarity means that (where 97 percent of South–South cross-border banks from the same region can lend more than Delivered by The World Bank e-library to: loans are intraregional) and Latin America (83 per-The World nonregional Bank banks and are more likely to expand cent) in 2005. Cross-regional South–South lending beyond the traditional focus on corporate banking IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 121 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Box 4.4 Determinants of South–South foreign bank entry T he economic literature on the determinants of foreign bank entry has not distinguished between foreign ownership by banks from industrial countries and devel- Determinants of foreign bank entry: northern versus southern foreign banks Northern bank Southern bank oping countries. (see, for example, Buch and DeLong 2004; Focarelli and Pozzolo 2000; and Galindo, Micco, Colonial linkages 0.757* 0.699* Border 0.297 0.297 and Serra 2003). However, country studies and anecdotal Common language 0.338* 0.338* evidence suggest that industrial-country banks invest in Distance –0.153* –0.123* developing countries for different reasons than do devel- Trade 0.014* 0.014* oping-country banks. To address this issue, we estimated GDP 0.040* –0.009* a model of decisions by foreign banks to enter develop- Financial sector depth –0.048* 0.008* Different legal system –0.045* –0.045* ing-country markets. We measure foreign bank penetra- Quality institutions 0.006 –0.060* tion, the dependent variable, in terms of the level of total Observations 5,532 assets owned by foreigners. The model is explained in detail in annex 3. Source: World Bank staff calculations. Note: Mean of dependent variable = 0.59 The results (see table) reveal some important simi- * = significant at level of at least 10 percent. larities and differences between the determinants of foreign bank investment in developing countries by industrial-country and developing-country banks: • After controlling for all of the above determinants of FDI, the quality of institutions does not appear to in- • FDI by both industrial-country and developing-coun- fluence the decision by an industrial-country bank to try banks is strongly related to bilateral trade flows, enter a developing country. However, banks from de- one indicator of integration between source and host veloping countries are more likely to enter developing countries. Essentially, banks tend to follow their cus- countries with weak institutions. This result seems to tomers. indicate that banks from developing countries, being • Colonial ties are an important explanation of foreign more familiar with working in domestic environments bank penetration by industrial-country banks, but where institutional development is low, are more less so for developing-country banks. suited to investing in such markets. • A common language, which reduces the cost of for- eign banking, is a significant determinant of foreign The coefficients in the table express the marginal ef- bank entry for both industrial- and developing- fects of the impact of the respective variable on foreign country banks. ownership by northern and southern banks. The marginal • Distance is negatively related to foreign bank entry, effects capture the combined effect of the impact of the ex- but the effect appears to be smaller for banks from planatory variable on the probability of entering the host developing countries than for banks from industrial country and on the amount of FDI. countries. Overall, the model provides support for the conclu- • After controlling for distance, a common border is sions in the literature that FDI in foreign banking is not a significant determinant of foreign bank entry. strongly related to economic integration, common lan- • Banks from industrial countries tend to go to large guage, and proximity; this holds true for both industrial developing countries, while banks from developing and developing-country banks. More interestingly, it ap- countries tend to enter the smaller developing coun- pears that developing-country banks are more likely to in- tries. In addition, the depth of the financial sector is vest in small developing countries with weak institutions, negatively correlated with foreign ownership by in- where industrial country banks are reluctant to go. These dustrial-country banks, but positively with ownership results indicate that FDI decisions are not so much influ- by developing-country banks enced by the absolute amount of risk faced by firms, but • Banks from industrial and developing countries are rather by a given firm’s ability to bear that risk better than equally likely to be deterred from entering a develop- other investors. ing country with a different legal system. For a more detailed discussion, see Van Horen (2006). Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 122 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S to sectors that require new and different sources of and Mulaisho 1999). The directors of the bank information. For example, in some developing were prosecuted for criminal charges for allegedly countries, foreign banks from the same region having received deposits while knowing that the have given more emphasis to providing retail fi- bank was insolvent.39 nancial services (mortgages, consumer loans, debt The available data, however, do not indicate and credit card services, and remittance services that, on average, developing-country banks invest- for expatriates) and loans to SMEs.37 ing in low-income developing countries are signifi- cantly weaker than industrial-country banks that Developing-country banks are not a do the same. The asset quality of developing-coun- significantly greater source of poor asset try banks in these countries is lower than that of quality or management banks from high-income countries, but the differ- Investments in the banking sector of developing ences are not statistically significant (table 4.6). countries by banks from other developing coun- Similarly, indicators of efficiency and operational tries could create instability if those banks were performance in low-income countries are slightly poorly managed or if their asset quality were low. better for northern banks, but not by enough to be As with industrial-country banks, however, the statistically significant. In middle-income countries record of entry into developing-country financial there is some indication that banks from high-in- systems by banks from other developing countries come countries seem to outperform developing- is mixed. For example, Ecobank, a successful pri- country banks, both in asset quality and in effi- vate sector banking group based in 13 countries in ciency and operational performance. However, West and Central Africa, has strengthened the since penetration of the banking sector by devel- banks it has taken over. Standbic, a South African oping-country banks is especially prevalent in low- bank, greatly improved the soundness and effi- income countries, the risks posed by southern for- ciency of the United Commercial Bank of eign banks to their host countries because of Uganda.38 By contrast, several branches of the possible poor capitalization or management are Meridian Bank of Zambia were liquidated after a not significantly greater than similar risks posed major run on its deposits (Rakner, van de Walle, by northern banks. Table 4.6 Performance indicators for northern and southern foreign banks, selected aggregates, 2000–4 Ratios in percentages Asset quality Efficiency and operational performance Memo Loan loss Loan loss Net Return on Cost-to- Net No. of reserves/ provision/net interest average income income/ countries gross loans interest revenue margin assets ratio total assets (banks) Low-income countries North foreign 7.05 15.54 9.47 1.88 65.80 1.77 30 (74) South foreign 6.92 26.11 8.94 0.84 90.90 0.77 30 (63) Middle-income countries North foreign 6.42 27.03 6.38 1.14 73.73 0.81 53 (439) South foreign 11.38 49.12 7.82 –0.35 76.04 0.17 53 (87) All countries North foreign 6.50 25.43 6.86 1.25 72.64 0.94 83 (513) South foreign 9.46 39.54 8.30 0.16 82.26 0.42 83 (150) Source: World Bank staff estimates based on Bankscope. Note: Ratios are calculated for each bank in each country and then averaged for North and South foreign banks separately within an income level. Host and source countries that are offshore banking centers are excluded from the sample. Pairs of entries that are significantly different from each other at the 10% level of significance are shown in bold. The ratio of loan-loss reserves to gross loans indicates how much of the total portfolio has been provided for but not charged off. Given a sim- ilar charge-off policy, the higher the ratio, the poorer the quality of the loan portfolio. Loan-loss provision over net interest revenue is the rela- tionship between provisions in the profit-and-loss account and interest income over the same period. This ratio should be as low as possible. Net interest margin is the ratio of net interest income to earning assets. The higher this figure, the cheaper the funding or the higher the margin the bank is commanding. Higher margins are desirable as long as asset quality is maintained. Return on average assets looks at the returns generated from the assets financed by the bank. The cost-to-income ratio measures the overheads and costs of running the bank as percentage of income generated before provisions. It is a measure of Delivered by efficiency, The World although Bank if the e-library lending to: margins in a particular country are very high then the ratio will improve as a result. Net income to total assets shows the The World Bank profitability of the bank. IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 123 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 South–South banking can strengthen domestic that in some countries the poor investment cli- financial services but may entail some risks mate severely reduces the availability of profitable Even if foreign bank entry does not generate a cap- investment opportunities; in such cases banks’ ital inflow (because subsidiaries may generate their cross-border lending may not reduce the effective funds locally), it can improve the quality and supply of domestic credit. availability of domestic financial services. In- Entry by developing-country foreign banks creased competitive pressure can lead to stronger may increase credit volatility. In general, foreign credit growth, more aggressive provisioning be- banks increase credit volatility if they quickly de- havior, and higher loss-absorption capacity—all of crease their exposure to the country when domes- which can help stabilize domestic banking systems tic conditions deteriorate (Caballero 2002) or re- (Crystal, Dages, and Goldberg 2001). Managerial duce their lending when deteriorating economic and technology spillovers may benefit domestic conditions in their home country reduce their cap- banks, as well. Foreign banks also can help stimu- ital. On the other hand, foreign banks may reduce late the development of the underlying supervisory credit volatility because they are less reliant on er- and legal system by pressuring host-country gov- ratic local deposits—their reputation for sound- ernments to improve institutions, thereby enhanc- ness may attract local deposits during a credit cri- ing the country’s access to the international capital sis, thus reducing outflows from the domestic market (see, for example, Levine 1996). Claessens, financial system. Demirguc-Kunt, and Huizinga (2001) find that Overall, developing-country banks may make greater presence of foreign banks (from high-in- a greater contribution to instability than industrial- come countries) is associated with reductions in country banks. Developing-country banks are profitability, lower noninterest income, and lower more likely to be subject to financial crises in their overall expenses of domestic banks. South–South home country than are industrial-country banks, foreign banking is too recent a phenomenon to and thus are more likely to reduce credit due to permit a judgment about whether entry by banks sharp changes in their capital. For example, banks from developing countries produces the same ef- from Latin America are more likely to react with a fects. It is possible that developing-country banks reduction in credit when they experience a reduc- are less sophisticated in technology and banking tion in real deposits than are banks from developed practices, so that they would not generate the countries (IDB 2002). Furthermore, the less secure same degree of competition and hence not lead to reputations of developing-country banks indicate the same efficiency gains. Alternatively, as argued that they may play a less important role in attract- elsewhere in this chapter, host countries may find ing local deposits during a domestic credit crisis. it easier to adapt technology from other develop- ing countries, thus increasing spillovers. In the ab- sence of empirical work, one can only speculate on which effect may be more important. Developing-country stock exchanges South–South banking has the potential to di- Emerging trends in regional versus rect capital away from the source country, thus international integration reducing the supply of credit available to market A feature common to many nations’ efforts to de- participants that are already credit deprived. This velop their financial sectors over the past several can happen when total lending by participating decades has been the establishment of a national banks is constrained by their available capital or stock exchange—or the expansion of an existing the availability of skilled staff (as opposed to one. It has been argued that such a development being constrained by the lack of investment op- can be an important step toward a modern, well- portunities in the domestic market). As capital is functioning financial sector—as a means of in- scarce in most developing countries, it is widely creasing and improving the allocation of savings presumed that domestic lending is constrained by and investments.40 Many international organiza- capital availability, at least in countries where the tions, including the World Bank, have supported Delivered by The World Bank e-library to: investment climate is adequate to support in- Bankthese efforts (IFC 1991). As a result, there are cur- The World creased economic activity. The fact is, IP however, rently some 85 stock exchanges operating in some : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 124 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S 75 developing countries.41 Many of the exchanges Table 4.7 Stock exchanges in selected developing countries, have very low ratios of market capitalization to December 2005 GDP and are characterized by lack of depth (low Market Market Annual Company capitalization capitalization turnover turnover), inadequate transparency, operational Market listings ($ millions) as % of GDP ratio (%) inefficiency, and poor regulation, calling into ques- Botswana 18 2,438 25 2.1 tion the notion that they contribute to efficient re- Ecuador 32 3,215 98 4.2 source mobilization and allocation. 42 Conse- Ghana 30 1,661 21 3.2 quently, in the past several years, there has been Latvia 45 2,527 16 7.9 Oman 96 15,269 45 31.5 growing interest in the possible advantages of con- Philippines 235 40,153 44 14.0 solidating national stock exchanges in developing Sri Lanka 239 5,720 26 18.3 countries and so addressing the impediments of Trinidad & Tobago 37 16,971 120 3.8 Tunisia 46 2,876 10 8.9 small size, illiquidity, and inadequate market infra- Ukraine 221 24,976 35 2.5 structure (table 4.7). Sources: Standard & Poor’s, Emerging Stock Markets Review (January 2006); Standard & Poor’s Global Stock Markets Factbook, 2005; World Bank database (for GDP data). Limited progress toward regional integration, Note: Annual turnover ratios are calculated by dividing the total value traded in 2004 by but some positive signs average market capitalization for 2003 and 2004. Stock exchanges across developing regions have introduced various initiatives over the past decade to forge closer regional links both intraregionally the focus of intraregional initiatives in recent years and, in some cases, extraregionally. Thus far, how- has been bond markets—via the ASEAN+3 initia- ever, actual progress toward merging or integrat- tives to develop an intraregional bond market. But, ing stock exchanges among developing countries so far, although issues of foreign currency–denomi- has been limited. nated bonds by Asian sovereigns and private firms Many developing-country capital markets re- have increased, most tend to be denominated in U.S. main more integrated with the major international dollars, and most of the investment in these issues is financial markets than with other developing sourced from Europe or the United States—albeit countries. In part, this is due to a lack of intrare- with a significant amount coming from Asian in- gional harmonization of tax, accounting, disclo- vestors residing there.46 sure, and other stock-market listing and trading In Latin America, by contrast, recent efforts regulations and procedures. In Asia, for example, to develop capital markets have focused on the eq- stock markets remain fragmented and poorly inte- uity markets and have included some plans that grated, and cross-border listings between develop- take an intraregional approach. The region’s two ing-country exchanges remain uncommon.43 Over- largest exchanges, in Mexico and Brazil, signed an seas listings by companies domiciled in Asian agreement in 2005 that will soon allow cross-bor- developing economies are still more likely to take der investments in shares on their exchanges. Since place via depositary receipt and other issues on de- the 1990s, the MERCOSUR countries have taken veloped-country exchanges, particularly in Hong steps to encourage more cross-border trading in Kong (China), Singapore, Japan, New York, Lon- the markets of Argentina, Brazil, Paraguay, and don, and, increasingly for South Asian firms in re- Uruguay. Nevertheless, the actual volume of cross- cent years, Luxembourg.44 Cross-border listings border listings and investment in intraregional se- by firms in southern Africa on the Johannesburg curities between developing countries in Latin and other national exchanges in the subregion are America remains small.47 45 not uncommon. However, many of the largest Steps to increase intraregional cooperation— South African companies moved their primary list- rather than outright integration—as a means of ings from Johannesburg to the London Stock Ex- developing national capital markets are increas- change (particularly during the 1990s), citing a ingly evident, particularly in the form of an in- need for access to a much larger capital market. crease in agreements between developing-country Neither Asia nor Latin America has taken a stock exchanges to encourage more cross-border Delivered by The World Bank e-library to: strong intraregional approach—at least in practice—The Worldlistings and investment, information and technol- Bank toward developing national equity markets. In Asia, ogy sharing, training, and staff exchanges. Some IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 125 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 4.6 Developing-country firms shift away At the same time, increased regulatory and from ADRs disclosure requirements in industrial-country mar- No. of new ADRs/delistings on U.S. stock exchanges kets, and their associated costs, are giving some 30 impetus to local initiatives to develop capital mar- Sub-Saharan Africa South Asia kets, including those taking an intraregional ap- 25 Latin America proach. More costly and complicated documenta- Middle East & North Africa tion requirements, and significantly increased 20 Europe & Central Asia human resource and other capacity requirements East Asia & Pacific for compliance with the more stringent reporting 15 standards of Section 404 of the U.S. Sarbanes- Oxley Act of 2003, have coincided with an appar- ent decline in the attraction of an overseas listing 10 on a U.S. exchange in recent years—particularly for companies based in Latin America, and also 5 for many companies based in Asia (see figure 4.6).48 New issues of depositary receipts by Latin 0 American firms on U.S. exchanges declined from Delistings Delistings Delistings Delistings Delistings Delistings New ADRs New ADRs New ADRs New ADRs New ADRs New ADRs 11 in 2000 to none in 2005.49 Moreover, there were six delistings of ADRs in 2005, five of which 2000 2001 2002 2003 2004 2005 involved Latin American firms. At the same time, more companies in middle-income countries in Source: Bank of New York Depositary Receipts Division. Latin America and elsewhere have made initial public offerings (IPOs) or other forms of share is- sues in recent years (see also figure 2.9). of these agreements also promote joint efforts to develop new financial products and develop the More must be done to improve financial stock-brokerage profession. A growing number of intermediation at the national level such cooperation agreements has been signed with Regional cooperation and, possibly at a later exchanges outside the region—in developed as stage, integration could improve the liquidity, effi- well as other developing countries. ciency, and competitiveness of securities exchanges in developing countries. But for many emerging Signs of a move away from American markets, further progress in developing well-func- Depositary Receipts (ADRs) and toward tioning national securities markets (and financial more local listings markets generally) is needed ahead of moves to in- Developing-country firms may be less likely in the tegrate those markets. Hasty integration of several future to list on major international financial cen- small, illiquid national stock markets would likely ters’ markets than on domestic markets. In part, create nothing more than a large, illiquid regional this is due to the recovery of trading activity and market. Short of full integration, underdeveloped share prices in developing-country stock mar- national exchanges could meanwhile benefit from kets—reversing the downturns of the late 1990s the steps they have been taking to encourage closer (see box 2.2). That recovery has been driven by cooperation, including through cross-border list- rapid economic growth and greater corporate ings and investment, and through information and earnings, as well as by local stock-market regula- technology sharing.50 More intraregional trading tory reforms to increase local trading activity, at- activity could also facilitate the privatization of tract more investors and issuers to local and re- large corporations, by providing a market for gional markets, and improve efficiency and large share issues that could not be absorbed on a competitiveness. There also is an ongoing effort— national basis. apparent across all developing-country regions— Beyond general progress in strengthening na- Delivered by The World Bank e-library to: to bring financial reporting and disclosure stan- Banktional financial markets, several steps are impor- The World dards more in line with international standards. tant at the national level to facilitate eventual IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 126 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S cross-border integration. Countries participating bling those data from available sources is an arduous in cross-border trades must have convertible cur- task (see annex 1). Moreover, very little research has rencies and would have to liberalize those remain- been done on South–South financial integration. In ing controls and other restrictions on capital flows part this reflects the relative novelty of developing that impede cross-border trading, payments, and countries as a significant source of capital, in part the settlements. Harmonizing regulatory and policy absence of data, and in part the desire of develop- frameworks would facilitate cross-border listing ment economists to focus their energies on the princi- and investment and would be a prerequisite to ac- pal source of capital flows to developing countries tual integration.51 (the high-income countries). We hope that this foray into South–South cap- ital flows will draw greater attention to developing countries as a source of capital. Greater efforts to Conclusion collect data are essential to progress. Further em- A vailable data indicate that more developing countries are lending to and investing in other developing countries. The expansion of South–South pirical research could focus on (1) the extent of spillovers from South–South FDI and how these differ from spillovers from North–South FDI; (2) capital flows reflects both the general growth of the impact of government impediments to, and in- cross-border financial transactions in the wake of centives for, outward investment in developing globalization and the increasing size and sophistica- countries; (3) the impact of developing-country tion of developing-country banks and multination- banks on macroeconomic instability in their for- als. Greater South–South flows promise greater re- eign markets, including the extent to which devel- sources for low-income countries, a more efficient oping-country banks transmit crises from source allocation of capital by lenders and investors famil- to host country and whether the quality of man- iar with developing-country conditions, and poten- agement and financial soundness of internation- tially greater transmission of technology and know- ally active developing-country banks differs how from FDI. greatly from high-income country banks; (4) the The potential benefits of greater South–South in- circumstances under which efforts to increase the tegration are supported by anecdotes, a few empirical integration of regional capital markets are likely to studies, and deduction and inference from the history improve their efficiency; and (5) circumstances of North–South capital flows, rather than by a large under which regional trade agreements and other body systematic research. The fact is that the data on forms of regional integration have a positive im- South–South capital flows are limited, and assem- pact on economic growth and development. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 127 (c) The International Bank for Reconstruction and Development / The World Bank Chapter 4 Annexes Annex 1: Data on South–South pation in a loan is taken directly from Loanware. capital flows Where loan-allocation details for a particular loan M ost countries do not routinely publish data on capital flows by source country. Thus it is not possible to rely on official sources to calculate transaction are not disclosed, an estimate of the cross-border South–South lending component for that transaction is derived by multiplying the total the portion of capital flows to developing countries transaction amount by the average share of non- that come from other developing countries. In con- local South–South lending in syndicated loan structing a database on South–South capital flows, transactions with some portion of developing- we have relied on a variety of sources, including the country bank participation arranged for borrow- Bank for International Settlements (BIS), Loanware, ers in the region that year.52 Bankscope, the United Nations Conference on Trade and Development (UNCTAD), the Organisa- Measuring South–South foreign tion for Economic Co-operation and Development bank ownership (OECD), the International Monetary Fund (IMF), Data on foreign banks in developing countries, as and the World Bank. well as related financial variables, are based on Bankscope and include all active commercial banks, Measuring South–South loans saving banks, cooperative banks, bank holding BIS recently has begun publishing data on lending by companies, and middle and long credit banks that banks domiciled in some developing countries. were available in Bankscope as of December 2005. However, data are available only since 2000, and When ownership information is not available in only for five countries (Brazil, Chile, India, Mexico, Bankscope, information is gathered from banks’ and Turkey). Moreover, data are available for all of Web sites or other Internet sources.53 We determine these countries only since 2003. While the BIS data whether each bank is foreign-owned, that is, do provide some indication of the role of banks in whether at least 50 percent of the bank’s shares are developing countries as lenders, they cannot provide owned by foreigners. In addition, the percentage of a very complete picture of South–South lending. shares are summed by country of residence of the Most of our analysis of South–South lending, shareholder, and the country with the highest per- therefore, is based on data on syndicated loans centage of shares is appointed as the source country. obtained from Loanware, although considerable Ownership is based on the direct ownership struc- work was required to calculate the share of ture; indirect ownership is not taken into account. South–South transactions. While many transac- Countries with fewer than five active banks in tion entries detail the allocation of loans among Bankscope were excluded from the sample. In ad- all participating banks, others do not, depending dition, Guatemala was excluded, as ownership in- on the disclosure practices of particular syndi- formation was available for only a small portion Delivered by The World Bank e-library to: cates. Where participation by all banks The dis- Bankof the country’s banks. We were left with a sample isWorld closed, nonlocal developing-country bank IP partici- of 103 developing countries. In total, the database : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 128 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S provides us with information on ownership and flows (to the 35 developing countries) are then ap- related financial variables for 2,297 banks, of proximated by FDI inflows in developing coun- which 35 percent are foreign owned. tries that are not from developed countries (Aykut and Ratha 2004). Measuring South–South foreign The estimation technique suffers from the sev- direct investment eral weaknesses, some of which will lead to an un- Developing countries do not report the source of derestimation, some an overestimation, of FDI inflows. Therefore, data on South–South FDI South–South FDI. First, FDI outflows to developing flows are calculated by comparing total FDI in- countries may be underreported by the high-income flows to developing countries with FDI outflows countries. It is likely that a portion of the FDI out- from high-income to developing countries; the dif- flows that are not identified by country go to devel- ference is South–South FDI flows. First, FDI out- oping countries, which would imply an overestima- flows from high-income countries to developing tion of South–South FDI. Second, FDI inflows are countries are calculated. For high-income OECD likely to be underreported by some developing coun- countries, the OECD provides data on FDI out- tries, which would imply that our data are underesti- flows to 35 developing countries that account for mates of South–South FDI. Third, round-tripping of 85 percent of all FDI inflows to developing coun- flows (the export of capital to a foreign country for tries. For high-income countries that are not part the purpose of investment back in the home country, of the OECD, including several offshore centers, often to benefit from tax incentives) will lead to data on FDI outflows are taken from the IMF and overestimation of South–South FDI flows. Fourth, UNCTAD. Since detailed destination data are not transactions channeled through offshore financial available, we assume that all of the FDI outflows centers may be misclassified as FDI. Fifth, FDI from from high-income non-OECD countries went to the North may be channeled through a developing developing countries. (This assumption leads to an country to another high-income country (indirect underestimation of South–South FDI flows.) Sec- FDI flows), causing an overestimation of ond, data on FDI inflows (to the 35 developing South–South flows. And finally, relying on a sample countries covered by the OECD database) are of 35 developing countries may lead to an underesti- taken from the World Bank. South–South FDI mation of the level of South–South flows. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 129 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Annex 2: Selected South–South M&A deals by southern multinationals in service sector, 2000–5 In services sector Value Year Acquiring company Country Acquired company Country Sector ($ millions) 2005 America Movil Mexico TIM Peru Peru Telecommunications 500 2004 Anglogold Ltd South Africa Ashanti Goldfields Ghana Gold ores 1500 2004 Sinergy Brazil Avianca Colombia Air transportation 400 2004 CEZA.S. Czech Republic Capital Electricity Colombia Bulgaria Electric services 400 2004 Teléfonos de Mexico Mexico Telecomunicaciones Colombia Telecommunications 400 2004 Teléfonos de Mexico Mexico Embratel Brazil Telecommunications 400 2004 Vempelcom Russia Kar-tel Kazakhstan Telecommunications 400 2004 YTL Power Malaysia Jawa Power Indonesia Electric services 200 2004 Teléfonos de Mexico Mexico Chilesat Chile Telecommunications 130 2004 Teléfonos de Mexico Mexico Techtel Argentina Telecommunications 100 2002 Vodacom South Africa Vodacom Mozambique Mozambique Telecommunications 260 2002 Ressano Garcia Railways company South Africa Caminhos de Ferro Mozambique Mozambique Cyclical services 78 2001 MTN South Africa MTN Nigeria Telecommunications 285 2001 Teléfonos de Mexico Mexico Comcel Columbia Telecommunications 257 2001 Industrial Development Corporation South Africa Mozal II Mozambique Basic industries 160 2001 Vodacom South Africa Vodacom Congo Republic of Congo Telecommunications 142 2000 Orascom Egypt Telecel 12 African countries Telecommunications 413 2000 Teléfonos de Mexico Mexico ATL Brazil Telecommunications 345 2000 Teléfonos de Mexico Mexico Conecel Ecuador Telecommunications 153 1998 Teléfonos de Mexico Mexico TelGua Guatemala Telecommunications 700 In extractive sector Location of the Value Year Acquiring company Country Acquired company Country acquired asset ($ millions) 2005 Andes Petroleum China EnCana Canada Ecuador 1420 2005 CNPC China Petro Kazakh Canada Mainly in Kazakhstan 4180 2005 CNOOC China MEG Energy Canada Canada 120 2005 Sinopec Group (50%) China-India National Iranian Oil Company Iran Yadavaran Oil Fields $70–100 and ONGC (20%) in Iran billion over 30 years 2004 CNPC China Plus Petrol Norte Peru 200 2004 Gazprom Russia Lietuvos Lithuania Lithuania 50 2004 Metorex South Africa Ruashi Mining D. R. Congo D. R. Congo 86 2004 Rangold Resources South Africa Loulo Concessions Mali Mali 80 2004 Rangold Resources South Africa Licences and assets Angola Angola 15 2003 CNOOC China Tangguh LNG project Indonesia 275 2003 CNPC China Oil field Kazakhstan N Buzachi 200 2003 Investor Group China Amerada Hess Indonesia 164 2003 Sinochem China Ecuador Block 16 Ecuador 100 2003 Lukoil Russia Beopetro Serbia Serbia 130 2003 AngloGold South Africa Ashanti Ghana Ghana 274 2003 Impala Platinum South Africa Zimbabwe Plat. Mes Zimbabwe Zimbabwe 85 2003 Impala Platinum South Africa Hartley Platinum Mines Zimbabwe Zimbabwe 80 2003 Impala Platinum South Africa Platinum mines Zimbabwe Zimbabwe 19 2003 Sasol South Africa Escravos gas to liquid plant Nigeria Nigeria undisclosed 2002 CNOOC China Repsol YPF SA Spain Indonesia 591.9 2002 PetroChina Corp China Devon Energy — Indonesia 262 2002 Escom Holding South Africa Grand Inga Falls D. R. Congo D. R. Congo 1200 2001 Saso Oil South Africa Pande Teemanegasfields Mozambique Mozambique 581 2000 AngloGold South Africa Ashanti Goldfields Tanzania Tanzania 83 1998 China National Petroleum Corp China Oil Field R. B. de Venezuela 240.7 1997 China National Petroleum Corp China Aktyubinskmunaygaz Kazakhstan 325 Source: UNCTAD and news sources. Note: — denotes not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 130 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S Annex 3: Model of determinants Notes of bank ownership 1. See annex 1 for the methods used to compile data on South-South transactions. T he following is an explanation of the model used in box 4.4. To test the differences be- tween determinants of foreign bank entry in devel- 2. Data on bilateral remittance flows are not available. The estimate in figure 4.1 assumes that bilateral remittances are a function of the stock of migrants in the sending coun- oping countries by banks from developing coun- try. This estimate is consistent with the fact that nearly half tries and from high-income countries, we estimate of the migrant stock from the South migrate to another country in the South. the following model using Tobit: 3. It is difficult to obtain data on foreigners’ purchases of stock issues. But see figure 2.14 on initial public offerings FCij = α1Collinksij + α2Collinksij * DS in emerging markets. + β1Borderij + β2Borderij * DS 4. The G-20 and G-90 groups were formed at the time of the WTO ministerial in Cancun in September 2003. The + γ1Comlangij + γ2Comlangij * DS G-20 includes some of the larger developing countries, + δ1Distij + δ2Distij * DS + κ1Tradeij while the G-90 is made up of countries from the African, + κ2Tradeij * D + λ1GDP + λ2GDP S Caribbean, and Pacific (ACP) group, the African Union, and * D + µ1Findepth + µ2Findepth S the least developed countries. * D S + ϕ1 Legaldif + ϕ2 Legaldif * D S 5. According to the index published by the Heritage + θ1Inst + θ2Inst * DS + ρ1Entryres Foundation. See http://www.heritage.org/research/ features/index/downloads.cfm. + ρ2GDPsource + ρ3GDPcapsource 6. These are unweighted averages. The average for + ρ4Dregion + τ1constant + εij high-income countries includes non-OECD countries. 7. The discussion of RTAs is taken from World Bank The dependent variable is defined as the ratio (2005b). 8. World FDI in services quadrupled between 1990 of the sum of assets of banks in host country i of and 2002 (UNCTAD 2004). By 2002, the services sector which a source country j owns 50 percent or more accounted for 70 percent and 47 percent of FDI stock in equity, divided by the total amount of banking as- developed and developing countries, respectively (World sets in host country i. Collinks is a dummy with a Bank 2004). 9. The services sector includes electricity, gas, water, value of 1 if the host and source countries have transport, communication, construction, wholesale and re- had colonial links either between colonizer and tail trade and repairs, hotels and restaurants, transport, colony or between those countries colonized by storage and communications, finance and insurance, real es- the same colonizer. Ds is a dummy with a value of tate, renting, and business services, public administration, 1 if both host and source country are a developing defense, education, health, social services, social and per- sonal service activities, and recreational, cultural, and sport- country. Border is a dummy with a value of 1 if ing activities. Not all services are nontradable or require the countries share a border. Comlang is a dummy physical proximity. with a value of 1 if the countries share the same 10. For example, America Movil (Mexico) bought out language. Dist refers to the log of the distance be- the shares of its partners (SBC and Bell Canada) in Brazil tween the host and source countries. Trade is the and of its partner (Bell Canada) in Colombia in 2002. 11. See Goldstein (forthcoming) and Pradhan (2005). log of exports plus imports in 2000 between the 12. Examples include the Indian R&D center of Chi- two countries. GDP is the log of the host country’s nese white goods producer Haier, and Russian design and GDP in 2000. Findepth is the log of M2 as a per- R&D centers for the shipping industry and drilling plat- centage of GDP in the host country in 2000. forms (Vahtra and Liuhto 2004). Legaldif is a dummy with a value of 1 if the origin 13. The extractive industries also attract a large share of developed-country FDI in Africa. In 2002, 53 percent of of the legal system of the host and source countries FDI from four major developed-country investors in Africa differs. Inst is the simple average of six indicators (France, the Netherlands, the United Kingdom, and the of quality of institutions in the host country in United States) was in the extractive sector (World Bank 2000 as measured by Kaufmann, Kraay, and Mas- 2004, figure 3.6). truzzi (2005). Entryres is a dummy with a value of 14. China has partnerships or investments in oil and gas exploration projects in Cuba, Peru, and República Boli- 1 if foreign bank entry is restricted. GDPsource variana de Venezuela. and GDPcapsource are the logs, respectively, of 15. For example, a Turkish soap and detergent pro- GDP and GDP per capita in the source country in The World Delivered by (Evyap) ducer Bank opened e-library to: factories in Egypt and Ukraine and is 2000. Dregion are dummies for each region. The World Bankto open one in Russia to escape uncompetitive planning IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 131 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 labor costs at home and growing competitive pressures in 30. A mandated lead arranger is a bank (or banks) re- these markets (IMF–World Bank 2005). Mauritius has re- sponsible for originating, structuring, and syndicating a ceived significant FDI in the textile and clothing sector but loan transaction. moved part of its production to lower-cost neighboring 31. This includes international transactions of the Madagascar and Mozambique in response to cost pressures banks with any of their own affiliates and with Panama, an from Asia (Goldstein 2003). offshore center. Excluding Panama, cross-border lending 16. In June 2005, India’s Ranbaxy won approval to originating from developing countries amounted to $77 bil- make lamivudine tablets for Africa under the U.S. Presi- lion in 2005 dent’s Emergency Plan for AIDS Relief. 32. Foreign claims include cross-border loans by the 17. Examples of SOEs in other sectors include bank’s head offices or its affiliates, and local loans by affili- Telekom Malaysia, Eskom, and Transet of South Africa. ates located in another country 18. For example, South African SOEs have invested in 33. Total assets are averaged over 2000-4. These num- Africa in part to promote the New Partnership for African bers include offshore centers. Excluding FDI in and from Development (UNCTAD 2005a). offshore centers, developing-country banks hold 3 percent 19. In July 2005, China’s CNPC was awarded four oil of foreign bank assets in developing countries. blocks in Nigeria in exchange for investing in the construc- 34. The data for South Asia reflect banks domiciled in tion of a hydropower plant (“China Goes Shopping,” Fi- Mauritius (an offshore banking center), most of which nancial Times, March 8-16, 2005). owned by banks from high-income countries that have set 20. SMEs have 1,000 or fewer employees (OECD up subsidiaries in India. 2005a). 35. Excluding FDI to and from offshore centers, Sub- 21. Since 1998, the Tata Group has been selling a fam- Saharan Africa shows the highest percentage of developing ily sedan for $4,000 to $6,000. It announced plans to intro- countries’ banks in total foreign bank entry (13.3 percent), duce a $2,000 car by 2008 (“Getting the Best to the followed by East Asia and the Pacific (10 percent), and the Masses,” Business Week, October 11, 2004). Middle East and North Africa (6.6 percent). In the other re- 22. Positive spillovers are benefits that the domestic gions South-South activity accounts for less than 2 percent economy enjoys but does not pay for, due to the presence of of FDI in the banking sector. foreign firms. Such benefits may include the availability of 36. Some banks, such as India’s Bank of Baroda and information and technology or the increased supply of the State Bank of India, Jordan’s Arab Bank, and the Bank trained workers (where, because of job mobility, the foreign of China have been active participants in cross-border syn- firm does not capture the full return to training). dicated transactions for borrowers outside their regions 23. Some of these initiatives are the OECD Guidelines since at least 1985. for Multinational Enterprises, the OECD Convention 37. Examples include plans by a number of Kazakh Against Bribery of Foreign Public Officials in International banks to offer financial leasing services (a growing financial Transactions, and various initiatives that promote trans- product geared to SMEs) in Eastern Europe and Central parency in the extractive industries. Asia. Evidence on the kinds of financial services provided by 24. A survey of 200 outward investors from Eastern developing-country banks can be found in Capital Intelli- Europe and Central Asia (Sevtlicic and Rojec 2003) showed gence (various country reports through the end of 2005), that most companies that have invested abroad—mainly in The Banker (various issues in 2005), and information pro- other developing countries—increased exports and im- vided on the banks’ Web sites. proved their financial performance. In India, outward in- 38. This discussion is based on conversations with vestment enhanced the export performance of SMEs in World Bank staff. manufacturing, compared with those that did not invest 39. The prosecution was reported in Zambia News On- abroad (Prahdan 2005). line. http://www.africa.upenn.edu/Newsletters/zno24.html 25. See The Banker (2005), Global Finance (2004 and 40. Engberg (1975) saw a role for capital markets in 2005), Capital Intelligence (2004 and 2005), EIU Country raising domestic savings and contributing to their more effi- Finance (2004 and 2005), Latin Finance (2005), and infor- cient allocation, even in less developed economies. Engberg mation posted on various bank Web sites. also argued that the broader range of financial assets associ- 26. Data reflect participation by nonlocal developing- ated with capital market development could raise personal country banks in cross-border syndicated lending to bor- savings rates. Levine (1990) showed that a stock market can rowers based in developing countries (see annex 1). positively impact growth by providing a means of trading 27. This increase is due in part to the rise in the num- the ownership of firms (shares) without disrupting the oper- ber of countries following the breakup of the Soviet Union. ating and productive processes within those firms and by Seven of the former Soviet republics received syndicated providing a way for investors to diversify their portfolios. lending in 2005. See also Demirgüc-Kunt and Maksimovic (1996); Boyd and 28. Examples include the State Bank of India and Smith (1998); Levine and Zervos (1998); Arestis, Demetri- Oman’s Bank Muscat. ades, and Luintel (2001). 29. In a sample of 1,143 cross-border syndicated loan 41. The number of countries with a stock exchange is Europe accounted transactions, local banks in eastern Delivered for Bank by The World e-librarygreater actually to: than 75, but several exchanges are inactive 13 percent of the total loan amount, and localThe banks in Bankor have negligible trading activity. World Latin America for 16 percent (Nini 2004). IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 132 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S 42. Forty-six of the 80 stock markets categorized as its reporting and disclosure requirements for companies that “emerging markets” in Standard & Poor’s Global Stock list on EU stock exchanges, including through a transparency Markets Factbook 2005 had a market capitalization of $10 directive slated to take effect in 2006. billion or less in October 2004. In contrast, just 3 of the 29 49. Although a Chilean firm issued new ADRs in developed-economy stock exchanges had a market capital- 2005, this transaction was an exchange of existing deposi- ization of $10 billion or less. Stock markets in many devel- tary receipts due to a company merger. oping economies rival those in developed economies when 50. The impact of South–South cross-border listings viewed in terms of the ratio of market capitalization to on developing countries’ stock exchanges is an important gross national income, however. Market capitalization is area for research, given the increasing number of agree- only one factor in determining the relative level of develop- ments between developing countries’ stock exchanges that ment of a stock exchange (Standard & Poor’s 2005). encourage cross-border listings and investment. 43. According to the IMF’s Asia-Pacific Outlook, Sep- 51. This would involve harmonizing not only stock- tember 2005, at least 95 percent of the listings on Asian na- market regulations, listing requirements, and procedures for tional stock exchanges are local listings. trading, clearing, and settlement, but also transaction fees, 44. Indian firms issuing global depositary receipts accounting and disclosure standards, corporate governance (GDRs) on the Luxembourg Stock Exchange (citing cost, standards, common standards for stockbrokers, and na- time, and marketing advantages) accounted for the majority tional rules for capital gains and withholding taxes. Such ef- (23) of the 42 total depositary receipts newly issued on the forts, as well as the development of common infrastructure main depositary receipt listing markets in 2005 (the United and systems, may have to address limitations in national States, the United Kingdom, and Luxembourg). The is- markets, such as poor institutional capacity for enforcing suance of GDRs by developing-country firms may improve regulations, rudimentary stock-market infrastructure, poor efficiency in the home market due to increased competitive and unreliable access to information and communications pressures on standards, procedures, and operations, but it technology, and exchanges at significantly different stages of may also impose costs due to diversion of order flow development. A regional securities regulatory body would abroad. The net impact on market liquidity and capitaliza- be essential if integration were to proceed to the point of tion from cross-border listings may depend on the propor- forming a regional exchange. tion of trading volume that shifts overseas, relative sizes of 52. For example, the South–South cross-border lend- the home and overseas markets, and changes, following the ing component of a qualifying syndicated loan (“loan A”) cross-border listing, in the extent of home-market segmenta- for a borrower in East Asia in 2005 that does not reveal tion due to investment barriers and intermarket information loan-allocation details is estimated by multiplying the av- transparency (Hargis and Ramanlal 1996; Hargis 1997; and erage share (15 percent) of nonlocal South–South lending Domowitz, Glen, and Madhavan, 1998). More recent re- in all qualifying transactions for East Asia that reveal loan- search (Karolyi 2004) found that an increase in issues of allocation details by the total “loan A” transaction American Depositary Receipts (ADRs) by firms in an amount. A qualifying transaction is defined for this pur- emerging market economy may be a result, rather than a pose as a syndicated loan disbursed to a borrower in a de- cause, of deteriorating local market conditions. veloping country, whereby one or more banks domiciled in 45. More than 70 percent of the equities listed on the other (nonlocal) developing countries participate in the Namibia Stock Exchange (NSX) are dual listed on the Jo- syndicate. In cases where loan-allocation details are un- hannesburg Stock Exchange, and the vast majority of NSX available for all qualifying syndicated transactions in a trading takes place in these dual listed stocks (Johannesburg particular region, as in Latin America in 1985 and 1995 Securities Exchange 2005). For a region-specific assessment and in the case of all regions in 1985 (with the exception of whether cooperation and integration of stock exchanges of two transactions), the estimate is derived from an aver- in southern and eastern Africa could offer a way of over- age of all transactions that provide loan-allocation data coming impediments to the development of these exchanges, for the region in the time series. see Irving (2005). 53. Currently our sample does not include Costa Rica, 46. Bank for International Settlements, 2005. the Dominican Republic, or Panama. 47. Despite a significant amount of foreign investment in securities traded on the region’s two largest exchanges, in Brazil and Mexico, the vast majority of it comes from devel- oped economies. References 48. In October 2005 China Construction Bank, which had reportedly been considering a listing on the NYSE, opted Aizenman, Joshua, and Ilan Noy. 2005. “FDI and Trade— instead to list on the Hong Kong, China exchange, with an Two Way Linkages?” NBER Working Paper 11403, Na- IPO of $8 billion—China’s largest to date and the largest tional Bureau of Economic Research, Cambridge, MA. worldwide since 2001. In the past few years, the international Albuquerque, R., N. Loayza, and L. Servén, L. 2005. “World financial press has contained numerous additional reports of Market Integration through the Lens of Foreign Direct firms domiciled in developing countries that have abandoned Investment.” Journal of International Economics 66 (2): plans to list on the major U.S. exchanges and, to some by The World267–95. extent, Delivered Bank e-library to: on the London Stock Exchange, because of more onerous list-The World Amorim,BankCelso. 2005. “Brazil Redraws the Trade Map.” ing requirements and associated higher costs. The European Global Agenda. http://www.globalagendamagazine.com/ IP : 192.86.100.36 of10 Mar 2009 Tue, Union also has been taking steps to increase the stringency 16:55:33 2005/celsoamorim.asp. 133 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Arestis, Philip, Panicos Demetriades, and Kul Luintel. 2001. Transition by Internalization: Outward Direct Invest- “Financial Development and Economic Growth: The ment from Central European Countries in Transition, Role of Stock Markets.” Journal of Money, Credit, ed. M. Svetlicic and M. Rojec. Aldershot, Hampshire, and Banking 33: 16–41. UK: Ashgate. Aykut, Dilek, and Dilip Ratha. 2004. “South–South FDI Engberg, H. L. 1975. “The Nairobi Stock Exchange: An Or- Flows: How Big Are They?” Transnational Corpora- ganized Capital Market in a Developing Country.” tions 13 (1). Journal of Management Studies 8. Baldwin, Robert, and L. Alan Winters, eds. 2004. Chal- Erdilek, Asim. 2005. “Case Study on Outward FDI by En- lenges to Globalization: Analyzing the Economics. terprises from Turkey.” Background paper for UNC- Chicago: University of Chicago Press. TAD conference on “Enhancing the Productive Capac- Banga, Rashmi. 2003. “Impact of Government Policies and ity of Developing Country Firms through Investment Agreements on FDI Flows.” Working Internationalization,” Geneva, December 5–7. Paper 116, Indian Council for Research on Interna- FIAS (Foreign Investment Advisory Service). 2005. “Survey of tional Economic Relations, New Delhi. Chinese MNCs.” International Finance Corporation, Bank for International Settlements. 2005. BIS Quarterly Re- Foreign Investment Advisory Service, Washington, DC. view (June). Financial Times. 2004. “Chinese Companies Acquire a Boyd, J., and B. D. Smith. 1998. “The Evolution of Debt Taste of Western Targets.” October 19, 2004 and Equity Markets in Economic Development.” Eco- Focarelli, D., and A. Pozzolo. 2000. “The Determinants of nomic Theory 12: 519–60. Cross-Border Bank Shareholdings: An Analysis with Brealey, R., and E. C. Kaplanis. 1996. “The Determination Bank-Level Data from OECD Countries.” Proceedings of Foreign Banking Location.” Journal of Interna- (Federal Reserve Bank of Chicago, May): 199–232. tional Money and Finance 15: 577–97. Galindo, Arturo, Alejandro Micco, and Cesar Serra. 2003. Buch, Claudia M., and Gayle Delong. 2004. “Cross-Border “Better the Devil That You Know: Evidence on Entry Bank Mergers: What Lures the Rare Animal?” Journal Costs Faced by Foreign Banks.” IADB Working Paper of Banking and Finance 28: 2077–102. 477, Inter-American Development Bank, Washington, Capital Intelligence. Various years. “Country Banking Re- DC. ports.” http://www.ciratings.com/. Giroud, Axele. 2005. “Chinese Outward FDI.” Background Caballero, R. 2002. “Coping with Chile’s External Vulnera- paper for UNCTAD conference on “Enhancing the bility: A Financial Problem.” Unpublished paper, Mass- Productive Capacity of Developing Country Firms achusetts Institute of Technology, Cambridge, MA. through Internationalization,” Geneva, December 5–7. Chudnovsky, Daniel, and Andres Lopez. 2000. “A Third Global Finance. Various issues. http://www.gfmag.com. Wave from Developing Countries: Latin American TNCs Goldstein Andrea. 2004. Regional Integration, FDI, and in the 1990s.” Transnational Corporations 9 (2). Competitiveness in Southern Africa. Paris: Organisa- Claessens, Stijn, Asli Demirgüc-Kunt, and Harry Huizinga. tion for Economic Co-operation and Development. 2001. “How Does Foreign Entry Affect Domestic ———. Forthcoming. Emerging Multinationals in the Banking Markets?” Journal of Banking and Finance Global Economy: Data Trends, Policy Issues, and Re- 25: 891–911. search Questions. London: Palgrave Macmillan. Claessens, Stijn, and Neeltje Van Horen. 2006. “Location Goldstein, Andrea, Nicolas Pinaud, Helmut Reisen, and decisions of foreign banks and competitive advantage.” Michael-Xiaobao Chen. 2006. “The Rise of China and Unpublished paper. World Bank, Washington, DC. India: What Is in It for Africa?” Development Center Crystal, J., G. Dages, and L. Goldberg. 2001. “Does Foreign Paper, Organisation for Economic Co-operation and Ownership Contribute to Sounder Banks in Emerging Development, Paris. Markets? The Latin American Experience.” Federal Re- Grosse, R., and L. G. Goldberg. 1991. “Foreign Bank Activ- serve Bank of New York Staff Reports 137, New York. ity in the United States: An Analysis by Country of Ori- Demirgüc-Kunt, A., and V. Maksimovic. 1996. “Stock Mar- gin.” Journal of Banking and Finance 15: 1092–1112. ket Development and Firm Financing Choices.” World Hallward-Driemeier, Mary. 2003. “Do Bilateral Investment Bank Economic Review 10: 341–70. Treaties Attract Foreign Direct Investment? Only a De Sol, Patricio. 2005. “Why Join a Chilean Firm to Invest Bit—and They Could Bite.” Policy Research Working Elsewhere in Latin America?” Unpublished paper, Pon- Paper 3121, World Bank, Washington, DC. tificia Universidad Católica de Chile. http://www.inge- Hargis, K. 1997. “ADRs in Emerging Equity Markets: Mar- nieriaindustrial.cl/documentos/182.pdf ket Integration or Fragmentation?” Working Paper Domowitz, Ian, Jack Glen, and Ananth Madhavan. 1998. D97-17, Center for International Business Education “International Cross-Listing and Order Flow Migra- and Research, University of South Carolina, Columbia. tion: Evidence from an Emerging Market.” Journal of Hargis, K., and P. Ramanlal, 1998. “When Does Interna- Finance 53 (December): 2001–27. tionalization Enhance the Development of Domestic EIU (Economist Intelligence Unit). Various dates. “Country Stock Markets?” Journal of Financial Intermediation 7 Finance Reports.” http://www.eiu.com. (3): 263–92. ———. 2005. “Country Report for Iraq.” September. Delivered by The World Bank ICICI to: 2005. http://www.icicibank.com. Bank. e-library Elteto, Andrea, and Antaloczy Katalin. 2003. “Outward For- BankIDB (Inter-American Development Bank). 2002. Beyond The World eign Direct Investments from Poland.” In IP Facilitating : 192.86.100.36 Borders: The New Regionalism in Latin-America. Tue, 10 Mar 2009 16:55:33 Washington, DC: Johns Hopkins University Press. 134 (c) The International Bank for Reconstruction and Development / The World Bank F I N A N C I A L I N T E G R A T I O N A M O N G D E V E L O P I N G C O U N T R I E S IFC (International Finance Corporation). 1991. “Financing OECD 2005a. “Centre for Entrepreneurship, SMEs, and Corporate Growth in the Developing World.” IFC Dis- Local Development.” http://www.oecd.org/depart- cussion Paper 12, International Finance Corporation, ment/0,2688,en_2649_33956792_1_1_1_1_1,00.html. Washington, DC. ———. 2005b. “Corporate Responsibility Practices of IMF (International Monetary Fund). 2005. “Asia-Pacific Emerging Market Companies—A Fact-Finding Study.” Outlook.” Washington, DC. September. OECD, Paris. IMF–World Bank. 2005. “FDI Monitoring Working Pro- Oil & Gas Journal Special Report. 2001. http://www.ogj.com ject—interviews with major investors both from devel- Padayachee, Vishnu, and Imraan Valodia. 1999. oping and developed countries.” “Malaysian Investment in South Africa.” Journal of Irving, Jacqueline. 2005. “Regional Integration of Stock Ex- Contemporary African Studies 17. changes in Eastern and Southern Africa: Progress and Portes, R., and H. Rey. 2005. “The Determinants of Cross- Prospects.” IMF Working Paper 122, International Border Equity Flows.” Journal of International Eco- Monetary Fund, Washington, DC. nomics 65 (2): 269–96. Johannesburg Securities Exchange. “Dual Listed Company Pradhan, Java Prakash. 2005. “Outward Foreign Direct In- Information.” http://www.jse.co.za. vestments from India: Recent Trends and Patterns.” Un- Kabelwa, George. 2004. “Technology Transfer and South published paper, Centre for the Study of Regional Devel- African Investment in Tanzania.” Globalisation and opment, Jawaharlal Nehru University, New Delhi. East Africa Working Paper 10, Economic and Social Rakner, L., N. van de Walle, and D. Mulaisho. 1999. Zam- Research Foundation, Dar es Salaam. bia. Washington, DC: World Bank. Karolyi, A. 2004. “The Role of American Depositary Re- Schiff, Maurice, Yanling Wang, and Marcelo Olarreaga. ceipts in the Development of Emerging Equity Mar- 2002. “North–South and South–South Trade-Related kets.” Review of Economics and Statistics 86: 670–90. Technology Diffusion: An Industry-Level Analysis.” Latin Finance. 2005. “Daily Briefs.” http://www.latinfi- Policy Research Working Paper 2861, Development nance.com. Research Group, World Bank, Washington, DC. Levine, Ross. 1990. “Stock Markets, Growth, and Policy.” Svetlicic, Marjan, and Matija Rojec. 2003. Facilitating Working Paper 484, World Bank, Washington, DC. Transition by Internalization: Outward Direct Invest- ———. 1996. “Foreign Banks, Financial Development, and ment from Central European Countries in Transition. Economic Growth.” In International Financial Mar- Aldershot, Hampshire, UK: Ashgate. kets, ed. E. B. Claude. Washington, DC: AEI Press. Shah, Ajay, and Ila Patnaik. 2005 “India’s Experience with Levine, Ross, and Sara Zervos, 1998. “Stock Markets, Capital Flows: The Elusive Quest for a Sustainable Banks, and Economic Growth.” American Economic Current Account Deficit.” NBER Working Paper Review 88: 537–58. 11387, National Bureau of Economic Research, Cam- Levy-Yeyati, Eduardo, Panizza Ugo, and Ernesto Stein. bridge, MA. 2002. “The Cyclical Nature of North–South FDI Standard & Poor’s. 2005. Global Stock Markets Factbook. Flows.” Business School Working Paper 15, Universi- New York: Standard & Poor’s. dad Torcuato Di Tella, Buenos Aires. ———. Various dates. Emerging Stock Markets Review. Lisitsyn, Nikita, Sergi F. Sutyrin, Olga Y. Trofimenko, and State Bank of India. 2005. “In the News.” http://www.state- Irina V. Vorobieva. 2005. “Outward Internationalisa- bankofindia.com. tion of Russian Leading Telecom Companies.” Elec- Stein, E., and C. Daude. 2001. “Institutions, Integration, tronic Publications of Pan-European Institute 1/2005, and the Location of Foreign Direct Investment.” Un- Turku School of Economics and Business Administra- published paper, Inter-American Development Bank, tion, Turku, Finland. http://www.tukkk.fi/pei Research Department, Washington, DC. Mathews, John. 2005. “Enhancing Productive Capacity of Sull, Donlad, and Martin Escobari. 2004. “Creating Value Developing Country Firms through Internationaliza- in an Unpredictable World.” Business Strategy Review tion.” Discussion paper for UNCTAD conference on 15 (3): 14–20. “Enhancing the Productive Capacity of Developing Swenson, D. L. 2004. “Foreign Investment and the Media- Country Firms through Internationalization,” Geneva, tion of Trade Flows.” Review of International Eco- December 5–7. nomics 12 (4): 609–29. Mirza, Hafiz. 2000. “The Globalization Business and East The Banker. Various issues in 2005. http://www.the- Asian Developing Country Multinationals.” In The banker.com/. Globalalization of Multinational Enterprise Activity Tobin, Jennifer, and Susan Rose-Ackerman. 2005. “Foreign and Economic Development, ed. Neil Hood and Direct Investment and the Business Environment in De- Stephen Young. New York: St. Martin’s Press. veloping Countries: The Impact of Bilateral Investment Narlikar, Amrita, and Diana Tussie. 2004. “The G-20 at the Treaties.” Research Paper 293, Yale Law Centre for Cancun Ministerial: Developing Countries and Their Law, Economics and Public Policy, New Haven, CT. Evolving Coalitions in the WTO.” World Economy 27 UNCTAD. 1998. World Investment Report 1998. Geneva: (7): 947–1148. UNCTAD. Nini, Greg. 2004. “The Value of Financial Intermediaries: ———. Delivered by The World 2003. Bank to: An Emerging FDI Outward Investor.” “China: e-library Empirical Evidence from Syndicated Loans to Emerg-The World Geneva:Bank UNCTAD. ing Market Borrowers.” U.S. Federal Reserve Interna- ———. 2004. World Investment Report—2004.” Geneva: IP : 192.86.100.36 DC.10 Mar 2009 tional Finance Discussion Paper 820, Washington,Tue, 16:55:33 UNCTAD. 135 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 ———. 2005a. World Investment Report—2005.” Geneva: ———. 2005a. Global Development Finance 2005. Wash- UNCTAD. ington, DC: World Bank. ———. 2005b. Country Case Studies presented at UNC- ———. 2005b. Global Economic Prospects 2006. Washing- TAD Expert Meeting on “Enhancing the Productive ton, DC: World Bank. Capacity of Developing Country Firms through Inter- ———. 2005c. “Developing Country Investors and Opera- nalization,” Geneva, December 5–7. tors in Infrastructure, Phase 1 Report.” Public Private Vahtra, Peeter, and Kari Liuhto. 2004. “Expansion or Exo- Infrastructure Advisory Facility, World Bank, Wash- dus? Foreign Operations of Russia’s Largest Corpora- ington, DC. tions.” Electronic Publications of the Pan-European In- ———. 2006. Information and Communications for Devel- stitute 8/2004, Turku School of Economics and opment. World Bank, Washington, DC Business Administration. http://www.tukkk.fi/pei WTO (World Trade Organization). 2003. World Trade Re- Van Horen, Neeltje. 2006. “Foreign Bank Entry in Develop- port. Geneva: World Trade Organization. ing Countries: Origin Matters.” Unpublished paper, Yamori, N. 1998. “A Note on the Location Choice of Multi- World Bank, Washington, DC. national Banks: The Case of Japanese Financial Institu- Williams, B. 1998. “Factors Affecting the Performance of tions.” Journal of Banking and Finance 22: 109–20. Foreign-Owned Banks in Australia: A Cross-Sectional Yi, Ren. 2004. “Motivations for Chinese Investment in Viet- Study.” Journal of Banking and Finance 22: 197–219. nam.” University of Melbourne Working Paper. World Bank. 2002. Global Economic Prospects 2003. Zedillo, Ernesto, Patrick Messerlin, and Julia Nielson. Washington, DC: World Bank. 2005. Trade for Development: Achieving the Millen- ———. 2004. Global Development Finance 2004. Wash- nium Development Goals. UN Millennium Project ington, DC: World Bank. Task Force on Trade. London: Earthscan. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 136 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank . 5 Challenges in Managing Capital Flows T he surging flows of international private borrowing has been reduced, thereby addressing a capital and favorable global economic en- major policy failure that accompanied the capital vironment present a significant opportu- surge of the mid-1990s. nity for developing countries, particularly for the These positive developments do not come middle-income countries that are the major recipi- without risk. Progress in macroeconomic stabiliza- ents of capital flows. These and other countries tion and reform since the Asian financial crisis has that have embraced sound macroeconomic funda- not been fully matched by improvements in corpo- mentals, open international trade, and financial rate governance; in many countries, adherence to integration must now find ways to leverage their global standards and norms is still a work in gains, while building an institutional and policy progress. Many countries still lack adequate capac- environment that will maintain the confidence of ity to manage risks associated with managed-float investors and insulate the economy from external exchange rate regimes and partially liberalized cap- shocks. Few policy decisions would appear as im- ital markets. The large buildup of official foreign portant to future growth and financial stability as exchange reserves by many countries, particularly those capable of preventing a recurrence of the in Asia, has resulted in a high concentration of cur- market and policy failures of the 1990s. Although rency and interest rate risks on central banks’ bal- initial conditions point to better management of ance sheets, with potentially adverse fiscal conse- capital flows this time around, significant down- quences. On the international front, growing side risks remain. uncertainty about the sustainability of the current At an annual average growth rate of 5.4 per- pattern of global capital flows, in which developing cent over the past four years (2002–5), economic countries export capital to the rest of the world, activity in developing economies has expanded particularly the United States, constitutes a major more than twice as fast as in high-income coun- vulnerability in international capital markets. The tries. And as authorities have increasingly adopted current episode of strong capital flows to develop- price stability—often in the context of inflation ing economies coincided initially with a consider- targeting—as an integral part of their macroeco- able easing of monetary policy in industrial coun- nomic management, inflation has fallen dramati- tries; that period came to an end in the United cally in virtually all developing countries, from an States in mid-2004 and in the Euro Area more re- annual median of 11.5 percent during 1993–6 to cently. Rising interest rates in the industrialized 4.5 percent during 2002–5. At the same time, world may keep some investors closer to home. greater autonomy in monetary policy, afforded by This chapter highlights the implications of re- the widespread transition to flexible exchange cent changes in the macroeconomic and financial rates, has allowed authorities to lower local inter- environment for policy makers in developing est rates, which, in many developing countries, are countries. It also maps out broad strategies for Delivered by The World Bank e-library to: now converging to international levels. With lowerThe World Bank the influx of capital to serve long-term managing local interest rates and greater exchange rate flexi- growth and development objectives. Given the dif- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 bility, the incentive to resort to short-term external ferences among developing countries in their stage 139 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 of economic development, and the considerable liquidity purposes reflects in part a clear pro- variation in the amount and impact of different clivity to self-insure against global financial kinds of private flows, policy makers will neces- shocks. As the volume of reserves increases, sarily be guided by country-specific considerations however, so does the importance of balancing in determining the course of policy. But overall, their use for intervention and insurance pur- the three core dimensions of managing capital poses against their domestic resource costs. flows at the current juncture are likely to be (i) en- Allowing local institutional investors to diver- suring macroeconomic stability and sustaining the sify their investment portfolio globally, while confidence of investors so that access to interna- ensuring more effective regulation, could pro- tional capital markets is sustained and enhanced; vide a viable channel of capital outflow, as (ii) implementing appropriate policies and risk- well as an opportunity to further diversify management strategies to encourage allocation of risk. Further, permitting such investments capital to long-term investment and growth; and would have the effect of transferring foreign (iii) designing appropriate safeguards to enhance exchange rate risks, currently concentrated on resilience through self-insurance and adherence to the books of central banks, to domestic insti- global norms and standards. tutional investors that have a long investment The key messages emerging from the analysis horizon and can benefit from a more diversi- presented in this chapter are: fied international portfolio. Moreover, open- ing up a channel for capital outflows would • Policy responses in the current period of in- also help to avoid the excessive exchange rate creased capital inflows have differed in impor- appreciations that can result from surges in tant respects from those that prevailed during capital flows. the previous boom in the mid-1990s. Govern- • As developing countries become more open to ments have generally managed to avoid exces- international financial markets, designing and sive expansion of aggregate demand and large building a sound regime of external financial current-account deficits. Their policies have policy making and regulation presents an ur- supported modest allocations of foreign capi- gent challenge. A consensus has formed tal resources to domestic investment, although around the three core components of such a the major chunk has been used to build up new regime—membership in a credible cur- foreign exchange reserves. So far, fewer coun- rency union, such as the European Union, or tries have seen their real exchange rate appre- an exchange rate that reflects market forces; ciate than during the 1990s boom. In many gradual opening of the capital account; and a countries, investment rates have not yet risen monetary policy framework that favors price to the peaks they reached before the East stability. These elements are present to varying Asian crisis. In Indonesia, Malaysia, and Thai- degrees in many developing countries involved land, for example, investment rates remain in private capital markets. Roughly one-half of lower than precrisis levels by 10 to 20 per- developing countries are now operating under centage points of GDP. At the same time, the a floating exchange rate regime (free or man- surge in portfolio inflows has been associated aged), while the 11 new and aspiring members with a dramatic escalation of stock market of the European Union are taking steps to peg prices and valuations in many developing their currencies to the euro. Priority now must countries, particularly in Asia, raising the risk be given to two points. First, the complex web of asset price bubbles—and of reversals of of capital controls and exchange rate restric- capital flows should those bubbles burst. For tions that persists in many countries should be oil-importing countries, higher oil prices and simplified and, as macroeconomic policies im- the consequent adjustment in the current-ac- prove and local capital markets develop, eased count balance have partly offset the impact of gradually over time. During the transition, strong capital inflows. curbs on short-term debt inflows may need to Delivered by The World Bank e-library to: • That many developing countries have The accu- World Bank be maintained, or even strengthened, while re- mulated foreign exchange reserves far in ex- IP : 192.86.100.36 strictions on outflows are eased. Second, au- Tue, 10 Mar 2009 16:55:33 cess of the level required for intervention and thorities must build a system of risk manage- 140 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S ment robust enough to respond to the needs of tion in the 1990s, many developing countries a more flexible exchange rate and open capital today have significantly lower external debt bur- account. dens, fewer currency mismatches in their debt • The development and partial application of a structures, higher reserves of foreign exchange, a set of international norms and standards on more flexible exchange rate regime, and more transparency, corporate governance, and regu- open capital accounts. But the benign external en- lation and supervision of national financial sys- vironment in which these improvements were tems has helped increase the confidence of for- made may become less so in the next few years, as eign investors in emerging market economies. the major industrial countries tighten their mone- To promote stability and maintain a financial tary policy and as markets come to reassess their environment conducive to a balanced expan- views and expectations regarding the evolution of sion and deployment of capital flows in devel- global interest rates and capital flows. oping countries, the international community Since the early 1990s, developing countries must be assiduous in promoting the further ap- have experienced two episodes of heavy influx of plication of those norms and standards. private capital. The first, occurring in the middle • The world economy is moving toward a multi- of the past decade (1992–7), resulted in an in- polar international monetary system in which crease in capital inflows from 3.2 percent of devel- policy interactions among the major industrial oping countries’ aggregate GDP in 1992 to 5.1 countries of the G-3—and with key emerging percent in 1997. The second began in 2002 and market economies—will be essential in securing continues to date. So far, it has brought a cumula- an orderly adjustment of the prevailing global tive total of $1,316 billion in capital to the devel- imbalances in external payments. One effect of oping world (approximately $350 billion annually inclusive interactions would be to lessen market averaged over 2002–5). This last episode has led anxiety over the course of global interest rates to an increase in private capital flows from 2.8 and capital flows. Emerging market economies, percent of developing countries’ aggregate GDP in which would suffer disproportionately from the 2002 to 5.1 percent in 2005. instability induced by a disorderly adjustment, The macroeconomic consequences and policy share with the industrial countries the desire for responses associated with the previous surge have a multilateral approach that will include correc- been explored in a large body of academic litera- tive actions in deficit and surplus countries ture (Johnson and others 2000; Radelet and Sachs alike. In addition, policy makers in emerging 1998; Corsetti, Pesenti, and Roubini 1998). The market economies should take advantage of the data from that period reveal several interesting opportunity presented by the current benign patterns for developing countries that had access global financial market environment to build to international capital markets: a considerable ac- institutions and mechanisms that will enable celeration in economic growth, a rise of two per- them to navigate their economies in a world of centage points in the ratio of investment to GDP, increasingly open capital accounts and market- and a considerable and widespread appreciation of based exchange rates. national currencies in real terms (19 percent). Moreover, about one-third of the inflowing capital was allocated to the accumulation of official re- serves of foreign exchange, which rose, in aggre- Two booms in capital flows—what gate, from $216 billion at the end of 1992 to $572 has changed? billion at the end of 1997. These facts provide a T he present surge in capital flows to developing countries differs substantially from the previ- ous episode in the mid-1990s. Greater global eco- good point of comparison for the current influx in private capital to developing countries. Looking at the cross-country distribution of nomic and financial integration, improved domes- capital inflows during current episode (see figure tic macroeconomic conditions, and sounder 5.1), 67 percent of developing countries received Delivered by The World Bank e-library to: domestic policies and institutions have enhancedThe World private flows within the range of 2 to 10 percent Bank the capacity of policy makers to deal with infu- of their GDP, and a further 16 percent received IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 sions of private capital. Compared with the situa- capital flows of more than 10 percent of their 141 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 5.1 Distribution of private capital flows total debt flows has declined for virtually all across developing countries, 2002–4 major debtors, particularly in crisis-affected No. of countries countries; second, the composition of flows 35 has rotated toward equity, particularly foreign Mean = 4.4 direct investment (FDI). 30 Median = 3.0 SD = 5.5 • The shift toward more flexible exchange rate 25 Skew = 1.4 regimes has helped overcome a major policy 20 failure underlying the financial crises of the 1990s. That shift, in conjunction with im- 15 proved macroeconomic conditions, has facili- 10 tated a continued process of relaxation or re- moval of formal controls on many 5 capital-account transactions in many develop- 0 ing countries, despite the severity and global –4.5 0 4 8 12 16 20 24 28 nature of the 1997 financial crisis. Private capital flows as a percentage of GDP • The current account in many developing Source: World Bank Debtor Reporting System and staff estimates. countries, particularly major oil exporters Note: 134 developing countries for which we had data were used. Private capital flows to GDP were averaged over the 3 years. and emerging Asia, has moved from deficit to sizable surplus, contributing to the accumula- tion of foreign exchange reserves. The initial GDP. The correlation between capital inflows and impetus came from countries’ strenuous ex- per capita income is positive but relatively low ternal adjustments to the crises of the 1990s, (0.18), reflecting the fact that many low–income but high commodity prices, robust global countries also have attracted private capital flows, growth over the past few years and interven- including The Gambia, Mozambique, Tanzania, tion to maintain undervalued exchange rates and Vietnam. for the purposes of export competitiveness The Asian financial crises of the mid-1990s have sustained and, in some cases, amplified provide a cautionary example of the potential the effect. These developments have com- macroeconomic effect on recipient countries of bined to improve the external debt burdens of large capital inflows. At that time, inflows gener- developing countries, as debt/export ratios ated a sequence of currency misalignment, asset and debt/GDP ratios have declined since their price escalation, excessive expansion of aggregate peaks in 1997–8. demand, inflationary pressures, current-account • The accelerated development of local bond imbalances, capital losses on central banks’ balance markets in many countries after the crises of sheets, and financial instability—a calamitous the 1990s has been helpful to the development chain of events that affected individual countries in of a more balanced financial structure, reduc- very different ways. A large body of theoretical and ing dependence on the banking sector and empirical research over the past decade has at- short-term foreign capital as sources of financ- tempted to identify confluences of global financial- ing. The presence of a well-functioning gov- market conditions and specific developing-country ernment bond market facilitates the conduct of characteristics that could lead to a recurrence of monetary policy through open market opera- that sequence (World Bank 1997; Calvo and others tions and helps improve debt management. 1996; Edwards 2001; Chinn and Ito 2002; Kletzer (This development is discussed in chapter 2.) and Spiegel 2004). That literature, combined with • External changes that are likely to affect the recent experience, points to five important trends, climate for capital flows include the euro’s domestic and global, distinguishing the present growing role as a major international reserve cresting of capital flows from the previous episode: currency, which widens policy makers’ choices. Higher international interest rates and Delivered by The World Bank e-library to: • The pattern of private capital flows to Thedevel- World Bank likely volatility in exchange rates, by contrast, oping countries has changed in two important IP : 192.86.100.36 will constrain policy making. The long and ag- Tue, 10 Mar 2009 16:55:33 respects: first, the share of short-term debt in gressive phase of monetary easing that started 142 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S in the United States in 2001 came to an end in Figure 5.2 Composition of financial flows to June 2004, with the Euro Area following suit a developing countries, 1992–7 and 2002–5 few months later. (See Chapter 1.) Billions of current $ 400 Debt The first three of those trends are discussed Portfolio equity below. 300 FDI 33% The composition of capital flows is changing 200 9% The composition of private foreign capital flowing 42% to developing countries during the current surge has shifted decisively toward equity, predomi- 100 11% 57% nantly FDI. The shift reflects government policies 47% that encourage equity and aim to reduce depen- 0 dence on external borrowing. Thus, on average, 1992–7 2002–5 FDI accounts for 57 percent of private capital Source: World Bank Debtor Reporting System and staff estimates. flows to developing countries (figure 5.2), much higher than portfolio equity (9 percent) and higher even than short- and long-term bank debt com- Figure 5.3 Ratio of foreign exchange reserves to bined (33 percent). In the mid-1990s, by contrast, short-term debt, by region the same figures were 47 percent for FDI, 11 per- Ratio of reserves to ST debt cent for portfolio equity, and 42 percent for debt. 8 The trend toward equity in the composition of 7 1992–7 private capital flows has been particularly pro- 2002–5 6 nounced in the two regions (Latin America and the 5 Caribbean and East Asia and the Pacific) that were 4 most directly affected by the string of financial 3 crises in the 1990s. 2 Greater reliance on equity financing also im- 1 proves countries’ external liability profile, because equity flows are more focused on long-term eco- 0 All East Asia Europe & Latin America nomic prospects and offer better risk-sharing char- & Pacific Central Asia & Caribbean acteristics than debt flows. Moreover, FDI tends to Source: World Bank Debtor Reporting System and staff estimates. be more stable than debt, in the sense that current FDI is strongly correlated to its past levels; the co- efficient of persistence of FDI, using a simple auto- term debt. Developing countries as a group are regressive estimation for a sample of developing now much better equipped than previously to deal countries, is found to be on average 0.62, while it with the potential volatility of private capital is 0.52 on debt (both short and long term).1 flows. Looking at reserve holdings on a regional An indication of the improvement brought basis, each of the regions holds in the form of re- about by the changing composition of capital serves at least 1.5 times their short-term debt (fig- flows is the significant reduction in the ratio of ex- ure 5.3). The ratio is particularly high in East Asia ternal debt to gross national income (GNI) for de- (8.3), largely because of China, whose accumu- veloping countries as a whole—from a peak of 44 lated reserves are 38 times greater than its short- percent in 1999 to about 34 percent in 2004—and term debt. The rising ratio of reserves to short- particularly for countries in East Asia and Latin term debt reflects not only the spike in reserve America. In Europe and Central Asia, however, holdings, but also the decline in short-term debt as the ratios remain relatively high compared with a percentage of total debt in most developing those seen in the early 1990s. countries since the mid-1990s (table 5.1). Delivered by The World Bank e-library to: A further sign of improved external liabilityThe World The Bank rotation towards equity and reduced re- positions in the developing world can be found in liance on short-term debt flows have significant IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 the ratio of foreign exchange reserves to short- policy implications for the management of capital 143 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 flows to developing countries, as they enhance the Table 5.1 Ratio of short-term debt to total debt in scope for monetary policy autonomy. Equity major borrowing countries, 1996–2004 Percent flows, in contrast to debt flows, tend to move countercyclically with local interest rates, increas- Short-term debt/total debt ing during periods of low domestic interest rates Country 1996 2004 Change due to the positive impact of low interest rates on China 19.7 47.2 27.5 domestic growth and corporate profitability and Poland 6.1 17.0 10.9 valuation. The classical Mundell-Fleming model Czech Rep. 28.5 37.5 9.0 Russian Fed. 9.5 17.8 8.4 (Mundell 1963, Fleming 1962) of the open econ- Hungary 12.3 19.5 7.2 omy and the implied impossible trinity—that Venezuela, R. B. de 7.9 12.2 4.3 countries can pursue only two of the three objec- Egypt 7.4 9.7 2.3 Algeria 1.0 2.0 1.0 tives of fixed exchange rates, free capital mobility, India 7.2 6.1 –1.1 and independent monetary policy—is predicated Turkey 21.7 19.7 –2.0 on the assumption that capital inflows are com- Argentina 21.2 16.2 –4.9 Nigeria 18.1 12.8 –5.3 posed predominately of short-term debt. In an eq- Pakistan 9.4 3.5 –6.0 uity dominated pattern of capital flows, authori- Malaysia 27.9 21.9 –6.0 ties have more autonomy in pursuing interest rate Colombia 20.4 14.2 –6.2 Indonesia 25.0 17.4 –7.6 policies geared toward domestic goals. Chile 25.7 17.5 –8.2 Brazil 19.8 11.4 –8.4 Countries now have more flexible exchange Philippines 18.1 8.3 –9.8 Mexico 19.1 6.6 –12.5 rates and more open capital accounts South Africa 41.6 27.8 –13.8 Policies on exchange rates and capital controls are Peru 22.2 8.0 –14.2 particularly important for developing countries, be- Thailand 42.3 22.4 –19.9 Averagea 18.8 16.4 –2.4 cause external developments have a greater effect on domestic inflation, monetary transmission, and fi- Sources: IMF, International Financial Statistics and World Bank staff estimates. nancial stability in developing countries than in in- Note: Major borrowing countries, based on the average volume of dustrial countries. Most developing countries are al- total debt stock over the period of 1996–2004 (in descending order). a. Excluding South Africa. ready more open to international trade in goods and services than are developed countries: from 2002 to 2004, developing countries’ trade averaged 54.5 per- cent of GDP, compared to 39 percent in developed bands), investors and borrowers may believe there countries. But developing countries as a group also is less need to hedge currency movements, and the face a potentially higher degree of volatility in capi- risk of borrowing in foreign currency appears to tal flows, and changes in the exchange rate may be reduced, encouraging excessive exposure. translate more quickly into domestic inflation than However, if a crisis does hit, and the central bank in developed countries.2 Even with their recent cannot maintain the peg or band, the costs to the progress in launching local-currency debt issues on banking system and corporate sector can be sub- global markets (see Chapter 2), developing countries stantial and damaging. still have much larger shares of their external debt Partly due to this experience, several develop- denominated in foreign currencies than do industrial ing countries have adopted greater exchange rate countries (Eichengreen and Hausmann 1999; flexibility, moving to a variety of managed-float Hawkins and Turner 2000). Such conditions predis- regimes, with central banks retaining the ability to pose an economy to greater vulnerability to external intervene in the market to influence the exchange financial shocks. rate and limit volatility. Since the early 1990s, Virtually all capital flow–related financial nearly 50 developing countries have abandoned crises of the 1990s involved a fixed peg or crawl- fixed or crawling pegs in favor of managed floats ing band exchange rate regime and considerable or fully flexible exchange rates (figure 5.4). No- currency mismatch on the balance sheets of both table examples are Mexico (1994), Indonesia Delivered by The World Bank e-library to: public and private borrowers (Fischer The2001; World Bank(1997), Colombia (1999), Brazil (1999), Chile Goldstein 2002). When countries maintain such (1999), and the Russian Federation (2002). In July IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 exchange rate regimes (fixed pegs or crawling 2005, the Bank Negara Malaysia adopted a man- 144 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S Figure 5.4 Changes in exchange rate flexibility, pressure on the exchange rate stems from tempo- 1991–2004 rary shocks or volatile capital flows, intervention No. of countries and interest rates, singly or in combination, should 80 73 be considered as tools to limit short-run exchange 1991 70 67 64 rate fluctuations. 1999 60 2004 There are institutional and microstructure re- 50 44 quirements associated with managing a flexible ex- 40 36 change rate regime. The key steps involve the devel- 30 27 27 27 opment of local money, capital, and cross-border 21 20 derivatives markets to provide the necessary depth, 10 sophistication, and hedging possibilities for manag- 0 ing currency risk, thereby providing stability for pri- Hard peg Intermediate Float vate agents and the economy as a whole. Sources: IMF Annual Report on Exchange Arrangements and Exchange Restrictions and World Bank staff estimates. Real exchange rate appreciation has been mild A significant, sustained, and rapid appreciation in a country’s real exchange rate is one of the precur- sors of a currency crisis.4 Figure 5.6 shows the aged float for the ringgit with reference to a cur- movements in real effective exchange rates in two rency basket and the People’s Bank of China reval- of the regions that experienced some of the largest ued the renminbi and announced that it would be exchange rate corrections during the crises of the determined with reference to a currency basket. 1990s. The appreciation in real exchange rates in Evidence also suggests that many developing the last few years has been much milder than dur- countries pursuing a managed float are tolerating ing that period. Latin America shows stronger ap- a greater degree of short-term fluctuation in their preciation over 2004–5 than does East Asia. currencies.3 Figure 5.5 displays the frequency dis- Looking at some individual countries, the real tribution of daily percentage changes in the bilat- exchange rate appreciated in 60 percent of devel- eral exchange rates of currencies in several crisis- oping countries over the period 1993–6, while affected countries against the U.S. dollar during only about one-third experienced an appreciation the current and previous surges in capital flows. in 2002–4. Moreover, the range of appreciations The left panel shows movements during the 1990s during the second surge has been significantly surge; the right panel shows current movements. smaller (figure 5.7).5 The bell-shaped daily fluctuations in exchange rates in the current episode indicate two-way Easing of capital controls movements in bilateral exchange rates. Since the 1990s, the shift to floating exchange Successful management and operation of a rates, the convergence of the currencies of Eastern flexible exchange rate regime requires proper pol- Europe toward the euro, and the deepening of local icy frameworks, market microstructure, and insti- capital markets have enabled many developing tutions to ensure smooth functioning of foreign countries to ease capital controls and foreign ex- exchange markets. Policy decisions must be made change restrictions. Progress in formulating and about whether to rely on interest rates and inter- implementing such liberalization measures across vention to stabilize exchange rates at times of high developing countries has been uneven, however, as volatility or uncertainty. Such decisions require an countries have moved at different paces and with assessment of the underlying sources of exchange different degrees of rigor (see box 5.1). The clearest rate volatility, which in the context of many devel- trend is in the liberalization of exchange rate re- oping countries often implies gauging the sustain- strictions. The number of countries that declared ability of capital flows. For example, policy mak- their currencies convertible on the current account, ers might ask whether a surge in capital flows was which often precedes capital-account convertibility, Delivered by The World Bank e-library to: composed primarily of volatile portfolio capital orThe World from approximately 62 in 1990 (or 40 percent rose Bank speculative debt, on the one hand, or more stable of the IMF’s membership) to 164 in 2004 (or al- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 and predictable FDI flows, on the other. When most 90 percent of the IMF’s membership). 145 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 5.5 Frequency distribution of daily percentage changes in exchange rates for selected developing countries, 1993–6 vs. 2003–5 1993–6 2003–5 Frequency Frequency 1,000 200 800 150 600 100 400 50 200 0 0 – 2.5 –1.5 –0.5 0.5 1.5 2.5 – 2.5 –1.5 –0.5 0.5 1.5 2.5 Thailand Thailand Frequency 1,300 Frequency 1000 160 700 120 400 80 100 40 – 200 0 – 20 –10 0 10 20 30 –4 –2 0 2 4 Turkey Turkey Frequency Frequency 250 1,000 800 200 600 150 400 100 200 50 0 0 –5 0 5 10 15 20 –6 –3 0 3 Brazil Brazil Frequency Frequency 1,000 100 800 80 600 60 400 40 200 20 0 0 –10 0 10 – 2.5 –1. 5 – 0.5 0.5 1.5 2.5 Mexico Mexico Sources: Bloomberg data service and World calculations. Bank staffby Delivered The World Bank e-library to: Note: The figures show the frequency distribution of daily Thepercentage World Bankchanges in the exchange rate between local currency and U.S. dollars. Increases in the exchange rate represent depreciations against the U.S. dollar, and decreases represent appreciation. IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 146 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S Figure 5.6 Movements in real effective exchange Figure 5.7 Real exchange rates for selected rates in East Asia and Latin America, 1993–2005 countries that receive higher-than-average private capital inflows as a ratio to GDP, 1994–7 and 2002–5 1990 = 100; regional GDP weights 165 1994–7 Latin America South Africa 155 Mexico & Caribbean Argentina 145 Rep. of Korea Indonesia 135 Thailand Hungary 125 Croatia Malaysia 115 Romania India Slovak Rep. 105 East Asia & Pacific Philippines Czech Rep. 95 Turkey Poland 85 Chile Brazil 75 Latvia Bulgaria 1993 1996 1999 2002 2005 China Estonia Sources: IMF, International Financial Statistics and World Bank Lithuania staff calculations. Russian Fed. –25 –15 –5 5 5 25 Average annual % change Three trends stand out in the liberalization of 2002–5 capital-account transactions: Argentina China Mexico • The easing or removal of quantitative restric- Malaysia Latvia tions on residents’ issuance of securities, in- Philippines Poland cluding debt, and outward FDI by private res- India Thailand ident entities Lithuania • The relaxation of limits on nonresidents’ ac- Chile Croatia cess to local money and securities markets Estonia Bulgaria • The reduction or elimination of taxes on Czech Rep. Brazil capital-account transactions. Rep. of Korea Romania Indonesia In Chile, for example, the limit on outbound Hungary Russian Fed. foreign investment by private pension funds was South Africa Slovak Rep. increased in 2003–4 from 16 to 30 percent, en- Turkey abling local investors to hold diversified portfolios –25 –15 –5 5 5 25 despite the small size of local capital markets. In Average annual % change Malaysia and Thailand, approved domestic insti- tutional investors may now invest up to 10 percent Sources: Bank for International Settlements and World Bank staff estimates. of their assets abroad. In the Republic of Korea, residents are encouraged to invest in overseas mu- tual funds to mitigate the impact of foreign in- flows. And in India, new measures have relaxed Many countries with open capital accounts overseas investment restrictions on banks and mu- have floating exchange rates tual funds, allowing banks to invest in money mar- The growing group of developing countries that ket and debt instruments abroad and raising from are considered relatively open to capital move- $500 million to $1 billion the limit on mutual ments appears in table 5.2. A variety of indices of funds’ investments in companies listed abroad. In financial openness were used to compile the list Delivered by The World Bank e-library to: Brazil this year, foreign investors were exemptedThe World (ChinnBank and 2002; Miniane 2004; Edwards 2005; from a 15 percent withholding tax on local gov- Quinn 1997; and Brune and others 2001). The IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 ernment debt investments. countries in the table all have achieved currency 147 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Box 5.1 Preconditions for capital-account liberalization B y the early 1990s, under the Code of Liberalization of Capital Movements of the Organization for Economic Co-operation and Development (OECD), developed The liberalization of capital accounts must be accom- panied by sound economic policies and institutions, so that governments are prepared to deal with the volatility countries had moved to open their capital accounts fully inherent in capital markets. The preconditions for a safe to cross-border financial transactions, including capital- transition to a more open capital account in most develop- market securities, money-market operations, and deriva- ing countries include a track record of fiscal prudence and tives instruments. Developing countries, by contrast, have stability (specifically, low inflation and a low fiscal continued to maintain, though in varying degrees, a wide deficit), a deep and well-regulated financial system, and range of administrative capital controls and foreign ex- adequate levels of reserves to provide the necessary buffer change restrictions. Capital-account regulation ranges against adverse external shocks. Against such a backdrop, from quantitative limitations on certain transactions (or a deliberate and sequenced opening will signal to financial on associated transfers of funds) to indirect measures in- markets the government’s commitment to sound finance, tended to influence the economic incentives of engaging thereby contributing to more stable capital flows. Once in certain transactions (IMF various years; Dailami 2000; capital-account liberalization has progressed, it is very and Eichengreen 2001). Although country circumstances costly to reverse, and the reinstitution of capital controls vary, controls generally have three goals: to discourage should be considered a last resort, appropriate only when short-term external debt flows in favor of longer-term in- alternative policy options have been exhausted. Even then, vestments, such as FDI (a motivation that gained momen- authorities would have to consider the reputational costs tum after the East Asian crises); to enhance monetary au- of invoking controls and carefully assess the likelihood tonomy and exchange rate stability; and to allow time for that the controls would meet their declared objectives in the establishment of an institutional and policy frame- today’s large and rapidly changing global financial envi- work within which capital-account liberalization will be ronment (Goldfajn and Minella 2005; Edwards 2005; successful (Rodrik 1999; Stiglitz 2002). Carvalho and Garcia 2005). convertibility on the current account of the bal- tively open to capital movements had adopted in- ance of payments—but they maintain some con- flation targeting regimes by the end of 2005—sev- trols on capital-account transactions. The table eral in the course of the year (table 5.2). Recent re- also reports on three other aspects of these coun- search (IMF 2006) indicates that a number of tries’ external financial profile: exchange rate developing countries that have pledged to use in- regime, monetary policy framework, and the num- flation targeting as their monetary policy frame- ber of years that currency convertibility on current work have had better macroeconomic perfor- accounts (signifying acceptance of IMF Article mance and in particular have outperformed VIII) has been in effect. It also indicates whether countries with other frameworks.7 there exists an offshore nondeliverable foreign ex- Six of the same 32 countries allow offshore change forward market (NDF)6 for each currency. trading in their currencies through NDFs, which Most countries that are largely open to capital- are similar to ordinary forward foreign exchange account transactions maintain a flexible exchange contracts, with the exception that at maturity rate arrangement. This affords policy makers a de- they do not require physical delivery of currencies gree of autonomy in setting interest rates to and are typically settled in U.S. dollars. NDFs are achieve price stability, something particularly de- largely short-term instruments—one month to sirable for countries such as Brazil, Chile, Mexico, one year—and are increasingly relied upon by for- the Philippines, South Africa, and Thailand, which eign investors to hedge their exposures against have adopted inflation targeting as an anchor for currencies that are not traded internationally and monetary policy. that are not convertible on capital-account trans- Along with the shift to greater exchange rate actions. Once a country permits convertibility and Delivered by The World Bank e-library to: flexibility, a number of developing countries have Bankdevelops onshore foreign exchange markets, NDF The World moved to inflation targeting regimes. Twelve of markets tend to diminish. Although NDFs are IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 the 32 developing countries considered to be rela- helpful instruments for managing cross-border 148 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S Table 5.2 Profile of external financial policy for developing countries considered relatively open to capital movements As of 2005 Years since Offshore currency Largely open countries Exchange rate regime Monetary policy article VIII assumed derivatives market Bolivia Intermediate Exchange rate anchor 38 Botswana Intermediate Exchange rate anchor 10 Costa Rica Intermediate Exchange rate anchor 40 Croatia Floating IMF program 10 Czech Rep. Floating Inflation target 10 Dominican Rep. Floating — 52 Ecuador Hard peg Exchange rate anchor 35 Egypt, Arab Rep. of Floating M aggregate 1 El Salvador Hard peg Exchange rate anchor 59 Estonia Hard peg — 11 The Gambia Floating — 12 Guatemala Floating Inflation target 58 Hungary Intermediate Inflation target 9 Indonesia Floating Inflation target 17 Yes Jamaica Floating M aggregate 42 Jordan Hard peg Exchange rate anchor 10 Kenya Floating IMF program 11 Latvia Intermediate Exchange rate anchor 11 Lebanon Intermediate Exchange rate anchor 12 Mexico Floating Inflation target 59 Yes Nicaragua Intermediate Exchange rate anchor 41 Panama Hard peg Exchange rate anchor 59 Peru Floating Inflation target 44 Yes Philippines Floating Inflation target 10 Yes Poland Floating Inflation target 10 Romania Floating Inflation target 7 Slovak Rep. Floating Inflation target 10 Yes Thailand Floating Inflation target 15 Yes Trinidad & Tobago Hard peg — 12 Turkey Floating Inflation target 15 Uruguay Floating M aggregate 25 Zambia Floating M aggregate 3 Sources: World Bank staff calculations based on Ito and Menzies 2002; Miniane 2004; Edwards 2005; Quinn 1997; Brune and others 2001 and Annual Report on Exchange Arrangements and Exchange Restrictions, IMF, various years. Note: Monetary policy: Inflation target = Public announcement of medium-term numerical targets for inflation with an institutional commit- ment by the monetary authority to achieve those targets. M aggregate = Monetary authority uses its instruments to achieve a target growth rate for a monetary aggregate that becomes the nominal anchor or intermediate target of monetary policy. Exchange rate anchor = Monetary authority stands ready to buy and sell foreign exchange at quoted rates to maintain the exchange rate at its predetermined level or range. IMF program = Implementation of monetary and exchange rate policy within the confines of a framework that establishes floors for international reserves and ceilings for net domestic assets of the central bank. currency risk, regulatory agencies in developing ure 5.8). This stands in marked contrast to the countries need to keep a close eye on them, given pattern observed in the first capital boom of the illiquidity of the currencies that underlie 1992–7, when developing countries as a whole ran NDF transactions and the potential for specula- an aggregate current-account deficit of 2 percent tive behavior. of GDP per year (or an aggregate deficit of $547.7 billion from 1992–7). Many countries now show surpluses on both Much of the current-account surplus accumu- their current and capital accounts lated during the present surge is attributable to oil Developing countries as a group have undergone a exporters and emerging Asia, which are benefiting significant turnaround in the past several years in from high oil prices and strong export growth, re- their external payment positions, moving from an spectively. The net oil-exporting countries as a Delivered by The World Bank e-library to: aggregate current-account deficit of $89 billionThe World group have seen large gains in their current- Bank (1.6 percent of GDP) in 1998 to a sizable surplus account surpluses, posting an aggregate surplus of IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 of $248 billion (2.6 percent of GDP) in 2005 (fig- close to $219 billion in 2005, up from $50 billion 149 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 5.8 Current-account balance, developing Figure 5.9 Value of oil imports, oil-importing countries, 1990–2005 countries, 2001–5 $ billions Percent $ billions Percent 250 3 250 11 200 2 10 Percentage of GDP (right axis) 200 150 1 9 150 Share of total 100 0 imports (right axis) 8 50 –1 100 7 0 –2 50 6 –50 –3 0 5 –100 –4 2001 2002 2003 2004 2005 –150 –5 Sources: World Bank Debtor Reporting System and staff estimates. 1990 1993 1996 1999 2002 2005 Sources: IMF, International Financial Statistics and World Bank staff calculations. Table 5.3 Current account aggregated by region, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries –84.7 –89.4 –4.1 47.1 18.8 69.8 122.3 153.1 248.4 East Asia and Pacific 17.2 59.8 60.3 53.7 39.8 61.2 74.9 93.6 143.4 Europe and Central Asia –27.7 –24.5 –1.3 16.3 17.6 5.6 –2.0 4.2 23.2 Latin America and Caribbean –65.3 –89.4 –55.4 –46.8 –51.9 –14.9 8.4 19.0 33.9 Middle East and North Africa 4.5 –9.7 6.2 25.3 15.4 12.0 28.3 41.0 76.0 Others –13.3 –25.4 –13.2 –0.7 –0.2 8.1 14.3 –2.3 –23.5 Memo item Oil exporting countries –32.5 –47.5 26.9 87.4 41.4 49.3 91.3 131.2 219.0 Oil importing countries –52.2 –42.0 –30.9 –40.3 –22.6 20.6 31.0 21.9 29.5 excl. China –89.2 –73.4 –52.0 –60.8 –40.0 –14.9 –14.9 –46.6 –97.2 Sources: IMF, International Financial Statistics and World Bank data reporting system. e = estimate. in 2002. By contrast, the current-account position gion. And in Latin America, thanks to favorable of oil-importing developing countries has in- prices for many non-oil commodity exports and creased from a surplus of $21 billion in 2002 to a relatively strong global economic growth, the re- surplus of $30 billion in 2005. The rise in their oil gion’s surplus increased in 2005 to $33.9 billion import bills from an aggregate value of $91.2 bil- (table 5.3)—the largest current-account surplus lion in 2001 to $229.8 billion in 2005 (now equal recorded for that region in 25 years. to approximately 10 percent of their total imports The overall surpluses appearing on the current of goods and services—figure 5.9) is substantially and capital accounts of the balance of payments of greater than the change in their current account, as many countries reflect an increase in holdings of the boom in non-oil commodity prices has cush- foreign currency due to net inflows from trade, ioned somewhat the impact of rising oil prices. workers’ remittances, and financial transactions Meanwhile, the Eastern Europe and Central (table 5.4). Delivered by The World Bank e-library to: Asia regions have recorded a large surplus, Thelargely World Bank For developing countries as a whole, these in- because of strong oil exports from theIPRussian flows have increased steadily since 2000. In 2005, : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 Federation that mask deficits elsewhere in the re- the combined current accounts and recorded capi- 150 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S Table 5.4 Sources of reserve accumulation, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e Change in reserves 52 16 33 45 82 172 292 405 392 Current account balance –85 –89 –4 48 21 72 124 158 246 Balance on goods & services –53 –44 33 76 48 86 107 128 146 Net workers’ remittances 71 73 77 84 96 113 141 160 167 Capital account 332 260 241 211 210 209 303 418 464 Net private capital flows 293 199 198 188 154 172 272 397 483 Net official capital flows 38 61 42 23 55 38 31 22 –19 Residents’ foreign asset accumulation and errors & omissions 195 155 204 213 148 109 136 172 318 Sources: IMF, International Financial Statistics and World Bank data reporting system. e = estimate. tal accounts of the developing world amounted to tive to recipient countries’ economies. Remittances $710 billion (7 percent of their aggregate GDP), of (which are the largest source of external financing which $392 billion was channeled into reserves by in many developing countries) may rise when the the official sector and the rest invested abroad by recipient economy suffers a downturn in activity, residents in the form of FDI, portfolio holdings, or because of macroeconomic shocks due to fi- and other vehicles. (The cited figures include er- nancial crisis, natural disaster, or political conflict rors and omissions in the balance-of-payments ac- (Clarke and Wallsten 2004, Kapur 2003, Yang counts.) The opening of capital accounts by many 2004 and 2005), as migrants may send more developing countries in recent years has increased funds during hard times to help their families and opportunities for capital outflows by firms and friends.8 According to official statistics, in 2005 other private investors seeking to improve their re- remittance flows are estimated to have exceeded turns through international diversification. $233 billion worldwide, of which developing Policy responses to such influx of liquidity countries received $167 billion. must take into account the difference in the dynam- ics and cyclical characteristics of current-account Current-account surpluses have fed foreign positions and private capital flows. Private capital exchange reserves flows to developing countries tend to move pro- Although the pace of foreign exchange reserve ac- cyclically, in line with global economic activity as cumulation slowed somewhat in 2005 in several expressed in GDP, trade, and commodity prices. developing countries, including India, Thailand, They increase during upswings in commodity and Malaysia, the conversion of current-account prices, for example, and decrease during down- surpluses into official reserves has continued. For turns, which tends to amplify balance-of-payment developing countries as a group, the stock of offi- swings from oil and other commodities. Current- cial foreign exchange reserves reached $2 trillion account positions, by contrast, are less volatile than by the end of 2005, compared to $1.6 trillion in capital flows; they move in a countercyclical fash- 2004 and $1.2 trillion in 2002. In 2005, 92 of 127 ion with respect to the business cycle (Lane 2003). developing countries increased their reserves, with Box 5.2 provides an estimate of the sensitivity of the largest accumulations occurring in China and private capital flows to international commodity oil-exporting countries (figure 5.10). In relation to price movements from 1980 to 2005. For develop- the size of their international trade, developing ing countries as a whole, private capital flows were countries’ reserve holdings are now twice as large twice as large during upturns as they were during as those in developed countries (figure 5.11). De- downturns, averaging $237 billion (in constant mand for official foreign currency reserves in U.S. dollars) during upswings in commodity prices, major industrial countries has been more subdued, and $109 billion during downswings. given their free-floating exchange rates, well-de- Delivered by The World Bank e-library to: While capital flows tend to rise during up-The World Bank capital markets, and less vulnerable veloped swings of economic cycles and decline in bad economies. At the end of 2005, the Euro Area re- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 times, remittances tend to be countercyclical rela- ported $167 billion in reserves (European Central 151 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Box 5.2 Capital flows are procyclical with respect to non-oil commodity markets C apital flows to developing countries tend to move pro- cyclically with world commodity prices, increasing when commodity prices are high and decreasing when they nificant portion of developing-country exports) averaged 16 years, while upswings averaged 8.5 years. For develop- ing countries as a whole, private capital flows were twice are low. Two factors account for this. First, commodity as large during upturns as they were during downturns, prices typically are negatively correlated with fixed income averaging $237 billion (in constant U.S. dollars) during and equity markets in advanced countries. Capital is pushed upswings in prices, and $109 billion during downswings. to the developing world when returns in mature capital This tendency is also confirmed by detailed regional analy- markets are low (typically during upturns), and vice-versa. ses using region-specific commodity price indices (exclud- Second, commodities still account for a large share of devel- ing energy) and capital flow data. The correlation between oping-country exports and production, affecting their terms private capital flows and commodity prices is particularly of trade and real exchange rates, and potentially influencing pronounced in East Asia, Europe and Central Asia, and business-cycle fluctuations, particularly in countries charac- Latin America. During the upturns in commodity prices, terized as having “commodity currencies” (Chen and Ro- private capital flows in East Asia, for example, were 3.1 goff 2002; Mendoza 1995; Cashin and others 2003). Thus times larger than they were during downturns. Similarly, the rise in aggregate demand increases domestic borrowing. in Europe and Central Asia, private capital flows were 3.2 Equally, as developing countries tend to face quantitative times larger during upturns than downturns. In the other constraints on their borrowing, the rise in creditworthiness three regions, private capital flows in total are more mod- that comes with higher earnings on commodity exports in- est, although they also tend to move procyclically. creases foreign lenders’ willingness to supply funds. The re- The recent surge in private capital flows is a good il- lationship between capital flows and commodity prices is lustration of this experience. Net private capital flows rose displayed in the figure below, which shows the behavior of from $154 billion in 2001 to an estimated $483 billion in net private capital flows (deflated by the U.S. GDP deflator) 2005, while non-oil commodity prices increased by 55 to developing countries, and the world price (in real terms) percent, and oil prices by 119 percent, in dollar terms. of their non-energy commodity exports from 1980 to 2005. This raises an important issue for oil importers: because This co-movement poses a problem for the manage- the non-oil commodity-price cycle may have reached a ment of capital flows in developing countries because, peak, while oil prices are likely to remain high (see chap- when commodity prices are falling (signaling a downturn ter 1), oil importers face the prospect of further declines in in economic activity), capital flows also tend to fall, po- their terms of trade, coupled with a fall in private capital tentially exacerbating the effects of an economic downturn flows. It remains to be seen whether the improved macro- for the developing country. economic environment achieved in recent years will be Over the period 1980–2005, downswings in world sufficient to cope with a substantial fall in both export commodity prices (for those commodities that form a sig- revenues and external finance. Private capital flows in line with non-oil commodity prices Volume of private capital flows during cycles, 1980–2005 Constant $ billions 109 All 237 165 305 Commodity prices Sub-Saharan 8 275 Africa 12 245 5 140 South Asia 12 215 Middle East & 5 North Africa 7 185 115 Latin America 44 155 & Caribbean 73 125 Europe & 20 Commodity market downswings Central Asia 64 Commodity market upswings 90 Capital flows 95 East Asia 28 & Pacific 88 65 0 50 100 150 200 250 65 35 1980 1985 1990 1995 2000Delivered by The World Bank e-library to: 2005 Constant $ billions The World Bank Sources: World Bank Debtor Reporting System and staff estimates. Sources: World Bank Debtor Reporting System and staff estimates. IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 152 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S Figure 5.10 Foreign exchange reserves, by region, Figure 5.11 Foreign exchange reserves as a share 1995–2005 of trade, 1970–2003 $ billions Percent 2000 40 1995 1999 35 1600 2004 30 Developing countries 2005 1200 25 20 800 World 15 400 10 Developed countries 5 0 Oil-exporting Emerging Developing countries Asia countries 0 1970 1975 1980 1985 1990 1995 2000 2004 Sources: IMF, International Financial Statistics and World Bank staff calculations. Source: World Bank staff calculations. Bank and Euro-System); the United States, $37.8 many high-reserve countries have so far managed billion (combined reserves of the Federal Reserve’s to contain expansionary outcomes through large- Open Market Account and the Treasury Depart- scale and routine sterilizations using open-market ment’s Exchange Stabilization Fund); the United operations and other means. In almost all countries Kingdom, $40.9 billion; and Japan, $828.8 billion, included in table 5.5, the change in net foreign as- the largest amount among the developed countries. sets on the central bank’s balance sheets between The large-scale reserve buildups in developing 2001 and 2005 has been largely offset by a de- countries reflect central banks’ policies of interven- crease in net domestic assets, leaving reserve money ing in foreign exchange markets. In practice, the largely unchanged as a percentage of GDP. central banks purchase from private and public en- tities part or all of their inward flow of foreign ex- The accumulation of reserves has concentrated change, paying for them with a mix of local cur- risks on central bank balance sheets rency and debt instruments. Massive foreign The effect of the sterilization of capital flows is to exchange intervention, therefore, is very likely to transfer much of the currency risk associated with have expansionary domestic monetary implications the intermediation of capital flows to the public in many developing countries. The authorities in sector, particularly to the central bank. When the Table 5.5 Changes in central bank balance sheets, 2001–5 % change relative to GDP Net foreign assets Net domestic assets Reserve money 2001 2005 Change 2001 2005 Change 2001 2005 Change Brazil 5.0 5.2 0.2 10.8 4.8 –6.0 6.6 11.1 4.5 China 19.6 34.4 14.8 22.0 6.7 –15.3 42.3 35.3 –7.0 Czech Rep. 22.5 25.0 2.5 –2.2 –5.0 –2.8 22.3 10.7 –11.5 India 10.2 20.2 10.0 8.0 0.2 –7.8 13.8 16.3 2.5 Malaysia 35.0 57.4 22.4 –6.8 –6.5 0.3 12.0 11.0 –1.0 Mexico 7.2 9.6 2.5 –1.7 –0.6 1.0 5.7 8.0 2.3 Poland 13.3 14.3 0.9 2.2 –1.5 –3.7 8.2 7.9 –0.3 Russian Fed. 9.9 24.3 14.3 5.0 –8.5 –13.5 10.8 13.7 2.9 Thailand 20.2 30.0 9.8 9.1 10.1 1.0 14.2 20.9 6.7 Turkey –3.1 6.2 Delivered 9.3 by The World Bank 22.9 e-library –18.1 4.9 to: 10.1 8.4 –1.8 Venezuela, R. B. de 10.5 24.9 14.5 The World 0.0 Bank –2.3 –2.3 7.3 9.2 1.9 IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 Sources: World Bank Data Reporting System and World Bank staff estimates. 153 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 central bank carries out an open-market sterilized The upward pressure on local interest rates in- intervention, it finances its purchase of foreign ex- duced by reserve accumulation could have the per- change reserves by issuing an equivalent amount verse effect of reinforcing the need for more re- of domestic public debt in the form of government serves, as higher interest rates could attract larger (or central bank) securities. Reserves are typically volumes of private inflows. Higher local rates may invested in certain classes of foreign assets deemed well conflict with the government’s policy of stim- to be of “reserve quality” or are used to pay down ulating investment and growth. And they almost existing external public debt. At the end of 2005, always cause an increase in the government’s pub- foreign exchange reserves accounted for about lic debt; such public finance issues arise even if three-fourths of the average assets of central banks these assets are held by agencies other than central of the countries with the largest reserve holdings, banks. The fact that governments tend to entrust ranging from 27 percent in Brazil to 93 percent in the responsibility for accumulation and manage- Malaysia (table 5.6). Since the interest rates on re- ment of official reserves to their central banks serve-grade assets are seldom as high as those on adds to the complexity of the problem at hand by domestic securities, the mismatch often represents bringing to the fore the unique institutional char- a significant loss of revenue, so that more debt has acter of central banks, their role in monetary and to be issued to cover the shortfall. exchange rate management, and their particular The chief domestic implication of high re- accounting and reporting norms and standards. serves is a large accumulation of public debt. As Central banks have a monopoly position in issuing domestic securities are the counterpart liabilities to domestic currency and the rules and agreements foreign assets on the central bank’s balance sheet, governing the distribution of their profits and divi- the bank must be concerned about the effects of a dends to the treasury vary considerably and are rise in local interest rates. Whether they are issued often determined by negotiation (Courtis and in the form of the central bank’s own obligations Mander 2003).9 or drawn from its existing inventory of govern- ment securities, the securities issued to balance out Countries are adjusting the currency foreign currency reserves must compete for the composition of their reserves available supply of domestic savings with securi- The range of foreign assets of reserve quality en- ties issued by the private sector. In some countries, compasses virtually all government securities is- such as China, the supply of domestic securities is- sued by large industrial countries that are denom- sued by the central bank has grown very rapidly in inated in major currencies and traded in deep recent years, from 2.2 percent of GDP in 2003 to liquid markets. The two key qualifying conditions 11 percent of GDP in 2005 (box 5.3). for reserve assets are that they need to be readily available to and controlled by national monetary authorities (IMF 2001). Official holders of re- Table 5.6 Foreign currency reserves and foreign serves may need to access them quickly and under assets as shares of total central bank assets in countries with high reserve accumulations, 2005 difficult market conditions, when the ability to Percent turn reserve assets into cash for intervention pur- poses at the prevailing market price is of the first Foreign reserves/ Net foreign assets/ Country Total assets Total assets importance.10 Almost 93 percent of developing countries’ Brazil 27.9 20.9 China 84.8 79.4 reported official reserve holdings as of the end of India 79.2 88.7 2005 were invested in three major currencies: the Malaysia 93.1 90.8 U.S. dollar, the euro, and the Japanese yen.11 The Mexico 86.4 86.3 Poland 91.8 93.1 euro’s share increased from 20 percent of re- Russian Fed. 84.8 89.7 serves held at end-2000 to 29 percent in 2005, Thailand 63.8 67.1 while the share of U.S. dollar reserves declined Turkey 62.9 27.3 Venezuela, R. B. de 73.9 95.1 from 68 percent to 60 percent during the same Delivered by The World Bank e-library to: Average 74.9 73.8 The World Bankperiod (figure 5.12). Sources: IMF, International Financial Statistics, and IP : 192.86.100.36 World The dominant role of the U.S. dollar is likely Tue, 10Bank Mar 2009 16:55:33 staff calculations. to have persisted into 2006, as much of the reserve 154 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S Box 5.3 Central bank debt in China I n the face of large capital inflows, the People’s Bank of China (PBC) has had to act to stabilize monetary growth, a challenge complicated by the fact that, until July sures, including reserve requirement ratios on domestic banks and credit ceilings on overheated sectors, such as real estate and infrastructure, in order to tighten monetary 2005, the PBC pegged the Chinese currency to the U.S. conditions and contain the inflationary consequences of dollar. A close examination of the PBC balance sheet re- large reserve accumulation. Such measures, coupled with veals a significant level of sterilization in the form of PBC the closed nature of China’s capital markets, have enabled securities issued to offset the domestic monetary conse- the PBC to follow a prudent course of monetary policy. quences of PBC’s purchases of foreign exchange. In 2004 The pace of growth in the money supply (M2) remained and 2005, the PBC issued bonds worth 805 billion and within PBC’s target of 15 percent for much of 2004–5, but 922 billion yuan, respectively, in local markets, raising the the rate of growth seems to have accelerated since the outstanding stock of such bonds from 303 billion yuan in third quarter of 2005, possibly because of PBC’s move to 2003 to 2,033 billion yuan in 2005 (figure at left). In ad- ease its efforts on sterilization so as to buffer the impact of dition, the authorities have relied on administrative mea- a currency revaluation (figure at right). Domestic bond issuance by China’s central bank, 2001–5 China’s money supply and reserve money, 2000–5 Billions of yuan % change year/year 2500 25 Money supply (M2) 2000 20 1500 15 1000 10 500 5 Reserve money 0 2001 2002 2003 2004 2005 0 2000 2001 2002 2003 2004 2005 Sources: IMF, International Financial Statistics and World Bank staff estimates. Sources: IMF, International Financial Statistics and World Bank staff estimates. accumulation during the year was done by Asian the pound sterling as the world’s major currency, and oil-exporting countries, whose main exports despite the fact that the U.S. economy had over- are priced in dollars and whose currencies are in taken Britain’s long before (Cohen 2000). many cases either linked to the dollar or to a bas- • First-mover risks. Choosing an alternative cur- ket of currencies in which the dollar is heavily rency is risky for any individual holder, since it weighted. Although models of optimal portfolio depends for its success on others also deciding investment allocation call for more euros in devel- to use that currency. In other words, there are oping countries’ reserve holdings (box 5.4), fur- network externalities, and such externalities ther shifts into euro reserves are likely to be ham- may justify historical dependence on the use of pered by several factors: that currency as a medium of exchange. • Effects on exchange rates. Switching out of • Inertia. Holdings of reserve currency reflect the the incumbent reserve currency may induce Delivered by The World Bank e-library to: currency’s importance in other areas, such asThe World adverseBank movements in dollar/euro exchange trade, which evolve slowly. A prime example is rates, so large holders may be reluctant to IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 the time it took for the U.S. dollar to overtake switch from the existing reserve holdings. 155 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Figure 5.12 Currency composition of developing changes associated with that surge have already countries’ foreign exchange reserves, 2000 and 2005 become clear. They are an increase in domestic in- 2000 2005 vestment in most recipient countries and a sharp 2% 2% escalation in asset prices in local equity markets. These effects must be considered to be the initial 20% manifestations of the current surge—longer term 29% consequences are still in the making. 68% 60% 4% Our analysis of monetary aggregates, based on a sample of 72 developing countries with access 6% to international capital markets, provided no clear 5% 4% signal of excess money supply growth associated U.S. dollar with the surge in private flows.13 Simple correla- Japanese yen tion and cross-country regression analyses re- British pound vealed no statistically significant relationship be- Euro tween private capital flows and indicators of Others domestic money and credit supply. One plausible explanation is the possibility of a shift in demand Sources: IMF, International Financial Statistics and World Bank staff calculations. for real money balances, brought about as many countries have lowered their inflation while simul- taneously experiencing robust economic growth. Such a consideration may be important in the Higher demand for money has absorbed some liq- current context for official holders of U.S. uidity reducing the pressure on domestic inflation. dollars in Asia and for oil exporters, and any Such findings are also consistent with the conclu- diversification is likely to be incremental sion that, to date, countries have elected to re- through purchase of non-dollar assets in the spond to the surge by accumulating (and steriliz- future, depending on the pace of their reserve ing) large quantities of reserves. This policy accumulation. response is understandable: authorities in recipient • Depth. No market in euro-denominated gov- countries see the surge as temporary and seek to ernment bonds, or indeed in the world, is as avoid adjustments in the current accounts of their deep and liquid as that for U.S. Treasury secu- balance of payments. Sustained access to capital rities. Although the aggregate issuance of flows over time, however, is necessary for capital Euro Area government debt is of the same inflows to have a tangible impact on economic order of magnitude as that of U.S. Treasury is- growth—to the extent that they increase domestic sues, Euro Area debt is the debt of 12 sover- investment or lead to increased domestic financial eign entities, rather than one. So far, there has intermediation (Bailliu 2000) or to enhanced do- been only limited coordination of the schedule mestic firm productivity. Reserve accumulation and structure of issues (Bernanke 2004). and sterilization cannot be a long-term solution to There is also a lack of debt instruments with capital inflows, particularly if developing coun- short maturities, since Euro Area governments tries remain attractive for foreign investment in issue relatively few short-term bills.12 the coming years. Capital flows are sometimes associated with increased domestic investment The effect of the recent influx of Private capital flows can contribute meaningfully capital flows on domestic investment to domestic investment, particularly if they are and asset prices sustained. The influx of private capital flows is I mproved macroeconomic fundamentals, in- creased exchange rate flexibility, and greater fi- associated with increased domestic investment, on average, as well as for most of the 72 develop- Delivered by The World Bank e-library to: nancial openness have enhanced the ability Theof na- Banking countries in our sample. Table 5.7 compares World tional policy makers to deal effectively IPwith the the investment performance (aggregate domestic : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 ongoing surge in capital flows. Two domestic investment as a percentage of GDP) of a large 156 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S Box 5.4 Optimizing allocations in reserve portfolios T he currency composition of reserves can be viewed in terms of a mean-variance, or capital-asset-pricing, model. Such models typically quantify the attractiveness of ling) for a representative country consuming a basket of goods with the same proportions as the SDR weights, on the basis of historical returns on government bonds since reserve assets in optimal portfolios over the long run, in the euro’s introduction. the absence of other factors. In the real world, the choice A comparison of real SDR returns on the major re- of reserve currency is subject to considerable inertia, that serve currencies since 1999 (table below) shows that the is, it evolves slowly. pound sterling had the highest ex post return. While the Thus portfolios based on an optimal reserve-portfolio euro’s mean return was higher than the dollar’s, its stan- model, when compared to actual reserve holdings, provide dard deviation was considerably larger. As a result, the an indication of long-run trends in the composition of re- representative country would hold a proportion of its re- serves (after inertia has worked itself out), rather than pre- serves in euros lower than its SDR weight, while the dol- dictions of near-term reserves changes. The table below lar’s proportion would be slightly higher. provides the optimal reserve allocation across four curren- cies (U.S. dollar, euro, Japanese yen, and the pound ster- Source: IMF Annual Report 2005. Real returns expressed in SDRs, January 1999–September 2005 % per annum Correlations Mean Standard deviation Dollar Pound Yen Euro U.S. dollar 1.98 15.88 1.00 –0.33 –0.09 –0.82 British pound 4.82 17.16 –0.33 1.00 –0.24 0.19 Japanese yen 1.55 26.53 –0.09 –0.24 1.00 –0.32 Euro 3.66 21.86 –0.82 0.19 –0.39 1.00 Optimal share SDR weight U.S. dollar 1.98 15.88 British pound 4.82 17.16 Japanese yen 1.55 26.53 Euro 3.66 21.86 sample of recipient countries during the first with domestic investment during 2002–4.14 This three years of the current surge (2002–4) with the result may reflect the higher share of FDI in capital preceding three years (1999–1). On average, flows in 2002–4, as compared with 1992–7, since across countries, investment rates stand approxi- inbound FDI adds directly to domestic investment mately at the pre–Asian crisis level, although (see box 5.5). In addition, FDI has the potential to many countries have not yet reached that level. In generate positive spillovers in the form of technol- Indonesia, Malaysia, and Thailand, investment ogy transfers, knowledge diffusion, and forward rates remain lower than pre-crisis levels by 10 to and backward linkages, potentially adding stimu- 20 percentage points of GDP, suggesting that the lus to overall domestic investment spending (Razin over-exuberance in investor behavior during the 2003; Alfaro, Chanda, and others 2004). previous capital flow surge has not yet material- ized, although a few countries, such as China, ex- The capital flows surge has not (yet) resulted hibit potential signs of overheating. in excessive demand expansion Simple cross-country regression of domestic One of the questions that arises during the current Delivered by The World Bank e-library to: investment on private capital flows or the compo-The World surge in capital flows, particularly in the quickly Bank nents of those flows reveals that the FDI compo- growing economies of China and India as well as IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 nent of capital flows has the strongest correlation in some of the oil exporting Eastern European 157 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table 5.7 Investment performance during the surge in capital flows, capital flows has resulted in the kind of overheat- 2002–4 ing of domestic economies seen just before the Investment as a % of GDP (averages) East Asian crisis. It is still early, however. Should Average over Change the surge continue, it could result in higher infla- Selected Countries 1994–6 1999–2001 2002–4 (2002/4–1999/2001) tion, currency appreciation, and declines in cur- rent-account balances over the next few years. Azerbaijan 22.7 22.6 45.6 23.0 Bangladesh 19.2 23.3 23.3 0.0 Botswana 25.4 23.7 27.6 4.0 Capital flows are associated with escalation Brazil 21.8 21.1 18.8 –2.3 in asset prices Chile 25.6 21.5 23.2 1.7 China 40.5 37.4 43.2 5.8 Although inflation as a whole has remained sub- Colombia 24.5 13.9 14.9 1.0 dued in most developing countries, one indicator Croatia 19.0 22.4 28.8 6.4 of potential demand pressures is the sharp rise in Ecuador 21.0 20.2 25.9 5.8 Egypt, Arab Rep. of 16.8 18.6 17.0 –1.6 stock prices. The stock market capitalization of El Salvador 18.4 16.7 16.4 –0.3 countries included in the Standard and Poor’s/IFCI Hungary 23.4 28.8 24.9 –3.9 index15 rose from $1.7 trillion at the end of 2002 India 23.9 22.9 22.8 0.0 Indonesia 31.2 18.3 20.4 2.1 to $4.4 trillion at the end of 2005 (figure 5.13). In Jordan 32.2 22.0 22.1 0.2 particular, market capitalization of Asian stock Kazakhstan 22.7 20.9 26.1 5.2 markets tripled during the same period, and stock Malaysia 42.1 24.5 21.9 –2.6 Mexico 21.7 22.8 21.0 –1.7 prices in other major emerging markets saw large Morocco 20.6 23.2 23.5 0.2 increases (more than 100 percent in some cases) in Nigeria 16.7 21.3 23.3 2.0 both local currency and U.S. dollar terms (table Pakistan 19.0 16.7 17.1 0.4 Peru 23.3 20.2 18.7 –1.4 5.9). For many countries, stock markets have now Philippines 23.5 19.6 17.1 –2.5 recovered to the levels they attained before the Poland 18.9 23.4 19.1 –4.3 East Asian crisis. Russian Fed. 24.9 18.5 20.7 2.2 South Africa 17.4 15.9 17.0 1.1 The sharp response of these markets to in- Sri Lanka 25.7 25.8 22.7 –3.0 flows of portfolio capital can be explained by their Thailand 41.4 22.5 25.3 2.9 small size, limited liquidity, and high concentra- Tunisia 24.8 27.1 25.0 –2.2 Turkey 23.8 21.5 23.3 1.7 tion in a few large issues. As shown in figure 5.14, Venezuela, R. B. de 16.3 26.1 19.3 –6.8 turnover ratios, as a percentage of market capital- Vietnam 26.9 29.5 34.2 4.7 ization, for most emerging stock markets in 2004 Zambia 12.3 18.8 24.6 5.8 Total 23.8 21.9 23.3 1.3 were less than 40 percent while for the NYSE and NASDAQ they were 90 percent and 249 percent, Sources: IMF, International Financial Statistics and World Bank staff calculations. Note: A selection of countries is presented; the overall average represents results for a sample respectively. India and Thailand were the excep- of 72 developing countries with access to international capital markets. The countries in the tions with turnover ratios over 100 percent. Trad- sample account for more than 95 percent of private capital flows to developing countries. ing in most emerging markets is also highly con- centrated; for example, in Mexico, trading in eight stocks accounted for 62.7 percent of total trades countries, is whether private capital flows are con- on the exchange. Therefore, relatively small for- tributing to overheating. Several traditional mark- eign portfolio inflows can have a major impact on ers of overheating (acceleration in inflation, rapid the stock prices in these exchanges. increases in domestic investment, and consumer One benefit of the rise in stock market valua- goods imports) have not been evident so far during tion has been its contribution to corporate restruc- this current surge. Inflation has decreased in many turing in several developing countries, especially in developing countries (table 5.8) and remained rel- East Asia. The high market valuations combined atively low, and currencies have not experienced with low local interest rates, have made it possible significant appreciation in terms of their real effec- for many firms to pay off debt, thus reducing tive exchange rate (as noted earlier). Moreover, leverage. The two most highly leveraged corporate there is no sign so far of a run-up in consumption sectors—those of the Republic of Korea and Thai- Delivered by The World Bank e-library to: and imports, and thus of current-account Thedeficits World Bankland—reduced their debt-to-equity ratios below or of sharp rises in domestic investment. IP :It does 75 percent by 2004, down sharply from nearly 192.86.100.36 Tue, 10 Mar 2009 16:55:33 not yet appear that the current surge in private 400 percent in 1997 (figure 5.15). 158 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S Box 5.5 Investment and private capital flows I n order to more carefully examine the relationship be- tween private capital flows and investment, a more rigor- ous analysis is required. In principle, both capital flows and • Taking into account financial development and trade openness, while controlling for other determinants of domestic investment, econometric analysis indicates domestic investment are endogenous variables affected by that for countries reaching a minimum threshold of third factors (such as the investment climate, productivity, financial development and capital-account openness, international interest rates, and economic growth). Because private capital inflows can have a positive and signifi- factors that stimulate domestic investment also tend to at- cant impact on investment. tract private capital flows (and vice versa), the high correla- • Financial development affects the ability of develop- tion of capital flows with investment is not surprising. The ing countries to attract private capital flows and use influence of third variables also suggests that the relation- them for domestic investment. For example, our esti- ship between capital inflows and domestic investment is mates indicate that in Ghana, where the ratio of M2 nonlinear, so that capital inflows have a positive and sig- to GDP is 17 percent, a one-percentage-point increase nificant effect on investment only once a threshold level of in private capital flows (as a share of GDP) would re- financial and economic development has occurred (Rioja sult in an increase in investment of 0.40 percent of and Valev 2004; Bailliu 2000; Alfaro and others 2004). GDP, but only if Ghana’s domestic financial size Econometric analysis offers a more rigorous explana- (ratio of M2 to GDP) was developed to reach 74 per- tion of the dynamics of capital flows and domestic invest- cent, a level comparable to Malaysia’s. ment in recipient countries. The underlying methodology • Similarly, a country like Brazil could experience an and estimation are summarized in the annex. Some key increase in investment of up to 1 percent of GDP as a findings are presented below: result of a one-percentage-point (of GDP) increase in private capital flows—if it became as open to finan- • There is strong statistical evidence that suggests pri- cial flows as Mexico (provided those resources were vate capital flows contribute to increased domestic in- channeled into domestic investment and not reserve vestment across developing countries with access to accumulation). international capital markets. Moreover, since the Asian financial crises, de- nancing (Pinkowitz and others 2003). Poor corpo- veloping countries have made some progress in es- rate governance limits the ability of firms to raise tablishing the institutional and regulatory founda- capital and grow, as capital markets place a lower tions they need to manage capital flows. At the value on poorly governed firms. Recent research same time, they have considerably improved cor- has also highlighted the importance of the coun- porate financial soundness, as firms in virtually all try-level dimension of corporate governance, in- crisis-affected countries have reduced leverage, en- cluding the relationship between the quality of a hanced profitability, and undertaken financial re- country’s institutions and the legal protection structuring. That progress needs to be set against given to investors’ rights, on the one hand, and the still evolving reforms in the areas of corporate effect on investors’ potential returns and overall governance, risk management, and transparency. decisions to invest in a particular country, on the Weak governance results in poor financial report- other (Doidge and others 2004). ing and disclosure, as well as insufficient manage- ment accountability, allowing resources to be used for personal or unrelated uses. It can also provide incentives for short-term gain rather than long- Lessons and policy agenda term stability. The links between financial soundness and good corporate governance are clear. Recent re- I n the last few years, many developing countries have deepened their integration into global capi- tal markets through greater exchange rate flexibil- Delivered by The World Bank e-library to: search has provided evidence that the quality ofThe World ity, development Bank of local capital markets, reduced corporate governance is positively related to dependence on short-term external debt, and grad- IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 growth opportunities and the need for external fi- ual liberalization of cross-border trade in financial 159 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table 5.8 Indicators of overheating in selected developing countries, 2002–4 Change from immediately preceding 3 years: annual period averages in % Current account balance GDP growth Inflation Selected countries 1999–2001 2002–4 Change 1999–2001 2002–4 Change 1999–2001 2002–4 Change Azerbaijan –5.7 –23.7 –17.9 9.5 11.0 1.5 –1.7 3.0 4.7 Bangladesh 3.3 4.0 0.7 Botswana 11.3 6.3 –5.0 6.1 4.8 –1.3 8.0 8.0 0.0 Brazil –4.5 0.4 4.8 2.2 2.5 0.4 6.3 10.0 3.7 Chile –0.9 –0.3 0.6 2.0 3.9 1.8 3.7 2.0 –1.7 China 1.8 3.4 1.6 7.5 9.0 1.5 –0.3 0.0 0.3 Colombia 0.1 –1.3 –1.4 0.1 3.3 3.2 9.3 6.3 –3.0 Croatia –4.4 –7.1 –2.7 2.1 4.4 2.2 4.3 2.0 –2.3 Ecuador 2.7 –2.7 –5.3 0.5 4.2 3.7 62.0 7.7 –54.3 Egypt, Arab Rep. of 2.7 6.3 3.7 El Salvador –2.1 –3.9 –1.8 2.4 1.9 –0.5 2.3 2.7 0.3 Hungary –7.5 –8.3 –0.7 4.4 3.5 –0.9 9.7 5.7 –4.0 India –0.5 0.9 1.4 5.4 6.5 1.1 4.3 4.0 –0.3 Indonesia 4.4 2.9 –1.5 3.2 4.8 1.6 12.0 8.3 –3.7 Jordan 1.3 2.3 1.0 Kazakhstan –1.8 –1.2 0.5 8.7 9.5 0.8 9.7 6.3 –3.3 Malaysia 11.2 11.0 –0.2 5.1 5.5 0.4 2.0 1.3 –0.7 Mexico –3.0 –1.5 1.5 3.4 2.2 –1.2 10.7 5.0 –5.7 Morocco 0.9 3.2 2.2 2.4 4.0 1.6 1.3 2.0 0.7 Nigeria 8.1 11.9 3.8 2.8 5.3 2.5 11.0 14.0 3.0 Pakistan 0.3 3.0 2.6 3.3 4.9 1.6 3.7 4.3 0.7 Peru –2.6 –1.1 1.5 1.3 4.6 3.2 3.0 2.0 –1.0 Philippines 6.5 3.3 –3.2 4.1 4.7 0.5 5.7 4.0 –1.7 Poland –5.5 –3.0 2.5 3.0 3.5 0.5 7.7 2.3 –5.3 Russian Fed. 13.9 8.9 –5.0 7.2 6.4 –0.8 42.7 13.7 –29.0 South Africa –0.2 –1.4 –1.3 3.1 3.4 0.3 5.3 5.7 0.3 Sri Lanka –3.8 –1.9 2.0 2.9 5.3 2.4 8.3 8.0 –0.3 Thailand 7.7 5.1 –2.7 3.8 6.1 2.3 1.3 2.0 0.7 Tunisia –3.5 –2.8 0.7 5.2 4.3 –0.9 2.7 3.3 0.7 Turkey 0.4 0.1 –0.3 –1.6 7.6 9.2 58.0 26.3 –31.7 Venezuela, R. B. de. 4.6 11.5 6.9 0.4 0.3 –0.1 17.7 25.0 7.3 Vietnam 3.2 –2.9 –6.2 6.2 7.3 1.1 0.7 5.0 4.3 Zambia –14.4 –6.3 8.1 3.6 4.4 0.8 24.7 22.0 –2.7 Total 0.4 1.7 1.3 3.4 4.9 1.5 15.9 10.4 –5.6 Sources: IMF, International Financial Statistics and World Bank staff calculations. Figure 5.14 Turnover on world stock exchanges, 2004 Figure 5.13 Market capitalization Percent $ billions 250 5000 Middle East & Africa 200 Emerging Europe 4000 Asia 150 Latin America 3000 100 50 2000 0 1000 hi an daq Au gh SE re SE SE la E In SE ky E Sh N SE Sã M SE an a E SE Sh he nge Lo en e Th do E g ing lia E n p SE E Ch SE e kar ica Ph xch sia ia E E M urs ao bai Sa pin ge ts ha . eu xc rp zh ¨ rs ai n S To a S ic M S n S on S tra i S nt e S ut a S ilip an D E Co Th Ja Afr E lay an Y B oP m en Bo Ko a bul nd o SE ta go Ko a n JS g, ore ex a lo , C Ist as s a di So in u N h c a n 0 Delivered by The World Bank e-library to: 2002 2003 2004 2005 E, an The World Bank BS iw H Ta IP : 192.86.100.36 Sources: World Federation of Exchanges and World Bank staff Tue, 10 Mar 2009 16:55:33 calculations. Sources: World Federation of Exchanges and World Bank staff calculations. 160 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S Table 5.9 Stock market performance in emerging markets, 2002–5 % increase in stock market valuation Local currency U.S. dollar % change, Average annual change, % change, Average annual change, Region/Country 2002–5 2002–5 2002–5 2002–5 Latin America Argentina 277.1 92.4 319.2 106.4 Brazil 167.4 55.8 305.3 101.8 Chile 107.3 35.8 143.4 47.8 Mexico 181.9 60.6 177.3 59.1 Peru 136.5 45.5 142.4 47.5 Asia China 92.2 30.7 97.1 32.4 India 165.8 55.3 183.1 61.0 Indonesia 183.6 61.2 158.0 52.7 Malaysia 36.6 12.2 37.4 12.5 Philippines 113.0 37.7 114.5 38.2 Thailand 128.3 42.8 140.0 46.7 Europe Czech Rep. 442.9 147.6 290.9 97.0 Hungary 175.2 58.4 189.1 63.0 Poland 111.6 37.2 149.1 49.7 Russian Fed. 181.9 60.6 213.4 71.1 Turkey 244.8 81.6 323.5 107.8 Middle East & Africa Egypt, Arab Rep. of 887.0 295.7 947.6 315.9 Morocco 68.6 22.9 84.6 28.2 South Africa 90.5 30.2 157.7 52.6 Sources: Standard & Poor’s IFCI index and World Bank staff calculations. Figure 5.15 Ratios of debt to equity in selected ated policy agenda for developing countries is broad countries, 1996–2004 and complex. However, several key themes are clear. Percent Policy responses to the latest surge in private 500 flows have included the buildup of large foreign Korea, Rep. of Indonesia exchange reserves. 400 Thailand Governments have attempted to minimize the Philippines Malaysia macroeconomic problems associated with large in- flows of foreign capital by recycling those resources 300 into official reserves. Central banks have purchased foreign exchange from local banks and other autho- 200 rized financial intermediaries and invested the pro- ceeds in liquid assets in major industrial countries, 100 particularly in U.S. Treasuries. Recognizing that this process cannot continue indefinitely, policy makers 0 in developing countries are exploring alternative 1996 1998 2000 2002 2004 policies, including improving the return on reserve Sources: Thomson Financial and World Bank staff calculations. holdings by asset diversification, transferring part of the currency risk to the private sector (notably by allowing institutional investors to invest some por- assets. Those developments, coupled with the shift tion of their foreign-currency earnings overseas, from potentially volatile short-term debt to more rather than selling them to the central bank), relying stable FDI, have improved the context for capital more on the stabilizing role of exchange rate Delivered by The World Bank e-library to: flows, raising the likelihood that the economic out-The World Bank and encouraging expansion in aggregate changes, comes of the present surge in capital flows will be demand (both consumption and investment). In IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 better than those observed in the 1990s. The associ- East Asia, efforts are being made to increase the size 161 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 and depth of regional financial markets to recycle change, face opportunities and challenges distinct reserves into productive investments within the re- from those of other developing countries.16 Oil is a gion. Such policy responses need to be orchestrated commodity with an active spot market, as well as a carefully, taking into account the potential threats growing liquid futures market that offers up to 5- of macroeconomic imbalances, overheating, and year contracts, affording oil-exporting countries a asset-price escalation, as well as the need to im- broad range of options and market instruments, prove risk management practices. such as oil derivatives, to manage the future stream For countries with large holdings of foreign of foreign exchange revenues. But, in practice, gov- exchange reserves, allowing local institutional in- ernments have been reluctant to enter futures and vestors to diversify their investment portfolio glob- derivatives markets for several reasons, including ally could provide a viable channel of capital out- their limited capacity for large-scale hedging, insuf- flow, as well as an opportunity for greater risk ficient expertise to trade successfully, and limited diversification. Allowing such investments would access for countries with poor credit. have the salutary effect of transferring foreign ex- A high concentration in a single export com- change risks, currently concentrated on central modity translates into a high degree of volatility in banks’ books, to domestic institutional investors, export earnings. In 2005, 14 of 31 oil-exporting which have a longer investment horizon and can countries depended on oil exports for more than benefit from a more diversified international port- 50 percent of their foreign exchange—among folio. Other vehicles for reducing the pressure on them Libya (94 percent), Saudi Arabia and Kuwait the central banks’ balance sheets might include the (85 percent), and Iran (73 percent). Several coun- creation of specialized investment vehicles similar tries have put aside a fraction of their oil revenues to the Government Investment Corporation of Sin- in so-called stabilization funds or funds for the fu- gapore, the Korea Investment Corporation, and ture. Experience with such funds has been mixed. Kazanah in Malaysia to manage a portion of for- To make the best of them, robust governance and eign exchange reserves for long-term investment. legal frameworks are required to insulate the The assets of institutional investors in several funds from political interference. The government developing countries, especially in East Asia and must set clear investment objectives, adopt sound Latin America, have been growing at a fast clip investment policies, and appoint professional due to rapid growth of pension funds and insur- managers to invest money with proper safeguards ance companies. The establishment of corporate and transparency. pension funds in countries such as the Republic of The development of international norms and Korea and Thailand has contributed to that standards on transparency, corporate governance, growth. Until recently, institutional investors in and regulation of national financial systems has most developing countries have followed very con- raised the confidence of foreign investors in servative investment policies, with government se- emerging market economies. curities accounting for the lion’s share of their as- A hallmark of efforts to improve the interna- sets. Institutional investors in most developing tional financial architecture in the late 1990s was countries are generally prohibited from investing the development, by the international financial in foreign securities. Exceptions include Chile, community, of a set of international norms and Malaysia, the Republic of Korea, and Thailand. At standards on transparency, corporate governance, the end of 2004, Chile’s institutional investors held and regulation and supervision of financial sys- 27.3 percent of their assets in foreign securities, tems. The new standards were designed specifically compared with just 2.8 percent for Thailand and to guide the countries affected by the Asian crises Korea, which only recently have gained the right of the late 1990s to return to international finan- to make limited overseas investments. cial markets, and more generally to pave the way Oil exporters face a different set of policy chal- for the gradual and sequential liberalization of in- lenges, including the need to design appropriate ternational capital movements. International schol- stabilization funds and to rely on market instru- ars have argued that the adoption of open-door fi- Delivered by The World Bank e-library to: market. ments to hedge against volatility in the oil The World Banknancial policies and practices tends to cluster in Oil exporters, most of which are heavily de- time and space (Simmons and Elkins 2004) and IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 pendent on a single commodity for foreign ex- that governments comply with international norms 162 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S and legal commitments if their peers do so and if times of financial market instability, policy coordi- the reputational cost of reneging is perceived to be nation may be needed to limit large swings in ex- high (Simmons 2000). Those arguments have pro- change rates.17 vided a strong intellectual basis for a standards- A second aspect of the multipolar world is that centered approach to bolster market confidence. a wider set of countries now matter in the resolu- Building on the success of earlier norms em- tion of policy imbalances. Developing countries, bodied in the IMF’s safeguards assessments and which would suffer disproportionately from the in- the Special Data Dissemination Standards stability induced by a hard landing, have a shared (adopted in 1996), international norms on trans- interest in seeing multilateral cooperation in inter- parency, financial infrastructure, and corporate national monetary relations. The scope of coopera- governance were formulated on the basis of volun- tion should cover global liquidity, the optimal tary compliance, with monitoring responsibility mode of adjustment, and the role of key currencies. assigned to multinational financial institutions. At The large size of the U.S. current account deficit the request of a member country, the IMF and the has as its counterpart large surpluses in Asia and World Bank assess compliance with the interna- among oil exporters. The anticipated need for a tional standards by preparing and publishing re- real effective depreciation of the dollar to help cor- ports on the observance of standards and codes rect that deficit will have to occur against a wider (ROSCs). International norms—standards of ap- set of currencies than those of the industrial coun- propriate and broadly accepted behavior—en- tries (the Plaza Agreement involved the G-5 coun- hance stability as investors are able to form accu- tries), which may well make policy coordination rate expectations of governments’ behavior. more difficult. However, it is clear that countries The world is moving toward a multipolar inter- with large reserve holdings have a shared interest in national monetary system in which the monetary a smooth adjustment of dollar’s exchange rate. and financial policies of the major industrial coun- Managing capital flows effectively will remain tries of the G-3—and of key emerging market critical to ensuring economic progress in develop- economies that are important players in global ing countries trade and finance—are of predominant importance. Private capital flows to developing countries hit One aspect of the new multipolar world is that an historic high in 2005, but there remains consid- the U.S. dollar is no longer without a serious com- erable room for growth, given developing countries’ petitor as an international currency. The emergence demographic profiles, per capita investment levels of a large and deep market for euro-denominated ($400 in 2004, compared with $6,000 in developed securities widens the opportunities for diversifica- countries), and economic prospects. Investors in de- tion available to developing countries as well as to veloped countries invest less than 3 percent of their other countries. Accumulating euro-denominated fi- portfolios of common stocks in developing coun- nancial assets in proportion to the Euro Area’s share tries; and only 5 percent of global bonds issued in of global production and trade allows governments recent years originated in developing countries. As to hedge against real-side fluctuations. The euro also developing countries’ financial markets become in- provides a potential anchor currency for economies creasingly integrated with global financial markets, closely linked to the existing Euro Area that wish to those percentages are likely to rise (as are develop- peg to a major and widely circulated currency. ing countries’ holdings of foreign assets). To take The emergence of the euro alongside the dol- advantage of those opportunities and protect mar- lar may introduce some instability, however, as the ket access, it will be essential for developing coun- lack of synchronization between the United States tries to vigorously maintain macroeconomic stabil- and the Euro Area may occasionally produce large ity. They also will need to strengthen domestic movements in exchange rates that could have seri- financial markets and institutions to cope more ef- ous consequences for developing countries. Policy fectively with the risks associated with growing cap- coordination may not be necessary in normal ital flows and to maximize the efficiency of capital times, when floating exchange rates and monetary allocation. Sustaining the economic policies and in- Delivered by The World Bank e-library to: policies oriented primarily to domestic targets forThe World Bank that can effectively deal with capital flow stitutions inflation and economic activity facilitate adjust- surges is likely to remain a key issue for developing IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 ment to the shocks hitting the two regions. But at countries for many years to come. 163 (c) The International Bank for Reconstruction and Development / The World Bank Annex: Capital Flows and Domestic Investment B ecause private capital flows may have a control variables are: government size (measured larger impact on investment where the fi- by government expenditure) and institutional de- nancial sector is well developed and restric- velopment (measured by the Freedom House index tions on capital movements are few (Bailliu 2000), of political freedom). Several other control vari- we studied interactions between private capital ables were tried (such as average years of school- flows, financial development, and capital controls. ing, inflation rates, and the extent of paved roads), We tested the relationship between private capital but they proved insignificant in the analysis. flows and investment in a simultaneous equation The motivation for including these control system, where we were interested in both the di- variables comes from several theoretical relation- rect effect of private capital flows on investment ships. Government size is a control for policy at and the indirect effect, which was determined the country level. Political freedom is a proxy for through the interaction of private capital flows institutional quality. The data set consists of a with financial development and capital account re- panel of observations for a sample of 72 develop- strictions, respectively. ing countries with access to international capital The dependent variables in our analysis are markets. The sample was drawn from all regions investment and private capital flows, each as a and includes countries in a broad range of devel- percentage of GDP. The explanatory variables in- opmental stages. China was excluded because of clude trade openness, financial development, capi- the size of its money supply in relation to GDP, tal controls, and a set of control variables. Trade which is far greater than any other developing openness (TO) is defined as exports plus imports country and might have biased the results. The divided by GDP. Financial development (FD) is data were averaged over five-year intervals over measured using M2. Restrictions on movements of 1980–2004 to produce a set of five observations private capital (CC) are measured by the Chinn- per country. The simultaneous equation model we Ito index (2002). The index is larger when there used in our analysis takes into account the endo- are fewer capital controls. Private capital flows geneity of investment and private capital flows (CF/GDP) include both debt and equity flows. The and is written as follows:  1   CF   FD   CC   CF   FD   CF   CC   = α i + β1  + β2  + β3  + β4  ∗ + β5  ∗ + γ i Xit + ε it (5.1)  GDP   it  GDP   it  GDP   it  GDP   it  it   GDP   GDP   it  GDP   it  GDP   it is the equation for investment and  CF   FD   CC   = φi + δ1 (growth )it + δ 2Bank + δ3  + θ i Xit + ε it (5.2)  GDP   it Delivered by The World   GDP   it to: e-library  GDP   it The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 is the equation for private capital flows. 164 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S Table 5A.1 Domestic investment and private capital flows Iterated 3SLS regressions Dependent variable is private capital flows Dependent variable is investment Variables Regression 1 Regression 2 Variables Regression 1 Regression 2 GDP per capita 0.000005* 0.000005* Private capital flows 1.37* 1.58* GDP growth 0.62* 0.62* GDP per capita –0.000008* –0.000008* Trade openness 0.12* 0.12* Trade openness –0.04 –0.15 M2 –0.05* –0.05* M2 0.10* 0.14* Capital controls –0.002 –0.002 Capital controls 0.001 –0.002 Gov’t spending 0.021 0.021 PCF × M2 –1.2 Political freedom 0.001 0.001 PCF × capital controls 0.11 Gov’t spending 0.12* 0.11* Political freedom –0.002* –0.002* Constant –0.03* –0.02* Constant 0.17* 0.15* Note: Regression 1 is without interaction effects; regression 2 is with interaction effects. Iterated 3SLS iterates over the estimated disturbance covariance matrix and parameter estimates until they converge. The technique does not require the assumption that errors are normally dis- tributed. PCF = private capital flows. * = significance at the 5-percent level or better. In each equation, X represents a vector of tic investment, whereas GDP per capita and polit- country specific characteristics: openness to trade, ical freedom (the absence of freedom) have small GDP per capita, government spending, and political negative effects. Turning to the interaction terms, freedom. We used an iterated three-stage least private capital flows have both a direct and indi- squares (3SLS) technique (Zellner and Theil 1962) rect effect on domestic investment. The indirect to estimate the simultaneous equation system to effect comes through the extent of financial devel- take into account the nonlinearity of the investment opment and capital controls, which is determined equation and the endogeneity of the regressors. by the coefficient estimates on the interaction terms First, estimation of private capital flows (column 1) (PCF × M2 and PCF × capital controls). showed that GDP per capita, GDP growth, and We then considered the marginal effects (ob- trade openness had positive and significant effects tained by differentiating investment with respect to on private capital flows, while financial develop- capital flows using the coefficient estimates from our ment measured by M2 had a small negative effect. estimations) of capital flows on growth and invest- For the baseline regression (shown in the first col- ment. We calculated the net effect (both direct and umn of the right-hand panel) of investment, we indirect) of private capital flows on investment as: found that private capital flows, government spend- ing, and financial development (measured by M2)  FD  β1 + β4  had a positive and significant effect on domestic in-  GDP   vestment. Political freedom also had a significant ef- fect—the coefficient is negative because higher val- for the interaction with financial development and ues of political freedom in this index imply less as freedom. Capital controls and trade openness were insignificant at the 10-percent level. (The coefficient  CC  β1 + β5  estimates from the 3SLS are presented.)  GDP   Next we performed a 3SLS regression that in- cluded, in the equation for investment, the inter- for the interaction with capital controls. From this, action effects reported in column (2) in the table, we determined the effect that deepening the finan- which shows first that when interaction effects cial sector or loosening capital controls might have are included, private capital flows and M2 have on investment through their interactions with pri- positive, significant, and direct effects on domes- vate capital. (An example is discussed in the text.) Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 165 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Notes tral bank may be counterparty to either for risk management 1. The coefficient of persistence referred to here is or foreign exchange intervention purposes (see, for example, measured as the coefficient on the lagged term in the regres- Hawkins 2003). sion of the annual ratios of FDI to GDP and debt to GDP, 10. In this regard, reserves need to be distinguished respectively, against a constant and their one-year lag values from other assets held by the official sector primarily for in- for each of the 72 developing countries with access to inter- vestment purposes, rather than for intervention in the for- national capital markets over the period 1980–2004. eign exchange market. 2. The conventional wisdom was that pass-through of 11. Swiss francs and several other currencies are used exchange rate changes into import prices is relatively rapid as foreign exchange reserves, but their shares are too small and more complete in developing than developed countries to be meaningful in this analysis. (Ho and McCauley 2003). Rapid pass-through was cited as 12. It is possible to imagine innovative solutions that a rational for exchange rate management, as changes in ex- would increase the liquidity of European markets, for in- change rates could translate into significant inflationary stance the creation of a single issuer of government short- pressure. However recent research has shown that pass- term paper, as proposed by Alexandre Lamfalussy (Speech at through underwent a transformation during the 1990s for the European Central Bank, April 29–30, 2002). However, many developing countries and now is much slower and less the prospect for such an institution, which presumably would complete (Frankel, Parsley and Wei 2005), although still buy up all the Euro Area governments’ issues, seems distant. faster and more pervasive than for developed countries. 13. The 72 countries in our sample account for more 3. Even when countries announce greater exchange rate than 95 percent of all private capital flows to developing coun- flexibility as a policy, their day-to-day practice may be quite tries. The countries in the sample range from large emerging different. See Calvo and Reinhart (2002) for a discussion. markets (such as China, Malaysia, and Thailand) to small 4. See for example, Schneider and Tornell (2004) and commodity-based economies. They were drawn from all re- Fischer (2001). The increased vulnerability from real ex- gions and from both mid- and low-income categories. change rate appreciation comes through loss of trade com- 14. The implication is that capital inflows and invest- petitiveness and possible worsening of current account bal- ment are correlated—at least some of the capital inflows are ances. going to domestic investment. As the regression excludes 5. During 2002–4, about half of the variation in the other determinants of investment, the degree of this rela- real effective exchange rate appears to have come from the tionship may be overstated. nominal exchange rate, rather than from movements in rela- 15. Excluding Bahrain, Israel, Republic of Korea, tive prices. A simple variance decomposition of the real ef- Saudi Arabia, and Taiwan (China). fective exchange rate into its components (nominal ex- 16. In the last two years, oil-exporting countries have change rate and differences between relative prices) shows benefited from the sharp increase in oil prices. In 2005, total that the nominal rate accounts for about 53 percent of the oil exports from developing countries increased to an esti- variation in the real rate during this period. mated $522.7 billion, up 37.6 percent from 2004. Oil ex- 6. The offshore nondeliverable forward market for se- ports from the Middle East were estimated at $242.7 billion, lected currencies is typically used to hedge currency risks in 46.4 percent of the total. In addition to the Middle Eastern markets where capital controls prevent effective onshore countries, the Russian Federation was one of the major bene- currency risk hedging. ficiaries of the hike in the price of oil. 7. The move to inflation targeting may be a conse- 17. In the mid-1980s, when the U.S. dollar was widely quence of the shift in many developing countries to policies perceived to be overvalued, the Plaza Agreement of Septem- that promote macroeconomic stability. If that is so, it can- ber 1985 helped bring it to a “soft landing”. In the current not be credited directly with improving macroeconomic per- environment a coordinated policy of intervention in foreign formance. As discussed in IMF (2006), the available evi- currency markets is neither desirable nor feasible, given the dence is only suggestive; the time series is too short and the changes in global finance market conditions and actors over number of countries with such targets are too few to make a the past two decades. definitive statement. 8. Yang (2005) found that the increase in remittances makes up for 13 percent of income losses in the current year References and 28 percent within four years of a hurricane. In contrast, increases in ODA and FDI make up for roughly 26 and 21 Alfaro, Laura, Areendam Chanda, Sebnem Kalemli-Ozcan, percent within four years. and Selin Sayek. 2004. “FDI and Economic Growth: 9. Also, despite considerable progress in recent years in The Role of Local Financial Markets.” Journal of Inter- achieving convergence of financial accounting standards be- national Economics 64 (1): 89–112. tween the United States and European Union, and in imple- Bailliu, Jeannine. 2000. “Private Capital Flows, Financial menting the IMF’s safeguards assessment policy, there is yet Development, and Economic Growth in Developing no accepted international accounting standards that are Countries.” Bank of Canada Working Paper 2000-15. suited to the nuances of central banks’ particular role and Bernanke, Ben. 2004. “What Have We Learned from the Euro mandate. Important questions remain on the proper Delivered treat- Bank e-library by The World to: Remarks made at the conference on “The Euro at Five?” ment of unrealized gains or losses, asset valuation, The re- Bank World and at Five: Ready for a Global Role.” Institute for Interna- porting and disclosure of derivatives contracts that 192.86.100.36 IP :the cen- tional Economics, Washington, DC, February 26. Tue, 10 Mar 2009 16:55:33 166 (c) The International Bank for Reconstruction and Development / The World Bank C H A L L E N G E S I N M A N A G I N G C A P I T A L F L O W S Brune, Nancy, Geoffrey Garrett, Alexandra Guisinger, and Edwards, Sebastian. 2001. “Capital Mobility and Economic Jason Sorens. 2001. “The Political Economy of Capital Performance: Are Emerging Markets Different?” Account Liberalization.” Paper prepared for presenta- NBER Working Paper 8076, National Bureau of Eco- tion at the 2001 Annual Meetings of the American Po- nomic Research, Cambridge, MA. litical Science Association, San Francisco, August 31. Edwards, Sebastian. 2005. “Capital Controls, Sudden Stops Calvo, Guillermo A., Morris Goldstein, and Eduard and Current Account Reversals.” NBER Working Hochreiter. 1996. Private Capital Flows to Emerging Paper 11170, National Bureau of Economic Research, Markets after the Mexican Crisis. Washington, DC: In- Cambridge, MA. stitute for International Economics. Eichengreen, Barry. 2001. “Capital Account Liberalization: Calvo, Guillermo A., and Carmen M. Reinhart. 2002. “Fear What Do Cross-Country Studies Tell Us?” The World of Floating.” Quarterly Journal of Economics 117 (2): Bank Economic Review 15 (3): 341–65. 379–408. Eichengreen, Barry, and Ricardo Hausmann. 1999. “Ex- Carvalho, Bernardo S. de M., and Márcio G. P. Garcia. 2005. change Rates and Financial Fragility.” NBER Working “Ineffective Controls on Capital Inflows under Sophisti- Paper 7418, National Bureau of Economic Research, cated Financial Markets: Brazil in the Nineties.” Paper Cambridge, MA. presented at the Inter-American Seminar on Economics Frankel, Jeffrey, David Parsley, and Shang-Jin Wei. 2005. entitled “Strengthening Global Financial Markets,” De- “Slow Pass-Through around the World: A New Import cember 1–3. http://www.nber.org/books/IASE05/. for Developing Countries?” NBER Working Paper Cashin, Paul, Luis Cespedes, and Ratna Sahay. 2003. 11199, National Bureau of Economic Research, Cam- “Commodity Currencies: Developing Countries Re- bridge, MA. liant on Commodity Exports See the Fate of Their Ex- Fischer, Stanley. 2001. “Exchange Rate Regimes: Is the change Rates Tied to Fickle Commodity Markets.” Fi- Bipolar View Correct?” Journal of Economic Perspec- nance and Development 40 (1): 45–48. tives 15 (2): 3–24. Cavoli, Tony, and Ramkishen S. Rajan. 2005. “Have Ex- Fleming, J. Marcus. 1962. “Domestic Financial Policies change Rate Regimes in Asia Become More Flexible under Fixed and under Floating Exchange Rates.” In- Post Crisis? Revisiting the Evidence.” CIES Discussion ternational Monetary Fund Staff Papers 9 (November): Paper 0503, Center for International Economic Stud- 369–72. ies, University of Adelaide. Goldfajn, Ilan, and André Minella. 2005. “Capital Flows Chen, Yu-chin, and Kenneth Rogoff. 2002. “Commodity and Controls in Brazil: What Have We Learned?” Currencies and Empirical Exchange Rate Puzzles.” NBER Working Paper 11640, National Bureau of Eco- IMF Working Paper 02/27, International Monetary nomic Research, Cambridge, MA. Fund, Washington, DC. Goldstein, Morris. 2002. “Managed Floating Plus.” Policy Chinn, Menzie D., and Hiro Ito. 2002. “Capital Account Analyses in International Economics 66. Institute for Liberalization, Institutions and Financial Develop- International Economics, Washington, DC. ment: Cross Country Evidence.” NBER Working Hawkins, John, and Philip Turner. 2000. “Managing For- Paper 8967, National Bureau of Economic Research, eign Debt and Liquidity Risks in Emerging Economies: Cambridge, MA. An Overview.” BIS Policy Paper 8, Bank for Interna- Clarke, George, and Scott Wallsten. 2004. “Do Remittances tional Settlements, Basel. Protect Households in Developing Countries Against Hawkins, Andrew. 2003. “Accounting for Financial Instru- Shocks? Evidence from a Natural Disaster in Ja- ments.” In Accounting Standards for Central Banks. maica.” Unpublished paper, World Bank, Washington, London: Central Banking Publications. DC. November. Ho, Corrinne, and Robert McCauley. 2003 “Living with Cohen, Benjamin J. 2000. “Life at the Top: International Cur- Flexible Exchange Rates: Issues and Recent Experience rencies in the Twenty-First Century.” Princeton Essays in Inflation-Targeting Emerging Market Economies.” in International Finance 221, Department of Econom- BIS Working Paper 130, Bank for International Settle- ics, Princeton University. http://www.princeton.edu/ ments, Basel. ~ies/old_series-win.htm. International Monetary Fund. Various years. Annual Report Corsetti, Giancarlo, Paolo Pesenti, and Nouriel Roubini. on Exchange Arrangements and Exchange Restric- 1998. “What Caused the Asian Currency and Finan- tions. Washington, DC: IMF. cial Crisis? Part I: A Macroeconomic Overview.” ———. 2006. “Inflation Targeting and the IMF.” Monetary NBER Working Paper 6833, National Bureau of Eco- and Financial Systems Department, Policy and Devel- nomic Research, Cambridge, MA. opment Review Department, and Research Depart- Courtis, N., and B. Mander, eds. 2003. Accounting Stan- ment, International Monetary Fund, Washington, DC. dards for Central Banks. London: Central Banking Johnson, Simon, Peter Boone, Alasdair Breach, and Eric Publications. Friedman. 2000. “Corporate Governance in the Asian Dailami, Mansoor. 2000. “Managing Risks of Global Fi- Financial Crisis.” Journal of Financial Economics 58 nancial Market Integration.” Managing Financial and (1–2): 141–86. Corporate Distress: Lessons from Asia. Washington, Kapur, Delivered by The World Devesh. Bank to: “Remittances: The New Develop- 2004. e-library DC: Brookings Institution Press. Bank Mantra?” G-24 Discussion Paper Series 29, The World ment Doidge and others 2004. UNCTAD/GDS/MDPB/G24/2004/5, United Nations IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 167 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Conference on Trade and Development, New York. Rioja, Felix, and Neven Valev. “Does One Size Fit All? A Re- http://www.unctad.org/Templates/Page.asp?in- examination of the Finance and Growth Relationship.” tItemID=2103&lang=1. Journal of Development Economics 74 (2): 429–47. Kletzer, Kenneth, and Mark M. Spiegel. 2004. “Sterilization Rodrik, Dani. 1999. The New Global Economy and Devel- Costs and Exchange Rate Targeting.” Journal of Inter- oping Countries: Making Openness Work. Washing- national Money and Finance 23 (6): 897–915. ton, DC: Overseas Development Council. Lane, Philip R. 2003. “Business Cycles and Macroeconomic Schneider, Martin, and Aaron Tornell. 2004. “Balance Sheet Policies in Emerging Market Economies.” Interna- Effects, Bailout Guarantees, and Financial Crises.” Re- tional Finance 6 (1): 89–108. view of Economic Studies 71 (3): 883–913. Mendoza, Enrique G. 1995. “The Terms of Trade, the Real Simmons, Beth A. 2000. “International Law and State Be- Exchange Rate, and Economic Fluctuations.” Interna- havior: Commitment and Compliance in International tional Economic Review 36 (1): 101–37. Monetary Affairs.” American Political Science Review Miniane, Jacques. 2004. “A New Set of Measures on Capi- 94 (4): 819–35. tal Account Restrictions.” IMF Staff Papers 51 (2): Simmons, Beth A., and Zachary Elkins. 2004. “The Global- 276–308. ization of Liberalization: Policy Diffusion in the Inter- Monetary Authority of Singapore. 2005. “Financial Stability national Political Economy.” American Political Sci- Review.” Monetary Authority of Singapore. December. ence Review 98 (1): 171–89. Mundell, Robert A. 1963. “Capital Mobility and Stabiliza- Stiglitz, Joseph. 2002. Globalization and Its Discontents. tion Policy under Fixed and Flexible Exchange Rates.” New York: W. W. Norton. Canadian Journal of Economics and Political Science World Bank. 1997. Private Capital Flows to Developing 29 (November): 475–85. Countries: The Road to Financial Integration. New Pinkowitz, Lee, Rene M. Stulz, and Rohan Williamson. York: Oxford University Press. 2003. “Do Firms in Countries with Poor Protection of Yang, Dean. 2004. “International Migration, Human Capi- Investor Rights Hold More Cash?” NBER Working tal, and Entrepreneurship: Evidence from Philippine Paper 10188, National Bureau of Economic Research, Migrants’ Exchange Rate Shocks.” Policy Research Cambridge, MA. Working Paper 3578, World Bank, Washington, DC. Quinn, Dennis P. 1997. “The Correlates of Changes in In- Yang, Dean. 2005. “Coping with Disaster: The Impact of ternational Financial Regulation.” American Political Hurricanes on International Financial Flows, Science Review 91: 531–51. 1970–2001.” Unpublished paper. Gerald R. Ford Radelet, Steven, and Jeffrey Sachs. 1998. “The East Asian School of Public Policy, University of Michigan. Financial Crisis: Diagnosis, Remedies, Prospects.” Zellner, A., and H. Theil. 1962. “Three-Stage Least Squares: Brookings Papers on Economic Activity 1, Brookings Simultaneous Estimation of Simultaneous Equations.” Institution, Washington, DC. Econometrica 30 (1): 54–78. Razin, Assaf. 2003. “FDI Flows and Domestic Investment.” CESifo Overview Economic Studies 49 (3): 415–28. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 168 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank . Statistical Appendix T he summary statistical tables have been sig- of each country’s future debt-service streams nificantly revised for this edition of Global (PV) to (a) gross national income (GNI) and Development Finance. The tables in this (b) to exports of goods and services. These statistical appendix are now divided into three sets variables are especially important in the Heav- (see the list of tables on the next page for full de- ily Indebted Poor Countries (HIPC) Initiative, tails): where countries are classified based on the ratio of the present value of public and pub- • External financing. These tables combine the licly guaranteed debt to exports of goods and IMF’s current account, foreign exchange re- services. These variables are averaged over serve, and net inward foreign direct invest- three years, 2002–4. ment data with the World Bank’s portfolio eq- uity and debtor reporting system (DRS) data These indicators do not represent an exhaustive to produce an overall tabulation of how re- set of useful indicators of external debt. They may gions finance themselves externally. not, for example, adequately capture the debt ser- • External liabilities and assets. These tables vicing capacity of countries in which government provide a summary of the DRS debt data that budget constraints are key to debt service difficul- is provided on a country-by-country basis in ties. Moreover, rising external debt may not neces- volume II. sarily imply payment difficulties, especially if there • Key external debt ratios and country classifica- is a commensurate increase in the country’s debt tions. These tables provide a summary of indi- servicing capacity. Thus these indicators should be cators typically used by country risk analysts used in the broader context of a country-specific to monitor and classify countries. The two key analysis of debt sustainability. ratios found in table A.29 are the present value Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 171 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Contents External financing Page A.1 External financing: all developing countries, 1998–2005 173 A.2 External financing: East Asia and Pacific, 1998–2005 174 A.3 External financing: Europe and Central Asia, 1998–2005 175 A.4 External financing: Latin America and the Carribean, 1998–2005 176 A.5 External financing: Middle East and North Africa, 1998–2005 177 A.6 External financing: South Asia, 1998–2005 178 A.7 External financing: Sub-Saharan Africa, 1998–2005 179 A.8 Net inward foreign direct investment, 1997–2005 180 A.9 Net inward portfolio equity flows, 1997–2005 181 A.10 Net inward debt flows to developing countries, 1997–2005 182 A.11 Net inward short-term debt flows to developing countries, 1997–2005 183 A.12 Net inward debt flows to public sector and publicly guaranteed borrowers, 1997–2005 184 A.13 Net inward debt flows to private sector borrowers, 1997–2005 185 A.14 Net inward debt flows from public sector creditors, 1997–2005 186 A.15 Net inward debt flows from private sector creditors, 1997–2005 187 A.16 Gross market-based capital flows to developing countries, 1998–2005 188 A.17 Gross international equity issuance by developing countries, 1998–2005 189 A.18 Gross international bond issuance in developing countries, 1998–2005 190 A.19 Gross international bank lending to developing countries, 1998–2005 191 A.20 Change in foreign exchange reserves, 1998–2005 192 External liabilities and assets A.21 Total external debt of developing countries, 1997–2005 193 A.22 Total external debt of developing countries: medium and long-term, 1997–2005 194 A.23 Total external debt of developing countries: short-term, 1997–2005 195 A.24 Total external debt of developing countries: owed by public and publicly guaranteed borrowers, 1997–2005 196 A.25 Total external debt of developing countries: owed by private sector borrowers, 1997–2005 197 A.26 Total external debt of developing countries: owed to public sector creditors, 1997–2005 198 A.27 Total external debt of developing countries: owed to private sector creditors, 1997–2005 199 A.28 Gross foreign exchange reserves of developing countries, 1997–2005 200 Key external debt ratios and country classifications A.29 Key external debt ratios for developing countries 201 A.30 Classification of countries by region and level of income 204 Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 172 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.1 External financing: all developing countries, 1998–2005 $ billions 1998 1999 2000 2001 2002 2003 2004 2005e Current account balance –89.2 –3.5 47.8 20.5 72.0 123.8 158.3 245.8 as % GDP –1.6 –0.1 0.8 0.4 1.2 1.8 2.0 2.6 Financial flows: Net equity flows 179.4 195.9 182.9 183.3 166.1 186.8 248.8 298.9 Net FDI inflows 172.4 183.3 168.8 176.9 160.3 161.6 211.5 237.5 Net portfolio equity inflows 6.9 12.6 14.1 6.4 5.8 25.2 37.3 61.4 Net debt flows 54.3 16.3 –1.0 –1.5 10.7 72.8 119.1 120.1 Official creditors 34.3 13.9 –5.7 27.4 5.2 –12.3 –28.7 –71.4 World Bank 8.7 8.8 7.9 7.5 –0.2 –0.9 1.3 0.7 IMF 14.1 –2.2 –10.7 19.5 14.0 2.4 –14.7 –41.1 Others 11.5 7.3 –2.9 0.4 –8.6 –13.8 –15.4 –31.0 Private creditors 19.9 2.5 4.7 –28.9 5.5 85.1 147.8 191.6 Net medium- and long-term debt flows 85.7 22.0 11.5 –6.2 1.2 30.2 77.8 122.3 Bonds 40.6 30.6 20.5 11.0 10.8 26.4 43.0 61.7 Banks 50.3 –7.1 –5.2 –10.8 –2.8 9.8 39.4 67.4 Others –5.2 –1.5 –3.8 –6.3 –6.8 –5.9 –4.6 –6.7 Net short-term debt flows –65.8 –19.6 –6.8 –22.7 4.2 54.9 70.0 69.3 Balancing item * –128.1 –175.6 –184.4 –120.6 –76.9 –91.8 –121.4 –271.9 Change in reserves –16.4 –33.2 –45.4 –81.7 –171.9 –291.6 –404.8 –393.0 (– = increase) Memo items: Bilateral aid grants 26.7 28.5 28.7 27.9 32.5 43.7 50.3 52.6 (ex technical cooperation grants) Net private flows (debt+equity) 199.3 198.3 187.7 154.4 171.5 271.9 396.6 490.5 Net official flows (aid+debt) 61.1 42.4 23.0 55.3 37.7 31.4 21.6 –18.8 Note: e = estimate. * Combination of errors and omissions and net acquisition of foreign assets (including FDI) by developing countries. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 173 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.2 External financing: East Asia and Pacific, 1998–2005 $ billions 1998 1999 2000 2001 2002 2003 2004 2005e Current account balance 59.8 60.3 53.7 39.8 61.2 74.9 93.6 143.4 as % GDP 4.5 4.2 3.4 2.4 3.2 3.5 3.8 4.9 Financial flows: Net equity flows 54.7 53.0 50.9 50.6 61.3 72.2 82.1 91.8 Net FDI inflows 57.8 50.8 44.3 48.5 57.2 59.8 64.6 65.3 Net portfolio equity inflows –3.1 2.2 6.6 2.0 4.0 12.4 17.6 26.5 Net debt flows –33.5 –11.7 –16.0 –8.2 –10.3 1.9 37.8 43.9 Official creditors 14.7 12.5 7.0 3.2 –7.9 –7.4 –5.5 –1.9 World Bank 2.8 2.4 1.8 0.9 –1.7 –1.5 –1.9 –1.2 IMF 7.0 1.9 1.2 –2.5 –2.7 –0.5 –1.6 –1.6 Others 4.8 8.2 3.9 4.8 –3.5 –5.4 –1.9 0.9 Private creditors –48.2 –24.2 –22.9 –11.3 –2.4 9.3 43.3 45.8 Net medium- and long-term debt flows –3.5 –10.9 –13.1 –13.0 –12.5 –9.2 9.1 15.8 Bonds 1.0 0.9 –0.7 0.4 0.1 2.5 9.7 12.3 Banks –4.8 –12.0 –11.3 –11.8 –10.3 –8.6 1.4 8.1 Others 0.3 0.2 –1.0 –1.6 –2.3 –3.1 –2.0 –4.6 Net short-term debt flows –44.7 –13.3 –9.9 1.7 10.1 18.5 34.2 30.0 Balancing item * –60.3 –72.3 –78.5 –34.5 –24.1 –12.3 23.4 –61.2 Change in reserves –20.7 –29.3 –10.1 –47.7 –88.0 –136.7 –237.0 –217.9 (– = increase) Memo items: Bilateral aid grants 2.5 2.5 2.5 2.2 2.2 2.5 2.8 2.7 (ex technical cooperation grants) Net private flows (debt+equity) 6.5 28.8 28.0 39.2 58.9 81.5 125.4 137.7 Net official flows (aid+debt) 17.1 15.0 9.5 5.3 –5.7 –4.9 –2.7 0.8 Note: e = estimate. * Combination of errors and omissions and net acquisition of foreign assets (including FDI) by developing countries. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 174 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.3 External financing: Europe and Central Asia, 1998–2005 $ billions 1998 1999 2000 2001 2002 2003 2004 2005e Current account balance –24.5 –1.3 16.3 17.5 5.6 –2.0 4.2 24.8 as % GDP –2.5 –0.2 1.8 1.8 0.5 –0.1 0.3 1.2 Financial flows: Net equity flows 31.4 31.7 31.5 33.0 34.8 36.4 66.5 77.9 Net FDI inflows 27.4 29.7 30.2 32.7 34.9 35.9 62.4 75.6 Net portfolio equity inflows 4.0 2.0 1.3 0.3 –0.1 0.5 4.2 2.3 Net debt flows 42.9 18.6 20.0 2.3 27.6 57.9 83.2 82.9 Official creditors 7.5 –0.6 0.0 2.2 2.7 –6.8 –10.5 –30.9 World Bank 1.5 1.9 2.1 2.1 1.0 –0.7 0.4 0.0 IMF 5.3 –3.1 –0.7 6.1 4.6 –2.0 –5.9 –9.7 Others 0.6 0.7 –1.4 –5.9 –2.9 –4.1 –5.0 –21.2 Private creditors 35.4 19.2 20.0 0.1 24.9 64.7 93.7 113.8 Net medium- and long-term debt flows 29.7 17.6 11.4 5.5 21.1 32.2 64.6 81.3 Bonds 16.0 8.2 5.3 1.6 3.9 10.4 25.7 34.6 Banks 14.6 10.1 7.7 6.1 18.7 23.4 40.6 48.5 Others –1.0 –0.7 –1.6 –2.2 –1.5 –1.6 –1.7 –1.8 Net short-term debt flows 5.7 1.6 8.6 –5.3 3.8 32.5 29.0 32.4 Balancing item * –44.7 –42.6 –51.1 –41.7 –24.2 –31.4 –74.6 –90.9 Change in reserves –5.1 –6.4 –16.6 –11.1 –43.7 –60.9 –79.3 –94.7 (– = increase) Memo items: Bilateral aid grants 5.4 8.2 8.6 7.1 8.5 8.6 10.3 9.7 (ex technical cooperation grants) Net private flows (debt+equity) 66.7 50.9 51.5 33.1 59.7 101.1 160.2 191.7 Net official flows (aid+debt) 12.9 7.6 8.5 9.3 11.2 1.8 –0.2 –21.2 Note: e = estimate. * Combination of errors and omissions and net acquisition of foreign assets (including FDI) by developing countries. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 175 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.4 External financing: Latin America and the Carribean, 1998–2005 $ billions 1998 1999 2000 2001 2002 2003 2004 2005e Current account balance –89.4 –55.4 –46.8 –51.9 –14.9 8.4 19.0 33.9 as % GDP –4.5 –3.1 –2.4 –2.7 –0.9 0.5 0.9 1.5 Financial flows: Net equity flows 71.9 84.7 78.8 73.6 49.6 44.5 60.2 73.9 Net FDI inflows 74.1 88.3 79.3 71.1 48.2 41.1 60.8 61.4 Net portfolio equity inflows –2.2 –3.6 –0.6 2.5 1.4 3.4 –0.6 12.5 Net debt flows 37.9 12.8 –4.7 6.2 –8.6 10.2 –11.4 –15.9 Official creditors 10.9 1.6 –11.1 20.4 12.8 4.9 –10.5 –36.5 World Bank 2.4 2.1 2.0 1.3 –0.3 –0.4 –1.0 –1.8 IMF 2.5 –0.9 –10.7 15.6 11.9 5.6 –6.3 –28.8 Others 6.0 0.4 –2.4 3.5 1.3 –0.3 –3.2 –5.9 Private creditors 27.0 11.2 6.4 –14.1 –21.4 5.4 –1.0 20.5 Net medium- and long-term debt flows 55.3 19.5 8.0 –0.6 –11.4 3.2 –3.9 18.5 Bonds 17.9 20.1 8.3 2.9 –0.4 11.3 –1.1 14.1 Banks 39.1 –1.4 0.5 –2.0 –9.1 –7.2 –2.8 4.4 Others –1.7 0.8 –0.8 –1.4 –1.9 –0.9 0.0 0.0 Net short-term debt flows –28.3 –8.3 –1.6 –13.6 –10.0 2.2 2.9 2.0 Balancing item * –29.6 –49.7 –24.3 –25.1 –25.4 –29.9 –42.9 –59.7 Change in reserves 9.2 7.6 –3.0 –2.9 –0.8 –33.2 –24.9 –32.1 (– = increase) Memo items: Bilateral aid grants 3.2 2.9 2.5 3.2 2.8 3.0 4.9 3.2 (ex technical cooperation grants) Net private flows (debt+equity) 98.9 95.8 85.2 59.5 28.2 49.9 59.3 94.4 Net official flows (aid+debt) 14.1 4.5 –8.6 23.6 15.6 7.9 –5.5 –33.3 Note: e = estimate. * Combination of errors and omissions and net acquisition of foreign assets (including FDI) by developing countries. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 176 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.5 External financing: Middle East and North Africa, 1998–2005 $ billions 1998 1999 2000 2001 2002 2003 2004 2005e Current account balance –9.7 6.2 25.3 15.4 12.0 28.3 43.8 67.2 as % GDP –2.9 1.8 7.0 4.0 3.1 6.7 9.3 11.9 Financial flows: Net equity flows 2.9 3.1 4.4 3.3 3.5 5.7 6.0 10.0 Net FDI inflows 2.7 2.4 4.1 3.4 3.7 5.6 5.3 9.1 Net portfolio equity inflows 0.2 0.7 0.2 –0.1 –0.2 0.1 0.6 0.9 Net debt flows 3.6 –3.0 –3.8 0.3 2.3 –0.4 –1.8 2.2 Official creditors –1.6 –2.5 –2.7 –1.2 –2.5 –2.5 –4.1 –2.4 World Bank –0.2 0.2 –0.3 –0.1 –0.3 –0.3 –0.6 –0.2 IMF 0.0 0.0 –0.2 –0.1 –0.3 –0.6 –0.5 –0.4 Others –1.4 –2.8 –2.2 –1.0 –1.9 –1.6 –3.0 –1.8 Private creditors 5.2 –0.5 –1.1 1.5 4.8 2.1 2.3 4.6 Net medium- and long-term debt flows 1.8 –1.4 0.8 3.8 4.5 0.2 2.3 3.7 Bonds 1.3 1.4 1.2 4.4 5.0 0.7 3.3 2.2 Banks 2.0 –1.6 0.4 –0.1 –0.3 –1.0 –0.8 1.6 Others –1.5 –1.2 –0.9 –0.4 –0.2 0.6 –0.2 –0.1 Net short-term debt flows 3.3 1.0 –1.9 –2.3 0.3 1.9 0.0 0.9 Balancing item * 1.5 –7.4 –21.1 –9.5 –5.8 –11.6 –33.7 –58.2 Change in reserves 1.7 1.2 –4.8 –9.5 –12.0 –22.0 –14.3 –21.3 (– = increase) Memo items: Bilateral aid grants 3.5 2.7 3.1 2.2 2.4 3.6 4.5 4.1 (ex technical cooperation grants) Net private flows (debt+equity) 8.1 2.6 3.3 4.8 8.3 7.8 8.3 14.6 Net official flows (aid+debt) 1.8 0.2 0.3 1.1 –0.1 1.2 0.4 1.7 Note: e = estimate. * Combination of errors and omissions and net acquisition of foreign assets (including FDI) by developing countries. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 177 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.6 External financing: South Asia, 1998–2005 $ billions 1998 1999 2000 2001 2002 2003 2004 2005e Current account balance –9.5 –5.3 –6.3 2.2 11.4 10.6 –4.6 –24.1 as % GDP –1.8 –0.9 –1.1 0.4 1.8 1.4 –0.6 –2.6 Financial flows: Net equity flows 2.9 5.5 6.7 8.8 7.8 13.7 15.9 20.6 Net FDI inflows 3.5 3.1 4.4 6.1 6.7 5.6 7.2 8.4 Net portfolio equity inflows –0.6 2.4 2.4 2.7 1.0 8.0 8.8 12.2 Net debt flows 4.7 0.5 3.5 –0.9 0.0 0.3 7.7 6.4 Official creditors 2.3 2.5 0.5 2.2 –2.4 –1.7 1.0 3.4 World Bank 0.8 1.0 0.7 1.5 –1.0 –0.2 2.0 1.6 IMF –0.4 –0.1 –0.3 0.3 0.1 –0.1 –0.3 –0.1 Others 2.0 1.6 0.0 0.4 –1.5 –1.5 –0.7 1.9 Private creditors 2.4 –2.0 3.0 –3.1 2.4 2.1 6.7 3.0 Net medium- and long-term debt flows 3.7 –2.1 3.9 –2.0 0.6 1.3 3.9 0.6 Bonds 4.2 –1.2 5.4 –0.2 –0.5 –3.0 4.1 –1.6 Banks 0.7 –0.5 –2.0 –1.4 1.2 4.4 0.0 2.2 Others –1.1 –0.4 0.5 –0.3 –0.1 0.0 –0.2 0.0 Net short-term debt flows –1.3 0.1 –0.9 –1.1 1.8 0.7 2.9 2.4 Balancing item * 4.8 4.3 0.8 0.1 7.8 10.4 8.2 3.4 Change in reserves –3.0 –5.0 –4.7 –10.2 –27.0 –35.0 –27.2 –6.3 (– = increase) Memo items: Bilateral aid grants 2.1 2.3 2.1 3.2 2.5 3.9 3.5 4.5 (ex technical cooperation grants) Net private flows (debt+equity) 5.3 3.5 9.7 5.8 10.1 15.8 22.7 23.6 Net official flows (aid+debt) 4.5 4.8 2.6 5.3 0.1 2.2 4.5 7.9 Note: e = estimate. * Combination of errors and omissions and net acquisition of foreign assets (including FDI) by developing countries. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 178 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.7 External financing: Sub-Saharan Africa, 1998–2005 $ billions 1998 1999 2000 2001 2002 2003 2004 2005e Current account balance –15.9 –7.9 5.7 –2.4 –3.4 3.8 2.4 0.6 as % GDP –5.1 –2.6 1.8 –0.8 –1.0 0.9 0.5 0.1 Financial flows: Net equity flows 15.5 18.0 10.7 14.0 9.1 14.3 18.0 24.7 Net FDI inflows 6.9 9.0 6.5 15.0 9.5 13.6 11.3 17.6 Net portfolio equity inflows 8.7 9.0 4.2 –1.0 –0.4 0.7 6.7 7.2 Net debt flows –1.3 –0.9 0.0 –1.4 –0.2 2.8 3.6 0.6 Official creditors 0.5 0.4 0.7 0.6 2.6 1.2 0.8 –3.2 World Bank 1.3 1.1 1.5 1.8 2.2 2.2 2.5 2.2 IMF –0.3 0.0 0.1 0.1 0.5 –0.1 –0.1 –0.4 Others –0.5 –0.7 –0.8 –1.3 0.0 –0.9 –1.5 –5.0 Private creditors –1.8 –1.3 –0.7 –2.0 –2.8 1.5 2.8 3.8 Net medium- and long-term debt flows –1.3 –0.7 0.4 0.1 –1.0 2.5 1.7 2.3 Bonds 0.3 1.2 1.0 1.9 2.7 4.6 1.2 0.0 Banks –1.3 –1.7 –0.6 –1.5 –3.0 –1.2 0.9 2.5 Others –0.2 –0.2 0.0 –0.3 –0.7 –0.9 –0.4 –0.2 Net short-term debt flows –0.5 –0.6 –1.1 –2.1 –1.8 –1.0 1.1 1.5 Balancing item * 0.2 –7.9 –10.2 –9.9 –5.2 –16.9 –1.7 –5.3 Change in reserves 1.5 –1.3 –6.2 –0.3 –0.3 –3.9 –22.2 –20.7 (– = increase) Memo items: Bilateral aid grants 10.1 9.9 10.0 10.0 14.0 22.0 24.2 28.4 (ex technical cooperation grants) Net private flows (debt+equity) 13.7 16.7 9.9 12.1 6.3 15.8 20.7 28.5 Net official flows (aid+debt) 10.6 10.3 10.7 10.7 16.6 23.3 25.1 25.2 Note: e = estimate. * Combination of errors and omissions and net acquisition of foreign assets (including FDI) by developing countries. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 179 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.8 Net inward foreign direct investment, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 168.7 172.4 183.3 168.8 176.9 160.3 161.6 211.5 237.5 East Asia and Pacific 62.1 57.8 50.8 44.3 48.5 57.2 59.8 64.6 65.3 China 44.2 43.8 38.8 38.4 44.2 49.3 53.5 54.9 53.0 Indonesia 4.7 -0.2 -1.9 -4.6 -3.0 0.1 -0.6 1.0 2.3 Malaysia 5.1 2.2 3.9 3.8 0.6 3.2 2.5 4.6 4.2 Philippines 1.2 2.3 1.7 1.3 1.0 1.8 0.3 0.5 1.1 Thailand 3.9 7.3 6.1 3.4 3.9 1.0 1.9 1.4 3.1 Europe and Central Asia 24.6 27.4 29.8 30.2 32.7 34.9 35.9 62.4 75.6 Czech Rep. 1.3 3.7 6.3 5.0 5.6 8.5 2.0 4.5 11.0 Hungary 4.2 3.3 3.3 2.8 3.9 3.0 2.2 4.6 4.0 Poland 4.9 6.4 7.3 9.3 5.7 4.1 4.1 12.6 7.7 Russian Fed. 4.9 2.8 3.3 2.7 2.7 3.5 8.0 12.5 14.6 Ukraine 0.6 0.7 0.5 0.6 0.8 0.7 1.4 1.7 7.8 Turkey 0.8 0.9 0.8 1.0 3.3 1.1 1.8 2.7 7.2 Latin America and the Caribbean 66.7 74.1 88.3 79.3 71.1 48.2 41.1 60.8 61.4 Argentina 9.2 7.3 24.0 10.4 2.2 2.1 1.7 4.1 4.7 Brazil 19.7 31.9 28.6 32.8 22.5 16.6 10.1 18.2 15.2 Chile 5.3 4.6 8.8 4.9 4.2 2.6 4.4 7.6 7.2 Mexico 12.8 12.4 13.4 17.1 27.7 15.5 12.3 17.4 17.8 Venezuela, R. B. de 6.2 5.0 2.9 4.7 3.7 0.8 2.7 1.5 3.0 Middle East and North Africa 2.1 2.7 2.4 4.1 3.4 3.7 5.6 5.3 9.1 Algeria 0.3 0.5 0.5 0.4 1.2 1.1 0.6 0.9 1.4 Egypt, Arab Rep. of 0.9 1.1 1.1 1.2 0.5 0.6 0.2 1.3 3.1 Morocco 0.0 0.0 0.0 0.2 0.1 0.1 2.3 0.8 1.0 South Asia 4.9 3.5 3.1 4.4 6.1 6.7 5.6 7.2 8.4 India 3.6 2.6 2.2 3.6 5.5 5.6 4.6 5.3 5.6 Pakistan 0.7 0.5 0.5 0.3 0.4 0.8 0.5 1.1 2.2 Sub-Saharan Africa 8.3 6.9 9.0 6.5 15.0 9.5 13.6 11.3 17.6 Angola 0.4 1.1 2.5 0.9 2.1 1.7 3.5 1.4 1.5 South Africa 3.8 0.6 1.5 1.0 7.3 0.7 0.8 0.6 6.3 Note: e = estimate. The data do not match GDF Volume II because of revisions of the estimates. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 180 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.9 Net inward portfolio equity flows, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 30.6 6.9 12.6 14.1 6.4 5.8 25.2 37.3 61.4 East Asia and Pacific 4.1 –3.1 2.2 6.6 2.0 4.0 12.4 17.6 26.5 China 5.7 0.8 0.6 6.9 0.8 2.2 7.7 10.9 19.0 Indonesia –5.0 –4.4 –0.8 –1.0 0.4 0.9 1.1 2.1 –0.2 Malaysia 0.0 0.0 0.0 0.0 0.0 –0.1 1.3 4.4 0.9 Philippines –0.4 0.3 1.4 –0.2 0.4 0.4 0.5 0.4 1.5 Thailand 3.9 0.3 0.9 0.9 0.4 0.5 1.8 –0.3 5.3 Europe and Central Asia 4.0 4.0 2.0 1.3 0.3 –0.1 0.5 4.2 5.8 Czech Rep. 0.4 1.1 0.1 0.6 0.6 –0.3 1.1 0.7 –1.5 Hungary 1.0 0.6 1.2 –0.4 0.1 –0.1 0.3 1.5 0.0 Poland 0.6 1.7 0.0 0.4 –0.3 –0.5 –0.8 1.9 1.3 Russian Fed. 1.3 0.7 –0.3 0.2 0.5 2.6 0.4 0.2 –0.2 Turkey 0.0 –0.5 0.4 0.5 –0.1 0.0 0.9 1.4 5.7 Latin America and the Caribbean 13.3 –2.2 –3.6 –0.6 2.5 1.4 3.4 –0.6 8.5 Argentina 1.4 –0.2 –10.8 –3.2 0.0 –0.1 0.1 –0.1 0.0 Brazil 5.1 –1.8 2.6 3.1 2.5 2.0 3.0 2.1 6.5 Chile 1.7 0.6 0.5 –0.4 –0.2 –0.3 0.3 0.0 1.7 Mexico 3.2 –0.7 3.8 0.4 0.2 –0.1 –0.1 –2.5 3.4 Venezuela, R. B. de 1.4 0.2 0.4 –0.6 0.0 0.0 0.1 –0.2 0.1 Middle East and North Africa 0.7 0.2 0.7 0.2 –0.1 –0.2 0.1 0.6 0.9 Egypt, Arab Rep. of 0.5 –0.2 0.7 0.3 0.0 –0.2 0.0 0.0 0.7 South Asia 2.9 –0.6 2.4 2.4 2.7 1.0 8.0 8.8 12.2 India 2.6 –0.6 2.3 2.3 2.9 1.0 8.2 8.8 12.2 Sub-Saharan Africa 5.6 8.7 9.0 4.2 –1.0 –0.4 0.7 6.7 7.2 South Africa 5.5 8.6 9.0 4.2 –1.0 –0.4 0.7 6.7 7.1 Note: e = estimate. The data do not match GDF Volume II because of revisions of the estimates. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 181 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.10 Net inward debt flows to developing countries, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 107.2 54.3 16.3 –1.0 –1.5 10.7 72.8 119.1 120.1 East Asia and Pacific 44.9 –33.5 –11.7 –16.0 –8.2 –10.3 1.9 37.8 43.9 China 18.5 –14.2 –1.6 –5.2 0.0 4.0 13.5 37.3 — Indonesia 10.1 –4.6 –3.8 –0.7 –6.0 –7.5 –5.5 –3.0 — Malaysia 8.4 –3.6 –0.7 0.3 5.2 3.4 –1.5 3.5 — Philippines 7.5 –4.1 3.7 2.6 2.1 –0.8 0.6 –0.7 — Thailand –1.3 –7.9 –9.4 –13.7 –10.0 –10.0 –7.5 –1.4 — Europe and Central Asia 35.6 42.9 18.6 20.0 2.3 27.6 57.9 83.2 82.9 Bulgaria 1.0 0.2 0.3 0.5 –0.2 0.6 1.0 1.7 — Czech Rep. 3.3 1.4 –0.2 –1.7 –0.5 1.0 3.5 5.2 — Hungary –1.4 2.7 2.0 0.4 1.7 0.5 4.4 14.6 — Poland 3.8 5.1 4.8 0.8 2.5 1.2 7.0 –2.2 — Russian Fed. 7.6 21.9 –4.2 –2.8 –3.9 –2.6 12.8 10.8 — Turkey 4.2 5.5 10.9 18.2 –4.5 13.2 4.9 13.4 — Latin America and the Caribbean 25.2 37.9 12.8 –4.7 6.2 –8.6 10.2 –11.4 –15.9 Argentina 17.1 11.7 6.4 4.3 –5.6 –1.8 –0.1 –4.1 — Brazil –1.3 6.7 –4.9 –1.0 5.5 –1.6 3.9 –11.2 — Chile 1.8 4.0 1.7 2.9 0.5 1.6 1.9 0.5 — Colombia 3.6 0.8 1.3 –0.2 2.8 –0.9 0.6 0.6 — Mexico –4.9 9.0 6.9 –15.8 –2.9 –8.2 –1.3 –2.9 — Venezuela, R. B. de 2.6 1.7 0.2 0.9 –1.1 –3.2 0.1 0.4 — Middle East and North Africa –3.5 3.6 –3.0 –3.8 0.3 2.3 –0.4 –1.8 2.2 Algeria –0.4 –1.7 –1.9 –1.6 –2.0 –1.4 –1.3 –2.5 — Egypt, Arab Rep. of 0.6 1.1 –0.6 –0.7 0.1 –0.7 –1.1 –2.0 — Lebanon 1.1 1.7 1.5 1.8 2.7 4.4 1.2 3.4 — South Asia 0.7 4.7 0.5 3.5 –0.9 0.0 0.3 7.7 6.4 India –1.6 3.0 –1.1 3.4 –1.9 –1.4 0.0 7.1 — Pakistan 1.6 0.7 0.7 –0.3 0.3 0.6 –1.1 –0.8 — Sub-Saharan Africa 4.4 –1.3 –0.9 0.0 –1.4 –0.2 2.8 3.6 0.6 South Africa –0.4 –0.3 –0.7 1.2 –0.8 –0.5 2.7 1.2 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 182 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.11 Net inward short-term debt flows to developing countries, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 9.2 –65.8 –19.6 –6.8 –22.7 4.2 54.9 70.0 69.3 East Asia and Pacific 4.7 –44.7 –13.3 –9.9 1.7 10.1 18.5 34.2 30.0 China 6.1 –14.1 –2.2 –2.1 1.8 9.6 18.4 29.3 — Indonesia 0.6 –9.7 –1.6 1.5 –1.0 0.2 –0.9 1.6 — Malaysia 3.9 –6.5 –2.5 –1.4 2.2 1.6 0.3 2.8 — Philippines 3.8 –5.9 –0.9 0.5 0.5 –0.4 0.6 –1.1 — Thailand –9.9 –8.2 –6.2 –8.5 –1.7 –1.3 –1.0 0.5 — Europe and Central Asia 10.9 5.7 1.6 8.6 –5.3 3.8 32.5 29.0 32.4 Bulgaria 0.8 –0.2 –0.3 0.2 –0.2 0.6 0.6 0.6 — Czech Rep. 2.4 –0.5 1.1 0.2 0.6 –0.3 1.6 3.2 — Hungary 0.0 1.4 –1.2 0.6 0.5 1.0 2.5 3.3 — Poland 2.5 3.3 2.8 –1.7 1.5 0.4 4.8 –2.7 — Russian Fed. –1.4 –0.5 –1.0 2.0 2.5 –1.6 9.6 5.0 — Turkey 0.6 3.2 2.3 5.4 –12.6 0.1 4.4 8.9 — Latin America and the Caribbean –7.8 –28.3 –8.3 –1.6 –13.6 –10.0 2.2 2.9 2.0 Argentina 8.5 –1.0 –1.5 –1.1 –8.3 –0.4 0.7 –0.3 — Brazil –16.0 –24.0 0.7 1.8 –2.5 –4.9 1.2 0.7 — Chile –1.5 –0.4 –0.8 1.9 –0.9 0.5 1.7 0.2 — Colombia –0.1 0.5 –2.3 –1.1 0.4 0.4 –0.1 1.8 — Mexico –2.0 –1.5 –2.3 –5.1 –4.4 –4.7 –0.7 –0.1 — Venezuela, R. B. de 1.5 –2.0 –0.1 2.0 0.7 –0.2 –0.2 0.0 — Middle East and North Africa 0.0 3.3 1.0 –1.9 –2.3 0.3 1.9 0.0 0.9 Algeria –0.2 0.0 0.0 0.0 0.0 –0.1 0.0 0.3 — Egypt, Arab Rep. of 0.6 1.3 0.0 –0.2 –0.7 0.1 0.3 –0.9 — Lebanon 0.1 0.2 0.2 0.3 0.1 –0.1 0.6 0.8 — South Asia –2.1 –1.3 0.1 –0.9 –1.1 1.8 0.7 2.9 2.4 India –1.7 –0.7 –0.4 –0.5 –0.7 1.4 0.9 2.5 — Pakistan –0.3 –0.5 –0.1 –0.3 –0.2 0.2 –0.3 0.0 — Sub-Saharan Africa 3.5 –0.5 –0.6 –1.1 –2.1 –1.8 –1.0 1.1 1.5 South Africa 0.1 0.5 –0.6 0.3 –1.2 –1.0 0.0 0.6 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 183 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.12 Net inward debt flows to public sector and publicly guaranteed borrowers, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 42.0 69.4 31.0 9.3 20.1 4.1 –11.2 –3.1 –36.5 East Asia and Pacific 29.0 19.1 11.1 4.7 –0.8 –11.7 –12.4 –2.5 2.5 China 11.1 2.5 1.6 –1.1 0.0 –5.3 –5.7 1.5 — Indonesia 3.6 9.0 2.0 0.9 –2.2 –3.1 –0.9 –4.4 — Malaysia 1.7 0.5 0.9 1.4 3.1 2.0 –2.1 0.0 — Philippines 1.8 1.6 4.6 3.0 0.2 0.7 0.8 1.4 — Thailand 9.4 4.6 1.9 –0.2 –2.7 –6.2 –5.4 –2.2 — Europe and Central Asia 15.6 21.8 7.0 5.2 –1.5 3.6 –4.6 –0.2 –13.8 Bulgaria 0.2 0.3 0.4 0.2 –0.1 –0.3 0.0 –0.7 — Czech Rep. 1.0 1.0 –1.0 –1.0 –0.9 0.2 0.5 1.7 — Hungary –1.8 –0.4 1.5 –1.4 –0.8 –0.8 0.5 3.2 — Poland 0.5 –0.1 –0.3 –1.4 –3.3 0.1 1.7 0.5 — Russian Fed. 7.1 16.2 –3.5 –3.9 –7.0 –4.1 –7.2 –7.1 — Turkey 2.5 –1.0 4.6 11.3 9.2 7.5 –1.7 –2.1 — Latin America and the Caribbean –2.0 25.2 12.4 –2.9 19.4 9.5 9.3 –2.5 –20.0 Argentina 4.9 8.4 8.7 6.4 6.7 –1.5 –0.9 –2.3 — Brazil –0.3 12.7 1.5 –3.4 9.6 10.4 3.1 –5.2 — Chile –0.3 0.6 0.6 –0.4 0.4 1.1 1.1 1.3 — Colombia 1.1 1.0 3.4 0.9 2.5 –1.3 1.7 0.4 — Mexico –9.9 0.7 –3.7 –9.1 –3.0 –1.9 –0.6 –1.4 — Venezuela, R. B. de 0.4 0.2 –0.6 –0.5 –1.7 –2.6 0.3 1.0 — Middle East and North Africa –4.1 –1.9 –2.9 –2.6 2.3 2.2 –2.3 –1.8 –1.4 Algeria –0.3 –1.7 –2.0 –1.6 –1.9 –1.4 –1.7 –2.8 — Egypt, Arab Rep. of –0.1 –0.5 –0.7 –0.6 0.8 –0.8 –1.1 –0.8 — Lebanon 0.5 1.7 1.4 1.4 2.5 4.7 0.6 2.5 — South Asia 0.8 5.5 1.4 4.5 0.5 –2.4 –2.8 0.8 –1.8 India –1.5 3.6 –0.1 3.8 –1.2 –3.4 –3.7 0.5 — Pakistan 1.6 0.9 1.2 0.3 0.9 0.4 –0.5 –0.8 — Sub-Saharan Africa 2.8 –0.4 1.8 0.4 0.2 2.9 1.6 3.1 –2.1 South Africa 1.1 –1.0 1.6 0.0 –0.4 1.4 0.0 1.1 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 184 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.13 Net inward debt flows to private sector borrowers, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 65.2 –15.1 –14.6 –10.3 –21.6 6.6 84.0 122.2 156.7 East Asia and Pacific 15.9 –52.6 –22.8 –20.6 –7.4 1.4 14.3 40.3 41.4 China 7.4 –16.7 –3.2 –4.1 –0.1 9.2 19.4 35.8 — Indonesia 6.5 –13.6 –5.8 –1.6 –3.8 –4.4 –4.6 1.3 — Malaysia 6.7 –4.0 –1.6 –1.1 2.1 1.4 0.6 3.6 — Philippines 5.8 –5.7 –0.9 –0.5 1.9 –1.5 –0.1 –2.1 — Thailand –10.7 –12.5 –11.3 –13.5 –7.3 –3.7 –2.0 0.8 — Europe and Central Asia 20.0 21.0 11.6 14.8 3.8 24.0 62.4 83.3 96.7 Bulgaria 0.8 –0.1 –0.1 0.3 –0.1 0.9 1.0 2.4 — Czech Rep. 2.3 0.4 0.8 –0.6 0.4 1.6 3.0 3.5 — Hungary 0.5 3.1 0.5 1.8 2.5 1.3 5.8 11.4 — Poland 3.3 5.2 5.1 2.2 5.8 1.0 5.3 –2.7 — Russian Fed. 0.5 2.4 –0.7 1.1 3.1 1.5 20.0 17.8 — Turkey 1.8 6.5 6.3 6.8 –13.7 5.7 4.7 15.5 — Latin America and the Caribbean 27.2 12.7 0.4 –1.8 –13.2 –18.1 0.9 –8.9 4.1 Argentina 12.3 3.4 –2.4 –2.1 –12.3 –0.5 0.8 –1.8 — Brazil –1.0 –5.3 –6.4 2.4 –4.2 –11.9 0.8 –6.0 — Chile 2.1 3.5 1.1 3.3 0.1 0.5 0.8 –0.9 — Colombia 2.5 –0.2 –2.1 –1.1 0.3 0.4 –1.1 0.2 — Mexico 5.0 8.3 10.5 –6.6 0.1 –6.2 –0.7 –1.5 — Venezuela, R. B. de 2.2 1.5 0.7 1.4 0.6 –0.6 –0.2 –0.6 — Middle East and North Africa 0.6 5.5 –0.1 –1.2 –1.9 0.1 2.0 0.0 3.6 Algeria –0.2 0.0 0.0 0.0 0.0 –0.1 0.5 0.4 — Egypt, Arab Rep. of 0.6 1.5 0.1 –0.1 –0.7 0.1 0.0 –1.2 — Lebanon 0.6 0.1 0.1 0.4 0.2 –0.2 0.6 0.9 — South Asia –0.1 –0.8 –0.9 –1.1 –1.4 2.3 3.2 6.9 8.2 India –0.1 –0.5 –1.0 –0.4 –0.7 2.0 3.7 6.6 — Pakistan 0.0 –0.2 –0.5 –0.6 –0.5 0.1 –0.6 –0.1 — Sub-Saharan Africa 1.6 –0.9 –2.7 –0.4 –1.6 –3.1 1.2 0.5 2.7 South Africa –1.5 0.7 –2.3 1.3 –0.4 –1.9 2.6 0.1 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 185 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.14 Net inward debt flows from public sector creditors, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 13.1 34.3 13.9 –5.7 27.4 5.2 –12.3 –28.7 –71.4 East Asia and Pacific 17.3 14.7 12.5 7.0 3.2 –7.9 –7.4 –5.5 –1.9 China 4.3 2.3 3.4 1.5 2.2 –1.2 –3.2 0.0 — Indonesia 3.6 8.5 4.8 2.9 –0.8 –1.4 –0.3 –3.7 — Malaysia –0.2 0.2 0.6 0.6 2.1 –0.2 –0.1 0.7 — Philippines 0.6 0.7 0.2 0.3 –0.3 –0.5 –0.6 –1.1 — Thailand 8.4 1.8 2.5 0.3 –1.5 –5.5 –4.5 –2.6 — Europe and Central Asia 6.7 7.5 –0.6 0.0 2.2 2.7 –6.8 –10.5 –30.9 Bulgaria 0.3 0.4 0.3 0.2 –0.3 –0.3 0.1 0.1 — Czech Rep. –0.1 0.0 0.0 0.1 0.2 0.0 0.1 0.0 — Hungary –0.1 –1.1 0.2 –0.2 –0.2 0.0 –0.5 0.4 — Poland –0.1 –0.5 –0.4 –0.5 –4.1 –1.1 –1.7 –2.8 — Russian Fed. 4.2 6.3 –3.0 –3.3 –4.8 –3.3 –4.3 –4.9 — Turkey –0.2 –0.4 –0.1 4.4 10.4 6.7 –1.3 –3.4 — Latin America and the Caribbean –8.7 10.9 1.6 –11.1 20.4 12.8 4.9 –10.5 –36.5 Argentina –0.1 1.1 –0.1 0.9 10.3 –1.4 –0.9 –2.2 — Brazil –1.2 9.5 4.5 –8.5 9.5 12.1 3.0 –7.3 — Chile –0.4 –0.1 –0.1 –0.1 –0.1 –0.3 –0.1 –0.1 — Colombia –0.5 0.2 1.0 0.1 1.1 0.0 2.1 0.1 — Mexico –8.0 –1.9 –5.3 –4.8 –0.7 0.2 –0.3 –1.2 — Venezuela, R. B. de –0.3 1.0 –0.1 –0.3 –1.1 –0.6 –0.6 –0.4 — Middle East and North Africa –4.0 –1.6 –2.5 –2.7 –1.2 –2.5 –2.5 –4.1 –2.4 Algeria 0.3 –0.4 –0.4 –0.4 –1.0 –1.3 –1.3 –2.3 — Egypt, Arab Rep. of 0.0 –0.2 –0.5 –0.6 –0.7 –0.8 –0.8 –0.9 — Lebanon 0.1 0.2 0.1 0.1 0.1 0.0 0.6 0.0 — South Asia 0.3 2.3 2.5 0.5 2.2 –2.4 –1.7 1.0 3.4 India –1.0 0.6 0.8 –0.3 0.4 –3.8 –2.7 0.9 — Pakistan 0.7 0.9 1.2 0.3 1.1 0.9 –0.3 –1.0 — Sub-Saharan Africa 1.4 0.5 0.4 0.7 0.6 2.6 1.2 0.8 –3.2 South Africa –0.4 –0.4 0.0 0.1 0.0 0.0 0.1 0.0 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 186 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.15 Net inward debt flows from private sector creditors, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 94.1 19.9 2.5 4.7 –28.9 5.5 85.1 147.8 191.6 East Asia and Pacific 27.6 –48.2 –24.2 –22.9 –11.3 –2.4 9.3 43.3 45.8 China 14.2 –16.5 –5.0 –6.8 –2.2 5.2 16.9 37.3 — Indonesia 6.5 –13.0 –8.6 –3.6 –5.2 –6.1 –5.1 0.7 — Malaysia 8.6 –3.8 –1.3 –0.3 3.1 3.6 –1.4 2.8 — Philippines 7.0 –4.8 3.5 2.3 2.4 –0.3 1.2 0.4 — Thailand –9.7 –9.6 –11.9 –14.0 –8.5 –4.4 –3.0 1.2 — Europe and Central Asia 28.8 35.3 19.2 20.0 0.1 24.9 64.7 93.7 113.8 Bulgaria 0.7 –0.2 –0.1 0.2 0.1 0.9 0.9 1.6 — Czech Rep. 3.3 1.4 –0.2 –1.7 –0.7 1.7 3.3 5.2 — Hungary –1.3 3.8 1.8 0.7 1.9 0.6 6.8 14.2 — Poland 3.9 5.6 5.2 1.3 6.6 2.2 8.7 0.6 — Russian Fed. 3.4 12.3 –1.2 0.5 0.9 0.8 17.0 15.7 — Turkey 4.4 5.9 11.0 13.8 –14.9 6.5 4.3 16.8 — Latin America and the Caribbean 33.8 27.0 11.2 6.4 –14.1 –21.4 5.4 –1.0 20.5 Argentina 17.3 10.7 6.4 3.4 –16.0 –0.5 0.8 –1.8 — Brazil –0.1 –2.1 –9.4 7.5 –4.0 –13.6 0.9 –3.9 — Chile 2.2 4.1 1.8 3.0 0.6 1.9 2.0 0.6 — Colombia 4.1 0.6 0.2 –0.3 1.7 –0.9 –1.4 0.5 — Mexico 3.1 10.8 12.2 –11.0 –2.3 –8.4 –1.0 –1.7 — Venezuela, R. B. de 2.9 0.7 0.3 1.2 0.0 –2.6 0.7 0.8 — Middle East and North Africa 0.5 5.2 –0.5 –1.1 1.5 4.8 2.1 2.3 4.6 Algeria –0.7 –1.3 –1.5 –1.2 –1.0 –0.1 0.1 –0.2 — Egypt, Arab Rep. of 0.6 1.3 –0.1 –0.1 0.8 0.1 –0.3 –1.1 — Lebanon 1.0 1.6 1.4 1.7 2.6 4.4 0.6 3.4 — South Asia 0.4 2.4 –2.0 3.0 –3.1 2.4 2.1 6.7 3.0 India –0.6 2.5 –1.9 3.6 –2.3 2.4 2.8 6.2 — Pakistan 0.9 –0.2 –0.6 –0.7 –0.7 –0.3 –0.8 0.2 — Sub-Saharan Africa 3.0 –1.8 –1.3 –0.7 –2.0 –2.8 1.5 2.8 3.8 South Africa 0.0 0.1 –0.7 1.2 –0.8 –0.5 2.6 1.1 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 187 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.16 Gross market-based capital flows to developing countries, 1998–2005 $ billions 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 160.4 149.7 201.8 135.9 131.6 186.0 249.4 385.2 East Asia and Pacific 25.5 27.0 48.5 19.6 41.1 50.6 55.8 84.8 China 9.2 8.3 29.1 7.5 15.8 24.8 31.9 50.1 Indonesia 0.7 2.4 1.1 1.0 1.6 6.5 4.1 7.2 Malaysia 3.4 7.4 6.8 5.1 12.8 7.3 10.5 8.2 Philippines 5.8 7.2 7.3 3.6 6.5 7.1 5.4 8.2 Thailand 6.4 1.5 4.2 2.5 3.6 4.3 3.6 8.4 Europe and Central Asia 35.3 29.4 40.4 24.9 32.3 50.0 81.4 142.3 Czech Rep. 3.0 1.4 1.4 0.8 0.7 2.5 4.6 3.4 Hungary 3.6 3.4 2.1 2.7 1.8 5.4 8.7 10.3 Poland 3.6 4.4 3.6 4.7 6.5 7.9 6.9 19.1 Russian Fed. 8.3 0.7 5.3 4.6 9.9 16.4 26.8 62.8 Turkey 9.9 12.8 22.3 6.8 7.3 7.5 14.9 24.1 Latin America and the Caribbean 81.8 72.5 87.7 71.9 40.8 60.6 68.2 94.9 Argentina 25.4 20.1 18.8 6.3 1.9 0.7 0.8 2.5 Brazil 17.8 15.1 27.9 23.7 13.2 17.8 23.4 30.5 Chile 4.2 9.6 7.7 6.1 3.6 5.2 8.0 10.2 Mexico 19.3 17.4 20.9 18.4 13.2 26.7 22.6 30.2 Venezuela, R. B. de 7.6 1.8 2.9 4.8 0.6 4.1 4.7 6.5 Middle East and North Africa 4.8 8.6 6.4 8.5 8.5 5.6 15.5 22.4 Egypt, Arab Rep. of 1.8 4.5 1.1 2.4 1.0 0.4 1.7 6.7 Lebanon 1.5 1.4 1.9 3.3 1.0 0.2 4.4 1.8 South Asia 5.3 4.1 5.8 3.0 2.1 5.8 16.7 25.8 India 4.3 3.7 5.4 2.6 1.6 4.1 15.1 24.1 Pakistan 0.9 0.0 0.0 0.2 0.4 1.5 1.3 1.3 Sub-Saharan Africa 7.7 8.1 13.0 7.9 6.9 13.6 11.9 15.2 South Africa 4.9 6.0 10.4 4.9 3.7 8.3 6.2 8.4 Note: e = estimate. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 188 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.17 Gross international equity issuance by developing countries, 1998–2005 $ billions 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 8.0 10.1 35.5 5.6 10.7 17.1 35.2 56.2 East Asia and Pacific 4.3 6.0 21.5 3.5 7.2 12.1 20.6 30.0 China 1.4 3.6 21.1 2.9 5.4 8.8 17.7 26.6 Indonesia 0.0 0.8 0.0 0.3 0.3 0.9 0.7 0.8 Malaysia 0.2 0.1 0.2 0.0 1.3 0.6 0.9 1.4 Philippines 0.5 0.4 0.1 0.0 0.1 0.1 0.2 0.9 Thailand 2.3 0.8 0.0 0.2 0.1 1.5 1.1 0.3 Europe and Central Asia 2.4 1.1 3.3 0.3 1.6 1.3 5.5 10.0 Czech Rep. 0.1 0.3 0.2 0.0 0.0 0.0 0.2 0.3 Hungary 0.3 0.2 0.0 0.0 0.0 0.0 0.8 0.0 Poland 0.8 0.3 0.1 0.0 0.2 0.6 0.9 0.9 Russian Fed. 0.0 0.1 0.5 0.2 1.3 0.6 2.7 6.5 Turkey 0.8 0.0 2.4 0.0 0.1 0.1 0.8 1.5 Latin America and the Caribbean 0.2 0.7 6.5 0.7 1.2 1.2 2.4 5.5 Argentina 0.0 0.3 0.4 0.0 0.0 0.0 0.2 0.0 Brazil 0.0 0.2 2.7 0.7 1.1 0.6 1.9 3.0 Chile 0.1 0.0 1.7 0.0 0.0 0.1 0.2 0.3 Mexico 0.0 0.2 1.6 0.0 0.0 0.5 0.1 2.0 Middle East and North Africa 0.4 0.3 0.4 0.0 0.0 0.0 0.2 1.2 Egypt, Arab Rep. of 0.1 0.3 0.3 0.0 0.0 0.0 0.1 0.7 South Asia 0.1 0.9 1.8 0.5 0.3 1.3 4.6 8.3 India 0.1 0.9 1.8 0.5 0.3 1.3 4.6 8.3 Sub-Saharan Africa 0.7 1.2 2.0 0.7 0.5 1.2 2.0 1.0 South Africa 0.7 1.2 2.0 0.7 0.5 1.2 1.9 1.0 Note: e = estimate. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 189 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.18 Gross international bond issuance in developing countries, 1998–2005 $ billions 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 57.2 61.5 60.5 60.9 51.7 82.2 102.4 130.9 East Asia and Pacific 4.2 8.2 5.1 7.1 12.5 11.6 15.7 20.3 China 1.9 1.4 1.3 2.6 0.9 3.3 6.2 4.9 Indonesia 0.0 0.0 0.0 0.1 0.8 1.5 1.4 5.0 Malaysia 0.1 2.6 1.4 2.4 6.0 1.4 3.5 3.2 Philippines 1.9 4.2 2.4 1.8 4.8 5.1 3.3 4.1 Thailand 0.3 0.0 0.0 0.3 0.0 0.3 1.4 2.2 Europe and Central Asia 15.3 12.9 14.4 10.3 13.8 26.5 38.2 54.7 Czech Rep. 0.6 0.4 0.0 0.1 0.4 0.3 2.6 1.7 Hungary 1.6 2.3 0.5 1.2 0.0 2.4 5.5 7.3 Poland 1.1 1.6 1.4 2.5 2.7 4.7 3.9 11.8 Russian Fed. 5.8 0.0 0.1 1.5 3.7 8.6 10.5 16.3 Turkey 3.4 5.7 8.6 2.2 3.5 5.5 6.4 9.8 Latin America and the Caribbean 34.7 36.8 36.3 36.7 21.1 38.8 35.9 43.0 Argentina 13.7 13.3 11.9 1.7 0.0 0.0 0.0 0.5 Brazil 6.4 7.8 11.4 12.8 7.4 13.6 11.6 14.6 Chile 0.5 2.4 0.5 3.0 1.7 2.8 1.3 2.7 Mexico 7.9 8.8 8.5 8.2 7.1 14.1 12.8 10.0 Venezuela, R. B. de 3.3 1.4 0.5 1.7 0.0 3.7 4.0 5.9 Middle East and North Africa 1.5 1.9 2.4 5.3 2.7 1.0 5.6 5.4 Egypt, Arab Rep. of 0.0 0.1 0.0 1.5 0.0 0.0 0.0 2.8 Lebanon 1.5 1.4 1.9 3.3 1.0 0.2 4.4 1.8 South Asia 0.4 0.0 0.6 0.1 0.2 0.5 5.1 5.3 India 0.4 0.0 0.6 0.1 0.2 0.5 4.5 4.7 Pakistan 0.0 0.0 0.0 0.0 0.0 0.0 0.5 0.6 Sub-Saharan Africa 1.0 1.8 1.7 1.5 1.5 3.9 2.0 2.3 South Africa 1.0 1.8 1.7 1.5 1.5 3.9 2.0 2.3 Note: e = estimate. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 190 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.19 Gross international bank lending to developing countries, 1998–2005 $ billions 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 95.2 78.1 105.9 69.3 69.1 86.7 111.8 198.1 East Asia and Pacific 17.0 12.9 21.9 9.0 21.5 26.9 19.5 34.5 China 5.8 3.3 6.7 2.0 9.6 12.6 8.0 18.5 Indonesia 0.7 1.6 1.0 0.5 0.5 4.2 2.1 1.4 Malaysia 3.2 4.7 5.2 2.7 5.6 5.3 6.2 3.5 Philippines 3.4 2.6 4.8 1.8 1.7 1.9 2.0 3.2 Thailand 3.8 0.7 4.1 2.0 3.5 2.5 1.1 5.9 Europe and Central Asia 17.6 15.4 22.7 14.4 16.8 22.2 37.8 77.6 Czech Rep. 2.3 0.6 1.2 0.8 0.3 2.2 1.8 1.4 Hungary 1.7 0.9 1.5 1.5 1.8 3.0 2.3 3.0 Poland 1.6 2.6 2.1 2.2 3.5 2.6 2.1 6.4 Russian Fed. 2.6 0.6 4.7 2.9 4.9 7.2 13.6 40.1 Turkey 5.7 7.1 11.3 4.6 3.7 2.0 7.6 12.7 Latin America and the Caribbean 46.9 35.0 44.8 34.6 18.5 20.6 29.9 46.3 Argentina 11.7 6.5 6.6 4.5 1.9 0.7 0.7 2.0 Brazil 11.4 7.2 13.8 10.2 4.7 3.7 9.8 13.0 Chile 3.7 7.2 5.5 3.1 1.8 2.2 6.6 7.3 Mexico 11.4 8.4 10.9 10.2 6.1 12.1 9.6 18.2 Venezuela, R. B. de 4.3 0.4 2.4 3.0 0.6 0.4 0.7 0.5 Middle East and North Africa 2.9 6.5 3.6 3.2 5.8 4.6 9.7 15.7 Egypt, Arab Rep. of 1.6 4.2 0.8 0.9 1.0 0.4 1.6 3.2 South Asia 4.8 3.2 3.4 2.4 1.7 4.0 7.0 12.2 India 3.8 2.8 3.0 2.0 1.2 2.3 6.0 11.0 Pakistan 0.9 0.0 0.0 0.2 0.4 1.5 0.8 0.7 Sub-Saharan Africa 6.0 5.1 9.4 5.6 4.9 8.5 7.9 11.9 South Africa 3.2 3.1 6.8 2.7 1.7 3.2 2.4 5.1 Note: e = estimate. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 191 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.20 Change in foreign exchange reserves, 1998–2005 $ billions (– = increase) 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries –16.4 –33.2 –45.4 –81.7 –171.9 –291.6 –404.8 –393.0 East Asia and Pacific –20.7 –29.3 –10.1 –47.7 –88.0 –136.7 –237.0 –217.9 China –5.1 –9.7 –10.9 –46.6 –74.2 –116.8 –206.7 –208.9 Indonesia –6.3 –3.8 –2.0 1.2 –3.7 –4.0 0.0 2.0 Malaysia –4.7 –4.9 1.0 –1.0 –3.7 –10.2 –21.9 –4.3 Philippines –2.0 –4.0 0.2 –0.4 0.2 –0.3 0.5 –2.8 Thailand –2.7 –5.4 1.9 –0.4 –5.7 –2.9 –7.5 –2.0 Europe and Central Asia –5.1 –6.4 –16.6 –11.1 –43.7 –60.9 –79.3 –94.7 Czech Rep. –2.8 –0.3 –0.2 –1.2 –9.1 –3.0 –1.6 –1.3 Hungary –0.9 –1.5 –0.2 0.6 0.6 –2.3 –3.3 –3.0 Poland –6.9 1.1 –0.2 1.2 –2.8 –3.8 –2.8 –5.9 Romania 0.8 1.4 –1.0 –1.4 –2.2 –1.9 –6.6 –5.3 Russian Fed. 5.0 –0.7 –15.8 –8.3 –11.5 –29.1 –47.6 –54.9 Turkey –0.8 –3.7 0.9 3.6 –8.2 –6.9 –1.7 –14.9 Latin America and the Caribbean 9.2 7.6 –3.0 –2.9 –0.8 –33.2 –24.9 –32.1 Argentina –2.3 –1.6 1.7 9.9 4.1 –2.7 –4.9 –4.7 Brazil 8.2 7.8 2.3 –3.2 –1.7 –11.7 –3.6 –0.8 Chile 2.0 1.1 –0.5 0.6 –0.8 –0.4 –0.3 –1.2 Mexico –3.3 0.5 –4.2 –9.2 –5.5 –7.8 –5.0 –10.2 Venezuela, R. B. de 2.4 –0.1 –0.9 3.8 0.8 –7.5 –2.3 –5.6 Middle East and North Africa 1.7 1.2 –4.8 –9.5 –12.0 –22.0 –14.3 –21.3 Algeria 1.2 2.4 –7.5 –6.1 –5.1 –9.8 –10.2 –13.1 Egypt, Arab Rep. of 0.6 3.6 1.4 0.0 –0.3 –0.2 –0.7 –6.4 Lebanon –0.6 –1.2 1.8 0.9 –2.2 –5.3 0.8 –0.2 South Asia –3.0 –5.0 –4.7 –10.2 –27.0 –35.0 –27.2 –6.3 India –2.6 –5.0 –5.3 –8.0 –21.7 –30.6 –27.5 –5.9 Pakistan 0.2 –0.5 0.0 –2.1 –4.4 –2.6 1.1 –0.3 Sub-Saharan Africa 1.5 –1.3 –6.2 –0.3 –0.3 –3.9 –22.2 –20.7 Angola 0.2 –0.3 –0.7 0.5 0.4 –0.3 –0.7 –1.8 Nigeria 0.5 1.7 –4.5 –0.5 3.1 0.2 –9.8 –11.3 South Africa 0.6 –1.9 0.3 0.0 0.2 –0.6 –6.6 –5.5 Note: e = estimate. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 192 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.21 Total external debt of developing countries, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 2,107.0 2,321.8 2,345.7 2,283.9 2,280.7 2,359.0 2,581.8 2,755.7 2,800.4 East Asia and Pacific 526.3 533.3 538.8 498.0 516.9 515.9 541.5 588.9 633.7 China 146.7 144.0 152.1 145.7 184.8 186.4 208.7 248.9 — Indonesia 136.2 151.2 151.2 144.4 134.1 132.2 136.9 140.6 — Malaysia 47.2 42.4 41.9 41.9 45.1 48.3 48.5 52.1 — Philippines 50.7 53.6 58.3 58.3 58.3 59.9 62.5 60.6 — Thailand 109.7 104.9 96.8 79.7 67.2 59.4 51.8 51.3 — Europe and Central Asia 390.4 490.3 502.9 511.0 508.1 561.4 680.5 794.9 870.1 Bulgaria 11.1 11.4 11.0 11.2 10.5 11.5 13.4 15.7 — Czech Rep. 23.2 24.3 22.8 21.6 22.8 27.7 34.8 45.6 — Hungary 24.6 28.5 29.9 29.5 30.3 35.0 47.4 63.2 — Poland 41.7 57.7 65.9 65.8 67.4 78.5 95.5 99.2 — Russian Fed. 127.6 177.8 174.8 160.0 152.5 147.4 175.5 197.3 — Turkey 84.8 97.1 102.2 117.3 113.4 131.2 145.4 161.6 — Latin America and the Caribbean 669.5 752.1 772.3 758.7 753.0 750.0 785.9 779.0 723.7 Argentina 128.2 141.4 145.7 147.4 154.1 149.9 166.1 169.2 — Brazil 198.0 241.7 245.2 243.4 231.1 233.1 236.6 222.0 — Chile 27.0 33.7 34.8 37.3 38.6 41.2 43.3 44.1 — Colombia 31.9 33.1 34.4 33.9 36.2 33.2 37.0 37.7 — Mexico 147.6 159.0 166.5 150.3 145.7 140.2 141.6 138.7 — Venezuela, R. B. de 35.7 37.8 37.6 38.2 36.0 34.0 34.8 35.6 — Middle East and North Africa 151.2 160.9 155.7 145.2 143.0 151.4 161.2 163.9 162.5 Algeria 30.9 30.7 28.0 25.3 22.6 22.9 23.6 22.0 — Egypt, Arab Rep. of 30.1 32.4 31.0 29.2 29.3 30.0 31.4 30.3 — Lebanon 5.0 6.8 8.2 9.9 12.4 17.1 18.6 22.2 — South Asia 149.6 157.6 162.0 160.0 156.2 168.7 181.5 193.9 194.8 India 94.3 97.6 98.3 99.1 97.5 104.8 112.6 122.7 — Pakistan 30.1 32.3 33.9 32.8 31.7 33.7 35.9 35.7 — Sub-Saharan Africa 219.9 227.7 214.0 211.0 203.6 211.7 231.2 235.1 215.6 South Africa 25.3 24.8 23.9 24.9 24.1 25.0 27.8 28.5 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 193 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.22 Total external debt of developing countries: medium and long-term, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 1,719.8 1,971.4 2,012.7 1,967.5 1,943.7 2,021.4 2,161.0 2,261.0 2,243.7 East Asia and Pacific 394.2 448.7 465.8 435.4 410.9 398.8 400.8 414.0 426.6 China 115.2 126.7 136.9 132.6 128.5 120.5 120.4 131.3 — Indonesia 103.3 131.1 131.2 121.8 112.3 109.4 114.0 116.1 — Malaysia 32.3 33.9 35.9 37.3 38.3 39.9 39.9 40.7 — Philippines 38.9 47.8 53.4 52.8 52.3 54.4 56.3 55.5 — Thailand 71.9 75.3 73.4 64.8 54.0 47.5 40.8 39.8 — Europe and Central Asia 331.2 415.2 423.5 424.4 425.5 474.2 547.6 634.4 675.9 Bulgaria 9.1 9.6 9.7 9.8 9.3 9.6 10.8 12.4 — Czech Rep. 15.1 16.6 14.0 12.6 13.2 16.9 20.8 28.5 — Hungary 21.2 23.7 26.3 25.4 25.7 29.3 38.4 50.8 — Poland 36.6 49.3 54.6 56.2 56.3 64.6 76.0 82.3 — Russian Fed. 121.7 163.1 159.0 144.4 133.5 131.1 145.1 162.2 — Turkey 66.8 75.9 78.8 88.4 97.0 114.8 122.4 129.7 — Latin America and the Caribbean 541.2 633.0 662.9 651.0 659.2 670.8 697.3 682.6 632.7 Argentina 96.2 110.5 116.2 119.1 134.0 134.8 143.1 141.8 — Brazil 163.2 211.8 216.0 212.5 202.8 209.7 212.0 196.8 — Chile 21.5 28.6 30.5 31.1 33.3 35.4 35.8 36.4 — Colombia 26.2 26.9 30.5 31.1 33.0 29.5 33.4 32.4 — Mexico 119.8 132.7 142.4 131.4 131.1 130.3 132.5 129.6 — Venezuela, R. B. de 31.5 35.5 35.5 34.1 31.2 29.4 30.5 31.2 — Middle East and North Africa 132.7 138.8 132.5 123.7 123.8 131.8 139.7 142.4 140.1 Algeria 30.7 30.5 27.8 25.0 22.4 22.8 23.4 21.6 — Egypt, Arab Rep. of 27.1 28.2 26.8 25.1 26.0 26.5 27.6 27.4 — Lebanon 3.2 4.8 6.0 7.3 9.8 14.5 15.5 18.2 — South Asia 141.4 150.5 155.0 154.0 151.3 161.9 174.1 183.6 181.9 India 89.3 93.3 94.4 95.6 94.8 100.7 107.6 115.2 — Pakistan 27.6 30.1 32.1 31.3 30.4 32.1 34.6 34.4 — Sub-Saharan Africa 179.1 185.2 173.1 179.1 173.0 183.9 201.5 203.8 186.6 South Africa 14.3 13.3 13.1 15.3 15.7 17.6 20.4 20.6 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 194 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.23 Total external debt of developing countries: short-term, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 387.2 350.4 333.0 316.4 337.0 337.6 420.8 494.8 556.7 East Asia and Pacific 132.1 84.6 73.0 62.6 105.9 117.0 140.6 174.8 207.2 China 31.5 17.3 15.2 13.1 56.3 65.9 88.3 117.6 — Indonesia 32.9 20.1 20.0 22.6 21.8 22.8 22.9 24.5 — Malaysia 14.9 8.5 6.0 4.6 6.8 8.4 8.6 11.4 — Philippines 11.8 5.9 4.9 5.5 6.0 5.6 6.2 5.0 — Thailand 37.8 29.7 23.4 14.9 13.2 11.9 11.0 11.5 — Europe and Central Asia 59.1 75.1 79.4 86.7 82.6 87.1 133.0 160.5 194.2 Bulgaria 2.0 1.8 1.3 1.5 1.2 1.8 2.7 3.2 — Czech Rep. 8.1 7.6 8.8 9.0 9.6 10.8 14.0 17.1 — Hungary 3.4 4.8 3.5 4.2 4.6 5.7 9.0 12.3 — Poland 5.1 8.4 11.3 9.7 11.1 13.9 19.5 16.8 — Russian Fed. 5.9 14.7 15.7 15.6 19.0 16.3 30.5 35.1 — Turkey 18.0 21.2 23.5 28.9 16.3 16.4 23.0 31.9 — Latin America and the Caribbean 128.3 119.1 109.5 107.7 93.8 79.2 88.6 96.3 91.0 Argentina 32.0 31.0 29.4 28.3 20.0 15.1 23.0 27.5 — Brazil 34.9 29.9 29.2 31.0 28.3 23.4 24.6 25.3 — Chile 5.5 5.1 4.3 6.2 5.3 5.8 7.5 7.7 — Colombia 5.8 6.2 4.0 2.9 3.3 3.7 3.6 5.3 — Mexico 27.9 26.3 24.1 18.9 14.6 9.9 9.2 9.1 — Venezuela, R. B. de 4.2 2.2 2.1 4.1 4.8 4.6 4.3 4.4 — Middle East and North Africa 18.6 22.1 23.2 21.5 19.2 19.6 21.4 21.5 22.4 Algeria 0.2 0.2 0.2 0.2 0.2 0.1 0.1 0.4 — Egypt, Arab Rep. of 3.0 4.3 4.3 4.1 3.4 3.5 3.8 2.9 — Lebanon 1.8 2.0 2.2 2.5 2.7 2.5 3.1 4.0 — South Asia 8.2 7.1 7.0 6.1 5.0 6.8 7.4 10.3 12.8 India 5.0 4.3 3.9 3.5 2.7 4.1 5.0 7.5 — Pakistan 2.5 2.2 1.8 1.5 1.3 1.5 1.2 1.2 — Sub-Saharan Africa 40.8 42.4 40.9 31.9 30.5 27.8 29.8 31.2 29.0 South Africa 10.9 11.4 10.8 9.6 8.4 7.4 7.4 7.9 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 195 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.24 Total external debt of developing countries: owed by public and publicly guaranteed borrowers, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 1,366.8 1,470.6 1,476.4 1,422.3 1,400.9 1,470.1 1,557.9 1,589.5 1,482.9 East Asia and Pacific 271.9 288.8 307.4 288.0 277.5 277.8 278.6 280.6 279.9 China 112.8 99.4 99.2 94.9 91.8 88.6 85.3 90.8 — Indonesia 58.8 76.4 83.9 80.6 77.9 79.4 84.3 82.6 — Malaysia 16.8 18.2 18.9 19.2 24.2 26.4 25.4 25.6 — Philippines 27.2 30.7 36.6 35.8 31.2 34.0 37.2 36.3 — Thailand 24.7 31.3 34.7 32.5 27.9 22.9 17.7 15.3 — Europe and Central Asia 288.6 320.8 315.7 304.5 292.1 309.5 336.3 352.5 327.8 Bulgaria 8.7 9.1 9.0 9.0 8.5 8.5 8.9 8.6 — Czech Rep. 12.9 11.6 7.7 6.6 5.7 7.1 8.7 12.0 — Hungary 15.3 15.9 16.9 14.4 12.7 13.6 16.3 20.7 — Poland 34.2 35.1 33.2 30.8 25.7 29.4 35.0 36.6 — Russian Fed. 119.8 140.9 136.4 122.6 111.2 102.6 103.9 103.2 — Turkey 48.1 50.6 51.6 60.6 68.4 82.3 88.6 89.7 — Latin America and the Caribbean 378.6 412.8 420.5 409.5 420.7 445.0 474.1 476.6 426.4 Argentina 72.8 82.6 88.9 93.2 102.4 106.3 114.7 117.9 — Brazil 87.4 103.6 102.4 99.8 106.5 122.3 129.9 122.9 — Chile 4.4 5.0 5.7 5.3 5.6 6.8 8.0 9.4 — Colombia 15.4 16.7 20.2 20.8 21.8 20.7 22.8 23.4 — Mexico 92.4 95.4 92.4 81.5 77.0 76.3 77.5 77.2 — Venezuela, R. B. de 29.0 29.6 28.7 28.0 25.2 23.4 24.5 25.9 — Middle East and North Africa 127.7 131.8 125.5 117.2 117.2 125.3 132.8 135.2 129.8 Algeria 30.7 30.5 27.8 25.0 22.4 22.7 22.9 20.9 — Egypt, Arab Rep. of 27.0 27.8 26.3 24.5 25.3 25.9 27.3 27.4 — Lebanon 2.3 4.0 5.3 6.6 9.0 13.8 14.8 17.5 — South Asia 129.7 139.3 144.6 135.4 132.8 141.1 150.2 155.3 147.7 India 80.1 84.9 86.4 80.1 78.8 82.3 85.6 88.7 — Pakistan 25.3 27.5 29.8 28.7 28.3 30.1 33.0 32.9 — Sub-Saharan Africa 170.3 177.0 162.7 167.7 160.6 171.4 186.0 189.2 171.3 South Africa 11.9 10.7 8.2 9.1 7.9 9.4 9.1 9.8 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 196 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.25 Total external debt of developing countries: owed by private sector borrowers, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 740.2 851.3 869.3 861.6 879.7 889.0 1,023.9 1,166.2 1,317.5 East Asia and Pacific 254.4 244.5 231.4 210.1 239.4 238.1 262.9 308.3 353.8 China 33.9 44.6 52.9 50.9 93.1 97.8 123.3 158.1 — Indonesia 77.3 74.8 67.3 63.8 56.2 52.8 52.7 58.0 — Malaysia 30.4 24.3 23.0 22.6 20.9 21.9 23.2 26.6 — Philippines 23.5 22.9 21.7 22.5 27.1 25.9 25.2 24.2 — Thailand 85.0 73.6 62.0 47.2 39.3 36.5 34.1 36.0 — Europe and Central Asia 101.8 169.5 187.1 206.5 216.0 251.9 344.3 442.5 542.3 Bulgaria 2.4 2.3 2.0 2.2 2.0 2.9 4.5 7.0 — Czech Rep. 10.2 12.7 15.1 15.0 17.1 20.6 26.1 33.5 — Hungary 9.3 12.6 13.0 15.2 17.6 21.4 31.0 42.4 — Poland 7.5 22.6 32.8 35.1 41.7 49.1 60.5 62.6 — Russian Fed. 7.8 36.9 38.3 37.4 41.3 44.8 71.6 94.1 — Turkey 36.7 46.6 50.6 56.7 45.0 48.9 56.8 71.9 — Latin America and the Caribbean 290.9 339.3 351.9 349.2 332.3 305.0 311.8 302.3 297.3 Argentina 55.4 58.8 56.7 54.2 51.6 43.6 51.4 51.3 — Brazil 110.7 138.0 142.8 143.7 124.6 110.8 106.7 99.1 — Chile 22.7 28.7 29.2 32.0 33.0 34.4 35.3 34.6 — Colombia 16.5 16.3 14.2 13.1 14.5 12.5 14.2 14.4 — Mexico 55.2 63.5 74.1 68.8 68.6 63.8 64.1 61.5 — Venezuela, R. B. de 6.7 8.2 8.9 10.2 10.8 10.6 10.4 9.7 — Middle East and North Africa 23.6 29.1 30.2 28.0 25.8 26.0 28.4 28.7 32.7 Algeria 0.2 0.2 0.2 0.2 0.2 0.2 0.7 1.1 — Egypt, Arab Rep. of 3.1 4.6 4.8 4.7 4.0 4.1 4.1 2.9 — Lebanon 2.7 2.7 2.9 3.3 3.5 3.2 3.8 4.7 — South Asia 19.9 18.3 17.4 24.6 23.4 27.6 31.3 38.6 47.0 India 14.3 12.7 11.9 19.0 18.7 22.6 27.1 34.0 — Pakistan 4.8 4.8 4.1 4.1 3.4 3.5 2.9 2.8 — Sub-Saharan Africa 49.7 50.7 51.3 43.3 43.0 40.3 45.2 45.8 44.3 South Africa 13.3 14.1 15.7 15.8 16.1 15.6 18.7 18.7 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 197 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.26 Total external debt of developing countries: owed to public sector creditors, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 790.1 865.6 880.0 839.5 827.1 875.7 926.3 924.1 820.8 East Asia and Pacific 152.5 179.1 200.2 188.1 180.4 183.7 190.1 193.6 190.1 China 39.8 45.1 50.4 50.4 50.6 50.8 51.4 57.2 — Indonesia 45.5 58.2 66.4 66.0 62.1 65.7 71.0 69.5 — Malaysia 4.0 4.5 4.8 5.0 5.9 5.8 6.2 7.0 — Philippines 19.6 22.1 23.6 21.9 19.7 20.9 22.2 21.6 — Thailand 17.8 21.4 25.3 23.9 20.8 16.6 13.1 10.7 — Europe and Central Asia 156.3 172.5 170.9 166.8 159.3 166.0 169.8 168.5 131.0 Bulgaria 3.4 3.9 3.9 3.9 3.4 3.5 4.1 4.5 — Czech Rep. 1.1 1.1 1.1 1.2 1.3 1.6 2.1 2.2 — Hungary 3.3 2.3 2.3 1.9 1.7 1.9 1.7 2.3 — Poland 26.6 27.1 25.1 23.7 17.8 19.7 20.4 18.6 — Russian Fed. 76.8 88.3 86.7 82.5 71.7 62.2 58.2 58.0 — Turkey 14.3 15.0 13.8 17.3 26.9 35.8 37.5 35.3 — Latin America and the Caribbean 145.3 160.6 162.8 149.7 162.3 182.4 196.0 187.4 148.2 Argentina 24.0 25.8 25.4 25.5 35.2 35.6 36.9 35.6 — Brazil 22.2 32.7 37.7 31.1 37.2 52.1 58.2 52.3 — Chile 2.2 2.2 2.1 1.9 1.7 1.5 1.4 1.3 — Colombia 5.6 6.0 7.8 7.7 8.6 8.9 11.3 11.5 — Mexico 32.1 31.4 26.3 20.8 19.9 20.5 20.6 19.6 — Venezuela, R. B. de 5.5 6.7 6.6 6.1 4.9 4.4 3.9 3.5 — Middle East and North Africa 99.5 103.8 98.2 90.7 88.2 91.5 96.4 95.6 90.2 Algeria 20.3 21.5 20.4 19.2 17.7 17.7 17.9 16.2 — Egypt, Arab Rep. of 25.9 26.9 25.7 24.0 23.4 24.7 26.4 26.4 — Lebanon 0.7 0.9 0.9 0.9 1.0 1.0 1.7 1.8 — South Asia 98.9 104.6 113.3 102.7 101.1 106.3 113.4 116.1 113.8 India 52.8 53.9 58.6 50.6 49.8 49.8 50.7 51.7 — Pakistan 22.8 25.1 27.7 26.7 27.0 29.3 32.3 32.0 — Sub-Saharan Africa 137.6 145.0 134.5 141.6 135.8 145.9 160.7 162.9 147.5 South Africa 0.4 0.0 0.0 0.1 0.1 0.1 0.2 0.3 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 198 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.27 Total external debt of developing countries: owed to private sector creditors, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 1,316.9 1,456.3 1,465.7 1,444.4 1,453.6 1,483.4 1,655.5 1,831.6 1,979.6 East Asia and Pacific 373.8 354.2 338.6 309.9 336.5 332.2 351.4 395.3 443.7 China 106.9 98.9 101.6 95.3 134.3 135.5 157.3 191.7 — Indonesia 90.7 93.1 84.9 78.5 72.0 66.5 65.9 71.1 — Malaysia 43.2 37.9 37.1 36.9 39.2 42.5 42.4 45.2 — Philippines 31.1 31.5 34.7 36.4 38.5 39.1 40.3 39.0 — Thailand 91.9 83.5 71.5 55.8 46.4 42.7 38.7 40.6 — Europe and Central Asia 234.1 317.8 331.9 344.2 348.8 395.4 510.7 626.4 739.1 Bulgaria 7.7 7.5 7.1 7.3 7.1 8.0 9.3 11.2 — Czech Rep. 22.1 23.1 21.7 20.4 21.5 26.1 32.7 43.4 — Hungary 21.3 26.2 27.6 27.6 28.6 33.1 45.7 60.9 — Poland 15.1 30.6 40.9 42.2 49.6 58.8 75.1 80.6 — Russian Fed. 50.8 89.5 88.1 77.6 80.8 85.3 117.3 139.3 — Turkey 70.5 82.2 88.4 100.0 86.5 95.4 107.9 126.2 — Latin America and the Caribbean 524.2 591.5 609.5 608.9 590.7 567.6 589.9 591.6 575.5 Argentina 104.2 115.6 120.3 121.9 118.9 114.3 129.2 133.7 — Brazil 175.8 209.0 207.5 212.3 193.9 181.0 178.4 169.7 — Chile 24.9 31.5 32.7 35.4 36.9 39.7 41.9 42.7 — Colombia 26.3 27.1 26.6 26.2 27.7 24.3 25.7 26.3 — Mexico 115.6 127.5 140.2 129.5 125.8 119.6 121.1 119.1 — Venezuela, R. B. de 30.2 31.0 31.0 32.0 31.1 29.6 30.9 32.1 — Middle East and North Africa 51.7 57.1 57.5 54.6 54.8 59.9 64.8 68.3 72.3 Algeria 10.6 9.2 7.6 6.1 4.9 5.2 5.7 5.7 — Egypt, Arab Rep. of 4.2 5.5 5.3 5.2 6.0 5.3 5.0 3.9 — Lebanon 4.3 5.9 7.3 8.9 11.5 16.1 16.9 20.4 — South Asia 50.7 53.0 48.7 57.3 55.1 62.5 68.1 77.8 80.9 India 41.5 43.7 39.7 48.5 47.7 55.0 61.9 71.1 — Pakistan 7.2 7.2 6.2 6.1 4.7 4.4 3.5 3.7 — Sub-Saharan Africa 82.4 82.7 79.5 69.4 67.8 65.8 70.6 72.2 68.1 South Africa 24.9 24.8 23.9 24.7 23.9 24.9 27.6 28.2 — Note: e = estimate; — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 199 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.28 Gross foreign exchange reserves of developing countries, 1997–2005 $ billions 1997 1998 1999 2000 2001 2002 2003 2004 2005e All developing countries 571.6 588.0 621.2 666.6 748.3 920.1 1,211.8 1,616.6 2,009.5 East Asia and Pacific 212.5 233.2 262.5 272.7 320.4 408.4 545.0 782.0 999.9 China 139.9 145.0 154.7 165.6 212.2 286.4 403.3 609.9 818.9 Indonesia 16.1 22.4 26.2 28.3 27.0 30.8 34.7 34.7 32.8 Malaysia 20.0 24.7 29.7 28.6 29.6 33.3 43.5 65.4 69.7 Philippines 7.2 9.1 13.1 13.0 13.4 13.2 13.5 13.0 15.8 Thailand 25.7 28.4 33.8 31.9 32.3 38.0 41.0 48.5 50.5 Europe and Central Asia 90.8 95.9 102.3 118.9 130.0 173.8 234.6 313.9 408.7 Czech Rep. 9.7 12.5 12.8 13.0 14.2 23.3 26.3 27.8 29.1 Hungary 8.3 9.2 10.7 10.9 10.3 9.7 12.0 15.3 18.3 Poland 20.3 27.2 26.1 26.3 25.2 28.0 31.7 34.6 40.5 Romania 3.7 2.9 1.5 2.5 3.9 6.1 8.0 14.6 19.9 Russian Fed. 12.8 7.8 8.5 24.3 32.5 44.1 73.2 120.8 175.7 Turkey 18.6 19.4 23.2 22.3 18.7 26.9 33.8 35.5 50.4 Latin America and the Caribbean 166.5 157.3 149.7 152.7 155.5 156.4 189.5 214.4 246.5 Argentina 22.2 24.5 26.1 24.4 14.5 10.4 13.1 18.0 22.7 Brazil 50.8 42.6 34.8 32.5 35.7 37.4 49.1 52.7 53.5 Chile 17.3 15.3 14.2 14.7 14.0 14.8 15.2 15.5 16.7 Mexico 28.1 31.5 31.0 35.1 44.4 49.9 57.7 62.8 73.0 Venezuela, R. B. de 14.0 11.6 11.7 12.6 8.8 8.0 15.5 17.9 23.5 Middle East and North Africa 43.7 42.0 40.8 45.6 55.1 67.1 89.1 103.3 124.6 Algeria 8.0 6.8 4.4 11.9 18.0 23.1 32.9 43.1 56.2 Egypt, Arab Rep. of 18.5 17.9 14.3 12.9 12.9 13.2 13.4 14.1 20.5 Lebanon 5.9 6.5 7.7 5.9 5.0 7.2 12.5 11.7 11.8 South Asia 30.0 32.9 37.9 42.6 52.8 79.8 114.8 142.0 148.3 India 24.3 27.0 32.0 37.3 45.3 67.0 97.6 125.2 131.0 Pakistan 1.2 1.0 1.5 1.5 3.6 8.1 10.7 9.6 9.8 Sub-Saharan Africa 28.1 26.6 27.9 34.1 34.4 34.7 38.6 60.9 81.6 Angola 0.4 0.2 0.5 1.2 0.7 0.4 0.6 1.4 3.2 Nigeria 7.6 7.1 5.5 9.9 10.5 7.3 7.1 17.0 28.3 South Africa 4.8 4.2 6.1 5.8 5.8 5.6 6.2 12.8 18.3 Note: e = estimate. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 200 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.29 Key external debt ratios for developing countries %, averages for 2002–4 Total external debt Present Value (PV) EDT as a % of (EDT) to exports of EDT as a gross national Total Debt service Interest service Country of G&S (XGS) % of XGS income (GNI) PV as % of GNI as a % of XGS as % of XGS Albania 74 51 25 17 4 1 Algeria 82 80 33 32 22 4 Angola 89 82 74 68 19 2 Argentina 451 510 141 159 33 6 Armenia 113 130 43 50 10 1 Azerbaijan 56 45 29 23 7 1 Bangladesh 179 124 37 26 6 2 Barbados 44 48 27 29 6 3 Belarus 30 30 20 20 3 1 Belize 188 208 99 109 65 15 Benin 268 113 56 24 9 3 Bhutan 431 432 99 100 9 4 Bolivia 275 136 77 38 23 6 Bosnia and Herzegovina 80 63 44 34 4 1 Botswana 14 12 8 6 1 0 Brazil 239 258 44 47 58 16 Bulgaria 139 143 81 83 22 5 Burkina Faso 432 203 48 23 13 4 Burundi 3,069 203 227 15 195 37 Cambodia 117 99 80 68 1 0 Cameroon 296 72 81 20 20 8 Cape Verde 144 100 67 46 7 2 Central African Republic 730 599 91 75 12 2 Chad 172 79 73 33 5 1 Chile 145 141 58 57 32 5 China 48 46 15 15 5 1 Colombia 189 204 45 49 38 12 Comoros 389 276 99 70 4 1 Congo, Dem. Rep. of 765 131 208 36 8 4 Congo, Rep. of 230 356 214 331 14 5 Costa Rica 66 70 34 36 8 2 Côte d’Ivoire 172 170 91 90 8 1 Croatia 200 194 113 110 33 8 Czech Republic 73 71 53 51 13 2 Djibouti — — 65 45 — — Dominica 176 159 93 84 14 6 Dominican Republic 62 61 39 39 7 3 Ecuador 191 205 65 70 42 14 Egypt, Arab Rep. of 123 108 36 32 9 3 El Salvador 114 123 50 54 10 5 Equatorial Guinea 10 9 — — — — Eritrea 260 154 90 53 7 3 Estonia 138 132 116 111 20 5 Ethiopia 460 144 97 30 7 3 Fiji 16 16 9 9 1 0 Gabon 124 117 80 75 7 1 Gambia, The 398 231 186 108 20 5 Georgia 133 100 49 37 14 2 Ghana 222 76 95 32 8 3 Grenada 232 213 115 106 16 10 Guatemala 85 88 22 23 8 4 Guinea 416 186 100 45 20 5 Guinea-Bissau 791 779 331 326 46 10 Guyana 171 65 189 72 6 2 Haiti 98 76 37 29 11 4 Honduras 171 68 95 38 9 3 Hungary 115 108 80 76 31 3 India 106 95 Delivered by The World 21 18 Bank e-library to: 16 3 Indonesia 181 63 175 The World Bank 61 26 6 IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (continued) 201 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.29 (continued) %, averages for 2002–4 Total external debt Present Value (PV) EDT as a % of (EDT) to exports of EDT as a gross national Total Debt service Interest service Country of G&S (XGS) % of XGS income (GNI) PV as % of GNI as a % of XGS as % of XGS Iran, Islamic Rep. of 33 31 10 9 5 1 Jamaica 126 141 79 89 16 7 Jordan 108 101 77 72 9 2 Kazakhstan 193 182 107 101 52 5 Kenya 185 136 47 34 10 2 Kyrgyz Republic 240 173 114 82 18 3 Lao PDR 365 276 101 76 9 2 Latvia 243 239 112 110 26 6 Lebanon 470 488 116 121 92 30 Lesotho 88 64 60 44 6 2 Liberia 1,891 2,133 674 760 0 0 Lithuania 95 96 53 53 18 3 Macedonia, FYR 107 94 45 39 13 3 Madagascar 330 170 74 38 8 3 Malawi 584 186 188 60 10 4 Malaysia 42 42 52 53 7 2 Maldives 58 46 52 42 5 1 Mali 251 98 83 33 8 2 Mauritania 526 186 161 57 13 5 Mauritius 70 69 44 43 8 3 Mexico 69 77 22 24 25 5 Moldova 117 108 81 75 15 4 Mongolia 1/ 143 108 114 86 4 2 Morocco 96 91 41 39 16 3 Mozambique 310 54 98 17 6 2 Myanmar 233 176 — — 4 1 Nepal 180 119 56 37 6 2 Nicaragua 283 78 127 35 7 3 Niger 452 156 74 25 12 3 Nigeria 141 140 72 71 9 4 Oman 30 29 18 18 8 2 Pakistan 194 156 44 35 23 4 Panama 106 129 77 94 16 7 Papua New Guinea 87 80 72 66 19 3 Paraguay 109 104 54 52 16 5 Peru 245 265 52 57 21 11 Philippines 120 124 71 73 23 7 Poland 126 121 47 45 44 4 Romania 138 136 52 51 22 5 Russian Federation 117 120 45 46 13 5 Rwanda 964 150 96 15 14 6 Samoa 451 400 177 158 17 13 São Tomé and Principe 1,655 459 666 185 44 14 Senegal 165 61 60 22 14 3 Serbia and Montenegro 216 209 80 77 13 5 Seychelles 101 104 94 96 9 2 Sierra Leone 903 188 177 37 14 7 Slovak Republic 87 86 68 67 20 4 Solomon Islands 169 129 76 58 16 6 Somalia — — — — — — South Africa 58 54 18 17 8 3 Sri Lanka 134 111 61 50 9 3 St. Kitts and Nevis 217 212 96 94 32 15 St. Lucia 109 105 64 62 7 4 St. Vincent and the Grenadines 143 129 71 64 12 4 Sudan 478 625 116 151 8 2 Swaziland 24 25 26 27 2 1 Syrian Arab Rep. 250 249 102 101 4 2 Delivered by The World Bank e-library to: Tajikistan 78 55 Bank The World 58 41 9 2 IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (continued) 202 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.29 (continued) %, averages for 2002–4 Total external debt Present Value (PV) EDT as a % of (EDT) to exports of EDT as a gross national Total Debt service Interest service Country of G&S (XGS) % of XGS income (GNI) PV as % of GNI as a % of XGS as % of XGS Tanzania 398 115 76 22 6 3 Thailand 52 50 36 35 12 2 Togo 242 191 106 83 3 0 Tonga 125 90 46 33 4 1 Trinidad and Tobago 49 53 29 31 7 3 Tunisia 148 147 79 79 16 6 Turkey 213 221 67 69 45 10 Uganda 379 162 78 33 8 3 Ukraine 70 71 42 42 14 2 Uruguay 338 351 104 108 42 16 Uzbekistan 130 123 48 45 22 4 Vanuatu 81 64 44 35 2 1 Venezuela, R. B. de 106 125 38 45 20 7 Vietnam 75 65 45 39 3 1 Yemen, Rep. of 95 66 53 37 4 1 Zambia 530 112 170 36 31 6 Zimbabwe 264 264 33 33 5 1 Note: For definition of indicators, see Sources and Definitions section. Numbers in italics include the effects of traditional relief and HIPC re- lief and are based on publicly guaranteed debt only. Under the Multilateral Debt Relief Initiative (MDRI), IDA, IMF and the African Develop- ment Fund are currently finalizing arrangements to provide debt stock cancellation to post-completion point HIPCs on debt owed to the three institutions. The IMF and ADF are providing 100 percent stock cancellation on debts outstanding as of year-end 2004, while IDA will provide 100 percent stock cancellation on debts owed as of year-end 2003. The present value of debt for HIPCs provided in the GDF does not incor- porate debt reduction under the MDRI and may include penalty charges. Exports comprise the total value of goods and services exported, re- ceipts of compensations of employees and investment income and worker’s remittances. In the ratios, the numerator refers to the 2004 data and the denominator is an average of 2002 to 2004 data. For exports and GNI averages, staff estimates are used when necessary. — = not available. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 203 (c) The International Bank for Reconstruction and Development / The World Bank G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 6 Table A.30 Classification of countries by region and level of income Sub-Saharan Africa Asia Europe and Central Asia Middle East and North Africa East and Eastern Income Southern West East Asia South Europe and Rest of Middle North group Subgroup Africa Africa and Pacific Asia Central Asia Europe East Africa Americas Low- Burundi Benin Cambodia Afghanistan Kyrgyz Yemen, Haiti income Comoros Burkina Faso Korea, Dem. Bangladesh Republic Rep. of Nicaragua countries Congo, Dem. Cameroon People’s Bhutan Moldova Rep. of Central African Rep. of India Tajikistan Eritrea Republic Lao PDR Nepal Uzbekistan Ethiopia Chad Mongolia Pakistan Kenya Congo, Rep. of Myanmar Lesotho Côte d’Ivoire Papua New Madagascar Gambia, The Guinea Malawi Ghana Solomon Mozambique Guinea Islands Rwanda Guinea-Bissau Timor-Leste Somalia Liberia Vietnam Sudan Mali Tanzania Mauritania Uganda Niger Zambia Nigeria Zimbabwe São Tomé and Principe Senegal Sierra Leone Togo Middle- Lower Angola Cape Verde China Maldives Albania Iran, Islamic Algeria Bolivia income Namibia Fiji Sri Lanka Armenia Rep. of Djibouti Brazil countries Swaziland Indonesia Azerbaijan Iraq Egypt, Arab Colombia Kiribati Belarus Jordan Rep. of Cuba Marshall Bosnia and Syrian Arab Morocco Dominican Islands Herzegovina Rep. Tunisia Republic Micronesia, Bulgaria West Bank Ecuador Fed. Sts. of Georgia and Gaza El Salvador Philippines Kazakhstan Guatemala Samoa Macedonia, Guyana Thailand FYRa Honduras Tonga Romania Jamaica Vanuatu Serbia and Paraguay Montenegro Peru Turkmenistan Suriname Ukraine Upper Botswana Equatorial American Croatia Turkey Lebanon Libya Antigua and Mauritius Guinea Samoa Czech Oman Barbuda Mayotte Gabon Malaysia Republic Argentina Seychelles N. Mariana Estonia Barbados South Africa Islands Hungary Belize Palau Latvia Chile Lithuania Costa Rica Poland Dominica Russian Grenada Federation Mexico Slovak Panama Republic St. Kitts and Nevis St. Lucia St. Vincent Trinidad and Tobago Uruguay Delivered by The World Bank e-library to: Venezuela, The World Bank R. B. de IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (continued) 204 (c) The International Bank for Reconstruction and Development / The World Bank S T A T I S T I C A L A P P E N D I X Table A.30 (continued) Sub-Saharan Africa Asia Europe and Central Asia Middle East and North Africa East and Eastern Income Southern West East Asia South Europe and Rest of Middle North group Subgroup Africa Africa and Pacific Asia Central Asia Europe East Africa Americas High- OECD Australia Austria Canada income Japan Belgium United States countries Korea, Rep. Denmark New Zealand Finland Franceb Germany Greece Iceland Ireland Italy Luxembourg Netherlands Norway Portugal Spain Sweden Switzerland United Kingdom Non- Brunei Slovenia Andorra Bahrain Malta Aruba OECD French Channel Israel Bahamas, The Polynesia Islands Kuwait Bermuda Guam Cyprus Qatar Cayman Islands Hong Kong, Faeroe Saudi Netherlands Chinac Islands Arabia Antilles Macao, Greenland United Arab Puerto Rico Chinad Isle of Man Emirates Virgin New Liechtenstein Islands (U.S.) Caledonia Monaco Singapore San Marino Taiwan, China Total 209 25 23 36 8 27 28 14 7 41 Source: World Bank data. Note: For operational and analytical purposes, the World Bank’s main criterion for classifying economies is gross national income (GNI) per capita. Every economy is classified as low income, middle income (subdivided into lower middle and upper middle), or high income. Other analytical groups, based on geographic regions and levels of external debt, are also used. Low-income and middle-income economies are sometimes referred to as developing economies. The use of the term is convenient; it is not intended to imply that all economies in the group are experiencing similar development or that other economies have reached a preferred or final stage of devel- opment. Classification by income does not necessarily reflect development status. This table classifies all World Bank member economies, and all other economies with populations of more than 30,000. Economies are divided among income groups according to 2004 GNI per capita, calculated using the World Bank Atlas method. The groups are: low income, $825 or less; lower middle income, $826–3,255; upper middle income, $3,256–10,065; and high income, $10,066 or more. a. Former Yugoslav Republic of Macedonia. b. The French overseas departments French Guiana, Guadeloupe, Martinique, and Réunion are included in France. c. On 1 July 1997 China resumed its exercise of sovereignty over Hong Kong. d. On 20 December 1999 China resumed its exercise of sovereignty over Macao. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 205 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 (c) The International Bank for Reconstruction and Development / The World Bank T he year 2005 was a landmark in global development finance. Private capital flows to developing countries reached a record net level of $483 billion, much of them going to middle- prospects for financial flows to developing countries. It discusses the structural transformation and mainstreaming of emerging debt markets, the increased importance of the euro, the fast- income countries that took advantage of the surge to improve growing phenomenon of credit default swap transactions on their external debt profile and build official reserves of foreign emerging sovereign and corporate names, and the greater reliance exchange. For the poorer countries, which have limited access in developing countries on local-currency financing. It also to market-based external finance, the development community highlights the difficult challenge that many emerging market stepped up efforts to enhance aid flows and to reduce debt economies and oil exporters now face in managing the large burdens through new multilateral debt-relief initiatives endorsed inflows of foreign exchange generated from capital flows and by the G-8 group of industrial countries in July 2005. Global trade transactions. Development Finance 2006 highlights this unique opportunity to bolster development efforts and place development finance on “Prospects for the Global Economy” is an online companion to a stable footing. Global Development Finance. It provides information on the global economic outlook, detailed regional forecasts, and additional The global economy grew by 3.2 percent in 2005, with the growth features such as interactive graphs, analytical tools, and access to rate in the developing world exceeding 5 percent for the third year underlying data.This online publication is available in English, running.While prospects are good for continued robust economic French, and Spanish at http://www.worldbank.org/globaloutlook. growth in developing countries, risks and vulnerabilities persist— including the possibilities of an abrupt and disorderly correction Global Development Finance 2006, II: Summary and Country Tables of global financial imbalances, external shocks leading to further includes a comprehensive set of tables with statistical data for 135 sharp hikes in oil prices, and a buildup of borrowing by countries countries that report debt under the World Bank Debtor Reporting of lower creditworthiness. As the world economy moves toward a System, as well as summary data for regions and income groups. multipolar international monetary system, multilateral cooperation It contains data on total external debt stocks and flows, aggregates, on international monetary relations and exchange rates have become and key debt ratios, and provides a detailed, country-by-country essential for addressing the current imbalances. Developing countries picture of debt. Global Development Finance 2006 debt data are also have much at stake. Policy makers in emerging market economies available in electronic format: GDF Online (an electronic subscription should give priority to strengthening institutions and promoting database) and the GDF CD-ROM. Each of these electronic databases policies and mechanisms that will improve their ability to navigate provides access to 217 historical time series indicators from 1970 in a world economy of increasingly integrated and interdependent to 2004, and country group estimates for 2005. global financial and production systems. The Little Book on External Debt is a new publication that provides Favorable financing conditions and robust growth coincide with a quick reference to key debt data in aggregate and individual increasing financial integration among developing countries with country tables prepared for the convenience of users. South-South flows of capital, now growing more rapidly than With analysis and data extending from short-term bank lending North-South flows, mirroring trends in trade flows. Although to long-term bond issuance in both local and foreign currency, South-South capital flows remain relatively small, they have the Global Development Finance 2006 is unique in its breadth of coverage potential to change the landscape of development finance over the of the trends and issues of fundamental importance to the financing next few years, particularly if growth in these countries continues of the developing world, including coverage of capital originating to outpace that in advanced countries. from developing countries themselves.The report is an indispensable Global Development Finance 2006, I: Review, Analysis, and Outlook resource for governments, economists, investors, financial consultants, is the World Bank’s annual review of recent trends in and academics, bankers, and the entire development community. THE WORLD BANK For more information on the analysis, please see 1818 H Street, NW www.worldbank.org/prospects; further detail about Washington, DC 20433 USA the Summary and Country Tables can be found at Telephone: 202 477-1234 www.worldbank.org/data. For general and ordering Facsimile: 202 477-6391 information, please visit the World Bank’s publications Web Internet: www.worldbank.org site at www.worldbank.org/publications, or call 703-661- E-mail: feedback@worldbank.org 1580; within the United States, please call 1-800-645-7274. Delivered by The World Bank e-library to: The World Bank IP : 192.86.100.36 Tue, 10 Mar 2009 16:55:33 I S B N 0 - 8 2 13 - 5 9 9 0 - 8 (c) The International Bank for Reconstruction and Development / The World Bank