Global 53098 Economic Prospects Crisis, Finance, and Growth 2010 Global Economic Prospects Global Economic Prospects Crisis, Finance, and Growth 2010 © 2010 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 13 12 11 10 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 volume 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. ISBN: 978-0-8213-8226-4 eISBN: 978-0-8213-8227-1 DOI: 10.1596/978-0-8213-8226-4 ISSN: 1014-8906 Cover photos: © iStockphoto.com/burakpekakcan (sky); © iStockphoto.com/mikeuk (sea) Cover design: Critical Stages The cutoff date for the data used in this report was January 8, 2010. Dollars are current U.S. dollars unless otherwise indicated. Contents Foreword xi Acknowledgments xiii Abbreviations xv Overview 1 The acute phase of the crisis is over 1 Impact of the boom period on developing-country potential 4 Medium-term implications of the bust for finance in developing countries 8 Medium-term impact on the supply potential of developing countries 12 Conclusion 14 References 14 Chapter 1 Prospects for Developing Economies 15 Recent developments in financial markets 18 Global growth 22 Prospects for high-income countries 23 Prospects for developing economies 25 Regional outlooks 26 Commodity markets 31 Inflation 35 World trade 36 Narrowing global external payment imbalances 38 Uncertain prospects 39 The impact of the crisis on the very poor 41 Policy implications for developing countries 42 Notes 43 References 44 Chapter 2 The Impact of the Boom in Global Finance on Developing Countries 45 Financial innovation, high-income finance, and the liquidity boom 47 Novel channels for credit creation 50 Developing-country finance during the boom 53 v C O N T E N T S Real-side consequences of the surge in global finance 61 Concluding remarks 70 Notes 71 References 72 Chapter 3 Medium-Term Impacts of the Crisis on Finance and Growth in Developing Countries 75 The impact of post-crisis regulatory and structural changes 77 Implications of a potential developing-country retreat from financial integration 88 The impact of higher borrowing costs 94 Strategies for dealing with a weaker international finance system 104 Implications for the global balance between savings and investment 107 Conclusion 108 Notes 109 References 112 Appendix: Regional Economic Prospects 117 East Asia and the Pacific 117 Europe and Central Asia 123 Latin America and the Caribbean 131 Middle East and North Africa 139 South Asia 146 Sub-Saharan Africa 154 Notes 163 References 164 Figures O.1 Financial conditions have stabilized 2 O.2 Developments in high-income countries have driven the industrial production cycle 2 O.3 The downturn in developing countries has been deeper and more broadly based than during previous recessions 3 O.4 Selected indicators of macroeconomic stability in developing countries, 2007 5 O.5 The cost of risk in high-income countries fell sharply during the boom 6 O.6 Developing-country potential output growth was boosted by low borrowing costs 7 O.7 Foreign participation in selected emerging equity markets 9 O.8 FDI as a share of investment in developing countries, 1995-2008 10 O.9 Very volatile external debt flows pose serious macroeconomic challenges 10 O.10 Indicators of regulatory quality 12 O.11 Private credit provision is strongly correlated with per capita incomes 12 O.12 Higher borrowing costs result in a permanent decline in developing-country GDP 13 1.1 Financial markets' stabilization has partially restored pre-crisis financial conditions in developing countries 19 1.2 Syndicated bank lending by region, 2008 and 2009 20 1.3 FDI flows to developing countries 21 1.4 External financing needs as a share of GDP, 2010 22 vi C O N T E N T S 1.5 Growth in industrial production 23 1.6 Change in stock building as a contribution to GDP growth in G-3 countries 24 1.7 Dispersion of GDP growth results in the first quarter of 2009 25 1.8 China's stimulus program yielded a pickup in import demand 27 1.9 Nonperforming loans rise across much of Europe and Central Asia 28 1.10 In Latin America EMBI-stripped spreads retreat as investor confidence returns 28 1.11 Lower oil prices and production yield sharp decline in oil revenues during 2009 30 1.12 South Asia's external position improves in most countries on lower oil prices and decline in domestic demand 30 1.13 Quarterly GDP data point to output stabilization in Sub-Saharan Africa 31 1.14 Real commodity price indexes 32 1.15 OPEC spare capacity 33 1.16 Global stock-to-use ratio of key agricultural markets (excluding China) 34 1.17 Real commodity prices 35 1.18 Inflation in low-, middle-, and high-income countries 35 1.19 Core inflation in high-income countries 35 1.20 Food prices in low-income countries 36 1.21 World trade is recovering 36 1.22 International tourist arrivals 37 1.23 An easing of global imbalances 39 1.24 U.S.-China imbalances have diminished markedly 39 1.25 Comparison of 2015 poverty forecast, GEP 2009 versus GEP 2010 43 2.1 Since the early 2000s, credit expansion has grown more than twice as fast as nominal GDP 49 2.2 High-income GDP and trade growth do not explain the acceleration in developing- country economic activity 50 2.3 Notional value of derivative transactions, 2002­08 52 2.4 Share of commercial bank and securitized assets in total credit held by U.S. financial sector, 1995­2008 52 2.5 The cost of risk in high-income countries fell sharply during the boom 55 2.6 Developing-country interest rates fell substantially during the boom period 56 2.7 Total capital inflows to developing economies 58 2.8 Portfolio equity flows to developing countries 59 2.9 FDI inflows to developing countries, 1980­2008 60 2.10 Distribution of capital flows as a percentage of GDP in 2007 61 2.11 The determinants of private finance 62 2.12 Private credit from banks and other financial institutions relative to per capita income, 2007 63 2.13 Rising investment rates contributed to an acceleration in potential output 70 3.1 Foreign participation in selected emerging equity markets 84 3.2 Debt financing of M&A transactions, 1995­2008 85 3.3 FDI as a share of investment in developing countries, 1995-2008 86 3.4 Net external debt flows from private sources, 1980­2008 88 3.5 The recent buildup in reserves was concentrated in East Asia and among oil exporters 89 3.6 Developing countries' average financial openness, 1990­2006 91 vii C O N T E N T S 3.7 The global synthetic price of risk versus the portion explained by economic fundamentals 95 3.8 Yields on selected U.S. government securities 97 3.9 Government debt is projected to rise in high-income, but not developing, countries 97 3.10 Falling capital costs were reflected in rising capital-output ratios 99 3.11 Impact of 110-basis-point shock on the capital-output ratio of a typical country 102 3.12 Higher borrowing costs result in a permanent decline in developing-country GDP 102 3.13 The effect of higher borrowing costs on the ratio of private-source debt to gross national income 104 3.14 Higher intermediation costs will reduce the risk-free interest rate relevant to savings decisions 108 A1 East Asian exports and production hard hit by downturn in capital goods demand 118 A2 Malaysia: A profile of recovery 119 A3 China's stimulus program supports exports from regional partners 119 A4 Equity markets have nearly recovered from earlier declines 120 A5 East Asian currencies continue to recoup against U.S. dollar 120 A6 Stimulus measures yield widening fiscal shortfalls across the region 121 A7 East Asia will enjoy a moderate rebound in 2010­11 122 A8 Gross capital flows to Europe and Central Asia picked up in mid-2009 125 A9 Many European and Central Asian economies post sharp current account adjustments 125 A10 Pace of contraction in industrial production moderates in Europe and Central Asia 126 A11 Nonperforming loans rise across much of developing Europe and Central Asia 130 A12 Industrial production and trade volumes are gaining momentum in Latin America and the Caribbean 132 A13 Central banks across Latin America eased monetary policy aggressively 134 A14 Real effective exchange rates depreciated in more integrated economies 134 A15 Quarterly GDP points to output stabilization in Latin America 134 A16 Capital flows return to Latin America 135 A17 Equity markets in the Middle East dropped quickly, and GCC recovery has been muted 140 A18 Oil prices, 2004­09 141 A19 Lower oil prices and lower Middle Eastern output yield sharp decline in oil revenues in 2009 141 A20 Reduced oil revenues reduce current account surplus positions in 2009 142 A21 Middle Eastern exports declined sharply as European demand fell 142 A22 Worker remittances fell by a moderate 6.3 percent in 2009 143 A23 Tourism receipts fall from record 2008 performance, but modest recovery likely 143 A24 Oil price and export revival hold key to recovery in the Middle East and North Africa 144 A25 Gross capital inflows to South Asia are recovering but remain below pre-crisis levels 147 A26 Local equity markets have generally returned to pre-crisis levels 147 A27 Recovery in industrial production has taken hold in South Asia ahead of most other regions 148 viii C O N T E N T S A28 Current account balances improve across much of South Asia in 2009 149 A29 International flows to South Asia 149 A30 Budget deficits in most South Asian countries outstrip developing-country average 150 A31 Interest payments represent a significant burden in South Asian economies 153 A32 Middle-income and oil-exporting countries of Sub-Saharan Africa hit hardest by global crisis 154 A33 Private consumption and trade contracted markedly in South Africa 155 A34 Current account balances of middle-income and oil-exporting countries deteriorated the most, while low-income countries remained aid-dependent 155 A35 EMBI-stripped spreads retreat as investor confidence returns 156 A36 Inflation in Sub-Saharan Africa down to low single digits on lower food prices 156 A37 Quarterly GDP readings point to output stabilization in Sub-Saharan Africa 157 A38 Growth in middle-income and oil-exporting countries in Sub-Saharan Africa will accelerate faster 162 Tables O.1 A modest recovery 3 O.2 Regional distribution of changes in financing conditions, 2000­07 6 O.3 Decomposition of increase in potential output growth directly attributable to capital deepening 7 O.4 Contribution of private-source debt inflows to external finance of developing countries with current account deficits, average 2003­07 9 1.1 The global outlook in summary 17 1.2 Prospects remain uncertain 40 1.3 Poverty in developing countries by region, selected years 42 2.1 Interest rates and inflation in industrial countries, January 2002­June 2007 49 2.2 Changes in domestic intermediation, 2000­07 56 2.3 Net capital inflows by region 60 2.4 Intertemporal changes in financial variables mainly reflected the cost of capital, but across countries institutional quality was most important 63 2.5 Regional distribution of changes in financing conditions, 2000­07 65 2.6 Rising investment rates by region 66 2.7 Intertemporal and cross-country influences on investment 67 2.8 Decomposition of increase in potential output growth directly attributable to capital deepening 70 3.1 Credit growth by foreign banks versus total credit growth, developing countries 82 3.2 Contribution of private-source debt inflows to external finance of developing countries with current account deficits, average 2003­07 88 3.3 Indicators of the quality of domestic financial systems 93 3.4 Historical and prospective costs of capital for developing countries 98 3.5 Possible impact of tighter financial conditions on developing-country capital costs 99 3.6 Impact on potential output of a return to normal pricing of risk and higher base rates 101 3.7 Higher borrowing costs and slower population growth imply slower growth in potential output over the longer term 103 3.8 The crisis could increase poverty by 46 million in the long term 104 3.9 Selected indicators of banking sector efficiency 105 ix C O N T E N T S 3.10 Potential impact of improved fundamentals on long-term growth prospects 106 A1 East Asia and Pacific forecast summary 118 A2 East Asia and Pacific country forecasts 123 A3 Europe and Central Asia forecast summary 124 A4 Europe and Central Asia country forecasts 129 A5 Latin America and the Caribbean forecast summary 131 A6 Latin America and the Caribbean country forecasts 137 A7 Middle East and North Africa forecast summary 141 A8 Middle East and North Africa country forecasts 145 A9 South Asia forecast summary 151 A10 South Asia country forecasts 152 A11 Sub-Saharan Africa forecast summary 154 A12 Sub-Saharan Africa country forecasts 159 Boxes 1.1 Prospects for remittances 38 2.1 Comparing this boom-bust cycle with other major cycles 48 2.2 Recent and systemically important financial innovations 51 2.3 Financial intermediation and economic development 54 2.4 The role of foreign banks in domestic intermediation 57 2.5 Capital flows can boost investment and efficiency 59 2.6 Determinants of cross-country differences in domestic and international financial intermediation 64 2.7 Understanding the increase in investment rates 68 2.8 Estimating potential output in developing countries 69 3.1 Likely directions of financial sector reforms 78 3.2 The impact of foreign bank participation on stability 81 3.3 Survey evidence on the decline in trade finance during the crisis 83 3.4 FDI and debt flows during crises 86 3.5 The debate over capital account liberalization 92 3.6 The synthetic price of risk 95 3.7 Higher borrowing costs will constrain domestic and external finance 100 x Foreword T HIS YEAR, Global Economic Prospects bubbles. The medium-term strength of the is being released at a critical juncture recovery will depend both on how well these for the world economy. A recovery challenges are met and on the extent to which from the financial crisis that rocked the world private-sector demand picks up. If policies are in the fall of 2008 is under way, but many adjusted too slowly, inflationary pressures and challenges remain and much uncertainty con- additional bubbles could develop; too quick of tinues to cloud the outlook. an adjustment could stall the recovery. In many respects, recent economic news Whatever the relative strength of the recov- has been encouraging. Industrial production ery in the next few months, the human costs of and trade, after falling by unprecedented this recession are already high. Globally, and amounts worldwide, are growing briskly; notwithstanding upward revisions to growth financial markets have recovered much of the projections for 2010, the number of people steep losses they incurred in late 2008 and living on $1.25 per day or less is still expected early 2009; and developing countries are once to increase by some 64 million as compared again attracting the interest of international with a no-crisis scenario. The recession has investors. However, the depth of the reces- cut sharply into the revenues of governments sion has left the global economy seriously in poor countries. Unless donors step in to fill wounded. Even as profitability returns to the gap, authorities in these countries may be many of the firms that were at the heart of the forced to cut back on social and humanitarian crisis, industrial production and trade levels assistance precisely when it is most required. have yet to regain their pre-crisis levels, and In addition to analyzing the immediate chal- unemployment has reached double digits in lenges for developing countries posed by the many countries and continues to rise. crisis, this year's Global Economic Prospects Given the depth of the crisis and the con- describes some of the longer-term implications tinued need for restructuring in the global of tighter financial conditions for developing- banking system, the recovery is expected to country finance and economic growth. While be relatively weak. As a result, unemploy- necessary and desirable, tighter regulation in ment and significant spare capacity are likely high-income countries will result in less abun- to continue to characterize the economic dant capital (both globally and domestically) landscape for years to come. This poses a and increased borrowing costs for developing real challenge for policy makers, who must countries. As a result, just as the very loose con- cut back on unsustainably high fiscal deficits ditions of the first half of this decade contributed without choking off the recovery. Similarly, to an investment boom and an acceleration in the extraordinary monetary stimulus needs to developing-country growth, so too will higher be scaled back to avoid the creation of new capital costs in coming years serve to slow xi F O R E W O R D growth in developing countries and provoke a borrowing costs--more than offsetting any decline in potential output. negative impacts from tighter international Countries should not respond passively. Ef- conditions. forts to strengthen domestic financial systems Overall, these are challenging times. The and expand regional cooperation (including depth of the recession means that even though regional self-insurance schemes) can help to growth has returned, countries and individu- reduce the sensitivity of domestic economies als will continue to feel the pain of the crisis to international shocks and counteract some for years to come. Policy can help mitigate the of the longer-term negative effects of tighter worst symptoms of this crisis. However, there international financial conditions. Such initia- are no silver bullets, and achieving higher tives are most likely to benefit middle-income growth rates will require concerted efforts to countries that already have reasonably well- increase domestic productivity and lower the developed regulatory and competitive envi- domestic cost of finance. ronments and healthy financial sectors. Fi- nally, both low- and middle-income countries Justin Yifu Lin should strengthen domestic financial regula- Chief Economist and tions. Over time, such steps can improve do- Senior Vice President mestic financial-sector efficiency and reduce The World Bank xii Acknowledgments T HIS REPORT WAS produced by staff from the World Bank's Development Prospects Group. The report was managed by Andrew Burns, with direction from Hans Timmer. The principal authors of the report were Andrew Burns, William Shaw, and Theo Janse van Rensburg. The report was produced under the general guidance of Justin Yifu Lin. Several people contributed substantively to chapter 1. Theo Janse van Rensburg was its main author. The Global Macroeconomic Trends team, under the leadership of Andrew Burns, was responsible for the projections. The projections and regional write-ups were produced by Dilek Aykut, Ivailo Izvorski, Eung Ju Kim, Annette De Kleine, Oana Luca, Israel Osorio-Rodarte, Theo Janse van Rensburg, Elliot (Mick) Riordan, Cristina Savescu, and Nadia Islam Spivak. These were produced in coordination with country teams, country directors, and the offices of the regional Chief Economists and PREM directors. The short-term commodity price forecasts were produced by John Baffes, Betty Dow, and Shane Streifel. The remittances forecasts were produced by Sanket Mohapatra, while Shaohua Chen from the Development Research Group and Dominique van der Mensbrugghe generated the long-term poverty forecast. Andrew Burns, William Shaw, and Nikola Spatafora were the main authors of chapter 2, with written contributions from Mike Kennedy, Oana Luca, and Angel Palerm. Chapter 3 was written by Andrew Burns and William Shaw, with written contributions from Dilek Aykut, Jean-Pierre Chauffour, and Mariem Malouche. Both chapters 2 and 3 benefited from the expert research assistance of Augusto Clavijo, Yueqing Jia, Eung Ju Kim, Irina Kogay, Sergio Kurlat, Sabah Mirza, and Nadia Islam Spivak. The accompanying online publication, Prospects for the Global Economy (PGE), was pro- duced by a team led by Cristina Savescu and composed of Cybele Arnaud, Augusto Clavijo, Sarah Crow, Betty Dow, Ernesto McKenzie, Kathy Rollins, Ziming Yang, and Ying Yu, with technical support from Gauresh Rajadhyaksha. The translation process was coordinated by Jorge del Rosario (French and Spanish) and Li Li (Chinese). A companion pamphlet highlighting the main messages of the commodities section of the report was prepared by Kavita Watsa and Roula Yazigi. Martha Gottron edited the report. Hazel Macadangdang managed the publication process, and Rebecca Ong and Merrell Tuck-Primdahl managed the dissemination activities. Book pro- duction was coordinated by Aziz Gökdemir along with Stephen McGroarty and Andrés Meneses, all from the World Bank Office of the Publisher. Several reviewers offered extensive advice and comments throughout the conceptualization and writing stages. These included Shaghil Ahmed, Daniel Benitez, Cesar Calderon, Otaviano Canuto, Kevin Carey, Rodrigo Chavez, Jeff Chelsky, Mansoor Dailami, Jeff Delmon, Shahrokh Fardoust, Alan Gelb, Jack Glen, Arvind Gupta, Fernando Im, Jacqueline Irving, Ada Karina xiii A C K N O W L E D G M E N T S Izaguirre, Ivailo Izvorski, Prakash Kannan, Auguste Tano Kouame, Robert Kahn, Doreen Kibuka-Musoke, Jean Pierre Lacombe, Atushi Limi, Justin Yifu Lin, Dominique van der Mensbrugghe, Celestin Monga, Mustapha Nabli, Fernando Navarro, Il Young Park, Maria Soledad Martinez Peria, Zia Qureshi, Hartwig Schafer, Luiz Pereira da Silva, Claudia Paz Sepulveda, Hans Timmer, and Augusto de la Torre. xiv Abbreviations ASEAN Association of Southeast Asian Nations BIS Bank for International Settlements CDOs collateralized debt obligations CDSs credit default swaps CPI consumer price index DECPG Development Economics Prospects Group (World Bank) ECB European Central Bank EMBI Emerging Markets Bond Index EMBI+ Emerging Markets Bond Index Plus FDI foreign direct investment GCC Gulf Cooperation Council GDP gross domestic product GIDD Global Income Distribution Dynamics model GNFS goods and nonfactor services GNI gross national income IBRD International Bank for Reconstruction and Development IDA International Development Association IMF International Monetary Fund IPO initial public equity offering Lao PDR Lao People's Democratic Republic LCU local currency unit LIBOR-OIS London Interbank Offered Rate-Overnight Indexed Swap rate M&A mergers and acquisitions MDG Millennium Development Goal MSCI Morgan-Stanley Composite Index NIEs newly industrializing economies ODA official development assistance OECD Organisation for Economic Co-operation and Development OPEC Organization of the Petroleum-Exporting Countries PPP purchasing power parity saar seasonally adjusted annualized rate T-bill Treasury bill (U.S.) TFP total factor productivity UAE United Arab Emirates xv Overview T he world economy is emerging from the throes of a historically deep and synchro- nized recession provoked by the bursting of a pace at which many developing-country econo- mies could grow without generating significant inflation. global financial bubble. The consequences of the While developing countries probably cannot initial bubble and the crisis have been felt in virtu- reverse the expected tightening in international ally every economy, whether or not it participated financial conditions, there is considerable scope directly in the risky behaviors that precipitated for reducing domestic borrowing costs, or in- the boom-and-bust cycle. And while growth rates creasing productivity and thereby regaining the have picked up, the depth of the recession means higher growth path that the crisis has derailed. that it will take years before unemployment and spare capacity are reabsorbed. This year's Global Economic Prospects ex- amines the consequences of the crisis for both The acute phase of the crisis the short- and medium-term growth prospects of is over developing countries. It concludes that the crisis and the regulatory reaction to the financial ex- cesses of the preceding several years may have T he immediate impacts of the crisis (includ- ing a freezing up of credit markets, a sharp reversal of capital flows, and the precipitous lasting impacts on financial markets, raising bor- equity market and exchange rate declines that rowing costs and lowering levels of credit and ensued) are largely in the past. Since March international capital flows. As a result, the rate 2009, stock markets in high-income and emerg- of growth of potential output in developing ing economies have recovered roughly half the countries may be reduced by between 0.2 and value they lost, with developing economies re- 0.7 percentage points annually over the next five bounding somewhat more strongly than high- to seven years as economies adjust to tighter fi- income ones. Interbank lending rates have nancial conditions. Overall, the level of potential returned to normal levels, developing-country output in developing countries could be reduced sovereign interest rate premiums have declined by between 3.4 and 8 percent over the long run, from a peak of more than 800 to around 300 compared with its pre-crisis path. basis points and stock market volatility has The report further finds that the very liquid receded (figure O.1). In addition, bond flows conditions of the first half of the decade contrib- to high-income corporate and emerging-market uted to the expansion in credit available in devel- sovereigns have returned to more normal levels, oping countries and that this expansion was and most developing-country currencies have responsible for about 40 percent of the approxi- regained their pre-crisis levels against the dollar. mately 1.5 percentage point acceleration of the However, bond markets and bank lending have 1 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Figure O.1 Financial conditions have Figure O.2 Developments in high-income stabilized countries have driven the industrial production cycle Change from recent peak to current levels Contributions to global industrial production growth (% saar) Basis points 16 90 450 12 left axis right axis 80 400 8 4 350 70 0 300 4 60 250 8 50 200 12 16 150 40 20 100 24 30 50 1 3 1 3 1 3 Q Q Q Q Q Q 07 07 08 08 09 09 20 20 20 20 20 20 20 0 MSCI EM index Vix index LIBOR-OIS High-income countries Developing countries Maximum decline: Decline as of China except China Oct. 2007 to Sep. 2009 Mar. 2009 (%) Current Source: World Bank. Maximum: Oct. 2008 Maximum: Oct. 2008 index point basis point Source: JP Morgan and Morgan-Stanley. level consistent with its pre-crisis trend growth rate. Overall, considerable slack remains in the global economy, with unemployment con- begun only recently to reopen themselves to pri- tinuing to rise, disinflation widespread, and vate sector borrowers in developing countries, commodity prices between 50 and 25 percent with syndicated loans to developing countries lower than their levels in mid-2008. totaling only $123 billion in 2009, compared with $236 billion during 2008. A subdued recovery The real side of the global economy is also Overall, after falling for two to three quarters, recovering, with industrial production at the global GDP has begun recovering; output grew global level growing at more than 12 percent rapidly during the second half of 2009 and is ex- annualized pace in the third quarter of 2009. pected to continue to do so during the first half The recovery, which was initially concentrated of 2010. However, as the positive contribution in developing countries, has become more bal- to growth from fiscal stimulus and the inventory anced recently as the drawdown of inventories cycle wanes, growth will slow, in part because in high-income countries slows and activity spending by households and the banking sector catches up to underlying demand trends (figure will be less buoyant as they rebuild their balance O.2). Nevertheless, the level of output remains sheets. As a result, global GDP growth, which is depressed worldwide, with industrial produc- projected to come in at 2.7 percent in 2010 (after tion still 5 percent below pre-crisis peaks in an unprecedented 2.2 percent decline in 2009), is October 2009. expected to accelerate only modestly to 3.2 per- Trade, which initially fell sharply, is also cent in 2011 (table O.1). recovering; the exports of developing coun- A weak recovery is also anticipated in devel- tries were expanding at a 36 percent annual- oping countries. Arguably the inventory cycle is ized pace in October, but the volume of world somewhat more advanced in East Asia and the trade remained 2.8 percent lower than its Pacific, and there are signs that the growth impact pre-crisis level and some 10 percent below the of fiscal stimulus in China may already be waning 2 O V E R V I E W Table O.1 A modest recovery (real GDP growth, percentage change from previous year) Region 2007 2008 2009e 2010f 2011f World 3.9 1.7 2.2 2.7 3.2 High-income countries 2.6 0.4 3.3 1.8 2.3 Euro Area 2.7 0.5 3.9 1.0 1.7 Japan 2.3 1.2 5.4 1.3 1.8 United States 2.1 0.4 2.5 2.5 2.7 Developing countries 8.1 5.6 1.2 5.2 5.8 East Asia and Pacific 11.4 8.0 6.8 8.1 8.2 Europe and Central Asia 7.1 4.2 6.2 2.7 3.6 Latin America and the Caribbean 5.5 3.9 2.6 3.1 3.6 Middle East and North Africa 5.9 4.3 2.9 3.7 4.4 South Asia 8.5 5.7 5.7 6.9 7.4 Sub-Saharan Africa 6.5 5.1 1.1 3.8 4.6 Memorandum items Developing countries excluding transition countries 8.1 5.6 2.5 5.7 6.1 excluding China and India 6.2 4.3 2.2 3.3 4.0 Source: World Bank. Note: e estimate; f forecast; growth rates aggregated using real GDP in 2005 constant dollars. (industrial production and import growth in the Figure O.3 The downturn in developing region are already slowing). Output is estimated countries has been deeper and more broadly to have picked up in virtually every other develop- based than during previous recessions ing region in the final quarter of 2009 and should Percent continue to do so early in 2010, before slowing 0 toward more sustainable rates later in the year. 1 The pace of the recovery is expected to be most subdued in the Europe and Central Asia region, 2 partly because the pre-crisis level of demand in 3 the region was well above potential and partly be- 4 cause the financial system in the region has been 5 more acutely affected by the crisis. Combined, GDP growth in developing 6 1982­83 1991­93 2001 2009 countries is projected to grow by some 5.2 per- cent in 2010, after a modest 1.2 percent rise in Change in GDP growth Output gap 2009 ( 2.2 percent if India and China are ex- Source: World Bank. cluded), and by a relatively weak 5.8 percent Note: Change in GDP growth is the percentage change in the in 2011. Despite these relatively robust growth growth rate of developing-country GDP between the crisis year(s) and the previous year. The output gap is the percentage difference rates, the unusual depth of the recession will between GDP and potential output during the crisis year(s). mean that spare capacity and unemployment will continue to plague economies in 2011 and some sectors may well still be shrinking. Over- significant spare capacity, high unemployment, all, the output gap (the difference between ac- and weak inflationary pressures will continue to tual GDP and what GDP would be if capital characterize both high-income and developing and labor were fully employed) in developing countries for some time. Already, the slowdown countries will remain elevated at about 4 per- in growth is estimated to have increased poverty. cent of potential output in 2011 (figure O.3). Some 64 million more people around the world The depth of the recession and the relative are expected to be living on less than $1.25 a day weakness of the expected recovery suggest that by the end of 2010 than would have been the 3 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 case without the crisis, and between 30,000 and a second recession. Indeed, in some middle-in- 50,000 children may have died of malnutrition come countries, very loose monetary conditions in 2009 in Sub-Saharan Africa because of the may already be generating asset price bubbles in crisis (Friedman and Schady 2009). Moreover, local real-estate and other asset markets. the slowdown is expected to cut heavily into government revenues in poor countries. Coun- tries eligible for soft loans and grants from the Impact of the boom period on International Development Association of the developing-country potential World Bank may require as much as $35 billion to $50 billion in additional funding just to maintain 2008 program levels, never mind the resources I n some ways, the crisis and recession from which the world economy is currently emerg- ing resemble previous boom-bust cycles. Like necessary to fund additional demands brought many other major crises, the current one is upon by the crisis. characterized by a sharp reduction in economic activity following an extended period of rapid and ultimately unsustainable credit expansion, The outlook remains clouded by accompanied by excessive risk taking by finan- uncertainties and the challenge of cial institutions. unwinding the stimulus At the same time the current crisis differs Many uncertainties continue to surround the from previous ones in fundamental ways. From short-term outlook for developing countries. a global perspective, this crisis is the most severe Principal among these is the extent to which and widespread downturn since 1945. Global private sector consumption and investment GDP is estimated to have contracted by 2.2 per- demand will respond to the pickup in activity cent in 2009 (the first absolute decline in global prompted by fiscal and monetary stimulus GDP among the postwar crises). Even in 2011 and the inventory cycle. Should the response demand is projected to remain 5 percent below be weaker than expected in the baseline the global economy's productive potential, projection or should the stimulus be with- which is almost twice the output gap during the drawn too quickly, the recovery could stall. next most severe recession (1982­83). Although a double-dip recession in the sense Moreover, in contrast with earlier down- of a return to negative global growth rates is turns, the current crisis struck virtually every unlikely, developing-country growth could developing country hard, even though, with come in as low as 5.1 percent in 2010 and the important exception of many in Europe and 5.4 percent in 2011, with some countries Central Asia, most countries did not exhibit potentially recording negative growth for unsustainable macroeconomic imbalances one or more quarters. (figure O.4). Outside of Europe and Central A related but opposite risk is that the stimu- Asia, regional inflation rates averaged about lus is not retracted quickly enough. In the case 6 percent or lower (well below the double- of fiscal policy, the risk is mainly one of in- digit rates in most regions during the early creased indebtedness and unnecessary crowding 1990s); most regional current account bal- out of private sector investment. On the mone- ances were near zero or strongly positive; tary policy side, the risk is that the vast mone- and ratios of debt to gross national income tary expansion that has been undertaken begins were modest. The importance of prudent to gain traction, potentially overinflating the macroeconomic policies was revealed during global economy. This could recreate liquidity the crisis, as the countries with the largest conditions similar to those that created the bub- imbalances suffered the biggest declines in bles that precipitated the crisis, causing global output (see chapter 3). imbalances to reemerge and forcing a much That the acute phase of the crisis was deeper more abrupt tightening of policy--possibly even than past ones may have important longer-term 4 O V E R V I E W so many countries are affected, unemployment Figure O.4 Selected indicators of macroeco- will remain high longer, skills will deteriorate, nomic stability in developing countries, 2007 otherwise healthy firms may go bankrupt, and Debt/GDP (percent) the overall level of economic dislocation and as- 45 sociated economic costs will remain high. Just 40 35 passing through the crisis may have a sustained 30 negative impact on productivity and the future 25 path of economic growth. In some economies, 20 prolonged weakness in demand could provoke 15 the disappearance of whole sectors instead of 10 just some companies. This could be especially 5 0 the case for declining sectors. Similarly, an un- even recovery with growth and economic dyna- l A and ric d ric n ci ia be a ia Af an Af ara ib ric Pa s As fic an a a a d tA mism concentrated in one region versus another ar e si tra e th t h C m or as h an Eas en p Sa ut d nA C ro N eE So b- Eu could sway the path of investment, making lag- an ti Su dl La id M ging countries look weaker and possibly creating Current account/GDP (percent) new comparative advantages in the leading re- 10 gions. Global trade patterns may be irrevocably 8 altered. 6 How these forces will play out and the poli- cies that should be put in place to respond to 4 them merit in-depth exploration. However, 2 dealing with all of the potential consequences 0 of the crisis for developing countries lies out- 2 side of the scope of this publication. The approach to the medium-term conse- 4 quences of the crisis described in the pages that l A and ric d ric n ci ia be a ia Af an Af ara ib ric Pa s As follow is more narrowly oriented toward the fic an a a a d tA ar e si tra e th t h C m or as h an Eas en p Sa ut d nA C ro N eE So b- consequences for developing countries of the Eu an ti Su dl La id M changes in financial conditions observed over Inflation (percent) the past decade and those that can be expected 12 in the next 5­10 years. Initially, the focus is on 10 how the boom in global financial markets af- 8 fected credit conditions, investment, and growth prospects in developing countries and on the 6 factors that help to explain which countries 4 benefited most from the boom. It then switches 2 to an examination of how changes in the regula- tory environment, risk aversion, and the policy 0 environment are likely to affect financial con- l A and ric d ric n ci ia be a ia Af an Af ara ib ric Pa s As fic an a a a d tA ditions, investment, and growth in developing ar e si tra e th t h C m or as h an Eas en p Sa ut d nA C ro N eE So b- Eu countries. an ti Su dl La id M Source: World Bank. Not all countries benefited fully consequences for growth, productivity, and even from the liquidity boom the structure of the world economy going for- The liquidity boom in high-income countries ward. Because the shock is so deep and because during the first seven years of the 21st century 5 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 created favorable financial conditions in both Figure O.5 The cost of risk in high-income high-income and developing countries. Much countries fell sharply during the boom more intensive use of a range of financial inno- Percentage Points vations, including the securitization of loans 16 and the development of off-balance-sheet Merrill high-yield corporate (US) Lehman investment 14 vehicles, allowed banks to off-load an im- grade (Euro) 12 portant portion of their loan portfolios onto Lehman investment 10 grade (US) capital and money markets. Effectively, these US 10-year government bond 8 innovations allowed unregulated securities to 6 support a portfolio of loans much like the tra- 4 ditional banking sector--but without capital Euro 10-year government bond 2 requirements and under a much less stringent Jan. 1998 Jan. 2000 Jan. 2002 Jan. 2004 Jan. 2006 regulatory framework. That permitted an un- Source: Datastream. precedented leveraging of equity capital, and the rapid expansion of liquidity that ensued helped to drive down interest rates, inter- est rate premiums, and the cost of capital in The ensuing investment boom both high-income and developing countries boosted the supply potential of (figure O.5). developing countries As a result, domestic banking sectors and The liquidity boom fed an investment boom the quantity of domestic credit available in developing countries that prompted a rapid within developing countries increased quickly. expansion in the supply potential of low- and At the global level, international banking sec- middle-income countries but with limited im- tor credits grew twice as fast as nominal GDP, pact on goods inflation in most countries. On and the quantity of capital flowing to low- and average, investment-to-GDP ratios in devel- middle-income economies surged. Overall pri- oping countries increased by 5.5 percentage vate sector lending increased by 5.5 percent points, ranging from a 1.4 percentage point of GDP; the ratio of international capital in- increase in Latin America and the Caribbean flows to GDP increased by about 5 percentage to an 8.1 percentage point rise in South Asia. points; and stock market capitalization in- As a result, between 2000 and 2007 creased by 79 percent of GDP (table O.2). capital-to-output ratios in developing countries Table O.2 Regional distribution of changes in financing conditions, 2000­07 Change between 2000 and 2007 in: Capital Stock market Private credit by Region Cost of capital inflows capitalization deposit money banks Investment (basis points) (percent of GDP) Developing countries 400 5.0 79 5.5 5.5 Low-income countries 2.3 Middle-income countries 5.6 East-Asia and Pacific 134 2.0 118 10.7 5.5 Europe and Central Asia 866 12.0 60 15.7 4.9 Latin America and the Caribbean 471 2.0 40 2.2 1.4 Middle East and North Africa 269 2.0 36 6.2 5.0 South Asia 142 7.0 107 14.8 8.1 Sub-Saharan Africa 685 4.0 59 6.8 3.6 Sources: World Bank; Beck and Demirgüç-Kunt 2009. Note: Regional values are simple averages of countries, except for investment rates, which are weighted averages. 6 O V E R V I E W Table O.3 Decomposition of increase in Figure O.6 Developing-country potential potential output growth directly attributable output growth was boosted by low to capital deepening borrowing costs Change in growth rate of potential Percent growth in potential output output (2003­07 versus 1995­2003) 8 Actual Due to Share due to 7 capital capital 6 Regions Total deepening deepening 5 Without capital 4 deepening (percentage points) (percent) 3 Developing 1.5 0.6 40.3 2 Middle-income 1.5 0.6 39.8 Low-income 1.3 0.8 63.7 1 0 East Asia and Pacific 1995 1997 1999 2001 2003 2005 2007 (excluding China) 0.4 0.1 19.8 China 0.3 0.9 283.5 Source: World Bank. Europe and Central Asia 3.1 0.6 18.7 Latin America and the Caribbean 0.3 0.1 46.6 were about 10 percentage points higher than Middle East and North Africa 0.8 0.5 66.7 they would have been had investment rates South Asia 1.4 1.1 78.5 Sub-Saharan Africa 1.9 1.5 79.5 held stable at their 2002 levels. The increase in capital services provided by the additional capi- Source: World Bank. tal contributed to about 40 percent of the 1.5 percentage point increase in the rate of growth of potential output (the level of output if capi- intermediation in developing countries between tal and labor were fully employed) during this 1998 and 2008. Nevertheless, other factors re- period (figure O.6 and table O.3). In so far as the main critical in understanding the cross-coun- rising share of new capital embodying the latest try differences in the level of intermediation technology contributed to the observed increase (and in the increases observed since the 1980s). in total factor productivity growth during this The quality of institutions and levels of market period, the actual contribution of the boom to openness are associated with 56 and 37 percent developing-country potential was even higher. of the variaion in the average level of interme- The notable exception was in the Europe and diation between developing countries in the top Central Asia region. Despite experiencing the and bottom quartiles according to the ratio of largest increase in intermediation, much of the domestic credit to GDP, respectively. In practi- additional resources went into consumption. As cal terms, this finding suggests that an improve- a consequence, investment-to-GDP ratios in the ment in institutional quality in Sub-Saharan region increased by only 4.9 percent, less than Africa to roughly the level observed in Latin the 5.5 percent average for all developing coun- America could generate an increase in the stock tries considered as a whole. And in contrast with of domestic credit to the private sector of about other regions, the expansion in domestic credit 12 percent of GDP and in international finance fueled a consumption binge that generated sig- of about 2 percent of GDP. nificant domestic and external imbalances and Countries with relatively open economies, ultimately unstable macroeconomic conditions. strong institutions, and supportive investment climates enjoyed the largest increases in external Economic policies are critical flows during the boom. Resource-rich countries to understanding cross-country also fared well in attracting external capital, in differences in intermediation part because their resources provided relatively Lower borrowing costs were the largest secure collateral that partially compensated for identifiable factor behind the increase in the weak quality of their institutions. 7 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Countries with good regulatory environ- · the introduction of rules and policies ments were also more successful in transform- designed to isolate developing coun- ing increased financing into increased investment tries from excessive financial market and, as a result, increased long-term supply po- volatility; tential. Inflows of foreign direct investment and · increasing reliance on domestic interme- domestic credit creation were associated with diation and efforts to deepen regional larger investment and growth effects than were financial markets; equity or debt-creating inflows. · a generalized increase in risk aversion; and Medium-term implications · a step backward from some of the in- novative financial instruments that were of the bust for finance in most associated with the financial crisis. developing countries T he short-term costs of the financial cri- sis have been severe and discouraging. In many countries, the sharp contraction in Anticipated regulatory changes in high- income countries are expected to broaden the range of financial institutions and activi- activity wiped out several years worth of the ties that come under supervision, increase re- additional GDP gains that the above-average porting criterion, reduce the scope for using growth of the preceding years had produced. derivatives and other innovative financial in- That a crisis rooted in regulatory failure in struments, and pay greater attention to inter- high-income countries has had such pro- bank dependencies and cross-border activities. nounced effects on developing countries may These changes, plus increased risk aversion have caused a backlash against financial and and the necessity for banks in high-income trade liberalization, particularly among the countries to rebuild their capital, suggest that many developing countries that implemented liquidity will be more scarce and expensive in stricter fiscal policy regimes, improved regula- the years to come. tory institutions, and introduced more flexible exchange rates during the 1990s and 2000s. Possible impacts of scarcer and Although these measures likely prevented the more expensive finance buildup of domestic vulnerabilities during the The extent to which international financial boom period, which would have made the cri- conditions impinge on developing-country fi- sis much more serious, they did not entirely nance goes well beyond the traditional current insulate developing countries from its effects. account financing of developing countries (see below). Indeed, in aggregate, developing coun- Tighter financial conditions are in tries are net lenders to high-income countries. the offing, implying reduced levels Once cross-border flows have been netted out, of finance developing countries invested more of their The lessons and fallout from the crisis are savings into high-income countries than high- likely to shape financial policies and market income countries invested in them between reactions for years to come. Beyond the im- 2000 and 2008. mediate and unprecedented global recession However, for many countries with capital that it has provoked, the crisis can be expected shortages, external savings are still a critical to significantly alter the global financial land- source of finance for investment. Excluding scape over the next 5 to 10 years. China and the oil exporters, the remaining de- These changes may include: veloping countries are, on average, net import- ers of capital. Of the 53 developing countries · a tightening and broadening of the scope that faced an external financing gap in 2009, of financial market regulation; most had current account deficits of 5 percent 8 O V E R V I E W Table O.4 Contribution of private-source Weaker and more expensive capital at the debt inflows to external finance of global level also will affect financial conditions developing countries with current account in developing countries indirectly by influenc- deficits, average 2003­07 ing conditions in domestic financial markets Net debt (changes in the cost of and rate of return on inflows Number of Current from external investment and borrowing, increased countries account private competitive pressure, technology, and knowl- with current deficit sources account (% of (% of edge transfers). deficits GDP) GDP) Tighter regulations, along with the transfor- mation of many investment banks into tradi- All countries 53 6.3 2.2 Low income 16 5.8 0.8 tional banks, may reduce the supply of financial Lower middle income 20 6.1 0.8 services, including the intermediation of devel- Upper middle income 17 7.1 5.3 oping countries' capital issuances (figure O.7). Of which: ECA 8 8.5 8.1 Over the past 10 years, American investment Source: World Bank. banks participated in 86 percent of the value Note: Data on current account deficits and debt inflows are simple averages of country numbers. Excludes small island of developing-country initial public offerings, economies. or 32 percent of the number of deals, and the operation of mutual funds and other investment of GDP or more, with private-source net debt vehicles allowed individual and institutional in- inflows equal to about 2.2 percent of GDP (or vestors in high-income countries to place money almost half the deficit)--0.8 percent if the coun- in developing markets. While developing-coun- tries in Europe and Central Asia are excluded try competitors could pick up some of these ac- from the mix (table O.4). For these countries a tivities and while high-income firms will almost significant withdrawal of external financing certainly continue their involvement in this busi- could have serious consequences for domestic ness, the likely result is that developing-country investment and long-term potential output. firms will have less access to capital. Moreover, Figure O.7 Foreign participation in selected emerging equity markets (portfolio equity inflows divided by stock market capitalization, percent) % of local stock market capitalization 50 2003 2007 45 40 35 30 25 20 15 10 5 0 Jo ey Ka Br a ge ic So on d n an de lic Po ru H ico an Re sia kh il ne M ia Tu d Tu n R ze Mal an h ia ov ilip ia R nes th a a Bu dia Pa nia M hile C o om a U bia om a Th ary La a za az si In lan tio n a Li tin ni c C hin R ari ric Ar ubl n ut es Sl Ph tv Pe Fe pub rk oc st ai la st rd ni ex y to ua In a g C ra ak pi n Af lg us ch a ki kr ai un ep or Es ol d si C Source: IMF. 9 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 overall productivity will be affected if less active Figure O.8 FDI as a share of investment in foreign investment banks have a comparative developing countries, 1995-2008 advantage in identifying firms and products with Percent of fixed investment strong potential in global markets. 25 Tighter regulation in high-income coun- 2000-09 tries and the need for parent banks to build up 20 1995-99 their capital may also impede foreign banks' 15 2005-08 participation in developing countries, which 10 could have negative consequences for their development--especially for poorer countries 5 with good regulatory regimes. Foreign banks 0 can serve as a conduit for foreign savings into pi ll E cif n A A d ea d ric d ia ric n lo A Pa sia tin al an M ribb an Af an Af ra As a developing country and can contribute to E ng C uro ic th m sia N le E n a ha So a d A La ntr pe ca ica th st h Sa or a an ast ut e er b- ve greater intermediation at lower cost by increas- Su de d e id ing competition. This can be especially impor- tant in less developed countries. However, the Source: World Bank. quality of domestic institutions is important. In the presence of weak institutions, foreign banks' participation may have no or even a net negative effect on intermediation and cost sav- ing, if, as has happened in some regions, they Figure O.9 Very volatile external debt flows cherry-pick the best clients and merely displace pose serious macroeconomic challenges domestic banks. US$ billion Share of GDP % Foreign direct investment (FDI) should be 600 Net external debt flows 4.0 less constrained than debt flows by the rise in 500 from private sources/GDP 3.5 risk aversion and more stringent regulation. Net external debt flows 3.0 400 from private sources However, parent firms will face higher capi- (in $US billion) 2.5 tal costs, and these are likely to reduce their 300 2.0 ability to finance individual projects. As a re- 200 1.5 sult, FDI inflows are projected to decline from 100 1.0 recent peaks of 3.9 percent of developing- 0.5 0 country GDP to around 2.8­3.0 percent of 0.0 GDP. The real-side consequences of such a 100 0.5 1980 1985 1990 1995 2000 2005 decline could be serious because foreign direct investment represents as much as 20 percent Source: World Bank. of total investment in Sub-Saharan Africa, Europe and Central Asia, and Latin America (figure O.8). Of course, access to foreign capital is not an system could have some benefits if it reduced unmixed blessing, as both this crisis and past developing countries' dependence on volatile crises serve to remind us. Historically, private capital flows. capital flows into developing countries, nota- Indeed, a central lesson from this boom-bust bly debt flows such as bank and bond lending, cycle is that although the very loose financial have been very volatile (figure O.9). Because conditions contributed to the growth boom in such capital flows can stop, or even reverse developing countries, the boom was not sustain- abruptly, countries that become heavily reliant able and the crisis, loss in output, and associated upon them can be very vulnerable. From this social dislocation were essential and arguably point of view, a less integrated global financial inevitable consequences of the boom. If better 10 O V E R V I E W regulation of financial flows going forward suc- accumulating and maintaining even higher ceeds in reducing volatility and the frequency of reserves. boom-bust cycles, the benefits of more stable Another strategy that has been followed and sustainable conditions could outweigh the after earlier crises has been the imposition of costs (see below) of more expensive and less capital controls or a slowdown in liberaliza- abundant capital. tion. While such steps may reduce the risk that an economy develops a level of external in- debtedness that makes it vulnerable to a rapid Countries may seek to insulate shift in sentiment, it does so at the expense themselves from global financial of the longer-term benefits (such as technol- markets. . . ogy transfers, increased investment, and fur- Of course, the extent to which a given country ther integration into the global economy) that experiences volatility in financial markets, as might have accompanied the excluded capital well as the consequences for the real economy, inflows. In addition, controls on capital are also depends on domestic policies. Despite often ineffective, particularly when they are the fact that countries with prudent and open used to support substantial exchange rate policies tended to benefit from the boom and misalignment. suffer least in the bust, the negative impacts of this crisis, which encompassed many devel- . . . and increase the role of oping countries that had managed the inflows domestic and regional alternatives associated with the boom period in a very Faced with a less active external financing sys- prudent manner, may induce authorities in de- tem, authorities and entrepreneurs in developing veloping countries to take additional steps to countries may take steps to promote domestic reduce their economies' vulnerability to large financial intermediation as an alternative to re- changes in conditions outside their control. liance on foreign capital. Given the importance In the past, developing countries have re- that intermediation has for development, such a acted to crises by increasing their official re- strategy could have significant benefits for those serves or imposed capital controls as a means middle-income countries that already have a of reducing the domestic consequences of ex- strong framework for financial intermediation, ternal shocks. Such self-insurance mechanisms by increasing the efficiency of domestic financial can be expensive. By some estimates, recent intermediaries through learning by doing and reserve holdings of developing countries have economies of scale. cost as much as 1 percent of GDP to maintain. For low-income countries, the longer-run ef- Nor are such reserves necessarily effective. For fects of a weaker international system may be example during the recent crisis, there was more serious. In the short run, low-income coun- only a limited correlation between the sever- tries may be less directly affected by the crisis- ity of the real-side downturn experienced by induced increase in borrowing costs--simply developing countries and the level of reserves because their economies are not well intermedi- they held going into the crisis period. This lack ated. However, a weaker international financial of correlation does not mean that reserves did system could deny them investments critical to not help cushion the shock--indeed, countries their development, particularly because defi- with low reserves and high current account ciencies in domestic intermediation systems are deficits tended to be hardest hit by the crisis. likely to prevent them from compensating for a However, it does suggest that beyond a point, reduced foreign presence (figure O.10). additional reserves offer little additional pro- The crisis is also likely to result in greater tection from this kind of international shock regional cooperation, which could strengthen observed and that countries should carefully financial services by capturing economies weigh the additional costs associated with of scale and facilitating risk sharing by 11 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 management strategies that contributed to Figure O.10 Indicators of regulatory quality boosting liquidity are all factors that are likely Index to increase borrowing costs in both high- 2.0 income and developing countries. Rule of law The overall expansion of investment and 1.5 growth during the boom period, without the 1.0 Control of corruption creation of significant inflationary pressures or external imbalances in many developing 0.5 Regulatory quality countries, suggests that in these countries the boom relieved what may have been a binding 0.0 capital constraint on growth, albeit in what proved to be a temporary and unsustainable 0.5 manner. The necessary and desirable tighten- 1.0 ing of regulations will hopefully reduce the frequency of boom-bust cycles and provide a r d Af an ia an ati sia d id be ca ia H fric an tri e So ica L lA n un om Pa s As M rib eri a A r Eu cific N -E n h- a es th t d tA ha more stable financial environment for devel- or as dl a tra e co inc C m h an Eas en p Sa ut d nA C ro e b- ig oping countries. However, higher borrowing a Su costs are likely to mean a temporary decline Sources: World Bank, Worldwide Governance Indicators Project. in the rate of growth of developing-country potential output. Financial services are criti- cal to the smooth functioning of an economy, and the level of domestic intermediation (for pooling reserves. Such cooperation may also example, the ratio of domestic bank credit to help strengthen South-South financial flows, GDP) is strongly correlated with economic de- which are likely to be important in sustain- velopment (figure O.11). ing FDI flows to many developing countries. The extent to which anticipated changes However, progress in regional financial co- in financial markets will increase borrowing operation has been slow in developing coun- tries. Further, such arrangements are likely to be of greatest benefit to regions that already Figure O.11 Private credit provision is have relatively robust domestic financial sys- strongly correlated with per capita incomes tems, such as East Asia and the Pacific. Poor GDP/capita in US$ countries with weak institutions can benefit 70,000 through integration with stronger regional 60,000 economies, but the promotion of regional in- tegration with other countries with weak insti- 50,000 tutions is unlikely to be beneficial. 40,000 30,000 20,000 Medium-term impact on the 10,000 supply potential of developing 0 countries 10,000 I ncreased risk aversion, the necessity for banks to recapitalize, increased borrowing require- ments from high-income governments, and 0.5 1.0 1.5 Private Credit/GDP 2.0 2.5 Sources: International Financial Statistics; World Bank. the falling into disrepute of many of the risk- 12 O V E R V I E W costs in developing countries will depend on Figure O.12 Higher borrowing costs result many factors, including the level of interest in a permanent decline in developing- rates in high-income countries. Currently, country GDP under the influence of the extraordinary Trillions of real 2005 US$ steps taken by the U.S. Federal Reserve Bank 3 and other central banks, medium (1-year) Pre-crisis path of potential output 3 and long-term (10-year) interest rates on U.S. government securities are 0.4 and 3.8 per- 2 centage points, respectively--some 290 and 2 60 basis points lower than during the boom Post-crisis path of potential output period. Similarly, developing-country risk 1 premiums have fallen and appear to have sta- 1 bilized at close to their pre-crisis levels or 2005 2010 2015 2020 2025 2030 about 150 basis points higher than during Source: World Bank. the boom period. If real interest rates in high-income countries return to their pre- boom levels and if the historical relationship between these base rates and interest rate Developing countries can mitigate spreads remain unchanged, the borrowing the effects of weaker international costs developing countries face could rise by conditions between 110 and 220 basis points compared Although there is little that developing coun- with their boom-period levels. tries can do to prevent a deterioration in global Just as the decline in borrowing costs dur- financial conditions, they should not stand by ing the first few years of this decade was as- passively. Much can be done to mitigate the sociated with a marked pickup in investment costs of a tightening of global financial condi- activity and potential growth rates, higher tions by reducing the domestic cost of inter- borrowing costs going forward will tend to re- mediation through strengthening regional and duce investment rates and result in lower lev- domestic institutions or by improving long- els of potential output than would have been term productivity growth. observed otherwise. Firms can be expected to Inefficiency of domestic financial sectors re- react to higher capital costs by employing less sulting from corruption, weak regulatory insti- capital and more labor and natural resources tutions, poor protection of property rights, and per unit of output, so economy-wide capi- excessive limits on competition can make bor- tal-to-output and investment-to-GDP ratios rowing costs in developing countries 1,000 basis will decline. During the transition period to points higher than in high-income countries the new, lower capital-output ratio, the rate (even more so if the even higher costs imposed of growth of potential output could slow by by informal lenders are taken into account). between 0.2 and 0.7 percentage points from Improvements in the policies and institu- the average of 6.2 percent rate observed dur- tions governing the financial sector can thus ing the 2003­07 period (figure O.12). Over have a significant impact in boosting domes- the long run, unless offset by other factors tic financial intermediation, one that can out- (notably improved domestic policies, see weigh any potential negative impact of higher below), this substitution away from capital- global risk premiums. Simulations suggest that intensive techniques could reduce the supply if developing countries continue to improve potential in developing countries by between policies and other fundamentals, so that their 3 and 5 percent and potentially by as much as interest spreads fall by an average of 25 basis 8 percent. points a year, they would more than offset 13 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 the long-term effects of the financial crisis-- highs, which is likely to be a source of frustra- producing a 13 percent increase in long-term po- tion for many countries. While some prudent tential output and increases in potential output reforms to reduce the sensitivity of domestic growth of about 0.3 percent per year by 2020. economies to some of the more volatile forms Efforts to increase domestic financial in- of international capital may be advisable, termediation should focus on strengthening policy makers need to remain mindful of the institutions, not on discriminating against for- benefits that financial openness and improved eign capital. Especially in countries with poor intermediation can bring. regulations or weak enforcement capacities, Looking forward, it is not desirable to discouraging foreign capital could have the recreate the unstable and unsustainable in- detrimental effect of forcing firms to rely on ternational conditions of the boom period. more expensive domestic sources of finance However, the domestic savings in developing and potentially reducing the overall level of in- countries represent an enormous growth po- termediation. Suppressing foreign capital also tential that is waiting to be released through could reduce firms' access to new technology, reforms aimed at reinforcing and growing do- expertise, and international market contacts. mestic intermediation. Although such reforms will take time to bear fruit over the longer term they may once again place developing Conclusion countries on the higher growth path that the T he international financial conditions of the boom period were unsustainable and resulted in the extremely disruptive and costly crisis has derailed. crisis from which the global economy is only References now emerging. At the same time they demon- Beck, Thorsten, and Asli Demirgüç-Kunt. 2009. "Fi- strated that, when exposed lower capital costs, nancial Institutions and Markets across Coun- developing countries were capable of sustain- tries and over Time: Data and Analysis." Policy Research Working Paper 4943. World Bank, ing significantly higher growth rates without Washington, DC. generating higher inflation. Friedman, Jed, and Norbert Schady. 2009. "How Over the medium term, international capital Many More Infants Are Likely to Die in Africa as costs are going to be higher than they were during a Result of the Global Financial Crisis?" Policy the boom period. As a result, developing-country Research Working Paper 5023. World Bank, growth potential will remain well below recent Washington, DC. 14 1 Prospects for Developing Economies T he acute phase of the financial crisis has passed and a global economic recovery is under way. Moreover, the recovery is frag- have recovered--with a rapid run-up during the last months of 2009. Also, borrowing costs for emerging market borrowers have stabilized ile and expected to slow in the second half of over the last few quarters, but remain elevated. 2010 as the growth impact of fiscal and mon- However, private sector firms remain shut etary measures wane and the current inven- out from international banking markets. tory cycle runs its course. Indeed, industrial Moreover, the Dubai World event and ripple production growth is already slowing (albeit effects to credit downgrades for Greece and from very high rates). As a result, employment Mexico can be expected to once again raise growth will remain weak and unemployment concerns about sovereign debt sustainabil- is expected to remain high for many years. ity and will impact risk assessments, capital The overall strength of the recovery and its flows, and financial markets in 2010. durability will depend on the extent to which The real economy is recovering as well. household- and business-sector demand Although global industrial production in strengthens over the next few quarters. While October 2009 remained 5 percent below its the baseline scenario projects that global level a year earlier, it is recovering, with out- growth will firm to 2.7 percent in 2010 and put in both high-income and developing 3.2 percent in 2011 after a 2.2 percent de- countries expanding at more than a 12 percent cline in 2009, neither a double-dip scenario, annualized rate (or saar) in the third quarter of where growth slows appreciably in 2011, or a 2009. Just as a sharp drop in inventories strengthening recovery can be ruled out. contributed to a precipitous initial decline in Financial markets have stabilized and are industrial production, the stabilization of recovering, but remain weak. Interbank li- inventory levels has contributed to a strong re- quidity as measured by the difference between bound in production, and this factor is ex- the interest rates commercial banks charge one pected to support industrial production, even another and what they have to pay to central as growth rates start to come down. bankers have declined from an unprecedented Trade too is recovering but remains de- peak of 366 basis points in dollar markets to pressed. Quarterly growth rates have moved less than 15 basis points--a level close to its into positive territory in recent months, but "normal" pre-crisis range. Currencies, which the U.S. dollar value of trade was still off fell worldwide against the U.S. dollar in the 17 percent from its September 2009 level. immediate aftermath of the crisis, have largely Lower commodity prices mean that the vol- recovered their pre-crisis levels. And interna- ume of trade has fared better, but it is never- tional capital flows to developing countries theless down by 3 percent from a year ago. 15 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 The most marked increases have been in de- 4.0 percent rate in 2010 and 2011, respec- veloping East Asia, and reflect, at least partly, tively, compared with 5.4 percent growth on the 4 trillion renminbi (or 12 percent of GDP) average between 2003 and 2008. Countries in fiscal stimulus put into place by the Chinese developing Europe and Central Asia have been authorities extending through 2010 (roughly hardest hit by the crisis and are expected to half spent). have the least marked recovery, with GDP ex- Much of that stimulus has found its way panding by only 2.7 percent in 2010 and by into imported raw commodities and investment 3.6 percent in 2011. goods. Indeed, partly because of restocking, The combination of the steep decline in Chinese demand for key metals has been sup- activity in 2009 and the relatively weak pro- portive of commodity prices, which have recov- jected recovery means that developing econo- ered about one-third of their earlier declines. mies will still be operating about 3 percent Nevertheless, international metal prices, mea- below their level of potential output1--and sured in U.S. dollars, are 20 percent below their unemployment, although on the decline will July 2008 levels, oil prices are 44 percent lower, still be a serious problem. Moreover, the im- and food prices 24 percent lower, with global pacts on poverty and human suffering in these oil demand some 2 percent lower than its peak countries will be very real. Some 30,000­ level of 87 million barrels a day in 2007. 50,000 additional children may have died of The combination of the abrupt fall in com- malnutrition in 2009 in Sub-Saharan Africa modity prices and ample spare capacity world- because of the crisis (UNSCN 2009; Friedman wide has resulted in median inflation in and Schady 2009), and globally by the end of developing countries falling from more than 2010, 90 million more people are expected to 10 percent in August 2008 to about 1 percent be living in poverty than would have been the in October 2009. case without the crisis. Global imbalances narrowed further during Few of the poorest countries will have the the crisis. This trend may be largely cyclical, as fiscal space to respond to the economic dislo- it relates to substantial declines in the U.S. trade cation caused by the crisis without significant deficit, the Chinese trade surplus, and the price additional financial assistance. It is estimated of oil. The durability of the narrowing will de- that IDA countries (those eligible for soft pend on the speed with which the United States loans and grants from the International Devel- can unwind its fiscal and monetary stimulus opment Association of the World Bank) will and the extent to which stimulus-based infra- require an additional $35 billion to $50 bil- structure investment in China contributes to lion in funding just to maintain current levels higher domestic demand rather than additional of programming, let alone come up with the export capacity. additional funding required to meet the needs Although the real-side effects of the crisis of those additional individuals thrown into have been large and serious, economic activity poverty.2 Worse, the recession may cause do- in most developing countries is recovering and nors to reduce aid flows precisely at the mo- overall growth is expected to pick up from the ment the flows need to rise. anemic performance of 1.2 percent in 2009 to Great uncertainty continues to surround 5.2 percent in 2010 and to 5.8 percent in 2011 future prospects. Even the weak recovery (table 1.1). Although much lower than the outlined above is not certain. If the private 6.9 percent growth rate that developing coun- sector continues to save in order to restore tries averaged between 2003 and 2008, these balance sheets, a double-dip, characterized by rates are well above the 3.3 percent average a further slowing of growth in 2011 is entirely performance during the 1990s. Excluding possible--especially as the growth impact of China and India, the remaining developing fiscal stimulus wanes. A stronger recovery is countries are projected to grow at a 3.3 and also possible, if the massive traditional and 16 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S Table 1.1 The global outlook in summary (percentage change from previous year, except interest rates and oil price) 2007 2008 2009h 2010i 2011i Global Conditions World trade volume 7.2 3.0 14.4 4.3 6.2 Consumer prices G-7 countriesa,b 2.0 3.1 0.2 1.1 1.7 United States 2.9 3.8 0.5 1.6 2.4 Commodity prices (US$ terms) Non-energy commodities 17.1 21.0 21.6 5.3 0.7 Oil price (US$ per barrel)c 71.1 97.0 61.8 76.0 76.6 Oil price (percent change) 10.6 36.4 36.3 23.1 0.8 Manufactures unit export valued 5.5 6.0 4.9 1.5 0.7 Interest rates $, 6-month (percent) 5.2 3.2 1.2 1.8 2.8 , 6-month (percent) 4.3 4.8 1.5 2.2 3.0 Real GDP growthe World 3.9 1.7 2.2 2.7 3.2 Memo item: World (PPP weights)f 5.0 2.7 1.0 3.5 4.0 High income 2.6 0.4 3.3 1.8 2.3 OECD Countries 2.5 0.3 3.3 1.8 2.3 Euro Area 2.7 0.5 3.9 1.0 1.7 Japan 2.3 1.2 5.4 1.3 1.8 United States 2.1 0.4 2.5 2.5 2.7 Non-OECD countries 5.4 2.6 2.3 2.9 3.9 Developing countries 8.1 5.6 1.2 5.2 5.8 East Asia and Pacific 11.4 8.0 6.8 8.1 8.2 China 13.0 9.0 8.4 9.0 9.0 Indonesia 6.3 6.1 4.5 5.6 5.8 Thailand 4.9 2.6 2.7 3.5 4.0 Europe and Central Asia 7.1 4.2 6.2 2.7 3.6 Russia 8.1 5.6 8.7 3.2 3.0 Turkey 4.7 0.9 5.8 3.3 4.2 Poland 6.7 4.9 1.6 2.2 3.4 Latin America and Caribbean 5.5 3.9 2.6 3.1 3.6 Brazil 5.7 5.1 0.1 3.6 3.9 Mexico 3.3 1.4 7.1 3.5 3.6 Argentina 8.7 6.8 2.2 2.3 2.4 Middle East and North Africa 5.9 4.3 2.9 3.7 4.4 Egyptg 7.1 7.2 4.7 5.2 6.0 Irang 7.8 2.5 1.0 2.2 3.2 Algeria 3.0 3.0 2.1 3.9 4.0 South Asia 8.5 5.7 5.7 6.9 7.4 Indiag 9.1 6.1 6.0 7.5 8.0 Pakistang 5.7 2.0 3.7 3.0 4.0 Bangladeshg 6.4 6.2 5.9 5.5 5.8 Sub-Saharan Africa 6.5 5.1 1.1 3.8 4.6 South Africa 5.5 3.7 1.8 2.0 2.7 Nigeria 6.3 5.3 4.3 4.8 5.1 Kenya 7.1 1.7 2.8 3.7 4.8 Memorandum items Developing countries excluding transition countries 8.1 5.6 2.5 5.7 6.1 excluding China and India 6.2 4.3 2.2 3.3 4.0 Source: World Bank. 2008 2009h 2010i 2011i Note: PPP purchasing power parity; h estimate; i forecast. a. Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States. Egypt 6.8 5.7 5.1 5.6 b. In local currency, aggregated using 2005 GDP Weights. c. Simple average of Dubai, Brent, and West Texas Intermediate. Iran 2.5 1.0 2.2 3.2 d. Unit value index of manufactured exports from major economies, expressed in USD. India 7.3 6.4 7.6 8.0 e. Aggregate growth rates calculated using constant 2005 U.S. dollar GDP weights. Pakistan 3.8 2.9 3.3 3.5 f. Calculated using 2005 PPP weights. Bangladesh 6.3 6.1 5.7 5.7 g. In keeping with national practice, data for Egypt, Iran, India, Pakistan and Bangladesh are reported on a fiscal year basis. Expressed on a calendar year basis, GDP growth in these countries is as in the table just above. 17 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 untraditional monetary stimulus that has been their intention to begin to do so soon. In the put into place in high-income countries begins United States, the Federal Reserve Board's to gain traction. federal fund rate has been hovering around 12 basis points, compared with close to 550 basis points in mid 2007. The European Central Bank's (ECB) policy rate remains in Recent developments in the 100-basis-point range, compared with a financial markets level of more than 400 basis points in 2008. T he unprecedented steps that were taken by policy makers in both developed and developing countries following the collapse Short-term market rates are also very low, reflecting the reduced opportunity cost of bor- rowing money from the monetary authority and of Lehman Brothers in September 2008 have increased confidence in the creditworthiness of gone a long way toward normalizing financial counterparties within the international banking markets and restoring capital flows to devel- system. Reflecting policy steps to recapitalize oping countries (see World Bank, 2009c for a banks and restore confidence in the interna- summary of such measures). tional financial system, the spread between the The immediate outflow of international price that commercial banks charge one another capital from developing countries to safe ha- for overnight lending and the overnight rate vens in the United States and Europe has re- charged by central banks--a common measure versed itself. As a result, a large number of of banks' confidence in one another--has fallen emerging-market exchange rates have recov- from an unprecedented 365 basis points at the ered their pre-crisis levels vis-à-vis the U.S. peak of the crisis to a more normal level of less dollar, equity markets have recovered much of than 15 basis points (figure 1.1, panel a). As a their initial losses, and, capital flows to devel- part of these efforts, central banks have taken a oping countries have begun to recover. number of extraordinary steps including lend- Toward the end of 2009, gross capital in- ing directly to private firms and intervening in flows to developing countries began to gain secondary mortgage markets. As a result, their momentum as uncertainty subsided and risk balance sheets have ballooned. aversion declined. On an annualized basis, As a result of these and other measures, total gross inflows to developing countries the freeze-up of financial markets that char- reached a $435 billion pace in the five months acterized the autumn of 2008 has eased ending November 2009, up from $218 billion considerably, and the spreads facing emerging- in the first half of the year. Although capital market borrowers have declined as well (fig- flows for the year as a whole remain 20 per- ure 1.1, panel b), with commercial borrow- cent below their 2008 levels and well below ers able to access funds for a premium of 359 their peaks in 2007, this recent surge in portfo- basis points and sovereign borrowers at a pre- lio flows has raised concerns that if sustained, mium of about 300 basis points. While these it could reinflate some of the asset bubbles spreads are higher than the pre-Lehman av- in stock, currency, and real estate markets erage of about 180 basis points, they remain among developing countries that the crisis had substantially lower than their long-term aver- only just begun to unwind. However, risk ap- ages, the fruit of improved fundamentals of petites may have been tempered by the Dubai many developing countries and years of policy World event and the credit rating downgrades reform. of Greece and Mexico at the end of 2009. As spreads declined and the acute risk Policy interest rates around the globe re- aversion of the immediate post-crisis period main very low, although some central banks eased, investors started moving back some of have begun tightening (e.g., Australia has al- the money that had been withdrawn from ready tightened by 75 basis points) or signaled developing-country capital markets. As a 18 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S Figure 1.1 Financial markets' stabilization has partially restored pre-crisis financial conditions in developing countries a. Liquidity in interbank markets has normalized b. Developing-country spreads have come down LIBOR-OIS spread key policy and regulatory responses; Basis points 400 1,200 G-7 meeting 1,000 300 800 200 600 400 100 200 0 0 Jun. 2008 Oct. 2008 Feb. 2009 Jun. 2009 Oct. 2009 07 08 09 7 8 9 07 08 09 10 00 00 00 20 20 20 20 20 20 20 .2 .2 .2 n. n. n. n. ay ay ay pt pt pt Ja Ja Ja Ja M M M Dollar Euro Pre-crisis US$ average Se Se Se EMBI global CEMBI global Pre-Lehman collapse average c. Most emerging currencies have recovered against d. Equity markets have rebounded the dollar percentage change (USD per LCU) (%) Equity market indexes (Jan. 2007 = 100) US reer 200 yen 180 US neer sterling 160 Mexican peso 140 Turkish lira euro 120 Canadian $ 100 Russian ruble Idonesian rupiah 80 Polish zloty 60 Korean won Hungarian forint 40 Brazilian real 20 South African rand Australian $ 0 09 08 09 20 10 0 10 20 30 40 50 08 10 09 08 07 07 07 20 20 20 20 20 20 20 20 20 20 n. Percentage change (USD per LCU) n. ay n. ay p. p. n. ay p. Se Se Ja Ja Ja Se Ja M M M Stronger local currency Gain since Mar. 2009 Emerging markets Asia Emerging markets Europe NET since Sep. 2008 Latin America Mature markets Sources: Thomson/Datastream; World Bank. result, beginning roughly in March 2009 restore global confidence by restoring some of developing-country currencies began appreci- the wealth initially destroyed in the crisis. ating against the dollar (figure 1.1, panel c), The revival in stock market activity has their stock markets began rebounding, recov- supported new equity placements by emerging ering between one-third and one-half of their economies, which totaled $98 billion in the initial losses (figure 1.1, panel d), helping to first eleven months of 2009, up sharply from 19 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 $66 billion during the same period in 2008. Although initial public offering (IPO) activ- Figure 1.2 Syndicated bank lending by region, 2008 and 2009 ity remained subdued during the first half of 2009, there were signs of a sharp rebound in US$ (billions) the third quarter, on the strength of large deals 120 by China, Brazil, and India, which together ac- 100 counted for about 85 percent of all emerging- market transactions year-to-date, compared 80 with an average share of 65 percent in the five 60 years through 2007. The relatively strong fun- damentals in these countries appear to have 40 raised investor preference for these econo- mies. Gross equity flows to the remaining 20 developing countries were still compressed at 0 0.15 percent of their GDP in 2009, versus ci ia l A nd be ca ric d ia ric n Af an Af ara Pa s As ib ri tra e a Eu ic a an So a a d st A 0.42 percent (of GDP) on average during the ar e si f th t h C Am or as h en p Sa ut an Ea C ro N eE b- n five years ending 2007. an ati Su dl L id d M Developing countries' access to interna- tional capital markets has also revived. Both 2008 2009 sovereign and corporate borrowers have ben- Sources: World Bank; Bankscope. efited from rising global liquidity, improved market conditions, and better long-term fun- damentals of emerging economies vis-à-vis ad- vanced economies. The recovery in corporate rebuild their balance sheets. In 2009, syndi- bond issuance by developing countries reached cated loan deals involving developing coun- $109 billion during 2009, up almost $45 bil- tries amounted to $123 billion, compared lion compared with 2008. During the first with $236 billion in 2008. There was a sur- trading week of 2010, Turkey and the prising jump in December 2009, when loans Philippines tapped international bond mar- amounted to $27 billion, mostly led by $10 bil- kets for $2 billion and about $1.5 billion, lion lending for energy-related projects in respectively, taking advantage of continuing Papua New Guinea and $6.5 billion trade favorable conditions. The improved bond and finance loan to the Brazilian government equity markets reflect a normalization of finan- (figure 1.2). Overall, banks' external claims cial markets and, to an unknown extent, the on developing countries reported to the Bank opening up of a carry trade precipitated by low of International Settlements (BIS) expanded real interest rates in high-income countries. by only $10 billion in the second quarter of Some middle-income countries (notably Chile 2009 (exchange rate adjusted), after contract- and Brazil) are attracting very large inflows, ing $126 billion in the first quarter of 2009 which if sustained at current rates, pose real and $279 billion in the fourth quarter of 2008. policy challenges and could generate signifi- Prospects for a resurgence in bank lending in cant stress. Some countries have sought to use the near term are likely to be muted (longer- increased intervention or other measures such term prospects are discussed in chapter 3), espe- as a financial operational tax (Brazil)--even as cially in regions such as Europe and Central the effectiveness of these measures is unknown. Asia where mounting nonperforming loans and In stark contrast to the recovery in bond large domestic adjustments are likely to restrain and equity markets, cross-border bank lend- both the demand and supply side for lending. ing remains weak as global banks continue At the same time, lending to natural-resource- to consolidate and deleverage in an effort to rich countries is likely to remain robust. 20 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S these stronger inflows have only partially off- Figure 1.3 FDI flows to developing countries set the sharp reduction in flows following the US$ (billions) crisis and have not re-created bubble condi- 140 tions. However, should these strong inflows 123 121 persist or strengthen, asset bubbles could 120 106 begin to reinflate, leaving countries vulnerable 100 86 to a second sudden stop in external finance. 73 77 80 69 60 Prospects and implications for 40 developing-country financing needs Overall, net private capital flows to develop- 20 ing countries in 2009 are estimated to have 0 fallen by $795 billion (relative to their high in 1 2 3 4 1 2 3 2007), or by almost 70 percent. Even with re- Q Q Q Q Q Q Q 08 08 08 08 09 09 09 20 20 20 20 20 20 20 covery on the horizon, projected flows in 2010 Source: World Bank. are only $517 billion, or 3.2 percent of GDP. Note: Countries include Brazil, Bulgaria, Chile, China (excluding some sectors), Croatia, the Arab Republic of Egypt, India, Lower-income countries will suffer the most Indonesia, Jordan, Kazakhstan, Malaysia, Mexico, Pakistan, from this shrinkage, because their already Philippines, Poland, Romania, the Russian Federation, South Africa, Thailand, Turkey, Ukraine, and R. B. de Venezuela. miniscule share of total private capital flows (i.e., 2.6 percent in 2007) is expected to dwin- dle to almost nothing in 2010. Even though small in absolute terms, the capital inflows to In contrast with debt-creating flows, for- these low-income countries represent a signifi- eign direct investment (FDI) has yet to show cant share of national income and investment, signs of rebounding. FDI tends to be the and their loss will certainly have a severe im- most stable source of international capital, pact on the ability of these countries to meet but inflows nevertheless have declined by their financing needs in the short to medium 40 percent since the first quarter of 2008 and term (see chapter 3). stood at $69 billion in 2009Q3 (figure 1.3). While capital inflows have declined sharply, Although these flows are expected to have re- the ex ante financing needs of developing covered during the last quarter, inflows to all countries have not changed significantly. Based developing countries for the year as a whole on the current account deficit projections for are expected to come in at $385 billion, only 2010, along with schedules of private foreign 30 percent of their 2008 values. While re- debt coming due, the total external financing source-related investment has picked up in needs of developing countries are expected to 2009 after the pause in late 2008, investment be on the order of $1.1 trillion in 2010, com- in the banking sector, which led the surge in pared with an estimated $1.2 trillion in 2009.3 recent years, remains limited. Countries in Europe and Central Asia and The recent decisions of the Dubai World Latin America and the Caribbean face the larg- holding company to ask its creditors for a est external financing needs in 2010, projected six-month standstill on debt payments, and at $447 billion and $280 billion, respectively of rating agencies to downgrade Greece and (figure 1.4). Although smaller in magnitude, Mexico's credit rating, remind that the echoes Sub-Saharan Africa's financing needs are also of the crisis continue to be felt. Global markets large, standing at nearly 12 percent of GDP. have largely been unaffected by these develop- Combining these projections with country- ments and capital flows to emerging markets specific estimates of the amount of private sec- have strengthened in recent months. So far, tor financing likely to be forthcoming suggests 21 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 However, since SDRs are allocated accord- Figure 1.4 External financing needs as a ing to country quotas, the benefits of this share of GDP, 2010 move for the neediest developing countries are Percent small. 16 14 12 10 Global growth 8 6 4 2 A fter a deep global recession, economic growth has turned positive, as a wide range of policy interventions has supported 0 demand and reduced uncertainty and systemic l A and ric n ric d be ca ia ci ia risk in financial markets. However, the recov- Af ara Af an As Pa s ib ri fic a a an a d tA ar e si tra e th t h Am or as h en p an Eas Sa ut C uro N eE So ery is expected to be slow, as financial markets b- n E an ati Su dl C id L d remain impaired, stimulus measures will need M to be withdrawn in the not too distant future, Maturing foreign debt Current account and households in countries that suffered asset- Source: World Bank. price busts are forced to rebuild savings while Note: The figure shows the external financing needs as a share of struggling with high unemployment. Although GDP of those countries with a financial need. global growth is expected to return to positive territory in 2010, the pace of the recovery will be slow and subject to uncertainty. After fall- that developing countries could face a total fi- ing by an estimated 2.2 percent in 2009, global nancing gap of as much as $315 billion in 2010. output is projected to grow 2.7 and 3.2 percent In 2009 those countries whose ex ante in 2010 and 2011, respectively ( 1.0, 3.5, and financing needs exceeded private capital inflows 4.0 percent when aggregated using purchasing- were forced to bridge the gap either by cut- power-parity weights). ting into domestic demand and via exchange The main drag on global growth is coming rate depreciation--thereby reducing their trade from the high-income countries, whose econo- deficits via lower imports, or by using other mies are expected to have contracted by 3.3 per- resources such as drawing down international cent in 2009. Japan, which felt the consequences reserves or drawing upon official assistance of the global crisis more severely than other (or both). Overall, developing countries con- high-income countries, experienced the sharpest sumed some $362 billion worth of their in- growth contraction ( 5.4 percent). Growth ternational reserves during the initial phases rates of 2.5 and 2.9 percent are expected in of the crisis, while a wide range of countries 2010 for the United States and for high-income increased borrowing from the International countries that are not members of the Organisa- Monetary Fund (IMF), the World Bank, and tion for Economic Co-operation and Develop- various regional and bilateral development ment (OECD), respectively. agencies. Currently, an overall count of the in- The global economic crisis affected devel- crease in official flows is unavailable, but the oping countries first and foremost through a World Bank (International Bank for Reconstruc- sharp slowdown in global industrial activity tion and Development, or IBRD, and IDA) alone due to a sudden cut in investment programs, increased its lending commitments by some consumer durable demand, and a widespread $12.8 billion, while the IMF made an additional effort to reduce inventories in the face of uncer- commitment of $70 billion by October 2009. tain future conditions. Falling export demand, The IMF's lending resources are being tri- commodity prices, and capital flows exacer- pled, to $750 billion, including a new special bated and extended the downturn. Overall, drawing right (SDR) allocation of $283 billion. growth in developing countries declined to an 22 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S estimated 1.2 percent in 2009, down from Figure 1.5 Growth in industrial production 5.6 percent in 2008. Among developing-country regions, econ- Industrial production (percent change, saar) omies in Europe and Central Asia were hit 40 hardest by the crisis, with GDP falling 6.2 per- 30 cent (with the Russian Federation contract- 20 ing 8.7 percent). The main causes were lower 10 oil prices (Russia) and difficulties in funding 0 large current account deficits in a risk-adverse 10 environment. 20 Growth in the East Asia and Pacific region 30 (particularly in China) as well as in South Asia 40 (particularly India) has been resilient, buoyed 08 09 09 8 9 9 08 8 00 00 00 00 20 20 20 20 .2 .2 2 2 by a massive fiscal stimulus package in China l. n. l. r. n. r. ct ct Ju Ju Ap Ap Ja Ja O O and by India's skillful macroeconomic manage- ment. Between 2008 and 2009, growth in the China Developing countries excluding China High-income countries World total East Asia and Pacific region is estimated to have eased by only 1.2 percentage points to 6.8 per- Source: World Bank calculations based on Thomson Datastream cent, while South Asian growth has remained data. stable at 5.7 percent. GDP growth in China is estimated to have slowed from 9 percent in 2008 to 8.4 percent in 2009, but is expected to recover toward 9 percent over the remainder of the forecast period. Prospects for high-income These developments have also been reflected countries in global industrial production, which declined sharply in the aftermath of the global financial crisis. In February 2009, world industrial produc- O utput among high-income economies in 2009 is estimated to have contracted by 3.3 percent, the first time since 1960 that the tion was falling at a 27 percent annualized pace, aggregate GDP of these countries has declined. but by the beginning of April/May, production Industrial production and trade flows among began recovering (figure 1.5), initially led by ac- high-income countries were particularly dis- celerating growth in China following the imple- tressed, with the former registering peak-to- mentation of the $575 billion (over five quarters) trough declines in excess of 20 percent in fiscal stimulus package. Increased import demand countries such as the United States, the United from China quickly spread to other countries, Kingdom, Germany, and Japan. with industrial production registering positive A pronounced growth rebound is under growth in emerging countries (excluding China) way. The initial turnaround was driven by an by March 2009 and high-income countries by investment rebound in developing countries, May 2009. As the benefits of the stimulus mea- particularly China and the newly industrial- sures and inventory restocking began to wane, ized economies of East Asia, which has spread industrial production growth rates have started to high-income capital-equipment-exporting to moderate. Whether this deceleration signals a countries such as Germany and Japan. High- transition to slower growth, more in line with un- income countries have now started making derlying demand patterns, or the beginnings of a larger contributions to world output and trade double-dip growth recession will largely depend growth, as the effects of stimulus measures on the extent to which consumer and business de- bear fruit in fostering domestic demand and mand picks up in the months ahead (see the Risks imports, and a turn in the inventory cycle un- section below). derpins production gains. 23 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Supported by large stimulus programs, albeit slowly. As a result, the inventory cycle Japan, Germany, and France all started grow- in Europe is expected to be shallower and ing in the second quarter of 2009, while GDP shorter-lived. in the United States expanded 2 percent in the In the United States, notwithstanding the third quarter. Recent data releases also indicate recovery of growth in the second half, whole a continued rise in output in Japan, growing by year GDP is estimated to have declined by 1.3 percent during the third quarter (seasonally 2.5 percent in 2009. The recovery is expected adjusted rate). to continue into 2010, supported by the inven- The growth rebound in high-income coun- tory cycle, the bottoming out of the housing tries is projected to remain relatively strong sector downturn, and fiscal and monetary over the next several months but should lose stimulus. However, the pace of recovery should strength during the course of 2010 as the slow toward the middle of 2010 as the growth growth impact of stimulus measures and the impact of these forces wanes and as banking rebuilding of depleted inventories cease to sector balance sheet consolidation, and still bolster growth. During the depths of the re- large negative wealth effects weigh on domestic cession, changes in stock building shaved off demand. Overall, growth is projected to come 2.4 percent (first quarter of 2009) from an- in at 2.5 percent in 2010 and stabilize at a rela- nualized growth (figure 1.6). The inventory tively modest 2.7 percent in 2011. cycle is expected to be an especially important The IMF estimates that even though element feeding the recovery in the United global bank write-downs amounted to States and the newly industrialized economies $1.3 trillion through the first half of 2009, because the destocking during the acute phase further write-downs of some $1.5 trillion of the crisis was particularly strong in these may be required as U.S.-domiciled banks economies. In Europe, although slower inven- have recognized only about 60 percent of tory accumulation had acted as a drag on anticipated write-downs. growth, inventories continue to build up, In high-income Europe, GDP is expected to decline by 3.9 percent in 2009 and to increase by only 1.0 percent in 2010. The support from fiscal and monetary policy for domestic de- mand, as well as improving global demand is Figure 1.6 Change in stock building as a contribution to GDP growth in G-3 countries likely to support growth in the region. How- ever, ongoing balance-sheet problems of Euro Percentage point contribution to growth Area banks are likely to remain a drag on fi- 4 nancing conditions. So far, commercial banks 3 have made little use of the governments' rescue 2 packages, and governments have yet to amend 1 rescue plans. As a result lending restrictions 0 are likely to remain a drag for capital expen- 1 diture. According to the latest ECB Financial 2 Stability Review (2009) only two-thirds of po- 3 tential losses in major European banks have 4 been provisioned or written off so far, with United States Japan Euro Area some 187 billion euros of potential losses still 2008 Q4 2009 Q1 remaining. 2009 Q2 2009 Q3 Outturns in Germany have been key to Source: World Bank. developments in the Euro Area more gen- erally. The German economy grew at an 24 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S annualized pace of 2.9 percent in the third Prospects for developing quarter of 2009, with growth largely driven by corporate investment and construction, economies while private consumption waned. Looking forward, the strong recovery in foreign or- M ost developing countries were not di- rectly involved in the risky behaviors that precipitated the financial crisis, and the ders for manufactured goods suggests that banking systems in most regions carried only net exports will come to support growth. In limited exposure to subprime loans. Nonethe- addition, increasing public capital expendi- less, economic activity in virtually all countries tures will support activity in the second half were sharply affected. By the first quarter of of 2009 and most of 2010. However, ris- 2009, 25 of 31 developing countries (for which ing unemployment will be a drag on private quarterly national account data are available) consumption. had reported negative growth rates (figure 1.7). In France output in the third quarter ben- Domestic demand in developing countries efited from a rise in exports even as private was particularly affected by the sharp slowdown spending remained stagnant and fixed invest- in fixed investment growth, which fell from ment continued to fall. In general, France suf- 13.4 percent in 2007 to 8.5 percent in 2008 fered less than other rich nations, because it and to an estimated 1.3 percent in 2009. In re- was neither a large supplier of international sponse to falling domestic and external demand, credit nor reliant on borrowing, and when industrial production came under pressure. By private demand suddenly plunged, the French the end of the first quarter of 2009, industrial government stepped in. Although the recession production was down by 12.9 percent from in the United Kingdom has been deeper than its level a year earlier, the volume and value most initially expected, positive GDP growth of developing-country exports had declined by is expected to resume in the fourth quarter, 30.2 percent and 17.6 percent, respectively, thereby ending six consecutive quarters of fall- and the commodity prices that had supported ing GDP. growth in the boom years in many countries Japan's economy grew by a revised 1.3 per- had fallen sharply. Moreover, the freezing of cent (saar) in the third quarter. The recovery capital flows in high-income countries and results mainly from stimulus efforts at home and abroad. Tax incentives, as well as a re- ward program for purchasing green products, are encouraging Japanese consumers to switch Figure 1.7 Dispersion of GDP growth results in the first quarter of 2009 to low-emission cars and energy-saving home appliances. At the same time, export volumes 10 have benefited from increased demand for 8 Japanese products from abroad, such as cars 6 and related products in the United States and 4 electronic goods in China. Over the next few quarters, growth is fore- 2 cast to benefit from an end to the sharp inven- 0 tory liquidation in both domestic and overseas l A and un pi er un om D ci ia be d ib an co -inc EC Pa s co elo th tri ng fic tri e an a d tA v O si tra e ar a gh -O markets and the continued support emanating es es an Eas en p C ic C ro hi on e er Eu th Am N de from the stimulus packages. Growth may come tin La under renewed pressure later in 2010/11, as the effect of the stimulus package ebbs and the Negative Positive anticipated recovery in major trading partners Sources: World Bank. remains modest. 25 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 increased borrowing costs generated a huge East Asia and the Pacific $690 billion financing gap that had to be met by East Asian economies were less adversely af- reduced imports, layoffs, and in some instances fected by the crisis than other regions, although substantial injections of foreign capital through as a key durable- and investment-goods- official agencies such as the IMF, World Bank producing region it experienced dramatic and various regional development banks. declines in trade and production between As a result, GDP growth in developing coun- September 2008 and March 2009. tries decelerated sharply, coming in at only The direct fallout from the financial cri- 1.2 percent for the year as a whole. Developing sis in high-income countries was limited. Europe and Central Asia, which went into the Although equity markets declined steeply and crisis period with large current account deficits rapidly, the region's financial system held rela- due to a consumption boom financed by inter- tively few toxic assets and its overall resilience national credit and FDI, was hardest hit. GDP had been improved by banking reforms fol- there fell an estimated 6.2 percent in 2009. Ex- lowing the East Asian financial crisis of the cluding these countries and China and India, 1990s. which were able to weather the worst effects of Regional industrial production was declin- the turmoil through large fiscal and monetary ing at a 9 percent annualized pace toward stimulus packages, GDP in the remaining de- the end of 2008, but started recovering early veloping countries fell by an estimated 2.2 per- in 2009 under the influence of the 4 trillion cent in 2009--well below the 3 percent trend renminbi (12 percent of GDP) fiscal stimulus growth rate of these countries going into the package and monetary easing introduced by crisis. While overall developing-country growth the Chinese government. Regional exports remained positive, the deceleration and dislo- plummeted more sharply still, falling at a cation that it has caused has been brutal. Un- 50 percent annualized pace in the first quarter employment is rising, an additional 90 million of 2009. Since then export volumes have been people are expected to remain in poverty (less recovering, up 18 percent during the third than $1 a day) by the end of 2010 as a result of quarter. Beginning in March 2009, regional the slower growth, and as many as 30­50 thou- trade partners started to benefit from the sand additional children are expected to have surge in Chinese imports associated with its died of malnutrition in 2009 in Sub-Saharan fiscal stimulus (figure 1.8), and overall export Africa (Friedman and Schady 2009). volumes have been growing at a 10 percent Prospects for developing countries are for a annualized pace in recent months. relatively robust recovery in 2010, with growth Overall, GDP resisted the global recession of 5.2 percent in aggregate or 3.7 percent if to a fair degree, expanding by 6.8 percent in China, India, and Europe and Central Asia are 2009. Excluding China, the deceleration excluded. Output should strengthen further in in growth was sharper. GDP in these countries 2011, but only modestly, rising to 5.8 percent grew at an estimated 1.3 percent, down from for the developing aggregate as a whole and 4.8 percent in 2008. Regionwide fiscal and 4.1 percent for developing countries excluding monetary stimulus plus the weakness of ex- China, India, and Europe and Central Asia. ternal demand have raised the contribution of domestic demand to overall growth. Looking forward, the stabilization of inter- Regional outlooks national financial markets and renewed capital M ore detailed descriptions of prospects for developing regions, including country- specific projections, are available in the regional inflows, coupled with a strong inventory cycle, particularly among the newly industrialized econ- omies in the region, are expected to boost growth appendix to this report and online at http:// in 2010 to 8.1 percent, with China leading the www.worldbank.org/globaloutlook. recovery with growth of 9 percent. Continuing 26 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S regionwide. In Russia, GDP in the third quarter Figure 1.8 China's stimulus program yielded a pickup in import demand was 9.0 percent lower than a year earlier, as the drop in commodity prices (particularly oil) and Export values (US$) seasonally adjusted, annual percentage change the sudden reversal of capital flows led to a sub- 80 stantial decline in revenues and fixed invest- 60 ment. Industrial activity is now recovering, growing at a 9.3 percent annualized pace in the 40 three-month period ending November 2009. In 20 November, industrial output shifted to positive 0 annual growth of 1.3 percent following 12 con- 20 secutive months of decline over prior year lev- els. In Poland industrial production was more 40 robust and GDP growth remained positive, 60 supported by relatively resilient domestic de- 80 mand and recent cuts to income taxes and improved retirement benefits. Production in 09 08 08 08 09 09 20 20 20 20 20 20 Turkey, while down 1.8 percent as of Novem- ay p. n. ay n. p. Se M Ja Ja Se M ber 2009 from a year earlier, has also been resil- East Asian exports Japanese exports ient and is recovering. In the Baltic States, Chinese imports industrial production declined between 15 and Sources: Haver Analytics and World Bank. 20 percent. Among other smaller countries, such as Armenia, Kyrgyz Republic, and Tajiki- stan, the depth of the slowdown in Russia has excess capacity in manufacturing and only mod- cut into remittances, whose U.S. dollar value erate advances in world trade growth (in historic declined by between 15 percent and 33 percent context) will restrain GDP growth from acceler- in the first half of 2009. ating much faster than 8.2 percent in 2011. The recovery in the region is expected to remain weak, given the substantial adjust- Europe and Central Asia ment in domestic demand levels required and Preexisting vulnerabilities in developing Europe extensive financial weakness. Because of the and Central Asia, including large current ac- depth of the crisis, the number of nonper- count deficits, excessive reliance on foreign forming loans has increased substantially and capital to finance domestic consumption, and can be expected to continue to rise for some sizable fiscal deficits in some countries, ex- time. This, in combination with higher inter- posed the region to a particularly sharp adjust- est rates and weak international capital flows, ment when international sentiment reversed may dampen investment growth (figure 1.9). with the onset of the crisis. Overall, the recovery is projected to be ane- Faced with the dramatic tightening of exter- mic, with GDP growth of only 2.7 percent nal financing conditions, authorities responded in 2010 and 3.4 percent in 2011. Moreover, with a mix of domestic macroeconomic adjust- given substantial weaknesses and remaining ment initiatives and extensive resort to official needs for balance-sheet consolidation, sub- financing from the IMF, the World Bank, and stantial downside risks persist, including the the European Union to replenish foreign reserve possibility of a secondary regional financial holdings, support budget initiatives, and resist crisis or a double-dip recession. downward pressure on local currencies. Notwithstanding these efforts, the crisis hit Latin America and the Caribbean the region hardest of all developing regions. Owing to sound macroeconomic fundamen- GDP is estimated to have fallen 6.2 percent tals, the Latin America and Caribbean region 27 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Figure 1.9 Nonperforming loans rise across Figure 1.10 In Latin America EMBI-stripped much of Europe and Central Asia spreads retreat as investor confidence returns % share of nonperformng loans in all loans Basis points Belarus 2,000 Estonia 1,750 Bulgaria 1,500 Slovak Republic Bosnia 1,250 and Herzegovina 1,000 Czech Republic Hungary 750 Turkey 500 Poland 250 Croatia 0 08 8 8 09 09 09 09 9 9 Macedonia, FYR 00 00 00 00 20 20 20 20 20 .2 .2 .2 .2 Russian g. b. r. n. g. ct ec ct ec Federation Ap Au Fe Ju Au O O D D Moldova Montenegro Argentina Brazil Chile Serbia Colombia Dominican Republic Panama Latvia Latin America Mexico Uruguay Lithuania Peru El Salvador Ukraine Source: JP Morgan. 0 5 10 15 20 25 30 Note: Spreads measure the credit risk premium over U.S. Treasury Share of total loans (percent) bonds. A stripped spread is a better measure for these comparisons as it accounts for collateral. In the calculation of the stripped 2007 spread, the value of collateralized flows (if any) is "stripped" from 2009 (most recent observation) the bond. Source: IMF Global Financial Stability Report, October 2009 Update, and World Bank. all of the countries in the region having has been able to weather the crisis much better regained their pre-crisis exchange rate level than earlier ones. Indeed, risk spreads in the relative to the U.S. dollar. Equity markets too region have declined to near pre-crisis levels as have been recovering, and spreads on regional investor confidence returned (figure 1.10). As sovereigns have come down. Partly as a result, elsewhere, both industrial production and many countries in the region have successfully international trade volumes declined sharply accessed international capital markets. in the face of the abrupt contraction in global However, stronger capital inflows, in part demand. By October 2009 industrial produc- attributable to still high interest rate differen- tion was 5.3 percent below its August 2008 tials, have put upward pressure on real effective level, even as it has been recovering at a exchange rates in some countries. In the case 9.8 percent annualized pace in recent months. of Brazil the government levied a 2 percent fi- Similarly, the volume of regional exports fell nancial transaction tax on foreign portfolio in- by 25 percent in the first several months of the flows, but that action may not be able to stem crisis and is only now beginning to recover. capital inflows or the real appreciation of the Falling commodity prices meant that the value real. Some economies may be faced with appre- of exports fell even more sharply, contribut- ciating real currencies, because of a weaker U.S. ing to lower incomes in many countries in the dollar, higher commodity prices, and strong region. capital inflows, losing external competitiveness Notwithstanding these circumstances, ex- at a time when external demand recovery is still change rates have held up well, with virtually fragile. 28 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S Within the region, the Mexican economy lag the overall recovery in output. As a re- suffered the deepest contraction, with quar- sult, regional GDP is projected to increase by terly GDP down 9.7 and 6.3 percent in the 3.1 percent in 2010, and 3.6 percent is ex- second and third quarters of 2009. The depth pected in 2011. of the fall in activity reflects Mexico's close Key challenges facing the region include ties to the U.S economy and its specialization winding down of monetary and fiscal stimu- in those sectors most deeply affected by the lus without undermining the recovery, pro- crisis (construction, automotive manufactur- viding adequately for the unemployed in a ing, and electric appliances). Moreover, it is fiscally sustainable manner, and maintaining estimated that the outbreak of the AH1N1 an open attitude toward international trade flu virus, which caused declines in air trans- and investment. port volumes of 80 percent in some months and hotel vacancies in tourist areas of 80 or more percent, shaved 0.7 percentage point off The Middle East and North Africa GDP. The developing economies of the Middle East In Brazil, GDP fell by 0.2 percent year-on- and North Africa region were adversely af- year in the first two quarters of the crisis period, fected by the crisis to varying degrees. At the but rebounded in the second and third quarter onset of crisis, equity markets among the high- of 2009. A robust fiscal policy package, includ- income Gulf Cooperation Council (GCC) ing support for the automotive sector and a re- economies and several bourses in the develop- versal of the inventory cycle, boosted industrial ing region plummeted--by more than the av- production, which was strong at a 22.2 percent erage for emerging markets. Recovery in these annualized pace in October 2009. At the same markets has been hesitant, given uncertainties time, lower policy interest rates and a decline surrounding financial conditions in Dubai and in interest rate spreads helped prompt a recov- the United Arab Emirates that have played a ery in private credit that has bolstered domestic major role in funneling FDI into the develop- demand. ing region. In Argentina, GDP increased by 0.5 and GDP growth in 2009 for the developing 0.2 percent on an annualized basis in the sec- economies is estimated to have eased from ond and third quarters of 2009. Policy-related 4.3 percent in 2008 to 2.9 percent. Despite uncertainties, including export restrictions, continuing large infrastructure development contributed to a sharp 17.3 percent decline in programs, the growth rate of developing oil industrial production and trade. The pace of exporters effectively halved from 2.9 percent recovery has also been held back by a severe in 2008 to 1.6 percent in 2009, mainly be- drought, which caused agricultural output to cause of oil production cutbacks to support plunge. OPEC (Organization of Petroleum-Exporting Output in the region is expected to Countries) price floors (figure 1.11). continue strengthening into 2010. Industrial Growth for the diversified economies fal- production is currently growing at a 22.2 per- tered by almost 2 percentage points in the year, cent annualized pace in Brazil and the con- from a strong 6.6 percent outturn in 2008 traction in Mexico is beginning to moderate. (powered by growth of more than 7 percent This should be further supported by ongoing in the Arab Republic of Egypt), to 4.7 percent. fiscal stimuli, the lagged benefits of strong The virtual collapse of key export markets (no- monetary policy support, a shift in the inven- tably the Euro Area), yielded sharp declines in tory cycle and improvements in the terms of goods exports from countries such as Egypt, trade. But many of the smaller countries in Jordan, Morocco, and Tunisia. At the same Central America, which are highly dependent time, remittances in 2009 dropped by 6.3 per- on migrant workers' remittances are likely to cent and tourism revenues by 5 percent--both 29 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Figure 1.11 Lower oil prices and production South Asia yield sharp decline in oil revenues during 2009 South Asia appears to have escaped the worst effects of the crisis, with GDP growth in the Oil revenues, US$ billion 800 region estimated at 6.0 percent in 2009, down from 6.9 percent in 2008. The slowdown in 600 GDP growth mainly reflected weaker invest- ment and private sector demand, which were only partially offset by an increase in public 400 expenditures. Several countries (Maldives, Pakistan, and Sri Lanka) faced serious chal- 200 lenges financing large current account deficits. Despite enduring a 5 percent decline in goods 0 and services export volumes, an even sharper 2004 2005 2006 2007 2008 2009 decline in import demand (partly explained by Other countries in region Other GCC weaker investment demand) and lower food and Saudi Arabia oil prices meant that trade and current account Sources: World Bank and International Energy Agency. balances improved in 2009 (figure 1.12). Re- Note: Revenues are for oil exporters in the Middle East and mittance flows to the region, which equal some North Africa region. 4.7 percent of GDP, fell an estimated 1.8 per- cent, representing a significant drop in house- hold incomes and foreign currency earnings. important sources of foreign income that sup- The region experienced sizeable capital out- port household consumption and job creation flows at the onset of the crisis, particularly in for these countries. Pakistan and Sri Lanka, where outflows were Looking forward, economic recovery will driven by investor concerns about rising do- depend on global demand for oil and gas, mestic and external imbalances. Improved which may not gain momentum until late in investor sentiment, particularly related to the forecast period. On balance, growth among relatively strong growth outturns (India) and developing countries in the region is antici- ongoing or new IMF stabilization programs pated to move moderately higher, to 3.7 per- cent during 2010, before firmer recovery sets in the following year. Nonetheless GDP gains of 4.4 percent in 2011 will remain below the Figure 1.12 South Asia's external position 5 percent growth attained earlier in the 2000s. improves in most countries on lower oil prices and decline in domestic demand The recent difficulties of Dubai World hold- ing company, an entity of the government of Current account balances as percent of GDP 6 Dubai, indicate that institutions in the region 4 were not entirely unaffected by the global fi- 2 nancial crisis. Given the very high investment 0 levels of the past several years, as well as asset 2 inflation (property prices increased particularly 4 sharply in Egypt and Morocco), there may be 6 large-scale financial losses within the region 8 that have yet to be realized. Should these mate- 10 Pakistan Sri Lanka India Bangladesh rialize, they could adversely affect market con- fidence, financial conditions, and employment 2008 2009 and investment in the region, to the detriment Source: World Bank. of medium-term growth prospects. 30 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S (Pakistan and Sri Lanka) as well as improved political stability (with the ending of the Figure 1.13 Quarterly GDP data point to output stabilization in Sub-Saharan Africa decades-old civil war in Sri Lanka) led to re- newed capital inflows during the second and Percent change (saar) third quarters of 2009. Partly as a result, ex- 20 change rates and local equity markets have re- 15 gained strength as both domestic and foreign 10 investors have become less risk averse. 5 The recovery in the region is expected to be 0 less marked than elsewhere, partly reflecting the 5 relative shallowness of the downturn. Growth 10 is expected to rebound to 7.0 and 7.4 percent 15 in 2010 and 2011, respectively, compared with South Africa Kenya 6.0 percent in 2009. 2008 Q1 2008 Q2 2008 Q3 2008 Q4 2009 Q1 2009 Q2 Sub-Saharan Africa 2009 Q3 The collapse of global trade slowed growth in Source: Haver Analytics. Sub-Saharan Africa markedly, to 1.1 percent in 2009 from an average of 5 percent in the preceding five years. Initially, as global capi- tal inflows reversed, the impact of the global crisis was felt most acutely by countries such sector is likely to be a drag on Nigeria's growth as South Africa, whose financial markets are in the near future. more integrated into global financial markets. The sharp decline in oil prices that accom- Subsequently, as trade collapsed, the impact panied the recession has caused current ac- spread to oil exporters (such as Angola) and count balances in the region's oil exporters to commodity exporters (such as Botswana and fall (by more than 10 percent of GDP in coun- Zambia). Lower tourism volumes, falling re- tries such as Angola, Gabon, and Nigeria). mittances, and lower levels of official devel- The recovery is expected to be modest, opment assistance also affected the region with GDP expanding by 3.8 and 4.6 percent adversely. Overall, GDP growth is estimated to in 2010 and 2011, respectively. However, the have decelerated by 4 percentage points, and outlook is very uncertain and the strength of gross national income (a measure that includes the recovery will depend to a large extent on terms-of-trade effects) fell by 3.7 percentage growth performance in key export markets. points between 2008 and 2009. A growth rebound there should translate into South Africa recorded three consecutive stronger external demand, while also trigger- quarters of negative growth, with output de- ing a recovery in FDI flows. Incomes in coun- clining at a 2.8 percent annualized pace in the tries dependent on workers' remittances are second quarter of 2009. But South African expected to remain subdued, largely owing to growth rebounded to positive territory in the continued high unemployment in the United third quarter, rising 0.9 percent, while in Kenya States and the European Union. GDP increased by 5.8 percent (saar) in the sec- ond quarter of 2009, suggesting that recovery is under way (figure 1.13). Nigeria's growth Commodity markets performance has been helped by a relatively strong performance in the agriculture sector as well as in gas and oil. Given the high level of A n extended period of strong developing- country growth, coupled with specific supply-side factors, resulted in uncomfortably nonperforming loans, however, the financial low levels of spare capacity in the oil sector 31 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 and low stocks of many grains and limited Figure 1.14 Real commodity price indexes stockpiles of many traded metals in the mid- 2000s (see World Bank 2009b for an in-depth Constant 2000 prices, trade-weighted and adjusted for currency fluctuations and inflation differentials (Jan. 2000 = 100) discussion of the commodity boom and bust). 350 These fundamental factors were also sup- 300 ported by financial investments associated 250 with the global liquidity boom (see chapter 2), 200 which may well have exacerbated the effect 150 of the tight supply situation on commodity 100 prices. Consequently, between 2003 and 2008 50 real non-energy commodity prices doubled, 0 while real energy prices rose 170 percent. Jan. Jul. Jan. Jul. Jan. Jul. Jan. 2000 2001 2003 2004 2006 2007 2009 Although commodity prices began fall- ing before the onset of the acute phase of the Energy Metals and minerals Agriculture financial crisis, both the financial contrac- Source: World Bank. tion associated with the crisis itself and the spectacular contraction in economic activity that it provoked generated a sharp decline in global demand for commodities. Between quarter of 2008 and the first quarter of 2009--a July 2008 and February 2009, the U.S. dol- result of reduced economic activity and induced lar price of energy plummeted by two-thirds, conservation and substitution toward other en- and that of metals dropped by more than ergy sources in response to high oil prices. Oil 50 percent, from earlier highs. Dollar prices demand in OECD countries began declining of agricultural goods retreated by more than in the fourth quarter of 2005 (when oil prices 30 percent, with the prices of fats and oils surged above $50 a barrel) and has been fall- dropping 42 percent. ing for more than four years now, with little Dollar prices of energy and metals commod- or no growth expected in 2010. Demand in ities began to recover in March 2009 broadly non-OECD countries also declined in the first in tandem with global economic activity. The quarter of 2009, but has since increased and price increases partly reflect the depreciation is projected to resume its trend growth rate in of the dollar that has since reversed almost all 2010. of the appreciation that was associated with OPEC responded to the fall in global de- the immediate flight of capital from the rest of mand by reducing its production by nearly the world to the United States. Indeed, the real 4 million barrels a day in an effort to main- local-currency price of international commod- tain prices at around $75 a barrel. As a result, ities (a measure that corrects for currency fluc- OPEC spare capacity--one measure of global tuations and inflation differentials) increased slack in oil markets--has increased to around much less than dollar prices (figure 1.14). For 6.5 million barrels a day, equivalent to about instance, although energy prices measured in five years of demand growth, and roughly the U.S. dollars rose 57 percent between February same level as in 2003 when oil prices were $20 and October 2009, the increase over the same a barrel (figure 1.15). Moreover, inventories period in trade-weighted real local-currency of already extracted oil and oil products re- terms was only 33 percent. main very high with some 150 million barrels currently being stored on ships at sea owing Crude oil to weak demand and saturation of some land- World oil demand, which grew on average by based storage facilities. 1.7 percent a year over the 2000­2007 period, While immediate-term supply is ample, the declined by nearly 3 percent during the last longer-term prospects are more clouded. Over 32 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S Figure 1.15 OPEC spare capacity Million barrels a day 7 6 5 4 3 2 1 0 Jan. 2001 Jan. 2002 Jan. 2003 Jan. 2004 Jan. 2005 Jan. 2006 Jan. 2007 Jan. 2008 Jan. 2009 Jan. 2010 Sources: International Energy Agency; World Bank calculations. the past decades non-OPEC supply (outside the estimates suggest that at current real oil prices, former Soviet Union where output rose strongly demand and supply should remain in balance in the early 2000s) has been fairly stagnant, with for the foreseeable future. increased production in Brazil, Canada, and West Africa offset by large declines in U.S. and North Metals Sea output. Although much higher prices now The global recession prompted a sharp decline have prompted increased investment, growth in demand for metals. During the first half of from new developments has been sluggish, 2009, global consumption of aluminum and partly because of high costs in 2007­08 caused copper, the most important metals in terms by shortages of equipment and skilled labor, of volume, fell by 19 percent and 11 percent, and because of numerous project delays. More- respectively. Restocking by Chinese compa- over, some three-fourths of known reserves are nies (China is the world's largest consumer of in the control of national oil companies (OPEC- metals) and the government's State Reserves controlled or otherwise), which forces major in- Bureau resulted in strong demand growth in ternational oil companies to invest in higher-cost the first half of the year, but during the second developments (such as oil sands and deepwater), half of the year the restocking waned, and a increasing their costs and the amount of lead time similar restocking in industrial countries has needed before projects come on stream. yet to materialize. As a result, global demand Given the large inventory overhang and for aluminum and copper in 2009 is estimated the modest increases in oil demand expected to have declined by 11 percent and 9 percent over the next few years, real oil prices are not respectively from 2007 peaks, with world expected to rise substantially. However, the demand outside China down by more than sector remains sensitive to both demand and 20 percent for both metals. supply developments, and a significant disrup- On the supply side, cutbacks at mines and tion to global supply could result in a sharp, if smelters were significant early in the downturn temporary, rise in prices once again. of the cycle. In addition, project cancellations, Unless significant additional reserves are tight scrap markets, and numerous strikes (in discovered over the longer term, OPEC's Canada and South America, for example) have pricing power will continue to increase. Ulti- helped tighten markets. Over the next two mately, however, alternative energy sources years, metals prices are expected to continue such as coal, natural gas, nuclear power, and to rise moderately as the global recovery pro- various renewables are likely to prevent real gresses and metal demand expands. Prices are, oil prices from rising without end. Industry however, not expected to rise substantially, 33 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 partly because of the large price appreciation Barring unforeseen production problems, to date, but mainly because of substantial idle agricultural markets are likely to remain well capacity in many sectors that can be profit- supplied. As a result, agricultural prices are ably brought back into production at current projected to decline by 13.8% in 2009, com- prices. Once demand growth returns to trend pared with 2008. Over the medium- to lon- and idle capacity is eliminated, the industry ger terms, agriculture prices are expected to will again be challenged to add sufficient ca- remain broadly stable in real terms, reflect- pacity in the face of strong growth in develop- ing two opposing forces. On the one hand, a ing countries--partly because new mines will stronger link between energy and agricultural be more expensive (underground versus open prices (higher costs of production plus de- pit, for example) and often in geopolitically mand for biofuel) will exert upward pressure difficult regions. The mining industry will also on prices; on the other hand, continued gains have to contend with declining ore grades, en- in total factor productivity (which tends to be vironmental and land rehabilitation, as well stronger in agriculture than in manufacturing) as water, energy, and labor pressures. How- should constrain production costs. ever, metals prices are not expected to reach Short-term food security concerns have the nominal peaks attained earlier this decade subsided, and most countries have reduced or over the forecast period. eliminated the export bans and other export restrictions that were put in place during the Agriculture commodity price spike of 2008. However, the Although agricultural prices have declined poverty challenges posed by higher food prices by 22 percent since their peak in June 2008, remain. Over the longer term, productivity gains they nevertheless remain almost twice as high at the global level should ensure long-term food as the lows reached in the early 2000s. The supply. However, advances in agricultural pro- recent fall in agricultural prices (relative to ductivity in many poor countries is not keeping previous peaks) reflects lower oil prices--a pace with population growth. As a result, there key cost component--and larger stockpiles of is a rising risk of increasing dependence on im- key agricultural commodities, including rice, ported food to meet basic needs. For example, maize, and wheat (figure 1.16), resulting from between 1980 and 2004, per capita agricultural favorable harvests and area expansion of key GDP in Sub-Saharan Africa grew by less than agricultural commodities. 1 percent a year (versus more than 3 percent per year in East Asia). Prospects Figure 1.16 Global stock-to-use ratio of key Over the medium term, real commodity prices agricultural markets (excluding China) are projected to remain relatively stable, with Stock-to-use ratio (percent) up- and downside risks more or less in bal- 35 ance (figure 1.17). Recent price rises reflect 30 dollar weakness and some overshooting asso- 25 ciated with the slowdown in global economic 20 activity. Long term there is some concern that non-industrial commodities may become more 15 procyclical and volatile than in the past. If the 10 influence of financial investors in commodity 5 markets rises, then the procyclical nature of 08 60 64 68 72 76 80 84 88 92 96 00 04 their activity could raise volatility in affected 20 19 19 19 19 19 19 19 19 19 19 20 20 Sources: World Bank calculations based on U.S. Agriculture Department data. markets. Similarly, the use of agricultural prod- ucts as an alternative fuel source may introduce 34 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S Figure 1.17 Real commodity prices Figure 1.18 Inflation in low-, middle-, and high-income countries Constant 2000$--manufacturing unit value (MUV) deflator 350 Percent 300 20 250 15 200 10 150 5 100 50 0 0 ­5 1960 1970 1980 1990 2000 2010 2020 06 06 6 7 07 08 08 08 09 9 00 00 00 20 20 20 20 20 20 20 2 2 .2 n. n. v. r. p. b. l. c. ay ct Metals Agriculture Ju Ap Energy No De Ja Ju Se Fe M O Source: World Bank. Low-income countries Middle-income countries Note: MUV deflator is the unit value index in U.S. dollar terms High-income countries of manufactures exported from the G -5 countries (France, Germany, Japan, United Kingdom, and the United States), weighted proportionally to the countries' exports to developing Source: World Bank. countries. an element of cyclicality into some food prices Figure 1.19 Core inflation in high-income that was not previously there. countries Percent Inflation 3 J ust as the sharp rise in food and fuel prices generated a rapid acceleration of headline inflation in both high-income and developing 2 1 countries during 2008, the fall of commodity 0 prices during the course of 2009 and the un- Jan. Jun. Nov. Apr. Sep. Feb. Jul. Dec. May Oct. precedented slowdown in the global economy 2006 2006 2006 2007 2007 2008 2008 2008 2009 2009 has led to a dramatic fall in headline inflation Euro Area United States (figure 1.18). The median rate of year-over-year consumer price inflation in high-income coun- Source: World Bank. tries, which peaked at 5.2 percent in mid-2008, turned negative in July, but was 0.6 percent in November 2009. The median inflation rate in developing countries has declined from a peak and Group of Seven countries is expected to of 12.4 percent in mid-2008 to only 2.6 per- average 1.6 percent and 1.1 percent in 2010, cent. Notwithstanding the declines in headline respectively. inflation, core inflation has remained relatively Inflation developments have changed stable in high-income countries (figure 1.19). drastically among middle- and low-income Only in Japan has core inflation dropped countries. Median inflation in low-income below zero. The bulk of the commodity price countries peaked at 15.4 percent in the middle deflation has now passed through the system; of 2008, but as of October 2009 it was therefore headline inflation can be expected 1.2 percent--well below the levels observed to rise toward core inflation rates in coming before the food and fuel boom. However, food months. Headline inflation in the United States inflation in developing countries has not been 35 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 benefiting from lower international food Figure 1.20 Food prices in low-income prices to the same degree as the poor in richer countries countries and that a significant portion of the Index, 2003 = 1.0 130 million pushed into extreme poverty dur- 1.12 ing the food-price spike (World Bank 2009b) 1.10 may not have exited poverty as might have 1.08 been expected given the fall in international 1.06 food prices. 1.04 1.02 1.00 0.98 World trade Jan. Sep. May Jan. Sep. May Jan. Sep. May Jan. 2003 2003 2004 2005 2005 2006 2007 2007 2008 2009 Low-income countries Limited sample I n general, global trade has followed a broadly similar pattern similar to industrial production, albeit the fall was deeper and the recovery lagged somewhat. The dollar value Sources: World Bank with data from the International Labour Organization (ILO). of world trade plummeted 31 percent between August 2008 and its low point in March 2009 (figure 1.21). The decline in volume terms was falling as rapidly as overall prices in the two- somewhat less pronounced when falling com- thirds of developing countries for which data modity prices and exchange rate fluctuations are available through May 2009 (figure 1.20). were taken out of the equation; nevertheless, As a result, by the end of May 2009, food by March 2009 global trade volumes were prices in developing countries had risen about down by 22 percent. Although global trade 8 percent faster than nonfood prices, when has recovered from these troughs, as of compared with January 2003. This suggests October 2009 it was still 2.8 percent below its that the poor in these countries may not be pre-crisis level. Figure 1.21 World trade is recovering a. Annualized quarterly increase in import volumes b. Annualized quarterly increase in export volumes Percent growth Percent growth 100 80 80 60 60 40 40 20 20 0 0 20 20 40 40 60 60 80 80 08 9 08 09 08 8 9 9 08 8 9 9 08 09 8 09 00 00 00 00 00 00 00 00 20 20 20 20 20 20 20 20 .2 .2 .2 .2 .2 2 .2 .2 n. n. n. r. r. r. l. l. l. n ct r ct ct ct l Ap Ap Ap Ap Ju Ju Ju Ju Ja Ja Ja Ja O O O O High-income countries East Asia and Pacific High-income countries East Asia and Pacific Other developing countries Other developing countries Source: World Bank. 36 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S The lag in the trade rebound does not appear to be wholly a reflection of weak trade finance Figure 1.22 International tourist arrivals (although doubtless it has played some role). Percent change in volumes, year over year Rather, the lag appears to reflect the still- 10 depressed level of investment activity (invest- 5 ment goods generally are heavily traded). Global investment fell by an estimated 9.7 percent in 0 2009 and even in 2010 investment is forecast to grow by only 4.9 percent. 5 The initial fall-off in import volumes was 10 relatively stronger in high-income coun- tries, partly reflecting the growth slowdown 15 that had already begun before the failure of 2008 Q1 2008 Q2 2008 Q3 2008 Q4 2009 Q1 2009 Q2 Lehman Brothers. With the crisis, the decline Source: World Tourism Organization. accentuated and broadened, with global im- port volumes falling at a 40 percent annual- ized pace in the first quarter of 2009. At the trough, imports in high-income countries were the second quarter, where the effects of the 24 percent off their August 2008 level; in de- global recession were magnified by the out- veloping countries they were also down by break of AH1N1 in that country and efforts 25 percent. by individuals worldwide to avoid infec- The trade slump was less marked in Asian tion. Most recently, global tourism arrivals countries, in part because of fiscal stimulus in appear to be picking up, with July volumes China. Most Chinese trade partners benefited only 4 percent lower than a year earlier. from the rebound in Chinese imports. By the Notwithstanding widespread efforts to sup- third quarter, import demand had strength- port tourism through special tax deductions, ened among most countries. After a period of the easing of visa restrictions, and invest- some weakness, reflecting faltering domestic ment plans, the World Tourism Organiza- demand, the United States' import volume tion expects global tourist volumes to have growth jumped to 29 percent in October declined by between 4 and 6 percent during (saar), in Germany to 27 percent, and in Japan 2009. to 31 percent as of November. Overall, world merchandise trade volumes In general, services trade has been more are estimated to have contracted by 17.6 per- resilient than merchandise trade, in part be- cent in 2009, with goods and services down cause a larger share is destined for personal some 14.4 percent. Given the expected weak consumption rather than investment expen- recovery and weak base effects, trade is pro- ditures. Tourism represents something of jected to expand by only 4.3 percent in 2010 an exception--such expenditures tend to be and by 6.2 percent in 2011. As a result even luxury goods and therefore more volatile. two years into the recovery, the overall volume The World Tourism Organization reports of goods and services traded is forecast to be that compared with 2008, tourism arrivals 5 percent lower than its 2008 peak. were off 7 percent in the first six months of Remittances are another important source 2009 (figure 1.22). Regionally, Central and of external currency for developing countries, Eastern European nations recording the larg- representing as much as 20 percent of GDP in est fall in tourism (11 percent), while Africa some countries. Remittances have been more registered a modest increase in tourist arriv- stable than capital flows and merchandise trade, als. Mexico was particularly hard hit, with but have nevertheless declined by an estimated arrivals down 19 percent (year-over-year) in 6.1 percent in 2009 (box 1.1). 37 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Box 1.1 Prospects for remittances O fficially recorded remittance flows to develop- ing countries reached a peak of $338 billion in 2008, up 16.7 percent from 2007, despite a slow- in flows ( 15 percent), while remittances to South Asia are expected to drop by a more modest 2 per- cent. Remittance flows to South Asia grew strongly down in several remittance corridors during the final in 2008 despite the global economic crisis, but now quarter of 2008. Based on monthly and quarterly there are risks that they may slow in a lagged re- data released by some central banks and in line with sponse to a weak global economy. East Asia and Sub- the World Bank's global economic outlook, it is esti- Saharan Africa also face similar risks. In contrast, mated that remittance flows to developing countries remittance flows to Latin America and the Caribbean will fall by 6.1 percent in 2009, before recovering and to the Middle East and North Africa have been in 2010. Weaknesses in the U.S. job market led to a weaker than expected in 2009; however, they appear significant decline in recorded remittances to Latin to have reached a bottom already, with the expecta- America, particularly Jamaica and Mexico, where tion of a recovery in 2010 and 2011. year-to-date remittances fell by 16 and 13 percent, Overall, migration and remittance flows are respectively. In contrast, remittances to some South expected to recover in 2010 and 2011, but the re- Asian countries such as Bangladesh and Pakistan covery is likely to be shallow, with remittances not have continued to record positive growth in 2009, expected to reach the level of 2008 even in 2011. In where year-to-date remittances have increased 16 all developing regions, remittance flows are likely percent and 27 percent respectively. The latter is in to face three downside risks: a jobless economic re- part a result of measures by the Pakistani authorities covery, tighter immigration controls, and unpredict- to increase flows through formal channels includ- able exchange rate movements. Despite these risks, ing subsidies for marketing expenses to providers remittances are expected to remain more resilient of remittance services (see the Pakistan Remittance than private capital flows and will become even more Initiative); in addition, some migrants are returning important as a source of external financing in many with accumulated savings.a developing countries. Policy responses could involve However, in all regions, remittance flows are ex- efforts to facilitate migration and remittances to pected to weaken from 2008 levels, with Europe and make these flows cheaper, safer and more productive Central Asia likely to record the largest deterioration for both the sending and the receiving countries.b a. See http://www.pri.gov.pk/. b. Ratha 2009. 3.9 percent in 2009. Given continued relatively Narrowing global external large gaps in global output and the absence of payment imbalances strong excess demand pressures (particularly I n aggregate, the crisis has prompted a nar- rowing of global imbalances attributable to an overall decline in the volume of trade (for in high-income countries)--which should keep oil prices in check--this narrowing is projected to be broadly stable, with imbalances rising a given percentage imbalance between im- only slightly to around 4.1 percent of GDP in ports and exports, weaker trade will reduce 2011 (figure 1.23). the global imbalance), falling oil prices, and The decline in the U.S. current account a narrowing of China's and the United States' deficit reflects the continuation of a preexist- trade imbalances. Overall, the absolute value ing narrowing of the non-oil trade balance of global current account balances (the sum of that was masked by rising oil prices. It also all surpluses plus the absolute value of all defi- reflects a sharp increase in household savings cits) is estimated to have declined from a peak rates as the wealth effect of the housing bub- of 5.9 percent of world GDP in 2008 to around ble ceased to support high levels of consumer 38 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S the fiscal and monetary stimuli in both of these Figure 1.23 An easing of global imbalances countries are unwound. In the case of the United Percent of world GDP States, if household savings rates fall as the re- 7 covery takes hold, but public sector spending is 6 not cut back, the trade deficit can be expected to 5 rise once again. Although growth in the United 4 States is still in the early stages of recovery, re- 3 cent improvements in the current account have 2 been maintained in the third quarter of 2009, 1 as the marginal increase in the current account 0 deficit from 2.8 to 3 percent of GDP is largely 00 01 02 03 04 05 06 07 08 09 10 11 related to higher oil import costs. 20 20 20 20 20 20 20 20 20 20 20 20 China United States Rest of world In China, the success of the authorities in stimulating domestic demand will determine Source: World Bank. whether or not its trade surplus begins to rise as world trade recovers. So far, most of the stimulus has gone into additional infrastruc- demand. The most recent small increase in ture investment. If, as some fear, this merely the U.S. trade deficit mainly reflects high oil bolsters the economy's export competitiveness prices as public sector deficits have only par- without promoting an expansion of domestic tially offset the increase in consumer saving. spending, then as global trade revives, China's The sharp narrowing of the Chinese surplus trade surplus could reemerge. While additional is more directly related to the drop in global exchange rate flexibility could help increase the trade and China's large fiscal package, which attractiveness of domestic markets for Chinese has bolstered imports at a time that global ex- producers, such a move is unlikely to eliminate port demand was weaker (figure 1.24). China's tendency toward large trade surpluses Whether the narrowing of these trade bal- unless it is accompanied by structural changes ances persists will depend importantly on how to decrease household and firm savings rates. In the baseline forecast, a further modest unwinding of global imbalances is projected Figure 1.24 U.S.-China imbalances have diminished markedly over the medium term. China's current ac- count surplus is expected to decline from an Total goods trade balance, US$ billion, saar 450 estimated 5.6 percent of GDP in 2009 to 4 per- 250 cent of GDP by 2011, while the U.S. current 50 account deficit is projected to rise only 150 marginally from an estimated 2.9 percent in 350 2009 to 3.1 percent of GDP by 2011. 550 Overall, the extent of global imbalances, mea- 750 sured as the absolute value of current account 950 balances as a percent of world GDP, is pro- 1,150 jected to decline from its peak of 5.9 percent in Jan. Aug. Mar. Oct. May. Jan. Aug. 2008 to 4.5 percent in 2011. 2000 2001 2003 2004 2006 2008 2009 China excluding oil China total balance Uncertain prospects A USA excluding oil USA total balance s emphasized above the economic re- Source: World Bank. bound that is currently under way is likely to continue for several months, supporting 39 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 relatively rapid growth. However, a great deal Table 1.2 Prospects remain uncertain of uncertainty clouds the outlook for the second Real GDP growth half of 2010 and beyond. The waning growth Region 2008 2009 2010 2011 impact of the fiscal stimulus, a progressive end to the inventory cycle, uncertainty about Baseline scenario World 1.7 2.2 2.7 3.2 the extent to which private sector confidence High-income countries 0.4 3.3 1.8 2.3 will step in and sustain the recovery, and the Developing countries 5.6 1.2 5.2 5.8 possibility of a second round of bank failures East Asia and Pacific 8.0 6.8 8.1 8.2 Europe and Central Asia 4.2 6.2 2.7 3.6 either in developed or developing countries are Latin America and the Caribbean 3.9 2.6 3.1 3.6 among the factors that could contribute to a Middle East and North Africa 4.3 2.9 3.7 4.4 more pronounced slowdown of growth in the South Asia 5.7 5.7 6.9 7.4 Sub-Saharan Africa 5.1 1.1 3.8 4.6 second half of 2010 and into 2011--potentially yielding a double-dip growth recession. Deeper recession scenario World 1.7 2.2 2.5 2.7 On the upside, if private sector confidence High-income countries 0.4 3.3 1.6 1.8 does return, there is a risk that the huge tra- Developing countries 5.6 1.2 5.1 5.4 ditional and nontraditional monetary stimu- East Asia and Pacific 8.0 6.8 7.9 7.5 Europe and Central Asia 4.2 6.2 2.6 3.2 lus that has been put into place will begin to Latin America and the Caribbean 3.9 2.6 3.0 3.2 gain traction, potentially reflating some of the Middle East and North Africa 4.3 2.9 3.7 4.4 bubbles that have only recently burst. Indeed, South Asia 5.7 5.7 6.9 7.3 Sub-Saharan Africa 5.1 1.1 3.8 4.4 some (Roubini 2009) are already arguing that very loose monetary policy in high-income Stronger growth scenario World 1.7 2.2 3.1 3.4 countries has produced a carry-trade oppor- High-income countries 0.4 3.3 2.2 2.4 tunity that is underpinning in an unsustain- Developing countries 5.6 1.2 5.8 6.3 able manner the resurgence of capital flows East Asia and Pacific 8.0 6.8 9.0 8.8 Europe and Central Asia 4.2 6.2 3.1 4.0 to developing countries, which may ultimately Latin America and the Caribbean 3.9 2.6 3.6 4.2 regenerate the kind of global imbalances that Middle East and North Africa 4.3 2.9 4.1 4.6 precipitated the crisis in the first place. South Asia 5.7 5.7 7.4 7.6 Sub-Saharan Africa 5.1 1.1 4.0 5.1 The following pages address some of these issues and present simulations designed to il- Source: World Bank. lustrate potential alternative outcomes should these pressures, which exist in the baseline, combination of a modest tightening of mon- hold greater or lesser sway in the months to etary and fiscal policy and ongoing restructur- come. These alternative scenarios are not ing in the banking sector causes investment meant to quantify the result of worst- or best- growth to be more subdued than in the base- case scenarios, but to illustrate a reasonable line. The results in the second panel of the table range of possible outcomes given the uncer- are based on a simulation that assumes that tainties prevailing today. 10 percent of the 2009 increase in structural deficits is withdrawn4 in 2010 and a further 20 percent in 2011 and that the rebound in in- A deeper growth recession vestment is only 80 percent as strong as in the Table 1.2 reports the results of two scenarios. baseline. The first outlines growth prospects for GDP In this scenario, global growth comes in under the assumption that the progressive im- about 0.2 percentage points lower in 2010 provement in global financial markets implicit and 0.5 percentage points lower in 2011 in the baseline scenario is not as strong as than in the baseline forecast. Growth rates projected and is exacerbated by the beginning in high-income countries decline by 0.2 and of efforts to unwind the fiscal and monetary 0.5 percentage points in 2010 and 2011, stimulus already in place. In this scenario, the respectively. This is more than the 0.1 and 40 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S 0.4 percentage points decline projected for 0.6 percentage points lower. Lower savings in developing countries, mainly because the the United States serve to push up its current fiscal stimulus in high-income countries was account deficit by about 0.2 percent of GDP. much bigger than in developing countries. Indeed in many developing countries financial constraints precluded any real countercyclical The impact of the crisis on the increase in spending. very poor Global imbalances narrow in this scenario mainly because of weaker consumer demand in high-income countries, and disinflation, T he financial crisis has taken its toll on achieving the 2015 poverty Millennium De- velopment Goal (MDG). Newly updated World unemployment and high output gaps become Bank estimates suggest that the crisis will leave even more pronounced problems. an additional 50 million people in extreme pov- erty in 2009 and some 64 million by the end A more buoyant private sector of 2010 relative to a no-crisis scenario.6 These reaction depressing statistics notwithstanding, the rela- The reaction of the private sector to the recov- tively rapid rebound in developing countries, ery is one of the major uncertainties underly- their future medium term prospects as described ing the outlook. In the baseline scenario, the in the first part of this chapter combined with negative wealth effect from the crash plus the the significant progress in most regions since indebtedness incurred during the boom period 1990, the poverty MDG is likely to be met at are expected to dampen consumer demand for the global level. several years. In addition, the weakened bank- The current projection of the percent- ing sector is not expected to be able to support age of developing-country population living the kind of investment rebound that normally on $1.25/day or less (a standard measure of follows a serious recession. poverty) in 2015 is 15 percent (table 1.3), well However, with monetary policy as loose as below the target rate of 20.8 percent (one-half it currently is in high-income countries, there the 1990 headcount index). This translates into is a reasonable probability that household around 920 million people living under the in- saving rates will decline more quickly than is ternational poverty line, which coincidentally assumed in the baseline. By the same token, is around 50 percent of the estimated number investment may react more forcefully to low of poor in 1990. There is significant regional interest rates and improved confidence.5 variation. East Asia and the Pacific will largely The third panel of table 1.3 reports the results surpass its regional target, in large part because of a simulation that assumes that the savings rate of the significant success in reducing poverty in in the United States declines over time from its China, by far the region's most populous coun- current level of 3.4 percent of household income try. Sub-Saharan Africa is projected to miss to about 2.7 percent of household income--close its target (by over 9 percentage points) as will to its average level in the 2000s. Savings rates in Europe and Central Asia. Africa's poor per- high-income European countries are assumed to formance reflects mainly weak growth in the fall by about the same amount. 1990s. The economic adjustments required by In this scenario, the combination of stron- the transition from planned economies to mar- ger consumption and investment throughout ket economies led to a rise in poverty in Europe the global economy increases GDP growth by and Central Asia, albeit from a low level. Sig- about 0.4 percentage points for high-income nificant progress in reducing poverty is antici- countries and 0.6 percent for developing coun- pated in both regions between 2005 and 2015. tries in 2010. As a result global trade is close to Progress on poverty using the broader $2/ 0.2 percentage points higher in 2011 than day definition is projected to be somewhat less in the baseline, and output gaps are about promising, with the headcount index dropping 41 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Table 1.3 Poverty in developing countries a new forecast of per capita income growth. by region, selected years Beyond the methodological advances in more Region or country 1990 2005 2015f 2020f recent surveys, they also reflect changes in the underlying distribution of income that are not Percentage of the population living on less than $1.25/day measured by changes in mean income (or con- East Asia and Pacific 54.7 16.8 5.9 4.0 sumption). Since last year's report, the new China 60.2 15.9 5.1 4.0 poverty forecast integrates 31 new household Europe and Central Asia 2.0 3.7 1.7 1.2 Latin America and the Caribbean 11.3 8.2 5.0 4.3 surveys. Combining these new surveys with last Middle East and North Africa 4.3 3.6 1.8 1.5 year's growth forecast implies a 0.5 percentage South Asia 51.7 40.3 22.8 19.4 point decline in the aggregate headcount index India 51.3 41.6 23.6 20.3 Sub-Saharan Africa 57.6 50.9 38.0 32.8 from 15.5 percent to 15.0 percent (figure 1.25). Total 41.7 25.2 15.0 12.8 The largest single change is for China where Percentage of the population living on less than $2.00/day the new survey causes the projected 2015 head- count index to drop by about 2 percentage East Asia and Pacific 79.8 38.7 19.4 14.3 China 84.6 36.3 16.0 12.0 points. Sub-Saharan Africa shows a small rise Europe and Central Asia 6.9 8.9 5.0 4.1 of 0.3 percentage point for the same reason. Latin America and the Caribbean 19.7 16.6 11.1 9.7 The new economic forecast, compared with Middle East and North Africa 19.7 16.9 8.3 6.6 South Asia 82.7 73.9 57.0 51.0 2008, has no net effects at the aggregate level, India 82.6 75.6 58.3 51.9 but raises slightly the headcount index for Sub- Sub-Saharan Africa 76.2 73.0 59.6 55.4 Saharan Africa and East Asia and the Pacific.7 Total 63.2 47.0 33.7 29.8 As 2015 is rapidly approaching, it is useful Number of people living on less than $1.25/day (millions) to look a bit further ahead and assess the needs East Asia and Pacific 873 317 120 83 of developing countries 10 years forward. This China 683 208 70 56 year's forecast for 2020 suggests that 826 mil- Europe and Central Asia 9 16 7 5 Latin America and the Caribbean 50 45 30 27 lion or 12.8 percent of developing-country Middle East and North Africa 10 11 6 6 citizens will be living on $1.25/day or less South Asia 579 595 388 352 and that there will be almost 2 billion poor India 435 456 295 268 Sub-Saharan Africa 296 387 366 352 people using the $2/day poverty line. The five Total 1,817 1,371 918 826 additional years would still leave Sub-Saharan Number of people living on less than $2.00/day (millions) Africa short of the poverty MDG. East Asia and Pacific 1,274 730 394 299 China 961 473 220 168 Europe and Central Asia 32 39 22 18 Policy implications for Latin America and the Caribbean 86 Middle East and North Africa 44 91 52 67 30 62 26 developing countries South Asia India Sub-Saharan Africa 926 1,091 973 926 702 828 728 686 391 555 574 595 A lthough the financial crisis has passed and the global economic recovery seems to be under way, many challenges for policy Total 2,754 2,557 2,060 1,926 makers and international financial institu- Source: World Bank. tions remain. Paramount among these is the f: Forecast. management of the unwinding of the fiscal and monetary stimulus that has played such a to a still high one-third of the total developing- critical role in avoiding a much more serious country population and more than 50 percent of downturn. the 1990 level, leaving some 2 billion people liv- Timing the tightening of fiscal and mon- ing with $2/day or less. etary policy to avoid killing off the recovery As is the case each year, the new poverty is one clear consideration. But so too is the forecast is a combination of two changes-- risk that the very loose monetary and fiscal new and more recent household surveys and conditions in high-income countries could 42 P R O S P E C T S F O R D E V E L O P I N G E C O N O M I E S Figure 1.25 Comparison of 2015 poverty forecast, GEP 2009 versus GEP 2010 $1.25/day headcount index, percent 40 35 30 25 20 15 10 5 0 East Asia and China Europe and Latin America Middle East South Asia India Sub-Saharan Total Pacific Central Asia and and Africa the Caribbean North Africa GEP 2009 GEP 2009 and new surveys GEP 2010 Source: World Bank. create dangerous conditions for developing measures succeed in generating additional output countries. Already very low interest rates in and government revenues, associated expendi- high-income countries are promoting carry tures will be more sustainable than more tradi- trades that may be promoting destabilizing tional expenditure-oriented ones. Countries with capital inflows into developing countries that sufficient fiscal space may seek to target measures could create new asset bubbles and the poten- to reduce (infrastructure) bottlenecks. Invested tial for future crises. For developing countries, wisely in human and physical capital, such steps the management of the recovery in capital flows can position a country to take better advantage is a critical challenge. Warding off new asset of the global recovery when it comes, by more ef- price bubbles may call for greater exchange fectively exploiting existing comparative advan- rate flexibility. If these inflows are enduring tages and helping to generate new ones. and effectively channeled into productive in- vestment, they could present a major boon to developing countries (see the analysis in chap- Notes ter 2). However if they exceed the absorptive 1. Potential output is the level of output commen- surate with the level of production when all factors of capacity of countries or are cut off abruptly, the production, i.e., labor, capital, and technology, are costs could be very high. fully employed. Given the slow growth and associated real- 2. Total revenues in IDA countries is estimated side adjustments that are expected over the me- (based on IMF 2009) to have fallen from an average dium term (see chapters 2 and 3), government of 26.2 percent of GDP over the 2000­08 period to policies should focus on productivity-enhancing 21.9 percent of GDP in 2009. This 4.3 percent (as growth strategies. For low-income countries, a share of GDP) decline in revenues is equivalent to nearly $35 billion. When 2009 revenues are compared these strategies may involve simultaneously ad- to the year before (when total revenues were equivalent dressing underlying structural problems such as to 28.1 percent of GDP), the fall in revenue is equivalent the quality of institutions, regulatory reform, and to nearly $50 billion. openness--all critical factors in promoting faster 3. The external financing need and gap projec- productivity growth. To the extent that these tions are based on the methodology developed in 43 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 World Bank (2009a) and assess the extent to which Friedman, Jed, and Norbert Schady. 2009. "How capital flows from private sources will meet develop- Many More Infants Are Likely to Die in Africa as ing countries' external financing needs defined as cur- a Result of the Global Financial Crisis?" Policy rent account deficit and scheduled principal payment Research Working Paper 5023, World Bank, on private debt. Private short-term debt is projected to Washington, DC. decline further in 2010, while medium and long term Horton, Mark, Manmohan Kumar, and Paolo Mauro. debt increase slightly. 2009. "The State of Public Finances: A Cross- 4. The fiscal stimulus for the G-20 countries was Country Fiscal Monitor." IMF Staff Position taken as the change in the discretionary measure of the Note SPN/09/25, International Monetary Fund, fiscal deficit from Horton, Kumar, and Mauro (2009). Washington, DC. For the remainder of developing countries, it was esti- IMF (International Monetary Fund). 2009. World mated as the change in the structural deficit (in those Economic Outlook. Washington, DC. countries where structural expenditures increased as International Energy Agency. 2009. Oil Market a percent of GDP)--using IMF estimates for general Report. Paris. government expenditures and revenues. Ratha, Dilip. 2009. "Dollars without Borders: Can 5. In the early 2000s investment grew about 1.6 the Global Flow of Remittances Survive the times as quickly as GDP in developing countries, but Crisis?" Foreign Affairs, October 16. http://www the elasticity has occasionally been as high as 2 dur- .foreignaffairs.com/articles/65448/dilip-ratha/ ing upswings. In the baseline, this investment-GDP dollars-without-borders. elasticity is roughly consistent with the trends earlier Ratha, Dillip, Sanket Mohapatra, and Ani Silwal. 2009. in this decade--about 1.4 times GDP growth in most "Migration and Remittance Trends 2009: A Bet- countries. Applying the larger historical elasticity of 2 ter-Than-Expected Outcome So Far, But Signifi- during upswings suggests that investment could grow cant Risks Ahead." In Migration and Development as fast as 10.9 percent on average over 2010­11 versus Brief. World Bank, Washington, DC. the 7.8 percent in the baseline scenario. Ravallion, Martin. 2009. "The Crisis and the World's 6. These calculations update those in Ravallion Poorest." Development Outreach, World Bank, (2009) and World Bank (2009c) and are consistent Washington, DC. December. with new survey evidence and the revised forecasts for Roubini, Nouriel. 2009. "Mother of All Carry Trades growth presented in this report. Faces an Inevitable Bust." The Financial Times, 7. We have not decomposed these changes into November 1. changes in the growth forecast itself and changes to UNSCN (United Nations Standing Committee on the poverty elasticity with respect to growth as emerges Nutrition). 2009. "Global Recession Increases from the new surveys. Malnutrition for the Most Vulnerable People in Developing Countries. Pregnant Women and Children Are Hardest Hit." In Nutrition Im- References pacts of the Global Food and Financial Crises. Agriculture, U.S. Department of. 2009. "Production, Geneva. Supply and Distribution" [online], available at World Bank. 2009a. Global Development Finance http://www.fas.usda.gov/psdonline/ [Accessed: 2009. Washington, DC: World Bank. December 9, 2009]. ______. 2009b. Global Economic Prospects: Commodi- Chen, S., and M. Ravallion. "The Impact of the ties at the Crossroads. Washington, DC. Global Financial Crisis on the World's Poor- ______. 2009c. "Protecting Progress: The Challenge est." Vox: Research-based policy analysis and Facing Low-Income Countries in the Global commentary from leading economists [Accessed: Recession," Background paper prepared for the December 9, 2009], www.voxeu.org/index G-20 Leaders' Meeting, Pittsburgh, PA, Sept .php?q=node/3520. 24­25, 2009. http://siteresources.worldbank.org/ ECB (European Central Bank). 2009a. "What Triggers NEWS/Resources/WorldBankG20PaperonLICs- Prolonged Inflation Regimes? A Historical Analy- Sept2009.pdf. sis," by Isabel Vansteenkiste. Working Paper Series. World Bureau of Metal Statistics. World Metal Statis- ______. 2009b. December 2009 Financial Stability tics Yearbook. London. Review. Frankfurt, Germany. http://www.ecb .europa.eu. 44 2 The Impact of the Boom in Global Finance on Developing Countries T he first seven years of the 21st century were very good for developing countries. GDP growth continued to accelerate as it had about 700 basis points in January 2003), and global credit expanded twice as fast as nomi- nal GDP.1 A range of financial innovations, done in the 1990s but at an even faster pace, including the securitization of loans and the while economic volatility was far lower than development of off-balance-sheet vehicles, in previous periods of rapid growth (IMF allowed banks to fund an important portion 2007). And while large countries with very fast of their loan portfolios through capital and growth rates, such as China and India, tended money markets, leveraging equity capital in a to attract the most attention, most of the accel- way never before possible. Partly as a result, eration in developing-country growth during the amount of finance--both domestic and this period occurred among smaller countries international--available to developing coun- that in the past had been growing much less tries expanded very rapidly, and countries en- quickly. joyed a sustained investment boom. Somewhat surprisingly and in contrast to That boom came to an abrupt end in the popular perceptions, this growth spurt oc- fall of 2008 with the failure of Lehman Broth- curred during a period in which external de- ers and the financial crisis that ensued (see mand conditions for developing countries chapter 1). Although clouded by uncertainty, were not that strong. Growth in high-income the longer-term consequences of the crisis countries was actually slower during the boom could be far-ranging. The sharp scaling back years 2003­07 than during the preceding of global production may result in permanent 13 years. Moreover, import demand from and long-lasting adjustments in global pro- high-income countries was growing only 5.6 duction patterns. Firms and regional special- percent a year, marginally slower than dur- izations may fail and disappear in a way that ing the preceding 13 years. More than all of they would not have had adjustment occurred the acceleration in developing-country ex- more gradually. Global trade patterns may be ports came from an expansion in their share in irrevocably altered, and the depth of the re- high-income country imports and very rapid cession in some regions and countries relative growth in South-South trade. to others may change the future pattern of Financial conditions were, however, very growth in the world. The temporary weakness favorable. Interest rates and interest rate of the financial sector in high-income coun- premiums were low (for example, the aver- tries may create opportunities for financial age secondary market spread on develop- firms in developing countries, allowing them ing countries' sovereign bonds fell to about to grow and expand in ways that might not 200 basis points by mid-2007, down from have been possible otherwise. Although each 45 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 of these possible consequences merits in-depth A number of key messages emerge from the exploration, dealing with all of the potential discussion in chapter 2: consequences of the crisis for developing coun- tries lies outside the scope of this publication. The acceleration in developing-country The analysis presented in this and the next growth during the 2003­07 period arose chapter focuses more narrowly on the medium- despite relatively lackluster GDP and im- term consequences of recent and anticipated port growth among high-income coun- changes in financial conditions for developing- tries. Developed-world GDP grew on country finance, investment, and supply po- average 0.2 percentage point slower than tential, both over the past decade and that during the 1990s and import demand in- can be expected in the next 5 to 10 years. This creased 0.4 percentage point less quickly. orientation was chosen partly because, con- The fall in borrowing costs during the trary to popular perceptions, real-side external 2003­07 period was associated with al- factors do not appear to have played a major most 70 percent of the increase in capi- role in the boom. Most important, this focus tal flows into developing countries and on the financial aspects of the crisis was cho- 80 percent of the increase in domestic sen because of the important role that finance intermediation. played in causing the crisis and because the While the biggest apparent contribution likely regulatory and market-based changes in to the changes in the extent of intermedia- the sector are somewhat less speculative than tion in developing countries was driven by those that might surround other important lower borrowing costs and the overall ex- elements of the post-crisis world. pansion of global liquidity, cross-country Within this overall context, this chapter ex- differences in the level of intermediation amines the link between the global expansion remain very large and are best explained of liquidity and the improvement in developing by fundamental factors such as the qual- countries' growth before the financial crisis. It ity of regulatory frameworks and the begins with a review of the credit boom and its business environment, inflation rates, and implications for the pricing of risk and borrow- levels of government debt. ing costs. It then describes how the global boom Country-specific differences in the contributed to the rapid expansion of domes- quality of institutions and the degree tically supplied credit and international capital of market openness of the top and flows in developing countries, discusses the fac- bottom performing 25 percent of tors that helped to determine which countries countries are associated with 56 and most benefited from the liquidity glut, and ex- 37 percent of the cross-country varia- amines the extent to which different countries tion in levels of domestic intermedia- were able to translate these more liquid condi- tion, respectively, and 1/3 and 1/5 of tions into increased investments. The chapter the cross-country difference in inter- concludes with some model-based measure- national capital flows. ments of the impact of the investment boom Countries with good regulatory envi- on growth and potential output in developing ronments were also more successful countries. All of this serves as a prelude to chap- in transforming increased financing ter 3, which analyzes the extent to which, in the into increased investment and GDP future, tighter financial regulation, increased growth. More than one-quarter of the risk aversion, and higher interest rates and 11.5 percent of GDP difference be- interest rate premiums are likely to constrain tween the investment rates of the top investment and potential growth in developing and bottom 25 percent of developing countries and the scope for developing countries countries appears to reflect differences to pursue policies to mitigate these impacts. in the quality of institutions. 46 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S Countries with high levels of finan- creation of significant inflationary pres- cial openness and well-developed do- sures or external imbalances in many de- mestic intermediation systems also veloping countries, suggests that in these had higher investment rates. About 3 countries the boom relieved what may percentage points of the difference be- have been a binding capital constraint on tween the investment-to-GDP ratio of growth. That in turn implies that such the top 25 percent of developing coun- stronger growth rates for developing tries and the bottom 25 percent is as- countries may be achievable over the long sociated with differences in the size of term if sufficient finance (domestic or foreign capital inflows. For domestic external) is forthcoming. Of course there intermediation, the same figure is just were exceptions, notably in the Europe under 2 percent of GDP. and Central Asia region, where the stron- These results suggest that if Sub- gest expansions in credit boom contrib- Saharan Africa could improve its insti- uted to macroeconomic instability. tutions to roughly the levels observed in Latin America, the overall extent of financial intermediation would rise substantially, perhaps by as much Financial innovation, high- as 12 percent of GDP in the case of income finance, and the domestic credit to the private sector liquidity boom and 2 percent of GDP in the case of international financial flows. Different forms of finance had different T he liquidity boom that preceded the financial crisis of 2008 was broadly based and rooted in a number of factors. effects on investment. Like other booms and busts, this one was Bond flows had significant impacts on prompted by a rapid increase in credit and investment in middle-income countries. investment that ultimately proved unsus- Bank lending, which dominated flows tainable and the ensuing bust provoked a into Europe and Central Asia, were sudden contraction in GDP (box 2.1). associated with a larger increase in Data from the Bank of International Settle- current account deficits and consumer ments (BIS) indicates that from 2002 through demand. 2007 international bank credit expanded Foreign direct investment (FDI) about twice as fast as nominal GDP and more funded as much as 20 percent of total than twice as fast as it had during the previ- investment in some regions, with low- ous decade (figure 2.1). Long-term interest income countries tending to be more rates were only between 1.5 and 2 percentage reliant on this form of financing than points higher than inflation in the major in- richer countries. dustrial countries (table 2.1), compared with Overall, more than half of the 1.4 per- about 3.5 percentage points (in the United centage point increase in potential out- States) during the global expansion in the sec- put growth rates in developing countries ond half of the 1990s. between 2003 and 2007 is directly at- The proximate cause of the credit boom is tributable to the capital deepening that a question of considerable debate--a debate was observed during this period, even that is unlikely to be resolved anytime soon. under the conservative assumption that Among the competing and not necessarily con- higher investment had no role in the rise tradictory explanations are: in productivity. The expansion of investment and growth A savings glut. According to this argument during the boom period, without the (see Bernanke 2005, among others), high 47 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Box 2.1 Comparing this boom-bust cycle with other major cycles T his boom-bust cycle shares many characteristics with earlier financial crises: an extended period of rapid and ultimately unsustainable credit expan- Box figure 2.1.2 Indicators of macroeco- nomic stability in developing countries, 2007 sion, accompanied by excessive risk taking by finan- a. Low inflation cial institutions, followed by a sharp reduction in Annual percent change in consumer prices economic activity. However, this crisis differs in three 12 important respects from earlier crises. 10 First, this crisis is the most severe and widespread 8 downturn since 1945. Global GDP is estimated 6 to have declined by 2.2 percent in 2009 (the only absolute decline in global GDP during the postwar 4 period), and GDP is projected to remain well below 2 potential output for years to come, with estimates 0 of the developing-country output gap peaking at East Asia Europe Latin Middle South Sub- 4.8 percent of GDP--almost 50 percent larger than and and America East Asia Saharan Pacific Central and and North Africa during the next most severe modern-day recession Asia Caribbean Africa (1982­83). b. Generally modest current account deficits Second, for the majority of developing coun- Percent of GDP tries this is a crisis that originated in high-income 10 countries. Moreover, with the notable exception of 8 many countries in Europe and Central Asia, it was 6 4 2 0 Box figure 2.1.1 GDP growth and output ­2 gaps in global crises since 1970 ­4 Percentage East Asia Europe Latin Middle South Sub- ­6 and and America East Asia Saharan Pacific Central and and North Africa Asia Caribbean Africa ­5 c. Relatively low debt levels Gross external debt/gross national income, percent ­4 Output Gap 45 40 ­3 35 30 ­2 25 GDP growth 20 ­1 15 10 0 5 1982­03 1991­93 2001 2009 0 East Asia Europe Latin Middle South Sub- Source: World Bank. and and America East Asia Saharan Note: GDP growth is the percentage change in GDP growth in the Pacific Central and and North Africa crisis year(s) compared with the preceding year. The output gap Asia Caribbean Africa is the percentage difference between GDP and potential output during the crisis year(s). Source: World Bank. 48 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S not preceded by the buildup of serious domestic and country hard, even though, outside Europe and external imbalances, and domestic actors largely did Central Asia, most countries did not exhibit unsus- not participate directly in the unsustainable activities tainable macroeconomic balances (box figure 2.1.2). that precipitated the crisis. In most countries regional inflation rates averaged During earlier global or large-scale crises trig- about 6 percent or lower (well below the double-digit gered by changes in high-income countries, major rates in most regions during the early 1990s); most impacts tended to be limited to developing countries regional current account balances were near zero or with preexisting vulnerabilities. The tightening of strongly positive; and ratios of debt to gross national U.S. monetary policy in 1979­80 boosted real inter- income were modest. However, the quality of policies est rates and brought on a global recession, which still affected the impact of the crisis--the countries hit hardest those developing countries with excessive with the largest imbalances suffered the most (see levels of private-source debt. The depreciation of the chapter 3). yen against the dollar in the mid-1990s reduced the Third, this crisis has struck many more countries competitiveness of East Asian economies that pegged than earlier recessions did, a factor that complicates their currencies to the dollar, which may have con- recovery for individual countries because there are tributed to the onset of the 1997 crisis. By contrast, few fast-growing external markets with which to the current crisis struck virtually every developing engage in an export-led recovery strategy. savings relative to investment in East Asia An excessive loosening of regulatory over- kept global interest rates low, fueling rapid sight. The reduction of regulatory barri- increases in investment and consumption in ers to speculation and excessive reliance high-income countries. on self-regulation of the banking sector in An extended period of loose monetary pol- industrial countries generated and failed to icy. Very loose monetary conditions in the curb excessive risk taking by financial insti- United States, Japan, and Europe over an ex- tutions (Crotty 2009). tended period of time bled through into lon- Financial innovations. In this loosely con- ger maturities, provoking an unprecedented trolled environment, the use of new finan- expansion in global credit (BIS 2006). cial innovations expanded rapidly; these innovations increased risk taking and helped to circumvent those regulatory bar- Figure 2.1 Since the early 2000s, credit riers that remained (Calvo 2009). expansion has grown more than twice as fast as nominal GDP Finally, in contrast to popular thinking, unusually strong developed-country demand Index, June 1990 = 100 600 Table 2.1 Interest rates and inflation in 500 industrial countries, January 2002­June 2007 400 Global Banking Assets (percent) 300 Consumer Average long-term price inflation interest rate 200 Euro Area 2.2 4.1 100 Japan 0.1 1.4 World nominal GDP United States 3.0 4.4 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 Source: OECD. Note: CPI inflation is expressed as the average annual percent- Sources: Bank of International Settlements; World Bank. age change over the period, and the average long-term rate is in percent. 49 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 was not a major factor behind the acceleration rose an average of 6.0 percent annually. The in developing-country growth or exports. strong performance of developing-country Indeed, the boom period 2003­07 was exports during this period reflected three main actually one of relatively slow growth for factors: rapidly expanding supply capacity in high-income countries in terms of both GDP developing countries, an increase in their share and imports. Developed-country GDP grew on of the imports of high-income countries, and average 2.3 percent during the period versus rapidly expanding South-South trade. an average of 2.5 percent between 1990 and 2003 (figure 2.2). Moreover, notwithstanding the somewhat heated rhetoric surrounding trade issues, high-income-country import Novel channels for credit demand, which grew an average of 5.6 percent creation during the boom period, actually expanded less quickly than during 1990­2003, when it W hatever the fundamental reason for the long credit boom, the increased availability of a number of new financial instruments (box 2.2) gave investors what ultimately proved to be a false sense that the risks of rapid credit expansion had Figure 2.2 High-income GDP and trade been reduced. This false sense of security growth do not explain the acceleration in contributed to the reductions of interest rates developing-country economic activity and interest rate spreads, thus facilitating the a. High-income GDP was slower than normal expansion of credit. Annual percent change The expanded use of a number of these 9 financial innovations boosted the growth of 8 Developing countries what has been called the "shadow banking 7 Period averages system"--comprising institutions that do not 6 have access to deposit insurance or central 5 bank rediscount operations and that are not 4 subject to the same prudential regulations as 3 banks (Farhi and Cintra 2009). These insti- 2 tutions nevertheless actively sold and mar- 1 High-income countries keted instruments that leveraged the savings 0 1990 1995 2000 2005 of households in a manner akin to the credit creation process of more traditional banks. b. High-income import demand was slower than during the 1990s The institutions involved included investment Annual percent change banks, hedge funds, investment funds, private 20 equity funds, special investment vehicles (in- Developing-country exports cluding those operated off balance sheet by 15 Period averages banks), pension funds, and insurance com- panies. The quasi-banking activities of these 10 entities were actively supported by ratings 5 agencies, which markedly increased their rev- enues by rating the structured products these 0 High-income imports entities sold. It is difficult to measure the contribution ­5 1990 1995 2000 2005 of the shadow banking system to the finan- Source: World Bank. cial boom, compared with more traditional balance-sheet transactions of the commercial 50 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S Box 2.2 Recent and systemically important financial innovations S ecuritization is not a recent innovation, but its use skyrocketed during the boom. It refers to the issuance of new securities backed by a pool of debt swapping payment obligations from one currency to another. These transactions are often used to protect a portfolio in the face of uncertain changes in interest instruments. By this mechanism, a relatively illiquid or exchange rates or to speculate on such changes. stream of future cash flows (such as a standard loan Collateralized debt obligations (CDOs) are securi- with fixed repayment terms) is converted into a secu- ties backed by collateral in the form of a portfolio rity that can be traded in the marketplace. of bonds, bank loans, or other debt (such as credit Credit default swaps (CDSs) are agreements in card debt). Repayments to the pool of investors are which the buyer makes a series of payments to the typically allocated according to some prioritization; seller, in return for which the seller is obligated to for example, senior CDO notes are paid first. Other compensate the buyer if the underlying bond or loan tranches earn higher returns but are only paid out goes into default. Effectively, these instruments pro- if funds are remaining. This structure permits issues vided insurance against default--although the regula- that satisfy differing trade-offs between risk and tory environment for such swaps and insurance are return: more speculative investors can purchase the very different. More extensive use of credit default lower-rated tranches, while more risk-averse inves- swaps also increased arbitraging opportunities by tors can purchase higher-rated tranches. making it easier for speculators to take positions in Other credit derivative products. U.S. financial securities that they did not own (Guttmann 2009). markets have generated several, more exotic ap- Interest rate and currency swaps are instruments proaches to securitizing debt transactions. For ex- that allow investors to effectively change the payment ample, credit-linked notes are sold with an embedded scheme associated with a loan or an asset. For exam- credit default swap, where the issuer is not required ple, interest rate swaps often involve contracting to to repay the debt if a specified event occurs (essen- make a fixed series of payments by one counterparty tially eliminating the need for third-party insurance). in exchange for receiving a second series of pay- Specialty finance companies have been created where ments based on a floating rate. Other swaps involve transactions involve both securitization and lending. banking system--in large part because it transactions, the notional value of which qua- faced much less comprehensive reporting re- drupled between 2002 and 2008 (figure 2.3), quirements and oversight. One indication of reaching more than 25 times U.S. GDP (figure its importance can be gleaned from the rise 2.4). The gross notional value of the derivative in the share in total U.S. domestic credit of market involves considerable double counting mortgage pools (issued by Fannie Mae and (the net exposure of counterparties is much Freddie Mac) and asset-backed securities. In smaller because of offsetting transactions); 1995 these securities accounted for 16 per- moreover the actual associated flows involved cent of credit assets held by the U.S. financial in these transactions are typically a very small sector or 30 percent of GDP. By 2007 the percentage of the notional values. Nevertheless, value of these securities had increased more the notional value provides a sense of how per- than fivefold, reaching 23 percent of credit vasive and far reaching these instruments had assets and 63 percent of GDP--almost as become in intermediating economic activity. large as the total of commercial bank assets Moreover, the notional values provide a (figure 2.3). sense of the systemic vulnerability represented The credit expansion was also reflected by these instruments, especially during the in the phenomenal rise in derivative swap acute phase of the crisis when the ability of 51 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 raising funds in the commercial paper market, Figure 2.3 Notional value of derivative for example). The instruments it employed transactions, 2002­08 (such as collateralized debt obligations--see US$, trillion box 2.2) had the perceived virtue (compared 450 400 Interest rate and currency swaps with bank loans) of spreading the risk of lend- Credit default swaps 350 Equity derivatives ing. Large, risky investments could be divided 300 efficiently among investors and thus increase 250 the feasibility of such projects. And investors 200 could more easily diversify their risk portfolio, 150 allowing them to undertake higher risk and re- 100 turn projects. 50 However, these instruments were either 0 loosely or not at all regulated and may have 2002 2003 2004 2005 2006 2007 2008 induced banks to reduce their lending stan- Source: International Swaps and Derivatives Association. dards more than they would have otherwise because the long-term risk associated with loans was being held by others. Moreover, counterparties to meet their commitments in the event, ownership was concentrated in was called into question and default payments some systemically important hands. Banks under derivative contracts mushroomed. were left with large holdings--often the lower- The impact of the expansion of the shadow quality, higher-yielding tranches. In addition, banking system was to greatly expand the banks that relied on secondary markets to buy amount of credit available and reduce its and sell loans tended to increase their leverage cost. Shadow banking effectively performed (Duffie 2007), which contributed to increas- the same functions as banks, increasing as- ing systemic risk to the extent that the buy- sets to several times their equity by funding ers of these securitized loans lacked sufficient long-term assets with short-term liabilities (by information to accurately evaluate the risks involved.2 The extreme complexity of some of these instruments and the lack of standardized exchanges made it difficult for both purchasers and sellers to evaluate them and exacerbated Figure 2.4 Share of commercial bank and securitized assets in total credit held by the difficulties in debt renegotiations in the U.S. financial sector, 1995­2008 case of financial distress. Ex post, it appears clear that these instruments generated substan- US$, trillion tial further systemic risks by multiplying in a 30 Commercial banks Securitized transactions nontransparent manner the interdependencies 25 in the financial system. 20 On balance, the growth of the shadow 15 banking system and the expanded use of securitization and derivatives products 10 worldwide (see box 2.2) contributed to the 5 expansion of credit in developing coun- 0 tries during the boom period. Several fac- 1995 2002 2003 2004 2005 2006 2007 2008 tors underpinned the increased use of these Sources: U.S. Flow of Funds Data, Federal Reserve accounts. instruments in developing countries. Their Note: Securitized transactions include mortgage pools issued by expanding use in high-income countries U.S. government-sponsored enterprises (principally Fannie Mae and Freddie Mac) and asset-backed securities. made more investors familiar with their ben- efits, while efforts to standardize derivative 52 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S contracts (by the International Securities Not all derivative transactions involving Dealers Association, for example, to de- developing-country instruments increased the velop standard documentation for credit availability of capital to developing countries. default swaps) helped reduce their costs and For example, synthetic collateralized debt ob- improved confidence in derivative transac- ligations were mainly a vehicle to facilitate tions. Growth in spot markets also encour- speculation on developing-country returns. aged greater use of derivatives for hedging Investors purchase a synthetic CDO, the re- purposes. In addition, the expansion of the turn on which was tied, say, to changes in the size and length of maturities in local currency credit default swap spread on bonds issued by bond markets facilitated the creation and the Brazilian government. Because these syn- pricing of developing-country interest rate thetic CDOs did not involve the repackaging derivatives (Saxena and Villar 2008). of existing bank loans, they did not reduce The expanded use of these products helped banks' exposure to developing-country debt to disperse risk, improve diversification among and therefore did not enable them to increase investors, and increase the pool of developing- lending. Indeed, some observers argue that by world investors, thereby increasing capital flows facilitating speculation, these instruments in- to developing countries. For example, banks creased volatility in developing-country finan- were able to expand lending to developing- cial markets.4 country borrowers--even high-risk borrow- ers--and transfer the risk to capital markets through credit default swaps (World Bank 2007) and by pooling loans and selling them Developing-country finance to investors in high-income countries. Be- during the boom tween 2003 and 2008, CDS spreads were quoted widely for 40 developing countries, in addition to a number of privately negoti- T he expansion of liquidity in high-income countries, the financial innovations, and the consequent fall in the price of risk dramati- ated deals that were not widely reported.3 cally changed developing-country finance. Net The proliferation of securitized and derivative capital inflows quintupled, and spreads on for- products enabled pension funds and insur- eign debt fell from 656 basis points in 2000 to ance companies, many of which face regula- 168 basis points at the end of 2007. Equally tory restrictions on the kinds of investments important, domestic credit as a share of GDP they can make, to take indirect positions in increased by 5 percentage points on average, developing-country loans by purchasing the with much larger increases in several regions, more highly rated tranches of securitized while domestic interest rates declined across loans. the board. These developments were accompa- The secondary sale of developing-country nied by an unprecedented tripling in the valua- loans to nonbank investors, or the banks' tion of equities traded on developing-economy own off-balance-sheet vehicles, contributed stock markets. to overall credit expansion by replenishing The rise in financial intermediation in- banks' reserves and allowing them to provide creased the supply of finance available to entre- new additional loans to developing coun- preneurs to undertake productive investment, tries. Increased availability of derivatives also thereby contributing to capital accumulation boosted the supply of FDI by providing inves- and the expansion of potential output. More- tors with a mechanism to hedge the short-term over, the influx of new investments, embody- foreign exchange risk involved in projects, ing newer technologies, facilitated an overall particularly in those targeting production for acceleration in technological progress in de- the domestic market (Griffith-Jones and Leape veloping countries that was also supported by 2002). macroeconomic and institutional reforms in 53 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 many countries.5 Finally, the acceleration in middle-income borrowers (and those low- growth itself likely triggered a further deepen- income borrowers with market access) may ing of financial markets (see box 2.3 for the see a surge of inflows that reverse especially positive interaction between financial interme- sharply when prospects deteriorate. diation and growth). Historically, this "stop-go" quality of fi- Of course, while a rapid increase in global nance, particularly external debt and portfo- liquidity can facilitate economic growth, in lio equity flows, has exacerbated booms and some circumstances it can also cause macro- painful busts in many developing countries. economic instability. Easy access to finance The source of instability is not always foreign, can lead to excessive consumption and unsus- however. In many instances, large swings in tainable current account deficits, as was the international capital flows have been ascribed case in many countries in emerging Europe to the behavior of domestic investors.6 In the and Central Asia. More generally, interna- East Asian financial crisis, much of the capital tional finance tends to be especially procycli- flight that contributed to the large currency cal for developing-country borrowers. Weak depreciations and macroeconomic instability institutions (including protection of property was the result of domestic investors fleeing rights) and low-income levels make them less local currency instruments in favor of foreign- creditworthy on average. As a result, when denominated instruments that were expected both global and domestic conditions are good, to be better stores of value (Kawai, Newfarmer, Box 2.3 Financial intermediation and economic development S everal empirical studies find that the size and efficiency of financial intermediation has a causal affect on growth: Measures of financial development But the literature is not unanimous in identifying a causal relationship between financial development and growth. Growth also has an impact on finan- are found to be correlated with growth in a subse- cial development. Moreover, third factors (such as quent period in a cross-section of countries (King technological innovations in communications and and Levine 1993, Levine and Zervos 1998 for devel- data processing, as well as the quality of institu- oped economies). Financial market deepening is found tions) affect both growth and financial development. to be related to productive efficiency in cross-section Several economists find a bidirectional relationship data, including both developed and developing coun- between financial development and growth (Luintel tries (Nourzad 2002). Financial development is associ- and Kahn 1999; Al-Yousif 2002; Demetriades and ated with poverty reduction (Jalilian and Kirkpatrick Hussein 1999). Hurlin and Venet (2008) find a ro- 2002; Beck, Demirgüç-Kunt, and Levine 2007) and bust causality from growth to financial development is found to precede growth in tests of Granger causa- in a sample of developed and developing countries tion on time series data (Neusser and Kugler 1998; but little evidence of causality from financial develop- Rousseau and Wachtel 1998). Instrumental variables ment to growth. Arestis and Demetriades (1997) find (English, French, German, or Scandinavian legal ori- that financial development causes growth in only a gin) as well as other econometric techniques are used few countries in their sample; Shan (2005) and Shan to isolate the causal impact of financial development and Morris (2002), using time series data (covering (Levine, Loayza, and Beck 2000). Financial develop- OECD countries plus China), find little evidence that ment also is found to raise growth principally through financial development leads economic growth; and its effects on total factor productivity (Beck, Levine, Al-Taimimi and others (2001) find no evidence of and Loayza 1999). Several country studies also show Granger causation between financial development or that financial development has a major impact on economic growth in either direction from a sample of growth over time (Levine 1997). Arab countries. 54 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S and Schmukler 2001; World Bank 1998). A Figure 2.5 The cost of risk in high-income similar dynamic underlay the crisis in Mexico countries fell sharply during the boom in 1994­95 (Frankel and Schmukler 1996). Percentage points In the case of Chile following the East Asian 16 and Russian crises, however, foreign investors Merrill high-yield corporate (US) were the main sources of capital flight (Cowan 14 and others 2005). 12 Lehman investment As discussed in chapter 1, during the recent grade (Euro) 10 crisis a rapid reversal in capital flows adversely Lehman investment grade (US) affected virtually every developing country, 8 US 10-year govt. bond even those that had pursued prudent mac- 6 roeconomic policies and accumulated large stocks of foreign currency reserves. That said, 4 the countries (notably many in the Europe and 2 Central Asia region7) that were hardest hit Euro 10-year govt. bond were precisely those in which the additional Jan. 1998 Jan. 2000 Jan. 2002 Jan. 2004 Jan. 2006 liquidity had been channeled into domestic Source: Datastream. consumption and that had accumulated signif- icant domestic and external imbalances during the boom period. The reduction in the price of risk markets had permanently reduced long-term The rapid expansion of global credit and the interest rates and risk premiums. low interest rates that accompanied it were Falling interest rates internationally, lower reflected in a sharp fall of secondary-market risk premiums, and, especially toward the end spreads on investment grade and high-risk of the boom period, rising commodity prices debt in industrial countries. For example, the also meant that financial conditions within risk premium on AAA corporate bonds in the developing countries relaxed. Reflecting both United States fell from 490 to 65 basis points these developments and the influence of policy between 2002 and 2007, while that on BBB improvements and political factors, interest grade European corporate debt fell from 390 rate premiums and the interest rates paid by to 55 basis points. The simultaneous fall of developing-country borrowers fell sharply in spreads on a wide variety of risky assets is con- several regions (figure 2.6). sistent with a significant reduction in the price of risk itself, either because of a decline in risk aversion on the part of investors or because of The expansion in domestic credit the emergence of a view that derivatives and The decline in borrowing costs was associ- other hedging mechanisms had lowered the ated with a rapid increase in financial flows, likely financial cost of holding a given level of domestic intermediation, and capital market risk (figure 2.5). valuations throughout the developing world The decline in interest rates and the fall in (table 2.2). Banking intermediation, as mea- the price of riskier assets at the beginning of sured by claims of deposit money banks on the decade were initially treated as a tempo- the private sector, expanded on average from rary cyclical phenomenon. However, as the 29 percent of GDP in 2000 to 35 percent boom period continued, commentators in- in 2007--greatly boosting the funds avail- creasingly began to argue that financial mar- able to firms for investment (see table 2.2). ket innovations such as credit default swaps In some regions, a growing participation by and the securitization of loans in secondary foreign banks in domestic financial systems 55 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 financial intermediation and the levels reached Figure 2.6 Developing-country interest rates varied significantly: fell substantially during the boom period Percentage points In Europe and Central Asia, bank credit 25 Jan. 2000 Jul. 2007 almost doubled and stock market capital- ization more than quadrupled (relative to 20 GDP), reflecting very low initial levels at- tributable to the region's communist past 15 (despite 10 years of transition), the pros- 10 pects for accession of several countries to the European Union, and the boom in oil 5 prices. Of the 25 countries with sufficient data, 12 registered increases in credit to the 0 private sector of more than 10 percent of GDP. er tes c ia an ati sia d be ca So ica t ia ric n Af as L lA n Af ra Pa s As ib ri a a C uro c M an a d tA Am Sta th E ha ar e E ifi ic tra e C m h or dle r an Eas en p Sa ut Financial intermediation also rose strongly d nA of ed id b- t ni Su U in South Asia. In India the ratio of bank N d an credit to GDP increased by 15 percentage Source: World Bank. points and the stock market capitalization nearly quintupled relative to GDP. Other supported the rapid rise in domestic financial countries in the region had more moder- intermediation (box 2.4). Firms in develop- ate increases (for example, the ratio of ing countries also benefited from a surge in bank credit to GDP increased 12 percent- stock market capitalization, which rose from age points in Bangladesh and 6 percentage 35 percent of GDP in 2000 to 114 percent points in Pakistan). in 2007.8 Moreover, lower interest rates and The increase in credit to the private sec- interest rate spreads reduced the cost of capital tor in the Middle East and North Africa facing investors in developing countries. Partly was smaller but still robust--partly reflect- as a consequence, ratios of investment to GDP ing the fact that as measured the cost of adjusted for inflation jumped by 5 percent of capital in the region actually increased (see GDP on average, with an 8 percent GDP jump table 2.2). Credit in Algeria, Morocco, and in South Asia. Tunisia registered gains of 5­6 percentage All regions participated in the financial points of GDP. Despite the near tripling of boom to some extent, although the increase in stock prices, the increase as a percentage Table 2.2 Changes in domestic intermediation, 2000­07 Private credit by banks Stock market capitalization Region 2000 2007 Change 2000 2007 Change (percent of GDP) (% points) (percent of GDP) (% points) Developing countries 29.3 34.8 5.5 35.3 113.9 78.6 East Asia and Pacific 66.1 55.4 10.7 47.1 165.1 118.0 Europe and Central Asia 16.8 32.5 15.6 17.5 77.3 59.8 Latin America and the Caribbean 24.9 27.1 2.2 31.6 71.4 39.8 Middle East and North Africa 33.0 39.2 6.2 19.9 56.1 36.2 South Asia 25.6 40.4 14.8 26.1 133.4 107.2 Sub-Saharan Africa 34.8 41.6 6.8 89.9 149.0 59.1 Source: World Bank calculations using Beck and Demirgüç-Kunt 2009. Note: For private credit, the regional numbers are simple averages of available country data. For stock market capitalization, the averages are weighted by GDP. 56 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S Box 2.4 The role of foreign banks in domestic intermediation F oreign banks play an important and growing role in domestic intermediation among developing countries. As of 2005, their share in total banking capital to expand lending, and if they are more ef- ficient and improve domestic bank efficiency (see below), they can reduce the cost of financial interme- assets in developing regions ranged from a low of diation and encourage higher volumes. In these in- 7.4 percent in South Asia to a high of 54.4 percent stances, foreign banks by stimulating intermediation in Europe and Central Asia (box table 2.4.1). More- may, in turn, encourage more rapid development. For over, during the boom period foreign banks increased example, in Europe and Central Asia, the acquisi- their share in total assets in all of the regions where tion of local banks by foreign banks was associated they already had relatively large presences. Indeed, with increased lending to small and medium-size the extent of the expansion in domestic credit is enterprises and retail markets (de Haas and Naaborg loosely related to the extent to which foreign banks 2006), even though foreign banks lent predominantly increased their market shares. The two regions with to multinational corporations, large domestic firms, the smallest foreign presence (East Asia and the and governments--potentially squeezing out smaller Pacific and South Asia) actually saw the market share players (see Gormley 2005 for the theoretical model). of foreign banks decline. Indeed, the entrance of foreign banks in a market The contribution of foreign banks to intermedia- tended to cause local banks to increase lending to tion in developing countries is not straightforward. small enterprises in part because of increased compe- In some countries they can serve as an important tition in lending to larger firms (Jenkins 2000). conduit that facilitates the importation of external In some cases foreign banks may reduce the level of financial intermediation. Research suggests that especially among low-income countries with weak regulatory frameworks and competition law, foreign Box table 2.4.1 Foreign bank participation banks may enter into a market and cherry-pick the and credit expansion best local clients (Detragiache, Tressel, and Gupta Change in 2006). In such circumstances, a larger presence of Share of Share of ratio of bank foreign banks may be associated with less credit to assets assets credit to the owned owned by private the private sector. by foreign foreign sector Overall, the evidence is mixed. Survey data in- banks, banks, over GDP, dicate that entrepreneurs in countries with larger Region 2001 2005 2000­07 participation by foreign banks face less binding East Asia and credit constraints (Clarke, Cull, and Martinez Peria Pacific 13.0 11.1 10.7 2001). Moreover, when domestic conditions are Europe and propitious (a solid local banking sector, and good Central Asia 42.0 54.4 15.7 regulatory and competitive protections), foreign Latin America and Caribbean 30.4 35.6 2.2 banks can contribute to an overall expansion of Middle East credit and a lowering of costs for borrowers. How- and North Africa 8.3 10.9 6.2 ever, foreign bank participation is not critical to South Asia 8.9 7.4 14.8 increasing financial intermediation in developing Sub-Saharan countries and can, in some regions with weakly Africa 46.2 49.5 6.8 contested and poorly regulated markets, result in Source: World Bank database on Financial Institutions and the crowding out of local providers and no net in- Structure. crease in intermediation. 57 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 of GDP and the level in 2007 were smaller The drop in private credit relative to GDP than in the other developing regions in East Asia and the Pacific stems in part The 7 percentage point increase in bank from adjustments following the East Asia credit (relative to GDP) in Sub-Saharan crisis, with particularly significant declines Africa mainly reflects a 12 percentage point in Malaysia (27 percentage points) and rise in South Africa, rather than a more the Philippines (14 percentage points). generalized increase in domestic financial However, East Asia is the developing re- intermediation. Of the 30 countries with gion with the deepest domestic financial complete data, 9 experienced declines in do- systems, and the region's ratio of bank mestic intermediation relative to GDP, and credit to GDP exceeded that of the United 12 countries experienced increases of less States (although remaining below that of than 5 percentage points. Sufficient data the more bank-based systems in Western on stock market capitalization are reported Europe). The further deepening of finan- for only 13 countries. The strong increase cial markets was reflected in the more than is attributable to capitalization more than tripling of stock market capitalization over tripling relative to GDP in Côte d'Ivoire, the period. Kenya, Mauritius, and Nigeria. The high level of stock market capitalization rela- tive to output, however, is attributable The rise in foreign flows to South Africa, where the level reached The increase in domestic financial intermedia- nearly three times output in 2006. Because tion during the liquidity boom was accompa- South Africa attracts investment from other nied by a rapid expansion of capital inflows economies in the region that lack stock (figure 2.7). Similar to increases in domestic markets and are hence not included in the credit, higher capital inflows can boost invest- average, the average tends to overstate ment and efficiency (box 2.5). the level of capitalization for the South While virtually every country saw inflows African economy per se. Excluding South rise, they did not rise by the same amount Africa, the region has the lowest level of in all countries, and not all forms of inter- stock market capitalization relative to out- national capital flow increased to the same put of the six developing regions. degree. Portfolio equity flows to developing The small average increase in credit to the countries increased rapidly before the financial private sector relative to output in Latin America and the Caribbean reflects very dif- ferent outcomes across countries, ranging Figure 2.7 Total capital inflows to developing from a decline of more than 26 percentage economies points in Bolivia and Uruguay to an in- US$ billion Share of GDP, % crease of 17 percentage points in Colombia 1,200 9 and Costa Rica. Macroeconomic policies in 8 1,000 Latin America have improved greatly since 7 Share of GDP their boom-and-bust experiences over the 800 6 last decades of the 20th century, and many 600 5 countries avoided an excessive buildup of 4 400 3 private credit and achieved steady growth 2 in incomes. Compared with most other re- 200 1 gions, the doubling of stock market capital- 0 Total capital inflows 0 ization was modest and may have reflected 1990 1992 1994 1996 1998 2000 2002 2004 2006 policy prudence by authorities in the region Source: World Bank. seeking to avoid an asset-price bubble. 58 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S Box 2.5 Capital flows can boost investment and efficiency M ost developing countries relied on external finance during the 2003­07 boom. Develop- ing countries' aggregate current account surplus External finance can improve efficiency by en- hancing the transfer of technology from more developed economies, helping firms achieve larger (which averaged almost $243 billion during this pe- size and thus benefit from economies of scale, build- riod) mainly reflected large surpluses of savings over ing reputations in global markets, and establishing investment in a few countries, notably China, and business and marketing contacts for developing coun- developing oil and mineral exporters. Three-fourths tries' exports (World Bank 2006). These effects can of the remaining developing countries for which be indirect or arrive more directly, as can be the case data are available were net importers of capital, with with some forms of foreign direct investment, if the current account deficits that averaged more than result is the importation of more sophisticated 6 percent of their GDP and 28 percent of their total machines or business techniques. investment spending (box table 2.5.1). Box table 2.5.1 Developing countries with current account deficits, 2003­07 Number of countries with Current account Current account deficit current account deficits deficit (% of GDP) (% of investment) All countries 53 6.3 26.8 Low income 16 5.8 29.1 Lower middle income 20 6.1 23.4 Upper middle income 17 7.1 28.3 Source: World Bank. Note: Data on current account deficits are simple averages of country numbers. Small island economies are excluded. crisis, from near zero in 2001 to $160 billion Developing countries' access to exter- in 2007, followed by a total collapse in 2008 nal bond markets and foreign bank lend- (figure 2.8). ing increased markedly during the liquidity boom, reaching a peak of 4 percent of developing-country GDP in 2007. Net FDI Figure 2.8 Portfolio equity flows to inflows increased from about 2.5 percent developing countries of GDP in 2001 to 3.9 percent in 2007 be- Portfolio investment, equity (US$ billion) Share of GDP, % fore falling slightly in 2008, along with the 160 1.0 reduction in global investment in general 140 Share of GDP 0.8 (figure 2.9). Official flows, in contrast, re- 120 100 0.6 versed from net inflows of $26 billion in 2001 80 to net outflows of $0.1 billion in 2007. 0.4 60 At the regional level, Europe and Central 40 0.2 Asia, East Asia and the Pacific, and Latin 20 0 America were the largest recipients of capital 0 Portfolio investment inflows, receiving more than 80 percent of net ­20 ­0.2 1980 1985 1990 1995 2000 2005 inflows over 2001­07, with the first two re- Years gions together accounting for 65 percent of the Source: World Bank. total. However, expressed as a share of GDP, the differences in inflows across regions were 59 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 1.5 percent of their GDP in 2001 to almost Figure 2.9 FDI inflows to developing countries, 1980­2008 7 percent in 2007, largely supported by the rise in resource-related FDI. FDI net inflows (US$ billion) Share of GDP, % 700 4.0 Across regions, the relative importance of 3.5 different types of capital flows varied some- 600 3.0 what. In most regions equity (especially FDI) 500 Share of GDP 2.5 accounted for both the bulk of capital inflows 400 2.0 in 2007 and most of the increase in inflows 300 1.5 over 2001­07 (table 2.3). In developing 200 1.0 Europe and Central Asia, however, net debt 100 0.5 flows grew from almost nothing in 2001 to 0 FDI 0 almost 10 percent of GDP. As such they rep- 1980 1985 1990 1995 2000 2005 resented about two-thirds of total inflows in Years 2007. Had Europe and Central Asia received Source: World Bank. the same increase in debt flows as other devel- oping regions, its overall inflows would have less pronounced--both in 2001, and in 2007 been closer to 8 percent of GDP, similar to when flows peaked. Flows to East Asia and those received by East Asia, South Asia, and the Pacific, relative to GDP, were only slightly Sub-Saharan Africa. Although many factors above the developing-country average in 2001 underpin the strength of debt inflows to the and were actually below average in 2007. In region--including enthusiasm for the region's contrast, while flows to developing Europe long-term prospects within the European and Central Asia as a share of GDP were below Union and the high share of foreign banks in average in 2001, they grew about fivefold by the overall banking sector--the population's 2007. South Asia also saw inflows increase willingness to take on exchange rate risk by very rapidly, from only about 1 percent of borrowing in foreign currencies helps to explain GDP in 2001 to more than 8 percent in 2007. why bank lending--including to private indi- Contrary to accepted wisdom, Sub-Saharan viduals--played such a prominent role. Africa actually received close to average flows At the country level, absolute flows are (relative to GDP) both in the pre-boom and extremely concentrated, with China, India, end-of-boom periods. Both middle-income the Russian Federation, and Brazil account- and low-income countries benefited from ing for about 50 percent of net inflows both the surge in capital flows. The flows to low- in 2007 and, on average, over 2001­07; the income countries more than quadrupled, from four also account for 73 percent of all flows Table 2.3 Net capital inflows by region Avg. 2001 2007 2001­2007 2001 2007 2001 2007 (% of total flows to (US$ billion) (% of region's GDP) developing countries) Developing countries 223 1,143 470 4 9 East Asia and Pacific 83 277 141 37 24 5 7 Europe and Central Asia 29 454 164 13 40 3 15 Latin America and the 87 215 87 39 19 4 6 Carribean Middle East and North 5 21 12 2 2 1 3 Africa South Asia 8 116 39 4 10 1 8 Sub-Saharan Africa 11 60 27 5 5 3 7 Source: World Bank. 60 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S The quality of domestic institutions (prox- Figure 2.10 Distribution of capital flows as a ied here by the Kaufmann­Kraay­Zoido- percentage of GDP in 2007 Lobaton index) is also correlated with both Percent of countries domestic and external finance. Demand for 35 32.1 capital will depend on the potential revenues 30 28.4 from a physical investment. Both domestic and 25 international investors operating in countries 20 with strong institutions and a well-functioning 16.5 regulatory environment, including reasonable 15 protection of property rights, will likely earn 10 9.2 higher real-side returns and therefore, all else 5.5 5 equal, be willing to take on more debt. Simi- 2.8 0.9 0.9 0.9 1.8 0.9 larly, lenders providing finance to borrowers 0 ­20 0 20 40 60 in countries with strong institutions and pro- Capital flows as % of GDP tection of property rights would be more likely Source: World Bank. to be able to enforce their claims for repay- ment and hence would be willing to lend more. Finally, the extent of real-side integration of an economy is also a good predictor of the (figure 2.10). However, relative to GDP, both extent of financial intermediation and private the flows and the change in flows are more capital inflows that a country receives (Fig- evenly distributed, with about 60 percent of ure 2.11 panel C. In the recent boom period, countries receiving flows of between 0 and external factors such as the high price of com- 10 percent of GDP in 2007 (figure 2.10). modities were also at play. Interestingly, while per capita income levels are highly correlated with the level of domestic intermediation Real-side consequences of the (figure 2.12), the size of capital flows is only surge in global finance weakly related to income. T he extent to which a given developing country benefited from the surge in global liquidity depended on a wide variety of factors, Although these correlations provide some insight into the differences in intermediation levels at a given point in time, they do not many of which are impossible or at best very speak to what drove the changes observed dur- difficult to measure in a consistent manner ing the boom (table 2.4). across countries. By far the biggest drivers of the observed Figure 2.11 reports simple correlations be- changes in the availability of domestic and tween private finance (as represented by do- international finance were changes in the cost mestic intermediation in the first column and of capital, here operating through the reduc- foreign capital inflows in the second column) tion of interest rates in high-income coun- and borrowing costs, the quality of institu- tries and interest rate spreads in developing tions, and the extent of real-side openness countries. Cross-country regressions (box 2.6) (all data are expressed in terms of the average suggest that for the average developing country from 2001 through 2007). Unsurprisingly, the a 500 basis point decline (roughly the mean de- levels of both domestic intermediation and pri- cline observed over the estimation period--as vate capital inflows are negatively correlated well as a standard deviation across the sample with borrowing costs--although the simple bi- of countries for which comparable data are variate correlation illustrated here is not very available) in borrowing costs resulted in an in- strong, mainly because of the interaction of crease in the level of domestic intermediation other factors (see below). equal to 4.5 percent of GDP and an increase in 61 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Figure 2.11 The determinants of private finance Domestic intermediation International private capital flows Panel A: vs. the cost of capital Domestic credit to private sector, % of GDP (avg. 2001­07) Private capital flows as % GDP (avg. 2001­07) 140 35 120 30 100 25 80 20 60 y= 1.56x 55.90 15 y= 0.22x 10.03 R2 = 0.04 10 40 R2 = 0.02 20 5 0 0 20 5 10 12 14 16 18 20 22 24 10 12 14 16 18 20 22 24 Cost of capital (avg. 2001­07) Cost of capital (avg. 2001­07) Panel B: vs. institutional quality Domestic credit to private sector, % of GDP (avg. 2001­07) Private capital flows as % GDP (avg. 2001­07) 140 30 120 y = 0.86x 0.93 25 R2 = 0.34 y = 0.13x 1.79 100 20 R2 = 0.13 80 15 60 10 40 5 20 0 0 5 0 20 40 60 80 100 0 10 20 30 40 50 60 70 80 90 KKZ index (avg. 2001­07) KKZ index (avg. 2001­07) Panel C: domestic credit and foreign private capital inflows vs. trade openness Domestic credit to private sector, % of GDP (avg. 2001­07) Private capital flows as % GDP (avg. 2001­07) Domestic credit and exports Private capital flows and exports 140 30 100 25 y = 0.13x 1.50 120 y = 0.50x 11.26 20 R2 = 0.15 80 R2 = 0.13 15 60 10 40 5 20 0 0 5 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 Exports of goods and services as % of GDP (avg. 2001­07) Exports of goods and services as % GDP (avg. 2001­07) Source: World Bank. foreign capital inflows of 0.5 percent of GDP. decline in the price of global risk (about the Likewise, panel estimates suggest that financial decline observed between 2003 and 2007) conditions in developing countries were even could result in an increase of 3.5 percent of more sensitive to international financial condi- GDP in foreign capital flows and an increase of tions. According to these estimates, a 1 point 7.5 percent of GDP in domestic intermediation 62 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S of capital. As a result, changes in the cost of Figure 2.12 Private credit from banks and other financial institutions relative to per capital (broadly understood to include the in- capita income, 2007 ternational price of risk) on average accounted GDP per capita in $US for almost one-half of the observed fluctua- 70,000 tion in capital inflows and about 60 percent of the increase in domestic intermediation (see 60,000 table 2.4), with domestic intermediation being 50,000 the only other quantitatively important factor 40,000 in the determination of net capital inflows. 30,000 However, other factors, including insti- tutional quality, overall economic openness, 20,000 and the extent of domestic financial sector in- 10,000 termediation (in the case of the capital flows 0 equation), were critical in explaining the wide 10,000 differences in the levels of intermediation 0.5 1.0 1.5 2.0 2.5 and inflows across countries both before and Share of private credit in GDP during the boom (lower panel of table 2.4). Sources: International Financial Statistics; World Bank. Cross-country differences in institutional quality (as measured by the Kaufmann­ Kraay­Zoido-Lobaton Index) explained al- (amounting to, respectively, more than two- most six-tenths of the variance in the level thirds and more than three-fourths of the mean of domestic intermediation across countries increase observed over the estimation period). and about one-third of the difference in net The association is especially strong for debt in- capital inflows. Indeed, a one-standard- flows, and, not surprisingly, it does not hold deviation improvement in institutional quality for official aid. (roughly equivalent to the average difference Changes over time in other important deter- in institutional quality between Sub-Saharan minants of domestic and international financial Africa and Latin America) could generate a intermediation were not as large as the changes 12 percentage point increase in the ratio of in international capital conditions and the cost private sector domestic credit to GDP, and an Table 2.4 Intertemporal changes in financial variables mainly reflected the cost of capital, but across countries institutional quality was most important Financial variables Net capital flows Domestic intermediation (As a percent of GDP) Change over 2001­07 in sample mean of dependent variable 4.9 8.6 Contributions of changes in (sample mean of): Global cost of risk 2.2 5.1 Institutional quality ... 0.08 Domestic intermediation 1.8 ... Difference in 2007 between top and bottom quartile in dependent variable Contribution of differences in: 1.61 34.5 Cost of capital ... 2.1 Institutional quality 3.7 19 Exports of GNFS 5.2 12.8 Source: World Bank. Notes: Calculations based on estimates reported in box 2.6. . . . Not estimated. 63 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Box 2.6 Determinants of cross-country differences in domestic and international financial intermediation D ata limitations among other constraints pre- vent a comprehensive modeling of the factors that explain the extent of the expansion of domestic Both as a robustness check and to explore the role of the country-invariant risk-premium variable (dis- cussed in chapter 3), panel regressions were also run and international finance in developing countries in for the period 2001­07, with net capital inflows and response to the global loosening of monetary condi- domestic intermediation as the dependent variables, tions. However, cross-country regressions that seek and the risk premium plus the full set of regressors to explain the average change in domestic intermedia- from the cross-sectional analysis as the independent tion (credit to the private sector) and international variables. All independent variables were lagged, to capital flows as a percentage of GDP provide im- diminish endogeneity concerns. These regressions portant insights into the role of the country-specific confirm a statistically significant association between potential explanatory variables (including changes both domestic intermediation and capital inflows on in the cost of capital, institutional quality, finan- the one hand and the international price of risk and cial development, exports, the budget surplus, and financial development on the other (box table 2.6.2, inflation). columns 2 and 3). While the variation in the level of These regressions confirm a statistically significant domestic financial intermediation was significantly association between the level of domestic intermedia- associated with institutional quality, the variation in tion in developing countries and institutional quality, international capital flows was not. Nor did the cost the share of exports in GDP, and their rate of growth of capital have an independent influence on either (box table 2.6.1, column 2). The results also indicate domestic or international intermediation beyond that a significant association between the level of inter- of the price of international risk, likely reflecting the national capital flows and institutional quality and strong link between variations in the two variables exports (box table 2.6.1, column 3). (see chapter 3). Box table 2.6.1 Cross-sectional regressions results Domestic intermediation Net capital inflows (Private sector credit, % of GDP, (% of GDP, Dependent variable Average 2001­07) Average 2001­07) Explanatory variables Coefficient Coefficient Cost of capitala 0.56* 0.02 Institutional quality (Kaufmann­Kraay­Zoido-Lobaton index)b 0.69*** 0.13*** Financial development (private sector credit, % of GDP) -- 0.05 Export of goods and nonfactor services (% of GDP) 0.28** 0.13*** Export growth 0.53** 0.244*** Budget surplus (% of GDP) 1.23*** 0.12 Inflation (logs, percent) 2.38 0.52 R2 0.46 0.36 Source: World Bank. Note: All regressions estimated using average values over the period 2001­07 for the dependent variables, and initial values for the independent variables; number of countries 103. Other controls include export growth, 1990­97 (percent, average annual rate); and indicators for countries in the upper quartile of both the fuel exports/GDP and the metals exports/GDP distribution. *, **, and *** denote significance at, respectively, the 10 percent, 5 percent, and 1 percent level. Significance is evaluated using robust standard errors. a. Measured as the U.S. T-bill rate, plus the country-specific spread, plus depreciation. b. Measured on a scale of 0 to 100, with a cross-sectional standard deviation of 19. -- Not applicable. 64 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S Box table 2.6.2 Panel regression results Domestic intermediation Net capital inflows Dependent variable (private sector credit, % of GDP) (% of GDP) Explanatory variables Coefficient Coefficient Global cost of riska 3.49*** 1.47* Cost of capitalb 0.03 0.03 Institutional quality (Kaufmann­Kraay­Zoido-Lobaton index)c 0.30** 0.01 Financial development (private sector credit, % of GDP) -- 0.27*** Exports of goods and non-factor services (% of GDP) 0.08 0.02 Budget surplus (% of GDP) 0.07 0.03 Inflation (logs, %) 0.12 0.13 R2 0.16 0.10 Source: World Bank. Note: All regressions estimated using annual data over the period 2001­07, with all independent variables lagged once; number of obser- vations 498, 493. Other controls include indicators for countries in the upper quartile of both the (fuel exports/GDP and the metals exports/GDP distribution; and a full set of country-specific fixed effects. *, **, and *** denote significance at, respectively, the 10 percent, 5 percent, and 1 percent level. Significance is evaluated using robust standard errors. a. See above for details. b. Measured as the U.S. T-bill rate, plus the country-specific spread, plus depreciation. c. Measured on a scale of 0 to 100. -- Not applicable. increase of 2 percent of GDP in private capi- average over 2001­07, an extra 0.5 percent tal flows after controlling for all other fac- of GDP in foreign capital inflows, and its tors. Countries with large export sectors and total domestic intermediation amounted to therefore a proven track record with foreign an extra 1.5 percent of GDP. Cross-country partners also tend to receive more foreign fi- differences in the extent of real-side openness nancing than those with weaker external ties. were associated with about one-third of the A country whose export sector was 5 percent- differences in net capital flows and in domes- age points larger than another's received, on tic intermediation. Table 2.5 Regional distribution of changes in financing conditions, 2000­07 Change between 2007 and 2000 in: Cost of Capital Stock market Private credit by capital inflows capitalization deposits money banks Investment (Basis points) (% of GDP) Developing countries 400 5.0 78.6 5.5 5.5 Low-income countries 2.3 Middle-income countries 5.6 East Asia and Pacific 134 2.0 118.0 10.7 5.5 Europe and Central Asia 866 12.0 59.8 15.7 4.9 Latin America and the Caribbean 471 2.0 39.8 2.2 1.4 Middle East and North Africa 269 2.0 36.2 6.2 5.0 South Asia 142 7.0 107.3 14.8 8.1 Sub-Saharan Africa 685 4.0 59.1 6.8 3.6 Sources: World Bank; Beck and Demirgüç-Kunt 2009; World Bank 2009. Note: Regional values are simple averages of countries, except for investment rates which are weighted averages. 65 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 The association between capital inflows and The liquidity boom and macroeconomic stability (as represented by the macroeconomic performance budget surplus and inflation) was in general The sharp increase in capital inflows to develop- not statistically significant, after controlling for ing countries and the rapid expansion of domes- the cost of capital, institutional quality, export tic finance were associated with a generalized intensity, and the extent of financial sector in- investment boom, although some countries were termediation. Although one would expect that more or less successful in transforming addi- macroeconomic stability would be an impor- tional finance into productive investments.9 On tant determinant of credit worthiness and as a average, between 2000 and 2007 investment-to- result the size of capital flows, the data suggest GDP ratios in developing countries increased by that the relationship is relatively weak. 5.2 percentage points, or 23 percent, compared Overall, ample global liquidity was a de- with their 2000 levels (table 2.6). termining factor in the surge in global capital Investment rates rose in all regions, most flows to developing countries, but where those markedly in South Asia, the Middle East, and flows went and in which form depended im- Sub-Saharan Africa. The very marked increase portantly on the characteristics of individual in investment rates in South Asia (up by more developing countries. Country-specific "pull" than 10 percentage points) partly reflects deep factors, such as the quality of the institutional structural reforms that were undertaken during environment and overall economic openness, the 1990s, the influence of which on investment shaped the direction of capital flows and the was redoubled by falling borrowing costs. In extent to which the domestic intermediation re- the rest of the developing world the rise in in- sponded by increasing the availability of credit. vestment rates was more modest. Rates in low- It follows that even in an international income countries rose by 6 percentage points environment in which capital may become versus 5.2 percentage points in middle-income scarcer and more expensive, countries can countries (inclusive of India). Despite the very take steps that can deepen their domestic strong capital inflows received by countries in capital markets and increase their access Europe and Central Asia, investment rates in to international capital. In particular, the that region rose by only 3.5 percentage points-- evidence suggests that improvements in the much less than the overall average for middle- regulatory environment, increased market income countries. By 2007, just before the onset openness, and more generally reforms that of the crisis, investment rates in East Asia and improve the business environment and re- duce the cost of capital can substantially Table 2.6 Rising investment rates by influence the level of capital inflows and region financial intermediation in a given country, Investment rate especially in Africa where the quality of in- stitutions remains well below the average 2000 2007 Change (%) (%) (% points) elsewhere. Indeed, in the expected tougher global environment, such factors are likely Developing countries 22.7 28 5.2 to be even more critical in determining the Middle-income countries 22.8 28 5.2 Low-income countries 21.1 27.1 6.0 direction of future flows--placing even East Asia and Pacific more value on forging ahead with further (excluding China) 22.1 26 3.9 reforms. Sufficient progress in these areas China 34.1 38.8 4.7 across enough countries could well mitigate Europe and Central Asia 19.9 23.4 3.5 Latin America and the Caribbean 18.6 22.1 3.5 to a large degree the expected increase in risk Middle East and North Africa 22.4 27.0 4.6 aversion, potentially allowing capital flows South Asia 22.0 32.8 10.8 in the longer run to regain more recent levels Sub-Saharan Africa 16.9 20.9 3.9 (see discussion in chapter 3). Source: World Bank. 66 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S Table 2.7 Intertemporal and cross-country Investment does not, of course, mechani- influences on investment cally translate into greater output and living Change over 2001­07 in investment/GDP standards: its efficiency must also be taken into (sample mean)a 5.4 account. In this context, additional economet- Contributions of changes in: Global cost of risk 1.9 rics suggest that increased financing was most Domestic intermediation 0.6 likely to lead to increases in growth in those Terms of trade 1.4 countries where the quality of institutions was Difference in 2007 between top and bottom quartile in investment/GDPb 11.5 high, a result that is consistent with the recent Contributions of differences: literature (Frankel 2009).10 Cost of capital 3.3 Net capital inflows/GDP 3.0 Source: World Bank. Impact of the investment boom on a. Based on panel regressions. b. Based on cross-sectional regressions. growth and potential output The prolonged reduction in interest rates during the Pacific exceeded 26 percent of GDP. Those the liquidity boom was associated with a rise in in Sub-Saharan Africa were much more modest potential output. Normally, the increase in in- (about 21 percent of GDP) but were neverthe- vestment from a fall in interest rates would be less 3.9 percentage points higher than in 2000. relatively short-lived (as would be the period of Many factors help explain the extent to low interest rates). During this most recent bub- which investment rates differ across coun- ble, however, interest rates remained low for tries and rise in some countries but not others a very long time, and as a result investors and (table 2.7; box 2.7). On average, about one- economists alike began to talk of a new regime third of the increase in investment rates ob- likely to be characterized by low interest rates. served between 2001 and 2007 is accounted If investors' expected interest rates (and with for by the reduction in the global cost of risk, a them the cost of capital) had decreased on a per- further 11 percent by increased domestic inter- manent basis,11 then economic theory suggests mediation, and about one-fourth by improve- that investors would have sought to increase the ments in the terms of trade in some countries. amount of capital they employed to produce a Cross-country differences were more than given level of output. As predicted by theory, twice as large as the changes over time, with during this transition period to a higher capital about 30 percent of the variation accounted output ratio, investment grew faster than usual for by differences in the cost of capital and a and the ratio of the stock of capital to GDP rose further 30 percent by the level of capital in- (figure 2.13). As a result, the rate of growth of flows a country attracts. The impact of do- potential output increased--see box 2.8 for a mestic intermediation and institutional quality brief description of the model of potential out- was not statistically significant here, possibly put employed here; the online technical annex reflecting difficulties in disentangling their (available at www.worldbank.org/GEP2010) to effect from that of the cost of capital. Con- this chapter provides further details--more rap- cretely, these results suggest that a reduction idly than normal during this period. in the cost of capital from the average level Overall, the rate of growth of potential found in Sub-Saharan Africa to that prevailing output among developing countries increased in Latin America would be associated with an by an average of 1.5 percentage points be- increase in investment equal to almost 2 per- tween 2003 and 2007 as compared with the cent of GDP. This reinforces the importance pre-boom period 1995­2002, with 40 percent of continuing with structural reforms aimed of that increase attributable to increased capi- at expanding still-underdeveloped financial tal services as a result of higher investment sectors (a point confirmed by the simulations rates.12 Table 2.8 breaks down this aggregate discussed in chapter 3). result across different regions. Although both 67 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Box 2.7 Understanding the increase in investment rates B ox table 2.7.1 reports cross-sectional regres- sion results that seek to describe differences in investment across developing countries in terms of development, and institutional quality, to capture the notion that domestic conditions may affect the ef- ficiency of investment. All independent variables were differences in the cost of capital, institutional quality, lagged, to diminish endogeneity concerns. domestic intermediation, and international capital The results (box table 2.7.2) confirm a statistically inflows, among other explanatory variables. These significant association between investment on the one regressions confirm a statistically significant associa- hand and both the global price of risk and domestic tion between investment ratios on the one hand and intermediation on the other. Even after controlling initial values of the cost of capital and international for the latter factors, the terms of trade have a sig- capital inflows on the other. Both as a robustness nificant impact on investment. In contrast, the impact check and to explore further the changes in invest- of the cost of capital, institutional quality, and inter- ments observed over time, including the role of the national capital flows is not statistically significant, country-invariant global risk premium, panel regres- possibly reflecting difficulties in disentangling their sions were also run for the period 2001­07, with effect from that of other variables. Additional regres- investment ratios as the dependent variables, and sion analysis, not reported here, indicates that equity the risk premium, plus the full set of regressors from capital inflows, notably FDI flows, have a stronger the cross-sectional analysis, as the independent vari- effect on investment rates than on international debt ables. Also included as possible explanatory variables flows (bonds and bank lending). were interactions between capital inflows, financial Box table 2.7.2 Investment to GDP ratio, panel regression results Coefficient Box table 2.7.1 Investment-to-GDP ratio, cross-sectional regression results Global cost of risk 1.33* Cost of capital 0.10 Coefficient Institutional quality (Kaufmann­Kraay­Zoido-Lobaton index)a 0.08 Cost of capital 0.59*** Financial development Institutional qualitya 0.00 (private sector credit, percent of GDP) 0.08* Financial development 0.04 Net capital inflows/GDP (percentage points) 0.34 Net capital inflows 0.45* Terms-of-trade index, weighted by trade ratio 0.06** Change in terms of trade 0.01 R2 0.24 R2 0.22 Source: World Bank. Source: World Bank. Note: All regressions estimated using annual data over the period Note: All regressions estimated using average values over the 2001­07, with all independent variables lagged once; number period 2001­07 for the dependent variable, and initial values for of observations 430. Other controls include trade-weighted the independent variables; number of countries 106. Other export market growth (percent); indicators for countries in the controls include trade-weighted export market growth (percent) upper quartile of both the fuel exports/GDP and the metals and indicators for countries in the upper quartile of both the fuel exports/GDP distribution; and a full set of country-specific fixed exports / GDP and the metals exports / GDP distribution. *, **, effects. *, **, and *** denote significance at, respectively, the and *** denote significance at, respectively, the 10, 5, and 1 per- 10, 5, and 1 percent level. Significance is evaluated using robust cent level. Significance is evaluated using robust standard errors. standard errors. a.Measured on a scale of 0 to 100. a. Measured on a scale of 0 to 100. 68 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S Box 2.8 Estimating potential output in developing countries T his Global Economic Prospects introduces new estimates of potential output based on a hybrid production-function model of potential output simi- as an expression for the rate of growth of potential output. For the purposes of this study, the capital stock lar to that used by the Congressional Budget Office was estimated using the perpetual inventory method (CBO) in the United States, the OECD, the Euro- from investment data (running from 1960 in the pean Commission and the Federal Reserve Board case of most countries) and assuming a deprecia- (CBO 2001; OECD 2008; Cournède forthcoming; tion rate of 7 percent (IMF 2005). Trend TFP was Denis and others 2006). In this model, which is calculated using an Hodrik-Prescott filter through described in more detail in the online annex to this spot estimates of TFP calculated by inverting the publication, the supply side of GDP is described by a above equation in level terms. The end-point prob- simple Cobb-Douglas function of the form lem was resolved by assuming that TFP growth from GDP AKaL1 a, 2008 through 2009 was equal to the average rate of growth of TFP during the period 1996­2006. The where GDP is gross domestic product, K is the capi- share of capital income in total output (alpha) was tal stock, and L is labor employed. Potential output assumed to be a uniform 40 percent in all developing is the level of output attained when the entirety of the countries. capital stock and effective labor supply is employed. An alternative approach used until recently by the Replacing L with the working-age population (P1565), OECD (it was recently abandoned in favor of one the labor force participation rate (Pr), and the unem- similar to that described here) calculates the capital ployment rate (UNR) gives stock on the basis of a smoothed investment rate series. This results in an estimate of potential that is GDP AKa(P1565 * Pr *(1 UNR))1 a. less sensitive to cyclical changes in investment behav- And stating everything in growth terms gives ior but has the disadvantage that full employment capital services are disconnected from the actual ob- y TF P aK (1 a) * (P1565 Pr (1 UN R)) servable capital stock. In the words of the U.S. Con- Assuming that all of the capital stock and all of gressional Budget Office, which also eschews using the labor force are fully employed (UNR and Pr the smoothed investment method, "unlike the labor equal their equilibrium values), that all of the services input, the capital input does not need to be cyclically of the available capital stock are used, and that total adjusted to create a `potential' level--the unadjusted factor productivity (TFP) is growing at its trend rate capital input already represents its potential contri- gives an expression for the rate of growth of poten- bution to output. Although use of the capital stock tial. For most developing countries, we do not have varies greatly during the business cycle, the potential reliable economy-wide data for Pr and UNR, so for flow of capital services will always be related to the the purposes of calculating the rate of growth of total size of the capital stock, not to the amount cur- potential, it suffices to assume that the equilibrium rently being used" (CBO 2001). unemployment and participation rates are constant, The use of actual rather than a smoothed capital which leaves us with stock means that the output gap fluctuates less over the cycle. y TF P aK (1 a) * (P1565) 69 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 results are sensitive to the level of investment Figure 2.13 Rising investment rates contri- buted to an acceleration in potential output rate used in the counterfactual exercise, the broad result that capital deepening accounted Investment to GDP ratio, % Growth of potential output, % 33 8 for almost half of the acceleration in potential 31 output observed during this period is robust to 29 7 different specifications.13 27 6 Growth of potential output 25 5 23 21 Investment to GDP ratio 4 Concluding remarks 19 17 15 3 2 F inance, whether it is delivered through the domestic banking system or originates from abroad, is an important enabler of eco- 1995 1997 1999 2001 2003 2005 2007 nomic development. At its best, it improves Source: World Bank. efficiency by funding potential-enhancing in- vestment projects that would otherwise not have been funded and by promoting and fa- middle- and low-income countries saw their cilitating the transfer of technologies and the potential growth rates increase by about the spread of best practices within an economy. same amount, with capital deepening account- However, the extent to which an increase in ing for a larger share of the total among low- intermediation is able to achieve these results income countries, with the remaining 60 per- depends importantly on the quality of domestic cent increase attributable to growth in popu- institutions, regulations, and overall absorptive lation and in total factor productivity. In capacity of an economy. Where the supply of the case of China, almost all of the increase credit, whether domestic or foreign in origin, in output during this period can be ascribed exceeds the absorptive capacity of an economy, to increases in the capital stock. While these it can lead to macroeconomic instability and thus make a negative contribution to long-term growth and potential output. Table 2.8 Decomposition of increase in potential output growth directly attributable For the vast majority of developing coun- to capital deepening tries, the period of 2000­07 was one of very liquid financial conditions. Both domestic and Change in growth rate of potential output (2003­2007 vs 1995­2003) international finance expanded rapidly, with those countries most open to world trade Due to Share due capital to capital and finance receiving the largest shares of Total deepening deepening the increase in credit. For most countries this expansion fueled an investment boom that Developing countries 1.5 0.6 40.3 contributed to faster productivity growth and Middle-income countries 1.5 0.6 39.8 Low-income countries 1.3 0.8 63.7 increased potential output through capital East Asia and Pacific deepening--without generating domestic in- (excluding China) 0.4 0.1 19.8 flation or serious external imbalances. That in China 0.3 0.9 283.5 Europe and Central Asia 3.1 0.6 18.7 turn suggests that for these countries a preex- Latin America and the isting capital constraint was at least temporar- Caribbean 0.3 0.1 46.6 ily relieved, ushering in a golden age of rapid Middle East and North Africa 0.8 0.5 66.7 and, at least at the country level, sustainable South Asia 1.4 1.1 78.5 growth. For a few countries, most notably a Sub-Saharan Africa 1.9 1.5 79.5 number in the Europe and Central Asia region, Source: World Bank. inflows and domestic credit creation either 70 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S exceeded the domestic economy's absorptive 2. The lack of information available to buyers of capacity or found its way into nonproduc- these instruments also should reduce their price. How- ever, sustained low interest rates during the 2002­07 tive hands, helping to feed an unsustainable boom appear to have eroded concerns over risk taking increase in consumer demand that generated on the part of many investors. Information asymme- large and ultimately unsustainable internal tries may also be mitigated by more stringent covenants and external imbalances. on loans sold on secondary markets than on loans held The financial crisis has brought an end to by the originating bank, although it is difficult for cov- these favorable conditions for both groups enants to anticipate all potential repayment issues. of developing countries. For the moment, the 3. Data on reported CDS spreads are taken from Datastream. most serious impacts have been felt in those 4. Over-the-counter derivatives played an impor- countries where the largest imbalances accrued. tant role in the excessive volatility affecting foreign Going forward as financial conditions improve, currency and asset markets during the East Asian crisis conditions in developing countries should also of 1997­98 (Kregel 1998). improve. But growth rates are unlikely to re- 5. Firms operating in countries at low levels of fi- gain their boom-period levels, if global liquid- nancial development are constrained from making the ity is both more expensive and less abundant in investments required to assimilate new technologies (Aghion and others 2004). Moreover, the intermedia- coming years, particularly over the next several tion services of a healthy financial sector also contrib- years as countries adjust to tighter international ute to development, efficiency, and economic growth conditions. International capital flows to devel- by enabling arms-length transactions that increase oping countries are not expected to reach their competition and the range of options available for both pre-crisis levels in the medium term. Competi- suppliers and buyers. Financial intermediation also tion among developing countries to attract in- helps to move resources from less productive uses to vestment flows (such as FDI) will be tougher more productive ones, and to reduce information and transactions costs, such as the cost of acquiring infor- than in previous years. Factors such as institu- mation on investments, monitoring of firms' managers, tional quality, trade openness, and regulatory and enforcing contracts (Levine 1997). framework will play an increasingly important 6. Rothenberg and Warnock (2006) find that nearly role in attracting these cross-border investments half the "sudden stop" crises in emerging markets can and financial intermediation. To what extent best be attributed to capital flight by local investors, financial conditions and developing-country while Cowan and others (2008) find that one in five growth potential will be affected will depend episodes are driven by surges in outflows rather than stops in inflows. importantly on the nature of the changes to 7. During the recent boom, the biggest expansion in come in the international financial architec- finance (both domestic and external) among the devel- ture, the extent that these changes impinge oping regions was in Europe and Central Asia, largely on financing conditions for developing coun- reflecting optimism about long-term prospects for the tries, and the success with which developing region given its quality labor force and its increasing countries are able to offset the less propitious political and economic integration with high-income external conditions by improving domestic fi- EU economies. Unlike other regions, the expansion in finance (increases of 12 percent of GDP in external nancial conditions. The nature of these changes flows and 15.6 percent in domestic intermediation) and their expected impact on growth and the exceeded the absorptive capacity of many countries, growth potential of developing countries are spilling over into increased consumption, inflation, and explored in more detail in chapter 3. rising current account deficits. 8. Bond markets also increased significantly in some of the middle-income countries, as discussed in World Notes Bank 2009. 1. Total claims on BIS-reporting banks increased by 9. At first blush, this appears to contradict some 21 percent a year on average between 2002 and 2007, of the evidence outlined in box 2.3 suggesting that compared with a 10 percent annual increase in nominal increased intermediation increases GDP and invest- world GDP. ment mainly by contributing to increased total factor 71 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 productivity. However, they are consistent but here the Beck, Thortsen and Asli Demirgüç-Kunt. 2009. "Financial expansion in potential output that drives an increase Institutions and Markets across Countries and over in investment derives (principally) not from increased Time: Data and Analysis." Policy Research Working TFP but from a fall in the cost of capital. Although Paper 4943, World Bank, Washington, DC. TFP growth did increase during the 2000s, the sharp- Beck, Thorsten, Asli Demirgüç-Kunt, and Ross Levine. est change observed was the decline in borrowing 2007. "Finance, Inequality and the Poor." World costs and with them the cost of capital. As a result, the Bank, Washington, DC. quantity of capital that could be supported by a given Beck, Thorsten, Ross Levine, and Norman Loayza. level of productivity and labor increased, inducing an 1999. "Finance and the Sources of Growth." increase in investment, an acceleration in GDP growth, Policy Research Working Paper 2057, World and an increase in potential output. Bank, Washington, DC. 10. These regressions also suggested a negative Bernanke, Ben. 2005. "The Global Savings Glut and the relationship between growth and the global price of U.S. Current Account Deficit." Sandridge Lecture, risk and the cost of capital. However, this link may be Virginia Association of Economics, Richmond, spurious: the estimated negative relationship between Virginia. April 14. http://www.federalreserve. growth and domestic credit expansion suggests that the gov/boarddocs/speeches/2005/200503102// data may be finding it hard to disentangle the impact BIS (Bank for International Settlements). 2006. 76th of the domestic credit expansion from that of some of Annual Report. Basel. its underlying determinants. Calvo, Guillermo A. 2009. "Financial Crises and 11. Indeed, some argue that long-term forces will Liquidity Shocks: A Bank-run Perspective." yet force interest rates back down to the levels observed NBER Working Papers 15425, National Bureau during the first half of the 2000s. of Economic Research, Cambridge, MA. 12. Calculations based on the results of coun- CBO. 2001. "CBO's Method for Estimating Potential terfactual simulations were conducted by assuming Output: An Update." CBO Paper, Congressional investment-to-GDP ratios between the period 2002 and Budget Office, Washington, DC. 2007 held constant, instead of increasing significantly, Clarke, George, Robert Cull, and Maria Soledad as they did in many countries. This has the effect of Martinez Peria. 2001. "Does Foreign Bank Pen- reducing the level of the capital stock and therefore the etration Reduce Access to Credit in Developing services of capital in the calculation of potential output. Countries? Evidence from Asking Borrowers." See box 2.8 for more on the model of potential output Policy Research Working Paper 2716, World employed here. Bank, Washington, DC. 13. The share attributable to capital deepening Cournède, Boris. Forthcoming. "Revised Estimates of ranges from a high of 57 percent when assuming a base Potential Output in the OECD." OECD Working year of 2000 to a low of 42 percent when assuming a Papers, Organisation for Economic Co-operation base year of 2003. and Development, Paris. Cowan, Kevin, Jose De Gregorio, Alejandro Micco, and Christopher Neilson. 2008. "Financial Diverisification, Sudden Stops and Sudden References Starts." In Current Account and External Financ- Aghion, Philippe, Peter Howitt, and David Mayer- ing, ed. Kevin Cowan, Sebastián Edwards, and Foulkes. 2004. "The Effect of Financial Develop- Rodrigo O. Valdés. Santiago: Central Bank of ment on Convergence: Theory and Evidence." Chile. Working Paper 10358, National Bureau of Eco- Crotty, James. 2009. "Structural Causes of the Global nomic Research, Cambridge, MA. Financial Crisis: A Critical Assessment of the Al-Taimimi, Hussein A. Hassan, Mouawiya Al-Awad, `New Financial Architecture.'" Cambridge Jour- and Husni Charif. 2001. "Finance and Growth: nal of Economics 33: 563­80. Evidence from Some Arab Countries." http:// De Haas, Ralph, and Ilko Naaborg. 2006. "For- www.fmpm.ch/docs/5th/Al_Tamini.pdf. eign Banks in Transition Countries: To Whom Al-Yousif, Khalifa Yousif. 2002. "Financial Develop- Do They Lend and How Are They Financed?" ment and Economic Growth: Another Look Munich Personal RePec Archive Paper. 6320, at the Evidence from Developing Countries." http://mpra.ub.uni-muenchen.de. Review of Financial Economics 11 (2): 131­50. Demetriades, Panicos O., and Khaled A. Hussein. Arestis, Philip, and Panicos Demetriades. 1997. "Financial 1999. "Does Financial Development Cause Eco- Development and Economic Growth: Assessing the nomic Growth? Time Series Evidence from 16 Evidence." Economic Journal 107(442):783­799. 72 T H E I M P A C T O F T H E B O O M I N G L O B A L F I N A N C E O N D E V E L O P I N G C O U N T R I E S Countries." Journal of Development Economics ment Discussion Paper 741, Harvard Institute for 51 (2): 387­411. International Development, Harvard University, Denis, Cécile, Daniel Grenouilleau, Kieran McMorrow, Boston, MA. and Werner Röger. 2006. "Calculating Potential Kawai, Masahiro, Richard Newfarmer, and Sergio Growth Rates and Output Gaps: A Revised Schmukler. 2001. "Crisis and Contagion in East Production Function Approach." Economics Asia: Nine Lessons." Policy Research Working Working Papers 247, European Commission, Paper 2610, World Bank, Washington, DC. Directorate-General for Economic and Financial King, Robert G., and Ross Levine. 1993. "Finance and Affairs, Brussels. Growth: Schumpeter Might Be Right." Quarterly Detragiache, Enrica, Thierry Tressel, and Poonam Journal of Economics 108 (3): 717­38. Gupta. 2006. "Foreign Banks in Poor Countries: Kregel, J. A. 1998. "Derivatives and Global Capital Theory and Evidence." Working Paper 06/18, Flows: Application to Asia." Cambridge Journal International Monetary Fund, Washington, DC. of Economics 22: 677­92. Duffie, Darrell. 2007. "Innovations in Risk Transfer: Levine, Ross. 1997. "Financial Development and Eco- Implications for Financial Stability." Working nomic Growth: Views and Agenda." Journal of Paper 225, Bank for International Settlements, Economic Literature 35 (2): 688­726. Basel. Levine, Ross, Norman Loayza, and Thorsten Beck. Farhi, Maryse, and Marcos Antonio Macedo Cintra. 2000. "Financial Intermediation and Growth: 2009. "The Financial Crisis and the Global Causality and Causes." Journal of Monetary Eco- Shadow Banking System." Revue de la regulation nomics 46: 31­77. 5 (1). http://regulation.revues.org/index7473.html. Levine, Ross, and Sara Zervos. 1998. "Stock Markets, Frankel, Jeffrey. 2009. "The Global Financial Crisis: Banks, and Economic Growth." American Eco- A Selective Review of Recent Research in the nomic Review 88 (3): 53758. International Finance and Macroeconomics Pro- Luintel, Kul. B., and Mosahid Kahn. 1999. "A Quan- gram." NBER Reporter, No. 2. http://www.nber. titative Reassessment of the Finance-Growth org/programs/ifm/ifm.html. Nexus: Evidence from a Multivariate VAR." Frankel, Jeffrey, and Sergio Schmukler. 1996. "Country Journal of Development Economics 60 (2): 381­ Fund Discounts and the Mexican Crisis of 405. December 1994." Open Economies Review 7. Neusser, Klaus, and Maurice Kugler. 1998. "Manu- Gormley, Todd A. 2005. "Costly Information, Foreign facturing Growth and Financial Development: Entry and Credit Access." Mimeo. Washington Evidence from OECD Countries." Review of University in St. Louis. Olin School of Business. Economics and Statistics 80 (4): 638­46. http://ssrn.com/abstract=896000. Nourzad, Farrokh. 2002. "Financial Development Griffith-Jones, Stephany, and Jonathan Leape. 2002. and Productive Efficiency: A Panel Study of De- "Capital Flows to Developing Countries: Does veloped and Developing Countries." Journal of the Emperor Have Clothes?" QEH Woking Paper Economics and Finance 26 (2): 138­49. Series 89, Queen Elizabeth House, Oxford Uni- OECD. 2008. "Beyond the Crisis: Medium-Term versity, Oxford, U.K. Challenges Relating to Potential Output Unem- Guttmann. 2009. "Asset Bubbles, Debt Deflation and ployment and Fiscal Positions." Chapter 4 in Global Imbalances." International Journal of Economic Outlook, No. 85. Paris. Political Economy 38(2): 46­69. Rothenberg, Alexander D., and Francis E. Warnock. Hurlin, Christophe, and Baptiste Venet. 2008. "Finan- 2006. "Sudden Flight and True Sudden Stops." cial Development and Growth: A Re-examination Working Paper 12726, National Bureau of Eco- Using a Panel Granger Causality Test." Labora- nomic Research, Cambridge, MA. toire d'economie d'orleans, Orleans, France. Rousseau, Peter L., and Paul Wachtel. 1998. IMF. 2005. World Economic Outlook. Appendix 2.1. "Financial Intermediation and Economic International Monetary Fund, Washington, DC. Performance: Historical Evidence from Five September. Industrial Countries." Journal of Money, Credit, Jalilian, Hossein, and Colin Kirkpatrick. 2002. "Fi- and Banking 30: 657­78. nancial Development and Poverty Reduction in Saxena, Sweta and Agustin Villar. 2008. "Hedging Developing Countries." International Journal of Instruments in Emerging Market Economies." Finance and Economics 7 (2): 97­108. In Financial Globalisation and Emerging Market Jenkins, Hatice. 2000. "Commercial Bank Behavior in Capital Flows. BIS Paper 44, Bank for Interna- Micro and Small Enterprise Finance." Develop- tional Settlements, Basel. 73 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Shan, Jordan. 2005. "Does Financial Development World Bank. 1998. Global Development Finance. `Lead' Economic Growth? A Vector Auto- World Bank. Washington DC. regression Appraisal." Applied Economics 37: ------. 2006. Global Development Finance. 1353­67. Washington, DC. Shan, Jordan, and Alan Morris. 2002. "Does Finan- ------. 2007. Global Development Finance. cial Development `Lead' Economic Growth?" Washington, DC. International Review of Applied Economics 16 ------. 2009. Global Development Finance. (2): 153­68. Washington, DC. 74 3 Medium-Term Impacts of the Crisis on Finance and Growth in Developing Countries T he lessons of the financial crisis are likely to shape financial policies and market reactions for some time to come. Beyond the costs, less abundant domestic finance, reduced availability of external finance, and a more constrained environment for domestic financial immediate and unprecedented global recession intermediation than during the boom. Efforts that it has provoked the crisis can be expected by developing-country governments to reduce to alter the global financial landscape signifi- dependence on foreign capital may also reduce cantly over the next 5 to 15 years in at least firms' access to finance and perhaps the effi- three important ways. ciency of domestic intermediation. First, authorities in high-income countries As businesses adapt to higher capital costs, will almost certainly strengthen financial the rate of growth potential output in devel- regulation to reduce excessive risk-taking by oping countries can be expected to slow by financial intermediaries, which will involve between 0.2 and 0.7 percentage points for broadening the coverage of regulation and the between five and seven years. And, unless imposition of higher capital requirements and offset by policy reforms to reduce domes- other limits on excessive and risky lending. tic borrowing costs, these constraints could Second, authorities in developing countries result in as much as an 8 percent decline in are likely to introduce rules and policies that long-term potential output (the level of output insulate them from excessive financial volatil- if the available labor and capital were fully ity, by placing greater emphasis on domesti- employed). This chapter outlines these ex- cally managed risk management strategies such pected changes and their consequences for fi- as capital controls and reserve accumulation. nance, investment, and economic activity over Third, market participants will likely be the medium term. The main conclusions are: more risk averse than they have been over the Necessary and desirable steps to rein in past decade, and the extent to which today's excessive risk taking by financial interme- risk management instruments are used to in- diaries, together with increased risk aver- crease leverage and global liquidity is likely to sion as a result of the financial crisis, will shrink. constrain global liquidity going forward. Stronger regulation in high-income countries Tighter regulations, an increase in the range is likely to reduce volatility in financial markets, of financial activities and firms that come which should contribute to more stable and sus- under regulatory scrutiny, increased cross- tained growth over the longer term. At the same border supervision, the necessity of banks to time, better regulation and higher risk aversion rebuild their balance sheets, and increased risk also imply that firms and governments in de- aversion on the part of investors will contribute veloping countries will face higher borrowing to a more stable global financial environment 75 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 but also result in less abundant and more ex- capital-to-output ratios, as firms economize on pensive capital. Although developing-country capital relative to labor and natural resources. borrowing costs have declined from their im- Because entrepreneurs will be working with mediate post-crisis highs, spreads appear to less capital than they would have had inter- have stabilized at levels that are about 150 est rates remained stable, the level of output basis points higher than during the boom pe- that developing economies will be able to sus- riod. Currently, the price of risk in both high- tain could fall by between 2 and 8 percent of income and developing countries is being held GDP. Moreover, during the transition period down by very low policy rates and quantitative to the new, lower capital-output ratio, the rate easing. As monetary easing is withdrawn, base of growth of potential output could decline by interest rates in high-income countries are ex- between 0.2 and 0.7 percentage points annu- pected to rise and with them the risk premiums ally for about seven years. and interest rates paid by developing-country Despite higher borrowing costs than during borrowers. Interest rates charged developing- the boom period, borrowing costs will remain country borrowers could rise 70 to 270 basis as much as 500 basis points lower than they points above boom-period levels. were in the 1990s (depending on the region), As a result of the crisis, external capital reflecting years of policy reform and improved flows will decline over the medium term, with economic fundamentals. Better debt manage- the extent of the decline dependent on the type ment, more flexible exchange rates, and more of capital flow. Unsecured bond and bank lend- stable political regimes are among the factors ing, as well as portfolio equity flows, are likely that mean borrowing costs will not return to to be severely constrained by the new global pre-boom levels. The biggest projected im- financial environment. Trade finance, which provements in borrowing costs compared often carries lower risk than other forms of with the 1990s are in Latin America and the bank lending because it is directly tied to col- Caribbean, Europe and Central Asia, and Sub- lateral, should be less affected. The rise in risk Saharan Africa. aversion and more stringent regulation is not Developing countries are likely to rely expected to hit foreign direct investment (FDI) more on domestic financial intermediation. to the same degree as debt flows, partly be- Given the severity of the crisis in high-income cause in the future, investors can be expected countries, the global financial system's to privilege less risky investment forms such as ability to provide ample credit is likely FDI. Nevertheless, FDI inflows are projected to to be impaired for an extended period of decline from recent peaks of 3.9 percent of time. Moreover, authorities in developing developing-country GDP to 2.8­3.0 percent countries may take a more skeptical view of GDP. And foreign bank participation in de- toward globalization and seek to promote veloping countries' domestic financial systems domestic financial intermediation as an al- may decline (or rise less quickly) because both ternative to reliance on foreign capital. This recent losses and tighter regulation will force strategy could ultimately benefit some of the parent banks to build up their capital. How middle-income countries that have a strong quickly and how durably these impacts will be framework for financial intermediation, by felt depends critically on how long it takes to increasing the efficiency of domestic finan- achieve sound, well-functioning financial sys- cial intermediaries through learning by doing tems in high-income countries and on the policy and economies of scale. However, a weaker reactions to the crisis in developing countries. international system may have undesirable Higher borrowing costs and tighter effects in many low-income countries, where rationing of credit will reduce potential output deficiencies in domestic intermediation sys- and growth in developing countries. Higher tems are likely to prevent them from com- borrowing costs will reduce firms' desired pensating for a reduced foreign presence. 76 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H The traumatic impact of the financial crisis may significantly boost domestic financial in- will encourage developing countries to take termediation, potentially more than offset- steps to insulate their economies from external ting any potential negative impact of higher shocks. Further efforts to raise international global risk premiums. For example, if the poor reserves are understandable in this context, institutional framework in many Sub-Saharan partly because high reserve levels may be use- African countries could be improved to the ful in deterring speculators and reducing the level of South Asia, borrowing costs would fall amplitude of smaller shocks. However, the by an estimated 275 basis points, more than benefits from holding reserves are subject to the most pessimistic estimate of the increase diminishing returns. This, combined with low in developing-country borrowing costs coming returns and risks from the concentration of re- from tighter global financial conditions. Were serves in dollars will make increasing reserves countries in Europe and Central Asia, Latin and holding high reserve levels costly. And even America, and Sub-Saharan Africa to reduce substantial reserves cannot prevent a country intermediation costs to the levels observed in from feeling the effects of a massive shock. Asia, they could see increases in potential out- Similarly, some countries may be tempted to put of 8 percent or more. pursue slower capital account liberalization or The remainder of this chapter analyzes in even to reverse past reforms, which could have more detail the qualitative and quantitative heavy long-term consequences. impact of the expected changes in global regu- The financial crisis is likely to encourage lations, risk aversion, and developing-country moves toward greater regional cooperation. attitudes toward financial volatility on inter- Such cooperation holds some potential to mediation, international capital flows, invest- help strengthen financial services by capturing ment, and growth. economies of scale and facilitating risk shar- ing through pooled reserves. It may also help strengthen South-South financial flows, such The impact of post-crisis as FDI, which are likely to be of increased im- portance for low-income countries. However, regulatory and structural progress in regional financial cooperation has changes been slow in developing countries. Further, such arrangements are likely to be of greatest benefit to regions that already have relatively T he financial crisis exposed a wide range of weaknesses in financial markets in high- income and some developing countries. A robust domestic financial systems, such as East laissez-faire implementation of the revised Asia and the Pacific. Poor countries with weak Basel II prudential guidelines that relied upon institutions are unlikely to strengthen their fi- large banks in some high-income countries to nancial systems by promoting integration with use internal risk assessments to help determine other poor countries with weak institutions. the necessary level of loan provisioning, com- Developing countries can mitigate the costs bined with very strong incentives to maximize of tighter global financial conditions by reduc- short-term profits, resulted in many of these ing the cost of intermediation domestically. institutions taking on far too much risk. They Inefficiency of domestic financial sectors, cor- may also have placed excessive confidence in the ruption, weak regulatory institutions, poor capacity of their hedging and risk-management protection of property rights, and excessive strategies to deal with their highly leveraged limits on competition can push borrowing costs portfolios. in developing countries as much as 1,000 basis In addition, banks reduced their capital points higher than in high-income countries. requirements by selling loans to subsidiaries, Improvements in the policies and institutions thus moving the loans off the banks' balance governing countries' domestic financial sectors sheets. While some of these loans were sold 77 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Box 3.1 Likely directions of financial sector reforms T he precise nature of the reforms that eventually will be instituted are as yet unclear. However, there is broad agreement on the directions that re- · banks and firms and that encompasses the inter- dependencies among them. Ensure that existing regulations are effectively form should take. Any eventual reform should: enforced. · Eliminate the pro-cyclicality of existing capital · Expand the perimeter of financial sector surveil- requirements and prudential norms, possibly lance to ensure that the systemic risks posed by even making these rules countercyclical. unregulated or less regulated financial sector · Recognize the central role played by private segments are addressed and introduce mecha- sector ratings agencies and bring them into the nisms to discourage regulatory arbitrage and regulated sphere. ensure effective enforcement of regulation. · Resolve the political and legal impediments · Bring on-book the off-balance-sheet transac- to the effective regulation of cross-border tions of major banks. institutions, develop special insolvency regimes · Recognize that a more systemwide approach to for large cross-border financial firms, and har- financial sector regulation is needed, notably monize remedial action frameworks. one that extends beyond the health of individual Source: This box draws heavily on Cortavarria and others (2009) and Brunnermeier and others (2009). to outside investors, banks often retained at healthy companies.3 Essentially, no regulator least a portion of the risk through such mecha- had the responsibility for overseeing all systemi- nisms as "implicit recourse," whereby banks cally important financial institutions (Jickling essentially reduced potential investors' losses 2009), and therefore none had the responsibility in securitized issues to preserve the banks' rep- to focus on how failures in one institution could utation and retain access to the market. The affect the system as a whole. exposure inherent in implicit recourse was not The broad outline of measures to be adopted subject to capital requirements.1 The combi- can already be seen in the changes to financial nation of securitization and implicit recourse sector regulation proposed by authorities in the increased banks' leverage (implicit loan to United States and Europe.4 Both a tightening of risk-weighted capital ratios) and short-term regulatory requirements (box 3.1) and the more profits (because investors would pay a higher effective enforcement of existing regulations price to buy loans where banks were willing to will be required. Regulators can be expected to make good a portion of subsequent losses), but increase capital requirements for deposit-taking it also increased their vulnerability to a down- institutions, in particular requiring higher capi- turn by effectively eroding capital adequacy.2 tal levels during booms to reduce the degree of Finally, neither public nor private sector procyclical lending. Regulators may also set quasi-regulators such as security exchange maximum leverage ratios and minimum levels commissions and ratings agencies foresaw the of liquidity, undertake more vigorous review fragility of the overall global financial system. of risk assessment procedures, and tighten re- This failure arose in part because these agencies' quirements governing off-balance-sheet trans- regulatory remit failed to provide adequately actions and nonbank subsidiaries. Authorities for the interdependencies that had crept into the also are likely to increase scrutiny and pru- system. As a result, the failure of one financial dential requirements for nonbank financial institution could, and did, lead to a rapid deteri- institutions, particularly those that are large oration in the balance sheets of other seemingly enough to pose a systemic risk, and ensure that 78 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H supervision takes into account the interdepen- significant further consolidation in the bank- dencies among financial institutions. ing sector in high-income countries, and the Supervisors may devote more scrutiny to transformation of leading investment banks how the structure of compensation affects from loosely regulated institutions to more tra- risk taking and accountability and may es- ditional banks. The first change may increase tablish stronger firewalls within institutions monopolistic tendencies within the sector, rais- where the potential for conflicts of interest is ing costs for borrowers and reducing returns high. More attention is likely to be placed on for savers, while the second may reduce the whether authorities have the appropriate tools supply of funds to developing-country firms and contingency plans for taking over insol- seeking additional equity financing. vent institutions and responding to a financial In addition, the financial crisis demonstrated crisis. Markets for derivatives and other in- that the ability of securitization and other exotic novative financial products will be subject to financial transactions to manage and reduce greater transparency requirements, and efforts the cost of risk had been grossly overstated. to protect consumers of financial products The complexity of these products and the in- from fraud and excessive risk will be redou- tricate interdependencies between financial bled. Finally it is likely that any new regime agents made individual risk exposures very dif- will include enhanced cross-border monitor- ficult to evaluate correctly--particularly because ing, regulation, and coordination mechanisms. these exposures depended to a large degree on In addition, regulators are likely to take a the even-less-well-understood risk exposure of much closer look at industry self-regulation counterparties. As a result, these products can be mechanisms such as securities exchanges and expected to play a much more modest role in the private ratings agencies, which are rightly or evolution of global finance in the years to come. wrongly blamed by many for failing to iden- The following subsections attempt to assess tify the elevated and dangerously unsustain- how these changes are likely to affect financial able level of risk that was being undertaken. conditions, investment, and growth prospects For example, some observers claim that the for developing countries in the years ahead. scandals earlier in this decade surrounding Enron, WorldCom, and some mutual funds, as well as failures of oversight tied to the current Impact on developing countries of banking crisis, have undermined public con- tighter financial sector regulations fidence in self-regulation by stock exchanges in high-income countries (Ellis, Fairchild, and Flether 2008). Some rat- Although the financial crisis was centered in ings agencies failed to implement required the banking sectors of high-income countries, procedures (for example, the rule that analysts it has affected near-term real-side prospects should not participate in discussions of fees throughout the world, including countries with issuers) to limit the conflict of interest in- that have followed very prudent micro- and volved in relying on fees from the institutions macroeconomic policies. The banking sectors that issued the rated financial instruments of most developing countries have not come (SEC 2008). Any erosion of the accuracy of under the same kind of stress, nor required ratings because of these conflicts had systemic the extensive national and international assis- implications, because many laws and regula- tance, as have those of the high-income coun- tions mandate the use of ratings to determine tries--with the notable exception of those in permissible investments (for example, by in- the Europe and Central Asia region. surance and pension companies) and bank Several features of the expected tighten- capital requirements (Jickling 2009). ing of regulation in high-inome countries is The financial crisis has brought on large, expected to benefit developing countries. endogenous structural changes, including a Improved transparency requirements (for 79 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 over-the-counter derivatives and commercial Consequences for the role of bank off-balance-sheet liabilities, for exam- foreign banks in developing-world ple) could ease investors' concerns over the intermediation risks involved, thus moderating the decline in The financial crisis could lead to lower for- flows. Regulatory changes intended to con- eign bank participation over the medium term trol the market segments that loomed large because stricter cross-country regulation and in the boom-and-bust cycle, such as subprime higher capital requirements (or the need to build mortgages in the United States, could encour- up capital after the huge losses during the crisis) age greater investment in markets where such will reduce not only banks' ability and willing- problems were much less important, such as ness to lend but also the attractiveness of foreign developing countries. Tougher disclosure rules operations. The implications for financial con- that reduce the attractiveness of using offshore ditions, investment, and output in developing centers for tax evasion could shift flows to de- countries extend beyond the direct impact that veloping countries. Higher capital requirements reduced activity might have on international in high-income countries may increase the rela- capital flows. The reliance by foreign banks on tive attractiveness of setting up subsidiaries in home-country deposits to fund lending implies a developing countries, where there is likely to net inflow of capital. However, loans by foreign be less pressure to raise capital requirements banks are also funded from domestic deposits (because most developing-country banking sys- that do not involve net capital inflows. More- tems did not suffer from the same weaknesses over, the benefits of foreign bank participation as banks in the high-income countries that were extend beyond the impact it might have on the source of the crisis). Developing countries credit creation and may include improved sta- may also benefit from a perception of increased bility and economic efficiency. risk in high-income countries, both because A reduced presence of foreign banks could they were the center of the crisis and because impair growth and efficiency in developing of the rise in public debt from efforts to support countries' banking systems. In an appropriate demand (see below). institutional environment, foreign banks tend However, the tightening of financial regula- to generate additional competitive pressures, tions in high-income countries will also have which induce cost reductions and quality im- negative consequences for finance in developing provements in domestic banks; improve ac- countries. Tighter rules will also increase bor- counting, auditing, and rating institutions; and rowing costs and reduce liquidity in developing increase pressures on governments to improve countries as compared with the boom period. the overall legal framework governing the fi- New regulations may affect the conditions under nancial system (Levine 1996). Except for Latin which foreign banks establish branch-bank ac- America, where results are mixed, foreign banks tivities in developing countries and may change in developing countries generally tend to be the nature and basis upon which trade finance is more efficient than domestic banks, and compe- arranged. Finally, the structural changes in the tition from foreign banks helps to improve the high-income financial world may have implica- efficiency of domestically owned banks, as mea- tions for equity markets in developing countries sured by reduced costs and spreads (Cull and and the ability of companies in the developing Martinez Peria 2007).5 As a result, a reduced world to raise equity in the high-income world. foreign bank presence can be expected to raise On balance, while tighter regulation will the cost of borrowing and lower the supply of fi- contribute to a more stable international en- nancial intermediation, thereby reducing firms' vironment and promote sustainable growth, access to resources for investment. tighter regulation also is likely to reduce loans A decline in foreign bank participation to all high-risk borrowers, including govern- could be less important for countries where ments and firms in developing countries. 80 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H Box 3.2 The impact of foreign bank participation on stability I n general foreign bank participation tends to reduce volatility in developing economies, depending on individual country (or bank) circumstances. Foreign banks (Cull and Martinez Peria 2007). Demirgüç- Kunt, Huizinga, and Claessens (1998) find that foreign bank participation reduces the likelihood banks are likely to be more diversified than domestic of a banking crisis. Martinez Peria, Powell, and banks. Partly reflecting the internal resources of Vladkova-Hollar (2002) find that foreign bank lend- international banking organizations and their greater ing is not significantly curtailed during crises, a view access to capital markets, they tend to fund more of supported by case studies for Malaysia (Detragiache their lending operations from deposits held abroad, and Gupta 2004), Argentina and Mexico (Goldberg, and their lending operations tend to be less procyclical Dages, and Kinney 2000), Latin America (Peek and than domestic banks (Cetorelli and Goldberg 2009). Rosengreen 2000), and 10 Central and Eastern Euro- At the same time, foreign banks may import instability pean countries (De Haas and van Lelyveld 2006). In as they reduce lending in response to shocks in their studies of selected Latin American countries, Gold- home countries. And foreign banks may reduce the berg, Dages, and Kinney (2000) and Crystal, Dages, franchise value of domestic banks by cherry-picking and Goldberg (2002) show that credit growth in for- the top clients, thus forcing domestic banks into lower- eign banks that had a long history in the host country return and riskier market segments (Hellman, Murdock, was less volatile than in domestic banks. Arena, and Stiglitz 2000). Thus the impact of foreign banks on Reinhart, and Vazquez (2007) find weak evidence stability is an empirical question. that foreign banks in 20 developing countries were The weight of evidence indicates that foreign less sensitive to monetary conditions than domesti- banks tend to be more stable lenders than domestic cally owned banks. barriers to competition--high bank concen- constrained by adverse conditions in their home tration, rules that strictly limit banks' activi- countries. This appears to be happening in ties, or excessive regulatory barriers to entry the current context, where huge bank losses in and exit--are significant. In such circum- industrial countries are reducing lending from stances, foreign bank entry may fail to exert parent and other overseas affiliates of foreign- competitive pressures on domestic banks. In- owned banks in developing countries (Cetorelli deed, research suggests that, faced with barri- and Goldberg 2009). ers to competition, high bank concentration, Preliminary data indicate that between June and profitability, foreign banks may invest and December 2008 lending by foreign-owned expressly to reap the rents available from the banks (at least those reporting to the Bank of distorted environment for financial intermedi- International Settlements, or BIS) fell by 13.6 ation. In these environments, reduced foreign percent, much more than the 4 percent decline bank participation would have limited impli- in total bank lending in developing econo- cations for efficiency. mies (table 3.1).6 In some regions, however, However, a lower foreign bank presence foreign banks have been a source of stability. could increase instability in developing-country In Latin America and the Caribbean lending financial systems. Foreign banks tend to be less by foreign banks declined by less than total dependent on domestic deposits and therefore lending, and in the Middle East and North less affected by domestic economic instabil- Africa (where foreign banks have a small share ity than domestically owned banks (box 3.2). of total claims) foreign bank claims rose more That said, the effect of foreign banks on sta- rapidly than total bank claims. In the other bility depends on the source of the shock: for- four regions, however, claims by foreign- eign banks may increase instability if they are owned banks fell considerably faster than did 81 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Table 3.1 Credit growth by foreign banks greater assurance that importers would pay versus total credit growth, developing on schedule; market disruption caused when countries critical market participants either collapsed (Percent) (Lehman Brothers and others) or encountered Foreign Growth Growth severe difficulties (many commercial banks); bank rate of rate of claims as foreign bank total bank and a drying up of the secondary market for a % of claims, claims, short-term exposure because of deleveraging total June­ June­ claims, December December by banks and other financial institutions. Region June 2008 2008 2008 Insofar as the decline in trade finance rep- resented a typical short-term crisis response, East Asia and Pacific 3.8 3.6 4.2 Europe and Central Asia 30.6 12.2 8.8 exacerbated (compared to previous devel- Latin America and the oping-country crises) by the breakdown of Caribbean 27.3 19.9 21.4 financial systems in high-income countries, Middle East and North Africa 9.7 11.7 5.0 trade finance should recover over the medium South Asia 10.2 7.7 2.4 term. However, the exact terms under which Sub-Saharan Africa 26.1 15.1 6.5 it will recover and at what cost remain to be All developing countries 14.4 13.6 4.0 determined (see the following section on bor- Source: Bank for International Settlements and International rowing costs). The extent of the recovery will Financial Statistics. depend on the precise rules adopted for bank- ing regulation. In particular, the Basel II Ac- total bank claims (or total bank claims actu- cord on banking laws and regulations does not ally rose, as in East Asia and the Pacific). The discriminate among different forms of short- more rapid decline in lending by foreign banks term credits (maturities of a year or less) with is attributable not to just a few large countries respect to maturity or risk profile. Thus a rigid with substantial weight in the total. Of the 89 application of the Basel II rules would imply developing countries with complete data on that banks would have to allocate as much both foreign bank claims and total claims, for- capital for trade finance as for other short- eign bank claims fell by more (or rose by less) term loans, even though trade finance often than total claims in 59. Despite these declines, has a relatively short maturity (perhaps a few as of December 2008 foreign bank claims re- weeks) and is lower risk (because the goods mained 40 percent above the level in Decem- being financed serve as collateral) than many ber 2006, so although foreign bank presence other forms of short-term exposure. has declined, that decline has yet to undo the Going forward, policy makers will need to substantial increase that preceded it during the take care that regulatory changes are sensitive boom period. to the importance of credit to trade, and that Trade credit is critical to sustaining the restrictions on the use of secondary markets international trading system.7 In the acute for securitized loans do not combine with phase of the financial crisis, a sharp decline blunt restrictions on country-specific loan ex- in the access of developing-country firms to posure limits to restrict trade in a permanent trade finance exacerbated the decline in ex- manner. To overcome this possibility, capital ports that reflected the fall in global demand requirements for secured trade transactions (box 3.3). As the crisis intensified, the availabil- should be made less onerous than those on ity of trade finance tightened and its cost rose unsecured loans. Until such time as Basel II for four main reasons: growing liquidity pres- regulations are revised, national authorities sure in mature markets; a perception of height- should take advantage of existing flexibility ened country and counterparty risks, resulting that allows them to establish different catego- in increased demand for letters of credit, insur- ries of short-term credit with different reserve ance, and guarantees because exporters needed requirements. 82 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H Box 3.3 Survey evidence on the decline in trade finance during the crisis S urvey data clearly indicate the deterioration in developing-country firms' access to trade finance during the financial crisis. In a World Bank survey of As a result, the cost of letters of credit was reported to have doubled or tripled for buyers in emerging coun- tries, including Argentina, Brazil, Bangladesh, China, 60 global buyers and suppliers in early 2009, 30 per- Pakistan, and Turkey (IMF/BAFT 2009). cent attributed the postponement or cancellation of By the end of 2008, trade finance deals were of- foreign sales to difficulties in obtaining trade finance fered at 300­400 basis points over interbank refi- (Shakya 2009). Another set of firm-level surveys un- nance rates--two to three times more than the rate dertaken in 2009 found that most firms, in particular a year earlier. More than 70 percent of respondents small and medium-size enterprises and new firms, indicated that the price of various types of letters have been affected by increased costs of trade finance of credit increased because of an increase in their and more stringent requirements, including guaran- own institution's cost of funds (80 percent of tees, to obtain more trade finance (Malouche 2009).a respondents), an increase in capital requirements Higher down payments, more stringent collateral re- (60 percent of respondents), or both. In the IMF/ quirements, and higher interest rates have reduced ac- BAFT survey of global banks (IMF/BAFT 2009) cess to preexport financing and become a particularly 71 percent of banks reported a decline in the value important obstacle for exporters. of their letter-of-credit business, with an overall Banks have tightened counterparty bank criteria, 8 percent decline in the first nine months of 2008 excluded certain banks and countries from lending (compared with the same period in 2007), accelerat- portfolios; demanded additional insurance and confir- ing to 11 percent during the period October 2008 to mation from banks; carried out more detailed due dili- January 2009. While 73 percent of banks recognized gence to preselect counterparties; adopted additional the role falling trade demand played in the decline in safeguards to loans through guarantees and higher trade finance lines, more than half also attributed the confirmation fees; and taken more restrictive attitudes trade finance decline to a decline in available credit toward new markets, new clients, and new products. (that is, a decline in supply). a. This survey covered 425 firms and 78 banks in 14 developing countries across Asia, Europe, Latin America, and Africa. Higher cost and reduced access to trade fi- Africa) where the banking system is compara- nance have particularly affected firms that are tively insulated from the international finan- highly dependent on external finance, nota- cial market. bly in several upper-middle-income countries such as Brazil that have enjoyed relatively easy access to international credit markets. Impact of re-regulation on Also affected have been firms that are highly developing-country equity markets integrated in global supply chains and mar- Tighter financial market regulation and the ginally creditworthy firms such as small and restructuring of investment banking activ- medium-size enterprises. By contrast, less af- ity in high-income countries, while reducing fected firms may include those that are less the likelihood that these firms contribute to reliant on the banking system for trade fi- financial instability, also may reduce the sup- nance (firms in a few low-income African ply of equity investment to firms in develop- countries, for example, that primarily rely on ing countries. This could directly affect the self-financing or cash in advance) and firms in capacity of innovative firms to expand or a few middle-income countries (such as South start up new product lines. Weighted by value, 83 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Figure 3.1 Foreign participation in selected emerging equity markets (Portfolio equity inflows divided by stock market capitalization, percent) % of local stock market capitalization 50 2003 2007 45 40 35 30 25 20 15 10 5 0 a H ico Th ary Tu d a kh il Tu n Jo y ze Ma an an Re sia de lic n ru d a La a ov ilip ia R nes ge ic th a a Bu dia R aria Pa nia an M hile co ol a U bia ne za az e ni n si a tio In lan ut esi ric Li tin ni C hin Ar ubl Sl Ph tv Pe Fe pub rk oc la st rd st ai ex ni us ch lay to ua In a om g Ka Br C ra Af ak pi n lg So on ai Po om ki C kr un ep or Es M h d si C R Source: IMF. approximately 86 percent of all developing- relative to total market capitalization may rep- country initial public offerings (IPOs) over resent very high levels of inflows relative to the the past 10 years have been at least partially market value of traded stocks. Despite these intermediated by American investment banks weaknesses, the data in figure 3.1 do suggest (accounting for 32 percent of the total number that in many markets foreign equity has played of deals).8 Moreover, investment banks have an important role in financing firms' activities. been instrumental in the creation of mutual The developmental impact of portfolio funds9 and other instruments that have al- equity inflows goes beyond the purely fi- lowed individuals and regulated institutional nancial aspect of the transactions. Empirical investors in high-income countries to include studies tend to find a positive impact of net firms in developing countries in their overall portfolio equity flows on growth, both in portfolio of assets. macroeconomic terms10 and in industry- and Comprehensive data on the contribution of firm-level data (Kose and others 2006).11 For- foreigners to developing-country equity mar- eign participation in equity markets increases kets are not available. However, portfolio in- their overall liquidity, which improves their vestment inflows as recorded in the balance of attractiveness to other investors (because li- payments accounted for as much as 40 percent quidity increases an investor's ability to divest of the market capitalization in some develop- in a timely manner) and encourages greater ing countries in 2007 (figure 3.1). Of course, investment in projects with a long time ho- inflows of funds to purchase stocks may be bal- rizon, because individual investors can eas- anced by outflows from the sale of stocks, so ily sell their holdings before the benefits are the figure on inflows certainly overstates the realized.12 Moreover, well-developed equity net new purchases of stocks by foreigners on markets contribute to transparency, because these markets. At the same time, the share of firms release information to attract capital, stocks traded in any one year in many emerging a process that ultimately improves the effi- equity markets is low, so high levels of inflows ciency with which investment is allocated.13 84 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H The participation of globally savvy investors below the 90 percent decline projected for net can also improve the identification of projects private debt flows. with global potential and lead to demonstra- The 1990s offer some insight into how tion effects that might otherwise be missed. the tighter financial environment could af- Openness to external financial flows can spur fect FDI. For example, during that period, equity market development by increasing li- banking sector difficulties in Japan translated quidity and pressures for improvements in es- into a sharp decline in Japanese investments sential infrastructure.14 in the United States. A single downgrade in credit ratings of Japanese banks resulted in a Impact on FDI 30 percent decline in the number of projects The sharp increase in global FDI flows be- initiated by Japanese investors in the United fore the financial crisis partly reflected a surge States (Klein and others 2002). And domes- in inexpensive debt financing: the value of tic financial markets and the availability of cross-border syndicated bank borrowing and credit were found to be important factors international bond issuance for the purpose in explaining investment outflows through of acquisition (including both domestic and cross-border M&As during the late 1990s cross-border) rose to almost $1 trillion in 2007, (Di Giovanni 2005). from $131 billion in 2003 (figure 3.2).15 In ad- Despite their sensitivity to financial condi- dition, almost 30 percent of global merger and tions, FDI inflows to developing countries are acquisition (M&A) deals between 2003 and unlikely to be as constrained as debt flows 2008 were carried out by high-income invest- over the medium term. South-South FDI flows ment banks, hedge funds, and other private may be more resilient than flows from high- equity firms (UNCTAD 2009). income to developing countries, owing to the The tightening of financial conditions significant role of state-owned enterprises with may thus affect firms' ability to finance FDI. softer budget constraints, limited reliance of Similarly, changes in the legal status of these Southern multinationals on international debt institutions and the expected expansion of markets, and continued efforts (particularly regulation to encompass more of their activi- by China) to gain access to energy and min- ties could further depress developing-country erals.16 In addition, tighter access to equity M&A deals. However, crises tend to affect and debt-creating credit for firms in margin- FDI to developing countries less than they af- ally creditworthy developing countries may fect debt flows (box 3.4), and the anticipated increase their willingness to enter into FDI- 30 percent fall in FDI flows in 2009 is well based mergers and acquisitions. Overall, FDI inflows to developing countries are expected Figure 3.2 Debt financing of M&A transac- to be sharply lower than they were during the tions, 1995­2008 boom period when they reached just under 4 US$, billions percent of developing-country GDP, although 1.8 they should recover to levels around 3 percent 1.6 of GDP, around the same levels observed dur- 1.4 Cross-border M&A ing the preboom period. 1.2 1.0 The effect of lower FDI inflows on growth 0.8 prospects for individual developing countries 0.6 will depend partly on the share of FDI in total 0.4 International investment. While FDI represents less than debt for 0.2 acquisition 5 percent of total investment in some regions, 0 in others, notably Europe and Central Asia, 1995 1997 1999 2001 2003 2005 2007 Latin America, and Sub-Saharan Africa, it Source: World Bank. can account for as much as 20 percent of all 85 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Box 3.4 FDI and debt flows during crises F DI inflows tend to be less subject to sharp up- swings and downswings during crises than debt flows. For example, net private-source debt flows to percent of GDP (box table 3.4.1). Net private debt flows continued to be negative four years after the crisis, while FDI remained stable. Similarly, during Latin America and the Caribbean fell from 5.7 per- the East Asian crisis net private debt flows to the cent of GDP the year before the onset of the 1980s most affected countries plummeted the year of the crisis to 1.6 percent the year after (and would have crisis while FDI was little affected, in part because fallen by more if not for concerted lending), while of the buying opportunities generated by the steep FDI flows only fell from 1 percent of GDP to 0.8 exchange rate depreciations. Box table 3.4.1 Net private-source debt flows versus FDI before and after crisis episodes (Percent of GDP) Crisis 4 Crisis 3 Crisis 2 Crisis 1 Crisis Crisis 1 Crisis 2 Crisis 3 Crisis 4 Latin American 1980s crisis Net private-source debt flows 4.50 5.39 5.57 5.67 3.88 1.63 0.24 0.38 0.57 FDI 0.75 0.89 0.87 1.01 0.85 0.76 0.60 0.79 0.61 East Asian crisis countriesa Net private-source debt flows 1.91 3.52 3.16 0.56 4.23 2.78 1.99 1.35 0.70 FDI 1.05 1.27 1.49 1.68 2.34 2.22 1.33 0.72 0.94 Mexico pesos crisis Net private-source debt flows 2.79 0.59 3.54 1.98 1.09 1.43 0.79 2.61 2.55 FDI 1.51 1.21 1.09 2.60 3.32 2.76 3.20 2.95 2.85 Argentina Net private-source debt flows 5.89 3.57 1.07 1.11 5.83 10.26 0.42 1.49 9.91 FDI 3.13 2.44 8.46 3.67 0.81 2.11 1.27 2.69 2.87 Source: World Bank. a. Indonesia, Republic of Korea, Malaysia, and Thailand. investment (figure 3.3). And in low-income countries, FDI represents an even larger share Figure 3.3 FDI as a share of investment in in total investment. In such countries, a 30 per- developing countries, 1995-2008 cent decline in FDI could represent as much as Percent of fixed investment a 6 percent of GDP decline in investment un- 25 1995-99 2000-04 2005-08 less domestic investment steps in to pick up the 20 slack. 15 Of course not all FDI constitutes invest- ment in the national accounting sense of an 10 addition to the productive capacity of a coun- 5 try, because much of what is recorded as FDI in the balance of payments represents merger 0 un pin ll E ci ia A d dl ean d ric d ia ric n and acquisition of already existing firms. Nev- co elo A tin ral an M ribb an Af an Af ra Pa s As tri g C uro fic th Am sia a d tA ha So a t e a a th st h es an Eas en p C ic Sa or a ut ertheless, a significant portion of FDI flows e er N eE b- v Su de id do represent both greenfield and brown- La field extensions of productive capacity--a Source: World Bank. point that is supported by the regression 86 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H analysis in chapter 2, which found that FDI global financial system could encourage some tended to have a greater impact than other lending to high-risk borrowers, which could capital flows on investment rates in develop- limit the decline in flows. ing countries. Moreover, as described in the The restructuring of high-income coun- 2008 edition of Global Economic Prospects tries' financial systems and tighter constraints (World Bank 2008), FDI can be an important on risk taking may also reduce interest in source of technology transfer in the form of project finance transactions, where the in- capital goods but also in the form of busi- volvement of foreign institutions (such as ness processes, which can have large spillover syndicated commercial banks, bond under- effects in developing countries.17 Finally, FDI writers, firms involved in leasing equipment, can provide the impetus for important regu- dedicated equity funds, and official export latory and governance changes that can have credit agencies) has helped to ensure access spillover effects throughout the economy, re- to international capital markets, to reduce ducing both the costs of doing business and financing costs through innovative financing the effective cost of capital--thereby spurring techniques that tie debt service obligations further investment (domestic and foreign). to the timing of expected project proceeds, to tap extended debt maturities (consistent Impact on private-source debt flows with the extended time required for con- The tighter financial environment expected in struction of many projects) that may not the post-crisis period, while reducing volatility be available in the domestic market, and in financial markets, is also likely to constrain to obtain expertise necessary to construct developing countries' access to international complex financial arrangements.18 Because debt markets over the longer term as well as the lead arrangers of project finance trans- in the short run (see chapter 1). The necessity actions are largely financial institutions in to rebuild banks' capital and greater concern industrial countries, a smaller profile of for- over risky loans will reduce both cross-border eigners could make it more difficult to ar- lending and foreign bank participation in de- range these complex financial transactions. veloping countries. Restrictions on financial in- However, this constraint is unlikely to be too stitutions' ability to assume risk may also limit binding. Institutions in developing countries developing-country borrowers' ability to issue have begun to take a more significant role bonds in international markets, as regulatory in arranging project finance transactions.19 strictures are extended to many of the institu- Moreover, even if regulatory constraints tions (such as hedge funds) that participate in limit industrial-country firms' participation this market. Finally the elimination of some in- in project finance, they could still provide struments may reduce the ability of some regu- financial expertise on a fee-for-service basis. lated institutional investors, notably public and Countries that have relied heavily (in rela- private sector pension funds, from taking both tive terms) on debt financing to meet their direct and indirect positions in some forms of current account obligations and finance in- developing-country debt. vestment may be most vulnerable to a change However, some aspects of the reaction to in the extent to which such financing is forth- the crisis could moderate the decline in debt coming. Notwithstanding that developing flows to developing countries. For example, if countries were net exporters of private capi- risk aversion among borrowers in high-income tal during the boom period, among develop- countries increases and they decide to repay ing countries with current account deficits, rather than roll over debt (as happened during net debt inflows on average financed about Japan's prolonged recession), room could be one-third of the overall deficit (table 3.2), freed up for banks to lend to developing coun- a ratio that rose to 90 percent in the case tries. Moreover, a more stable and transparent of countries in Europe and Central Asia. 87 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Table 3.2 Contribution of private- source debt inflows to external finance of Figure 3.4 Net external debt flows from private sources, 1980­2008 developing countries with current account deficits, average 2003­07 US$ billion Share of GDP % Number of Net debt 600 4.0 countries inflows 3.5 500 with Current from current account private 3.0 400 % of GDP account deficit (% sources 2.5 Income category deficits of GDP) (% of GDP) 300 2.0 200 1.5 All countries 53 6.3 2.2 Low income 16 5.8 0.8 1.0 100 Lower-middle income 20 6.1 0.8 0.5 Upper-middle income 17 7.1 5.3 0 0 Of which: Europe Net external debt flows and Central Asia 8 8.5 8.1 100 0.5 1980 1985 1990 1995 2000 2005 Source: World Bank. Note: Data on current account deficits and debt inflows are Source: World Bank. simple averages of country numbers. Excludes small island economies. Unsurprisingly, it is the upper-middle-income developing Europe and Central Asia--where, countries (excluding minerals exporters) that at least in ex post terms, the inflow of capital relied most on debt inflows from private exceeded these countries' capacity to absorb sources; such inflows financed three-quarters or manage it. More generally, debt-creating of their current account deficits, on average. flows tend to contribute to volatility in de- Of these countries, those in the Europe and veloping-country consumption rather than Central Asia region account for the largest facilitating consumption smoothing, as might private-source debt inflows (averaging 8 per- be expected from a theoretical point of view cent of GDP); private-source debt inflows for (Kose, Prasad, and Terrones 2005; Kose and the other countries in this group averaged 2.5 others 2006).20 percent of GDP. Impact on macroeconomic Implications of a potential stability developing-country retreat from A more closely regulated international finan- financial integration cial sector--especially if provisions to reduce the procyclical nature of prudential require- ments are adopted--may help to reduce the A lthough the global liquidity expansion in the early 2000s ushered in a lengthy period of rapid and arguably sustainable ex- volatility of private debt flows. In addition to pansion of the supply potential of most devel- the vertiginous drop in private debt flows dur- oping countries, the financial crisis brought ing the current crisis, such flows also dropped that boom period to an end--generating a from almost 3 percent of developing-country substantial, sharp, and painful deceleration GDP just before the Latin American debt crisis in economic activity throughout the develop- in the 1980s to 1 percent of GDP in the depth ing world. The speed of the global downturn of the crisis in 1983 (figure 3.4). Arguably and the costs that it has imposed on countries their rapid increase during the recent boom that otherwise had followed exemplary poli- period was a principle cause of the imbalances cies and whose own financial systems did not that arose in some countries, notably those in engage in unsustainable practices has quite 88 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H naturally led many observers to question the Figure 3.5 The recent buildup in reserves relative benefits and costs of global financial was concentrated in East Asia and among oil integration. exporters Policy and behavioral changes following Months of import cover earlier crises suggest that this crisis may gen- 16 erate a number of important changes in the 14 12 nature and behavior of financial markets in 10 developing countries over the next 5 to 10 8 years. For example, countries and national 6 4 authorities can be expected to be much more 2 cautious in the management of their reserves. 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 Following the Asian crisis of 1998, many Asian economies placed a greater emphasis on Oil exporters Middle East and reserve accumulation as a self-insurance strat- North Africa East Asia and Latin America and egy against external instability. A more skep- Pacific the Caribbean tical and slow approach to capital account East Asia and Pacific Europe and except China Central Asia liberalization, or even a reversal of reforms Sub-Saharan Africa South Asia made to date, is also possible. Countries in both Latin America and East Asia took such Source: International Finance Statistics, IMF. steps following crises in their regions in the 1970s and 1990s. Finally, countries may seek to reduce their vulnerability to external finance by developing their own domestic fi- in 2007. Other developing countries also in- nancial sectors or seeking to establish regional creased reserves during this period but to a financial mechanisms. lesser extent (figure 3.5). In some of these countries, a desire to have a larger cushion against a financial reversal was Reserves policies one motivation for increasing reserves (Jeanne Developing countries may decide to increase 2007). Issues of sound macroeconomic man- their reserves to provide themselves some in- agement also played an important role in surance against future fluctuations in global reserve increases in many countries. Sharp in- financial conditions. An ample supply of in- creases in capital inflows (see chapter 2) and ternational reserves can help countries com- very strong commodity revenues during the pensate temporarily for a sudden reversal boom period may have exceeded the domestic in external capital flows, thereby avoiding a economy's capacity to invest productively. In sharper real-side adjustment than might oth- such cases, a sterilized accumulation of excess erwise be required. Large reserves can also capital inflows may have been the main reason dissuade speculators from provoking a capi- for reserve accumulation. tal crisis by demonstrating the authorities' Although the existence of high reserves capacity to prevent a disruptive adjustment. might prevent a run on a country's currency, Following the East-Asian crisis, oil-importing the extent to which reserves actually helped countries in the region (excluding China) in- lessen the real-side impact of this crisis is creased their reserves from an average of not clear. In fact, although accumulation of about 3 months worth of imports before the reserves stopped at the peak of the crisis, crisis to more than 9 months worth in 2007. relatively few countries actually used their In China the increase was even more marked, reserves to smooth adjustment. Thus, while from about 7 months worth of import cover developing-country reserves fell from $3.7 in the pre-crisis period to almost 18 months trillion in December 2007 to $2.2 trillion in 89 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 December 2008 (in part owing to the appre- about 8 percent of GDP, most developing ciation of the dollar), the precipitous fall in countries enjoyed a net benefit from reserves imports meant that import-cover ratios actu- (Hauner 2005). By 2004, however, with re- ally rose. serves equal to 19 percent of GDP on average, Moreover, countries with large reserves most developing-country groups were losing suffered downturns almost as severe as did money--as much as 0.2 percent of GDP in the countries with small reserves. The correlation Middle East and Central Asia, and 0.6 percent of the slowdown in GDP growth in 2008 with of GDP in Sub-Saharan Africa and emerging ratios of reserves to imports in 2007 is only Europe. And the cost must have increased sub- 0.15, which is not significantly different from stantially by 2007, when reserves represented zero at the 5 percent level.21 Reserves levels 26 percent of developing-country GDP.23 also played no role in supporting exchange Jeanne (2007) estimates the annual cost of rates: the correlation between reserves to im- reserves holdings for 20 emerging markets at ports and the change in exchange rates during about 1 percent of GDP over 2000­2005, and 2008 is 0.04 (the same result is obtained if Rodrik (2006) obtains roughly the same result one excludes countries in Europe and Central for developing countries. Asia, which suffered very large exchange rate Exchange rate movements represent a further changes). Finally, there is little relationship potential cost for developing countries that between reserves levels and output declines, react to the financial crisis by increasing their even if one controls for other determinants reserve holdings. As of the second quarter of of the impact of the crisis on growth, includ- 2009, the U.S. dollar accounted for 63 percent ing trade and financial openness, exposure of the known composition of official interna- to falling commodity prices, initial current tional reserves,24 with the euro representing account imbalances, and other financial vul- 27 percent, and other smaller currencies and nerabilities (Blanchard, Faruqee, and Klyuev special drawing rights the remainder. Large 2009). At least in the case of a generalized fi- currency swings in exchange rates can be ex- nancial crisis, very high reserve levels appear tremely expensive for countries with extensive to make little contribution to easing real-side reserves. For example, if developing countries adjustment. in East and South Asia hold the same propor- Significant costs are associated with main- tion of reserves in dollars as the global aver- taining higher reserves. Most reserves are held age, a not unusual 20 percent decline in the in the form of low-yield securities that are dollar would reduce the value of their interna- issued by high-income-country govern- tional reserves by approximately 10 percent of ments. To accumulate foreign currency re- their GDP. serves, central banks must issue debt in local currencies--usually at much higher interest Capital account restrictions rates than can be earned on the foreign cur- The financial crisis may also cause countries rency being purchased, with the difference to reconsider plans or recent moves to liber- between the two (after exchange rate move- alize their capital accounts. As discussed in ments) having to be paid by the treasury. The chapters 1 and 2, the abrupt reversal of in- cost of holding reserves can be disaggregated ternational capital flows at the onset of the into the forgone return from alternative uses, crisis caused not only local equity markets to typically increasing public investment or re- crash (losing as much as 50 percent of their ducing debt; minus the financial return on value in days) but also virtually every non- reserves and lower spreads on foreign debt pegged currency in the world to depreciate by if higher reserves improve creditworthiness; between 10 and 30 percent against the dol- plus the current cost of past sterilizations.22 lar. Imposing some sort of capital controls During the 1990s when reserves averaged might have prevented or at least moderated 90 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H the pace of the outflows, reducing the domes- Figure 3.6 Developing countries' average tic disruption and wealth loss that ensued. financial openness, 1990­2006 Of course, tighter controls, had they been in Index place, might well have reduced the benefits 0.8 that countries enjoyed from the boom period 0.6 (see chapter 2). 0.4 Middle-income 0.2 The empirical evidence on the effective- 0 ness of these controls is mixed. Controls have ­0.2 Developing countries ­0.4 had little success in reducing capital outflows ­0.6 ­0.8 (except in Malaysia during the East Asian cri- ­1.0 Low-income sis), although controls on inflows have had ­1.2 1990 1992 1994 1996 1998 2000 2002 2004 2006 some effect in increasing the independence of monetary policy (Magud and Reinhart 2006). Source: Chinn and Ito 2007. Moreover, with the exception of a few coun- Note: Simple average of countries in each group. Excludes all countries that have missing data within time period, as well as tries with unsustainable current account posi- economies with population of less than 1 million. tions, large capital outflows do not appear to have been a major cause of financial distress during the current crisis. Indeed, develop- depend on the type of restrictions introduced, ing countries reduced their deposits in BIS- the international economic environment, in- reporting banks by about $260 billion (ad- vestors' views on economic prospects for the justed for exchange rate changes) during the country, and the quality of macroeconomic last quarter of 2008--suggesting a repatria- policy and institutions. In general, evidence tion of capital rather than flight.25 Neverthe- tying economic growth or real-side invest- less, developing-country policy makers may ment to capital account openness is weak both take a more skeptical stance toward large ex- because of technical reasons surrounding es- ternal debt inflows during booming economic timation methods and because of complex in- conditions and toward allowing domestic teractions that make generalizations difficult banks and firms to take speculative positions (box 3.5). Some studies point to beneficial in foreign currencies. effects from increased competition and better Indeed, one of the effects of the Asian cri- economic policies, while others focus on the sis was to slow the pace of capital account increased instability that can accompany capi- liberalization. Following that crisis, some of tal account openness, particularly with rapid the most affected countries tightened restric- liberalization in a context of weak domestic tions on capital account transactions. The institutions. Chinn-Ito index of capital account liberal- ization declined in Malaysia and Indonesia following the crisis, and the pace of improve- Increased reliance on domestic and ment in Thailand and the Philippines slowed regional financial arrangements markedly. More generally, even countries not A weaker international financial system will directly involved in that crisis slowed or re- almost certainly result in a movement toward versed the pace of liberalization (figure 3.6). increased reliance on domestic and regional While the pace of liberalization eased for sources of capital. Indeed, a weaker interna- middle-income countries in the middle 2000s, tional intermediation system will increase reli- it came to an almost standstill among low- ance on other sources of intermediation, even income countries.26 if the domestic or regional markets do not How a tightening of restrictions on capital grow in size. However, the reduced competi- account transactions would affect growth and tive pressure from abroad can be expected to stability in individual developing countries will increase the potential private economic return 91 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Box 3.5 The debate over capital account liberalization M aintaining an open capital account can con- tribute to increasing financial depth, and hence spur development, but mostly in industrial countries to define the quality of institutions, or how "strong" they would need to be to benefit from capital account liberalization.b and developing counties with high institutional qual- Capital account openness may also promote ity (Klein and Olivei 2008). Open capital accounts instability. The East Asian crisis of 1997­98 and can strengthen domestic financial systems by pro- subsequent episodes increased concerns over the moting competition and economic policies through potential for open capital accounts to raise instabil- market discipline (Kaminsky and Schmukler 2002), ity. Calvo (1996) emphasizes the risk that investors' as well as improve returns and diversification for herd behavior leads to excessive volatility in cross- developing country residents' savings. Despite these border capital flows, and Furman and Stiglitz (1998) theoretical benefits, empirical studies of the relation- emphasize the important role that contagion played ship between capital account openness and growth in exacerbating capital flight from open financial sys- have shown decidedly weak results.a tems during the East Asian crisis. Some cross-country Several reasons are offered for the limited empiri- empirical tests find little evidence that open capital cal support for the theoretical benefits of capital accounts increase the incidence of crises (Glick and account openness. The efficiency effects can be hard Hutchison 2001; Edwards 2005), while others find to measure and thus may not be detected by panel that capital controls help insulate countries from the regressions (Kose and others 2006). It is difficult to negative impact of widespread financial instability define exactly how open economies are to capital on growth (Eichengreen and Leblang 2002) and that account transactions, given the myriad forms of con- capital account liberalization increases output volatil- trols that are possible and their varied effectiveness. ity in less financially developed economies, but not in These benefits provide only a temporary, but not more developed economies (Mukerji 2008). permanent, impact on growth, so the effect may not Whatever the average relationship between capital be found in cross-section analyses (Henry 2007). The account liberalization and crises, or between liber- endogeneity of financial integration with respect to alization and long-term growth, crises are traumatic growth also can complicate measurements of the re- and costly, both in terms of welfare and politicians' lationship of the two. The benefits of capital account careers. It is fair to say that the economic literature openness may be greater in environments with strong does not provide a convincing argument that the policies and institutions (for example, external com- welfare benefits of liberalization are extremely large, petition may strengthen sound financial intermediar- or that the risks of crises are extremely small, in ies, but destroy weaker ones). However, it is difficult developing-country environments. a. Kose and others (2006) summarize the results of 25 empirical stud- b. Kose, Prasad, and Taylor (2009) define threshold levels of financial ies, in which the majority find that international financial integration development, openness to trade, institutional quality, and macroeco- has either no effect on growth in developing countries or mixed effects nomic stability that make it possible to realize the benefits of financial depending on model specifications. openness, while recognizing that some of these threshold levels have relatively wide confidence intervals. from these activities and from public incentives regional cooperation in an effort to buffer to strengthen the surrounding infrastructure-- shocks originating in high-income countries. including the domestic and regional competi- Indeed, in the period following the financial tive environment. As a result, domestic and crises in both Latin America and East Asia, regional financial intermediation may well ex- there was a concerted effort to build regional pand compared with a no-change scenario. and domestic institutions that could comple- The severity of the financial crisis and its ment or even replace international financial transmission through global capital markets institutions and thus reduce reliance on what is likely to turn policy makers' interest toward was perceived to be an excessively volatile 92 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H source of intermediation services. Among these exchanges, along with attendant standardiza- efforts were the pooling of financial resources tion of requirements and regulatory coopera- and reserves (for example, currency swaps tion; and improve coordination of fiscal and under the Asian Chiang Mai Initiative and the monetary policies among closely linked re- Latin American Reserve Fund), initiatives to gional economies. promote or create a new reserve currency, and To what extent these potential benefits can the closer integration of local financial mar- be realized is uncertain. In countries or regions kets. The latter included the harmonization of with strong domestic institutions and inter- bond market standards and conventions across mediation, greater reliance on domestic and ASEAN (Association of Southeast Asian regional intermediation may have long-run Nations) countries (Miyachi 2007) and the positive effects. In others, where local institu- Manila Framework to monitor domestic tions are weak, the absence of a foreign option policies, to provide technical assistance for may reduce investors' overall access to savings strengthening domestic financial sectors, and and reduce competitive forces that serve to to establish cooperative arrangements to stabi- keep borrowing costs low. lize Asian currencies (Jeon 2002). Because of a poor legal and institutional To the extent that efforts to promote re- framework for the financial sector or lim- gionally based intermediation solutions are ited financial sector development, many do- successful--and assuming that regional in- mestic (and regional) financial systems are vestors understand better than international poorly prepared to substitute for foreign fi- investors the differences across countries nancial services. As discussed in chapter 2, within regions--regional capital markets domestic intermediation in much of the de- should help reduce domestic borrowing costs, veloping world remains low; equity markets, both by reducing transaction costs27 and by although growing, are thinly traded and dis- reducing risk, notably by lowering the like- persed; and significant domestic debt markets lihood of contagion (when investors with- are restricted to just a few countries (World draw from countries based on problems in Bank 2006). Sub-Saharan Africa, for ex- other countries that are perceived to share ample, has the weakest domestic framework similar features) affecting all countries in a for financial intermediation--the worst lev- region when one runs into fundamental dif- els of corruption, poorest legal framework, ficulties. Regional financial integration could lowest regulatory quality, largest levels of also support regional trade integration; help capital flight, and smallest banking sectors reap the benefits of scale economies achiev- of the developing regions (table 3.3). The able through the amalgamation of regional development of regional equity markets in Table 3.3 Indicators of the quality of domestic financial systems Offshore deposits as a Control of Regulatory Liquid liabilities Private credit % of domestic Region corruption Rule of law quality as a % of GDP as a % of GDP bank deposits East Asia and Pacific 0.6 0.4 0.1 70.3 54.1 4.5 Europe and Central Asia 0.3 0.3 0.2 42.7 36.9 6.5 Latin America and Caribbean 0.4 0.6 0.1 38.2 32.5 60.2 Middle East and North Africa 0.3 0.2 0.2 76.3 45.5 13.7 South Asia 0.6 0.4 0.5 56.9 36.0 4.7 Sub-Saharan Africa 0.6 0.6 0.5 27.3 21.4 82.1 High-income countries 1.5 1.4 1.4 94.8 122.9 13.4 Source: World Bank databases: Governance Matters and Financial Institutions and Structure. Note: Values for the first three indicators range from 2.5 to 2.5, with higher values indicating better performance. 93 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Sub-Saharan Africa could boost liquidity and costs and the premiums associated with a reap economies of scale in the provision of given level of risk can be expected to rise. In services and supervision, particularly if the addition, the discrediting, disappearance, or weaker economies can integrate with more reduced scope for many of the financial instru- developed regional financial markets (such as ments that contributed to the credit boom (see South Africa). However, pursuing integration chapter 2), as well as the large losses sustained among the poorest African countries with by investors throughout the world, are likely weak institutions could simply create a large, to reduce demand for speculative financial in- illiquid, and poorly functioning market. Thus struments going forward. As a result, banks, it is important for these countries to reach firms, and individuals can be expected to out beyond the region to the international in- carry less leverage, reduce their exposure to vestment community. currency and maturity mismatches, and take In any event, increasing regional coop- greater care in evaluating counterparty risk. eration and integration will require time. A sharp increase in risk aversion on the part For example, progress on the pooling of in- of investors was apparent in the immediate ternational reserves in Asia has been slow, wake of the crisis, and the spread on riskier although in May 2008 the ASEAN 3 min- assets such as corporate bonds and equities isters reached agreement to set up the pool spiked sharply in both high-income and devel- with $120 billion in reserves by the end of oping countries--as did the premium required this year.28 Regional integration of Asian of developing-country sovereign borrowers. bond markets appears to have slowed since Since then, borrowing costs in high-income 2002, at least as shown by the correlation countries have been suppressed by monetary of returns across countries (Fung, Tam, and and fiscal actions, notably the efforts by the Woo 2008). Low levels of liquidity and in- U.S. Federal Reserve to keep securitized mort- sufficient administrative expertise have lim- gage markets liquid and to make direct pur- ited steps toward integration in Sub-Saharan chases of corporate bonds. Risk premiums on Africa. The Caribbean Community has failed many assets have declined as the crisis has sta- to implement its commitment to a mon- bilized. The fall in risk premiums partly reflects etary union (Ocampo 2006). While several the very loose monetary conditions in high- regional financial institutions have emerged income countries, which have lowered yields in the Arab world, these have largely been there and prompted the kind of search for yield devoted to the provision of aid to low-in- behavior that characterized the boom period it- come countries (Corm 2006). And even the self. Developing-country risk premiums appear creation of procedures to enhance macroeco- to have reached new "normal" levels that are nomic dialogue in Latin America has made about 150 basis points higher than during the only limited progress (Machinea and Rozen- boom period. Some further moderation may be wurcel 2006). possible, but this tendency will be countered as today's loose conditions tighten, and high- income-country yields rise. Assuming past rela- tions continue to hold, developing-country risk The impact of higher borrowing premiums can be expected to rise with them. costs The remainder of this section suggests that T he preceding discussion suggests that over the medium term the supply of both do- mestic and international finance could be sig- the balance of these competing forces will likely cause borrowing costs in developing countries to be higher than during the boom nificantly lower (relative to economic activity) period, and it explores the implications that than it was during the pre-crisis boom period. increase may have for investment and growth With a reduced supply of credit, borrowing in these countries. 94 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H Weaker financial intermediation Figure 3.7 The global synthetic price of risk and increased risk aversion in high- versus the portion explained by economic income countries fundamentals Changes in risk management techniques, low Synthetic price of risk 6 short-term interest rates (which reduce the cost of financing a given level of risk and tend, at 5 least temporarily, to reduce the likelihood of 4 default), and an increasingly held but ultimately Risk measure 3 misplaced view that overall risk had declined contributed to a significant decline in the price 2 of risk worldwide during the early 2000s. As 1 a result, interest rates and spreads on a wide 0 range of riskier assets in both developed and de- veloping countries fell sharply (see discussion in 1 Predicted risk chapter 2, including figures 2.5 and 2.6). 2 measure Econometric work first undertaken by the Jan. Jan. Jan. Jan. Jan. Jan. Jan. 1998 2000 2002 2004 2006 2008 2010 Organization for Economic Co-operation and Development (OECD 2004; Sløk and Kennedy Source: Kennedy and Palerm forthcoming. 2005) demonstrated that as much as 82 per- cent of the decline in risk premiums during the first half of the 2000s for a wide range of assets given level of risk in high-income countries (high-income investment grade and below-in- declined sharply between 2002 and late 2003 vestment grade corporate bonds, equities, and and remained depressed straight through to developing-country sovereign and corporate mid-2007 before it spiked with the onset of bonds) could be explained by a single common the financial crisis (figure 3.7). factor, which the authors termed a synthetic Since then, the price of risk has declined; an price of risk (box 3.6). update of the measure commissioned for this This synthetic-price-of-risk measure, which study (Kennedy and Palerm forthcoming) put continues to be updated by the OECD (2009), it at about 0.2 points in September 2009, or is used to monitor financial conditions in high- about 1 point (a full standard deviation) higher income markets. It suggests that the cost of a than its average rate during the boom period. Box 3.6 The synthetic price of risk risk.a which is constructed to have mean zero and T he synthetic price of risk was calculated as the first principal component computed from a set of returns on corporate bonds (high grade and high standard deviation one. Thus an increase in this measure of 1 unit does not represent a 100 basis yield) for the United States and the Euro Area ver- point increase in risk premiums. One goal in con- sus their respective government bonds, the implied structing this statistical measure is to explore how risk on equities for each economy, and the global much of the change in the price of risk can be ex- EMBI (Emerging Markets Bond Index Plus) spread plained by economic fundamentals, and how much (versus U.S. Treasury bonds). This synthetic price may result from changes in investors' appetite for of risk is a purely statistical measure of the cost of risk. a. The first principal component is a statistical technique that analyzes the covariance of a group of data series by decomposing the series into orthogonal subgroups composed of weighted averages of the original series. The first principal component is the index of the weighting scheme that explains the largest share of the overall variance of the series being examined. 95 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Despite being a purely statistical construct, than 45 percent of the changes in country- the OECD's synthetic-price-of-risk measure specific risk premiums over the period January can be modeled as a function of fundamental 2002 to April 2009. economic factors, including real policy interest Although changes in these country-specific rates in the United States and the Euro Area, the factors contributed to the rise in country- outlook for the cyclical position of the OECD specific interest-rate spreads, by far the larg- (measured as the 12-month rate of change in est source of the rise resulted from the more the OECD's leading economic indicators); generalized repricing of risk: the increase in and global bond default rates. In normal times the synthetic price of risk explains 75 percent changes in these economic determinants do a or more of the rise in interest-rate spreads be- reasonable job of explaining changes in the tween July 2007 and December 2008 for all synthetic-price-of-risk measure. However, they but two of the developing countries in the were unable to explain much of the fall in the sample (Kennedy and Palerm forthcoming).31 synthetic risk premium during the early 2000s. The estimated relationship between The difference has been attributed to reduced developing-country risk premiums and the risk aversion among investors stemming from global-price-of-risk measure suggests that when financial innovations and "animal spirits" dur- global conditions tighten, developing-country ing this period (Sløk and Kennedy 2005). Funda- risk premiums tend to rise faster than risk pre- mentals only explain about half of the increase miums in high-income countries. Among the in the synthetic price of risk in the immediate main determinants of the global price of risk post-crisis period, with the remainder explained are monetary and fiscal conditions. The OECD by an increase in risk aversion.29 Similarly, the estimates that an increase of 100 basis points analysis of Kennedy and Palerm (forthcom- in policy interest rates in high-income coun- ing) suggests that--through September 2009-- tries will result in a 0.41 increase in the global- the global price of risk fell further than was price-of-risk measure. Combining this with the warranted by fundamentals. results for developing countries, a 100 basis point increase in the base interest rate would The influence of the global price increase the risk premiums of countries with of risk good credit histories by a further 11.1 basis Kennedy and Palerm (forthcoming) extend the points and of countries with less good records earlier OECD work by linking changes in the by an additional 57 basis points. These results global price of risk to changes in the risk pre- are somewhat higher than the estimates derived miums of specific developing countries.30 They by Dailami, Masson, and Padou (2008) in a show that much of the increase in developing- slightly different context.32 country risk premiums during the post- crisis period can be explained by this updated Medium-term developing-country synthetic-price-of-risk measure. Simple corre- interest rates lations indicate that for a sample of 17 devel- The rapid decline in the synthetic-cost-of- oping countries, between 10 and 80 percent of risk measure and in developing-country risk the changes in these countries' risk premiums premiums since they peaked in early 2009 between 1998 and 2008 can be explained by has reduced borrowing costs substantially. fluctuations in the OECD's synthetic-price-of- However, as monetary policy in high-income risk measure, with 40 percent or more of the countries tightens from its currently very loose variation explained in half of the countries in state, interest rates and interest premiums can the sample. A more comprehensive modeling be expected to rise. To what extent they in- of country-specific factors (including foreign crease in the long run is uncertain. indebtedness, macroeconomic stability, politi- Several factors point to higher borrowing cal risk, and economic growth) explains more costs for developing countries in the future. 96 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H Figure 3.8 Yields on selected U.S. government Figure 3.9 Government debt is projected to securities rise in high-income, but not developing, Yields, % countries 9 Percent of GDP High-income countries 8 1-year T-bill 120 7 10-year government bond 100 6 80 5 Middle-income countries 60 4 3 3-month T-bill 40 2 20 1 Low-income countries 0 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Jan. 1 Jan. 1 Jan. 1 Jan. 1 Jan. 1 Jan. 1 Jan. 1 Jan. 1 1995 1997 1999 2001 2003 2005 2007 2009 Source: IMF. Source: Datastream. · The global price of risk is currently sup- enough to lower risk premiums by more pressed by very loose monetary policy than the increased costs the regulations conditions (see figure 3.7). impose, resulting in a net reduction in · Base rates used to calculate developing- borrowing costs. country borrowing rates are also low · Increased indebtedness in high-income because of post-crisis flight to quality countries is likely to raise their borrow- (figure 3.8), and this effect can be ex- ing costs and developing-country base pected to dissipate over time. rates, but it also may result in lower · Base rates are likely to rise further interest rate premiums for developing because of the sharp increase in high- countries. Developing countries' govern- income countries' government debt, ment debt is not expected to increase from under 80 percent of GDP in 2007 as much as in high-income countries, to perhaps as much as 109 percent of largely because of a limited access to GDP by 2014 (figure 3.9). Reducing finance (see chapter 1). As a result, rela- these debt levels will take some time; tive to high-income-country debt, the and as long as they and deficits remain inherent riskiness of developing-country high, so too will governments' demand debt may decline, along with their risk for funds to cover interest payments. premiums, although probably not by Moreover, high debt levels will increase as much as they rise in high-income investors' concern over the sustainabil- countries. Overall, developing-country ity of fiscal policy, which also will tend borrowing costs would still be higher to increase base interest rates. but the relative attractiveness of invest- ments in developing countries would be At the same time, several factors suggest enhanced, and they could be expected that developing-country borrowing costs may to attract a larger share of a reduced not rise as much as otherwise might have been quantity of total lending (see McKibbin expected. and Stoeckel 2009 for a modeling of this effect in a slightly different context). · Improved regulation, by increasing transparency, may lower the uncertainty In the modeling work that follows, neither surrounding investment instruments by the negative nor positive implications of these 97 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Table 3.4 Historical and prospective costs of capital for developing countries Real Federal Real base Funds rate East Europe Latin Middle rate (deflated (deflated Asia and America East and Sub- Base by core by core and Central and North South Saharan Costs rate inflation) inflation) Pacific Asia Caribbean Africa Asia Africa (Percent) (Basis points) Interest rates and spreads Average 1995­2003 5.8 2.7 2.0 274 1135 802 434 276 832 Average 2004­May 2007 4.5 2.4 0.1 224 196 384 351 215 289 Level in August 2008 3.9 1.4 3.3 286 273 350 474 287 301 Current (Oct. 2009) 3.0 1.6 0.4 304 412 362 419 286 408 Consistent with long-term price of risk 2.0 318 523 478 496 363 470 Cost of capital (Percent) Average 1995­2003 9.7 2.0 12.4 21.1 17.7 14.0 12.5 18.0 Average 2004­May 2007 9.4 0.1 11.7 11.4 13.3 13.0 11.6 12.3 Level in August 2008 8.4 3.3 11.2 11.1 11.9 13.1 11.2 11.4 Current (Oct 2009) 8.6 0.4 11.7 12.7 12.3 12.8 11.5 12.7 Spreads constant at October 2009 level U.S. base rate at avg 04-07H1 level 9.4 12.5 13.6 13.1 13.6 12.3 13.5 U.S. base rate at avg 95-03 level 9.7 12.7 13.8 13.3 13.9 11.9 12.6 Spreads at level consistent with average price of risk U.S. base rate at avg 04-07H1 level 9.4 12.6 14.7 14.2 14.4 13.1 14.1 U.S. base rate at avg 95-03 level 9.7 12.9 14.9 14.5 14.7 13.3 15.4 Source: World Bank. factors on borrowing costs are taken into ac- pre-crisis period, notwithstanding significant count explicitly. Rather, a range of estimates declines from their peak levels during the most for future borrowing costs is presented, based acute phase of the crisis. upon different historical "norms" for real base As discussed in chapter 2, the widespread interest rates and recent relationships between fall in borrowing costs before the crisis was developing-country spreads and the price of associated with an investment boom and a sig- global risk. nificant increase in capital-to-output ratios in The top panel of table 3.4 reports histori- developing countries (figure 3.10). Many fac- cal real interest rates and interest rate spreads tors, including the tax regime, inflation, the for selected developing-country regions. Dur- rate of depreciation of the asset in question, ing the boom period 2004­May 2007 the real and other regulatory features that impinge on yields on long-term U.S. treasury bills were profitability (Jorgenson 1963; Lau 2000) go 30 basis points lower than during the previ- into determining the cost of capital in a given ous nine-year period. Spreads on developing- country (or firm for that matter); however, country debt were about 370 basis points most of these factors are relatively constant lower, implying that during the pre-crisis pe- over time, and there is very limited informa- riod, borrowers coming to international mar- tion on their appropriate values at the country kets were paying some 400 basis points less level. As a result, the second panel of table 3.4 for U.S. dollar loans than in the preceding converts the historical interest rates and inter- nine-year period. As of early October 2009, est rate spreads of the first panel into estimates risk premiums remain on average about 110 for the cost of capital based on very strong basis points higher than they were during the simplifying assumptions. Specifically, the cost 98 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H that spreads remain at their current levels but Figure 3.10 Falling capital costs were that real base rates in the United States return reflected in rising capital-output ratios to their pre-boom-period levels. In the second Change in capital-output ratio (2000­08), % of GDP scenario, the same assumption about base 20 rates is applied but spreads are assumed to rise 15 in response to a return to the levels consistent with the long-term price of global risk. And 10 the final scenario builds on the previous two by assuming that U.S. base rates rise by a fur- 5 ther 100 basis points above their pre-crisis av- erage levels, reflecting the potential impact of 0 higher high-income-country debt and tighter regulations. ­5 In the first scenario, the cost of capital d La cif ia be ca ric d ia ric n l A rop g Af an Af ara Pa s As tra u pin ib ri th tin ic an a a d tA si e ar e rises a great deal in some regions (notably by th t h en t E lo C Am or as h an Eas Sa a) ut C p e N eE d ce ev So b- as much as 240 basis points in Europe and an (ex D Su dl id e M Central Asia), while in others it changes rela- an tively little. In the second scenario, interest Source: World Bank. rate spreads rise from current levels to levels consistent with fundamentals. This implies a of capital at the country level is defined as the further capital cost increase of 50­100 basis real dollar borrowing cost (deflated by core points in every region other than East Asia inflation from the U.S. consumer price index) and the Pacific. In the final scenario, the cost plus depreciation, which is assumed to be increases are augmented by a 100-basis-point 7 percent throughout the developing world.33 increase in the base interest rate--assumed In addition to abstracting away from many to be caused by concerns about debt servic- of the determinants of the cost of capital, ing, long-term inflation, and the influence of this definition also assumes that developing- tighter regulation raising the cost of capital in country borrowers (both domestic and exter- high-income countries. nal) pay the U.S. Treasury-bill rate plus the in- terest rate spread. In reality, formal domestic interest rates tend to be higher, and informal Implications of higher borrowing borrowing rates significantly higher still. costs for growth and potential GDP Table 3.5 calculates the change in the cost in the medium term of capital by region between the boom period Economic theory suggests that higher capital and three alternative futures. The first assumes costs would cause the desired capital-output Table 3.5 Possible impact of tighter financial conditions on developing-country capital costs East Asia Europe and Latin Middle East Sub- Real base and Central America and and North South Saharan Alternative scenarios rate Pacific Asia Caribbean Africa Asia Africa Change in cost of capital compared to the boom period (percentage points) Spreads contant at today's level 0.3 1.1 2.4 0.0 0.9 0.3 0.3 Spreads consistent with long-term average price of risk 0.3 1.2 3.5 1.2 1.7 1.7 2.1 Higher base rates and spreads consistent with long-term price of risk 1.3 2.2 4.5 2.2 2.7 2.7 3.1 Source: World Bank. 99 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Box 3.7 Higher borrowing costs will constrain domestic and external finance T he expected rise in risk aversion and in the cost of capital will likely slow, possibly even reduce, the gains in domestic intermediation that the low in- Box figure 3.7.1 The growth of bank credit to the private sector dropped following the collapse of Lehman Brothers terest rates of the boom period facilitated (see chap- ter 2). Already, the sudden increase in borrowing Average monthly percentage change in real terms costs and increased risk aversion that followed the 1.8 1.6 crisis have cut into lending (and borrowing) through- 1.4 out the developing world (see also chapter 1). Since 1.2 September 2008 the pace of credit expansion in the 1.0 69 developing countries for which data are avail- 0.8 0.6 able fell by almost half, from a monthly increase of 0.4 around 1.1 percent between January and September 0.2 2008 to a much more modest 0.6 monthly pace be- 0 tween September 2008 and October 2009 (box fig- All developing Low Lower-middle Upper-middle countries income income income ure 3.7.1). The decline was particularly pronounced in middle-income countries, perhaps because their Jan. 2008­Sep. 2008 Sep. 2008­May 2009 more integrated financial systems were most directly affected by the change in global financial conditions. Sources: International Financial Statistics, IMF. While sovereign bond flows have picked up recently, Note: Refers to 69 countries with data through May 2009. Bank credit to the private sector is deflated by the U.S. Consumer corporate bond issues and especially bank lending Price Index. remain sharply lower than during the boom period. As policy interest rates in high-income countries rise, the differential between long-term yields on Central Asia, that were unable to sustain the very developing-country debt and short-term dollar- strong capital inflows of the first part of this decade. denominated borrowing costs will decline, and the As seen in the discussion of the boom in chapter 2, financial incentive that is currently boosting capital higher financing costs will affect international capital flows to developing countries will ease. As a result, flows as well. Higher interest rates will reduce the borrowing costs are expected to be relatively high demand for external finance, while banks are likely over the medium term, and therefore the growth of to lower the supply of loans because marginally cred- domestic financial intermediation is also likely to itworthy developing countries will be able to reliably moderate on a sustained basis. While these events service a smaller quantity of debt. Thus the expected are likely to have negative consequences for develop- rise in borrowing costs will compound the reductions ment in many countries, they may have some benefits in external finance owing to tighter regulation in in those countries, including many in Europe and industrial countries (described above). ratio in the economy to fall. As a result, if costs Table 3.6 reports on three simulations do rise in the longer run, entrepreneurs will have undertaken to provide some insight into the less capital to work with than if interest rates likely quantitative impact of the higher cost of had remained low (box 3.7).34 This implies capital on growth. The interest rate scenarios that the level of output that the economy will that drive the simulations are those outlined be able to sustain is likely to decline and that in table 3.4. during the transition period to the new lower In these simulations, higher borrowing capital-output ratio, the reduced rate of growth costs serve to lower long-term developing- of potential output will be temporarily lower. country potential output (measured as the 100 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H Table 3.6 Impact on potential output of a return to normal pricing of risk and higher base rates Spreads return to "normal" Spreads remain at October Spreads return to "normal" levels and base rates rise 2009 level, but base rates levels, and base rates rise 100 basis points above rise to pre-crisis levels to pre-boom levels pre-boom period levels Short-term Short-term Short-term impact on Long-term impact on Long-term impact on Long-term potential level of potential level of potential level of growtha potentialb growtha potentialb growtha potentialb (percent) Developing countries 0.2 3.4 0.4 5.2 0.7 8.0 Middle-income countries 0.3 3.5 0.4 5.2 0.7 8.0 Low-income countries 0.1 1.0 0.5 5.7 0.7 8.2 East Asia and Pacific 0.3 3.9 0.4 4.3 0.6 7.5 Europe and Central Asia 0.5 6.5 0.8 9.2 1.0 11.4 Latin America and Caribbean 0.0 0.0 0.3 3.6 0.6 6.3 Middle East and North Africa 0.2 3.0 0.4 5.4 0.7 8.2 South Asia 0.1 1.0 0.4 5.1 0.6 7.8 Sub-Saharan Africa 0.1 0.5 0.4 3.4 0.5 4.9 Source: World Bank. a. Short-term impact on potential growth rate change is measured in percentage points during the transition to new lower level of potential output (about 5­7 years). b. Percent change in long-term level of potential output. difference between the simulated 2050 level of The second scenario assumes that de- potential in the base case versus the scenario)35 veloping-country spreads rise somewhat as by between 3.4 and 8.0 percent overall, with monetary conditions tighten, in line with the regional declines as large as 11.4 percent. The historical relationship between developing- impact on trend growth during the period in country spreads and high-income policy rates. which economies transition down to these Overall, potential output falls 5.2 percent in lower long-term levels of potential (about the long run, and transitional growth rates are 5­7 years in these simulations) is less marked, expected to be about 0.4 percentage points averaging between a 0.2 and 0.7 percentage lower than they would have been otherwise. point annual decline. In this second scenario, low-income countries The first simulation can be understood as are hit somewhat harder relative to the first a relatively benign outcome. It assumes that scenario because their interest rate premiums developing-country spreads remain at to- are projected to rise by more than those of day's levels--somewhat lower than might be middle-income countries. expected when high-income monetary policy In the third scenario, weaker fundamen- tightens and interest rates rise--and that inter- tals are assumed to raise borrowing costs in est rates in high-income countries rise to their high-income countries an additional 100 basis pre-crisis average level between 2002 and points. This causes borrowing costs to rise 2005. Even with this modest tightening of li- globally, with developing-country long-run quidity conditions, long-term potential output potential output falling by around 8 percent in developing countries falls 3.4 percent com- and annual potential growth rates falling by pared with a scenario where the cost of capi- 0.7 percentage points for about seven years. tal remained at the very low levels observed The impact of higher interest rates, invest- during the boom period. During the transition ment, and the capital stock is illustrated in to new lower capital ratios, the annual rate of figure 3.11, which is drawn from the second growth of potential output declines by about scenario. The declining capital stock causes the 0.2 percentage points for about seven years. level of output that the economy can sustain 101 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Figure 3.11 Impact of 110-basis-point shock Figure 3.12 Higher borrowing costs result in on the capital-output ratio of a typical a permanent decline in developing-country country GDP Capital-output ratio, postshock level divided by baseline level (Potential output) 1.05 Trillions of real 2005 U.S. dollars 1.00 6 0.95 Low capital costs growth path 0.90 5 0.85 0.80 4 Higher capital costs 0.75 growth path 0.70 3 0.65 0.60 2 2005 2013 2021 2029 2037 2045 1 Source: World Bank. 2005 2010 2015 2020 2025 2030 Note: Results from the second scenario in table 3.6. Source: World Bank. in the long run to be lower than it would movement of both people and capital from less have been had borrowing costs remained to more productive uses, especially if it suc- unchanged. Overall, the capital-output ratio ceeded at the same time in boosting aggregate falls by about 12 percent compared with the productivity, would help to speed the transi- baseline--a baseline that already includes the tion toward a new equilibrium. post-crisis recession but assumes that the long- The long-run declines in potential output vis- term cost of capital returns to its boom-period à-vis the baseline reported in the first and sec- level. Initially the capital-output ratio rises ond scenarios of table 3.6 are broadly consistent in response to the shock because the decline with results produced by McKibbin and Stoeckel in investment it provokes causes GDP to fall (2009) in their analysis of the global financial more quickly than the capital stock. Relatively crisis--even though their focus is not on the quickly the combination of depreciation of the impact on potential output in developing coun- existing capital stock and slower investment tries and the shock is not framed in precisely the growth causes the capital stock to fall.36 same manner. In part, this similarity reflects the As the new capital-output ratio is achieved, fact that their model includes many of the same the pre-crisis potential output growth rate is mechanisms as the one presented here. These re- regained. Nevertheless, as presented in table sults are also very similar in size to the average 3.6, there is a permanent 5.2 percent loss in post-financial-crisis decline in potential output GDP (figure 3.12). identified by the International Monetary Fund Of course, this process may occur more in its September 2009 World Economic Out- quickly in the real world owing to a more look (IMF 2009), although that work, which rapid depreciation of the existing capital stock was based on a decomposition analysis of the stemming from the closing down of industries economic outturns following past crises, attrib- that are no longer competitive. An important uted a smaller share of the total decline to the implication of this logic is that a policy de- decline in the capital output ratio. The analysis signed to assist existing firms that are placed presented here differs by explicitly modeling in in long-term difficulty by less liquid global a forward-looking manner the main mechanism conditions will tend only to prolong the period through which lower potential output is reached of slow growth through which an economy and relating it to the primary driver of both the must pass in these circumstances. In contrast, boom and the bust--the decline and subsequent a policy that combined efforts to facilitate the sharp rise in the global price of risk.37 102 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H Table 3.7 Higher borrowing costs and slower population growth imply slower growth in potential output over the longer term Latin Middle East Sub- Developing Middle Low East Asia Europe and America and and South Saharan countries income income and Pacific Central Asia Caribbean North Africa Asia Africa (Average rate of growth of potential output, percent) Baseline scenario: spreads and base interest rates constant at boom period levels 1980­95 2.6 2.6 2.6 8.3 0.6 2.4 2.6 5.1 1.8 1996­2002 4.6 4.6 3.9 7.8 3.1 3.0 4.2 5.8 3.4 2003­08 6.2 6.2 5.4 8.7 5.9 3.5 5.0 7.4 5.6 2009­15 6.3 6.3 5.8 8.1 5.7 3.8 5.4 7.4 6.5 2016­50 4.5 4.5 5.2 5.5 4.5 1.6 2.9 4.9 5.1 Pre-crisis base rates and return to normal spreadsa 1980­95 2.6 2.6 2.6 8.3 0.6 2.4 2.6 5.1 1.8 1996­2002 4.6 4.6 3.9 7.8 3.1 3.0 4.2 5.8 3.4 2003­08 6.2 6.2 5.4 8.7 5.9 3.5 5.0 7.4 5.6 2009­15 5.9 5.9 5.4 7.8 5.2 3.3 4.9 7.0 6.3 2016­50 4.5 4.4 5.2 5.4 4.4 1.5 2.8 4.8 5.1 (Percentage point difference in average potential output growth rates) Difference compared with baseline 2003­08 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2009­15 0.4 0.4 0.4 0.3 0.5 0.5 0.4 0.4 0.3 2016­50 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.0 Source: World Bank. a. Results from the second scenario in table 3.6. The top panel of table 3.7 illustrates the Slower future growth and lower income influence of slower population growth and levels imply potentially large long-term im- an aging population on potential output in pacts on poverty and disease in developing the baseline scenario, where borrowing costs countries. Recent estimates suggest the reces- are assumed to remain unchanged at their sion will likely result in some 30,000­50,000 2008 level and where total factor productiv- additional childhood deaths in 2009 and 2010 ity trends remains unchanged. In this scenario, in Africa (Friedman and Schady 2009). And, developing-country growth is set to slow by assuming the 4.2 percent of GDP decline in 1.7 percentage points between the boom pe- potential output reported in the second sce- riod and the 2016­50 period, mainly because nario of table 3.7, by 2015 the crisis and its of slower growth of the working-age popula- aftermath can be expected to have prevented tion reflecting both aging and slower popu- some 46 million poor people from emerging lation growth. Total factor productivity is from poverty (table 3.8). assumed to be constant.38 The second panel The higher borrowing rates under the sce- shows average potential output growth rates narios outlined above are unlikely to have a assuming the same increase in capital costs as major impact on debt sustainability, even for in the second scenario of table 3.6. Although the most highly indebted developing countries. the long-term growth rate of potential out- Assuming that borrowing costs rise by put after adjustment of the capital stock to 2.7 percentage points (roughly equivalent its new levels is broadly the same as in the to the most pessimistic of the scenarios in baseline case, potential output growth is 0.4 table 3.6), private-source-debt-to-GNI ratios percentage points slower during the transi- among developing countries would rise by at tion period and potential output in the end most 6 percentage points. In most of the 20 period is some 5.2 percent lower than in the developing countries with the highest ratio of baseline. private-source debt to GNI, the rise is about 103 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Table 3.8 The crisis could increase poverty Strategies for dealing with a by 46 million in the long term Impact on poverty of a weaker international finance 5.2 percent decline in potential output as of 2015 system Change in head count Percent change W hile individual developing countries can do little to influence developments in the global financial system, they can do a great Region (millions) in head count deal at the local level to influence intermedia- East Asia and Pacific 6.3 19.3 tion costs and offset, possibly even reverse, the Europe and Central Asia 0.9 27.7 expected increase in capital costs--perhaps not Latin America and Caribbean 3.4 14.3 Middle East and North Africa 0.8 46.8 immediately but over the longer term. South Asia 16.3 31.2 As discussed in chapter 2, the most important Sub-Saharan Africa 18.2 11.7 influence on changes in borrowing costs in de- Developing countries 46.0 17.0 veloping countries during the past 10 years has Source: World Bank. been changes in the international cost of risk. Note: Estimates based on the GIDD model. However, inefficiencies in domestic financial systems are more important than international financial conditions in explaining the levels of 1 percentage point (figure 3.13). However, borrowing costs in many developing countries. the impact of higher interest rates for coun- Fundamentals still matter--in determining both tries with external borrowing constraints the risk premium attached to financial transac- could be more disruptive if it occurs quickly. tions with individual countries and the relative If borrowing costs rise 270 basis points, a cost effectiveness of a given investment project. country with a private-source-debt-to-GNI Even the financial systems of the most advanced ratio of about 40 percent (the median of the developing countries underperform the financial 20 most indebted countries) would have to re- systems in high-income countries. For example duce its foreign exchange expenditures by 1.1 11 of the 14 developing countries included in percent of GNI. Assuming a 4 percent import an aggregate index of financial system effi- elasticity, this would imply a 4.4 percent cut ciency scored in the bottom half of the rankings in domestic demand.39 (Dorrucci, Meyer-Cirkel, and Santabarbara 2009). Figure 3.13 The effect of higher borrowing costs on the ratio of private-source debt to gross national income Percent of GDP 7 6 5 4 3 2 1 0 ria oa n ia s ia a va ia e ia nd YR le ca nd n ey a na lle ta i in io in tv ar an rb an hi do ai rk be m la la vi a at nt hs ,F he La C lg Se m go rit kr om Po ai Tu Sa ol er ge Li ak Bu ia yc Ja Th au U M ze d R Ar on z Se Fe M Ka er ed H an ac a si ni M us s Bo R Source: World Bank. 104 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H Table 3.9 Selected indicators of banking sector efficiency (Operational ratios of commercial banks, average 2006­08) Net interest revenue/ Other operating Noninterest Profit margin Net interest average assets income/average expenses/ (before tax and non- Region margin (%) (%) assets (%) average assets (%) operational items East Asia and Pacific 3.85 3.34 1.52 3.09 1.77 Europe and Central Asia 6.94 5.86 11.05 14.18 2.73 Latin America and Caribbean 8.38 6.69 3.51 8.44 1.77 Middle East and North Africa 3.21 2.58 1.66 2.66 1.58 South Asia 3.47 3.05 1.88 3.26 1.68 Sub-Saharan Africa 7.17 5.80 4.23 6.58 3.45 Developing countries 6.62 5.50 7.12 10.22 2.40 High income 3.57 3.25 1.94 4.06 1.13 Low income 6.16 5.02 3.46 5.47 3.01 Lower-middle income 5.57 4.37 2.42 5.21 1.58 Upper-middle income 7.00 5.91 9.05 12.39 2.57 Source: World Bank calculations using data from Bankscope. Table 3.9 reports information on net inter- reducing the overall level of intermediation that est revenues (the difference between interest banks might otherwise accomplish. income paid on deposits and earned on loans), Although interest rate margins and over- other income, and noninterest expenditures head costs provide only rough indicators of (principally overhead), all expressed as a per- the economic costs imposed by inefficient fi- cent of the total value of banks' assets. On av- nancial systems, they do suggest considerable erage, the difference between interest paid on potential for reducing borrowing costs by im- deposits and charged on loans by developing- proving efficiency. For example, if the spread country banks is 225 basis points higher than between deposit and lending rates in Sub- in high-income countries. Overhead expendi- Saharan Africa, Latin America, and Europe tures are more than twice as large as in high- and Central Asia were reduced to the levels income countries, and profit rates are twice as observed in East Asia or South Asia, borrow- high. In addition, in some developing countries ing costs in these regions would be lowered by banks rely much more heavily on nonbanking between 300 and 450 basis points, much more activities for revenues and profits. Indeed, if than the estimated increase from tighter global developing-country banks did not have these financial conditions in even the most pessi- other revenues, they would be making losses in mistic scenario discussed earlier. Based on the all but one of the developing regions. econometric work reported in chapter 2, such High overhead costs (as a share of total asset a reduction in borrowing costs could prompt ratios) reflect a variety of factors, and their rela- an increase in the level of domestic intermedia- tive weight varies across banks and countries. In tion equivalent to 2.5 percent of GDP. countries and regions that also show high profit Moreover, the differences in banks' net rates (Sub-Saharan Africa and to a lesser extent interest margins between developing and Europe and Central Asia), these high overheads high-income countries are considerable under- may reflect a lack of competitive pressure as statements of differences in the costs of bor- well as a thin market. High levels of nonin- rowing. Particularly in low-income countries, terest income such as in Europe and Central small formal banking systems tend to drive Asia, Latin America, and Africa suggest that borrowers to informal lenders. Because the shortcomings in the regulatory environment, interest rates charged by informal lenders are including in the protection of property rights, often several multiples of interest rates in high- may make nonbanking activities more profit- income countries, improvements in efficiency able than traditional deposit and loan making, that expanded developing countries' banking 105 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 systems could result in very large reductions other than through higher borrowing costs. in interest rates facing less-creditworthy bor- Thus, according to the regression analysis in rowers. In addition, many potential borrow- chapter 2, in addition to a lower cost of capi- ers are denied finance altogether, so efficiency tal, a one-standard-deviation improvement in improvements that expanded the volume of the business environment would increase the credit would have large welfare effects. profitability of investments by enough to in- The impact of improvements in the regula- crease investment-to-GDP ratios by another 4 tory environment would likely be even more percentage points. positive, because weak protection of property Table 3.10 gives a sense of the potential rights and excessive levels of corruption reduce growth implications if developing countries the profitability of investment projects in ways were successful in reducing borrowing costs Table 3.10 Potential impact of improved fundamentals on long-term growth prospects Middle East and Sub- Developing Middle Low East Asia Europe and Latin America North South Saharan countries income income and Pacific Central Asia and Caribbean Africa Asia Africa (Average rate of growth of potential output, percent) Baseline scenario: spreads and base interest rates constant at their 2008 level 1980­95 2.6 2.6 2.6 8.3 0.6 2.4 2.6 5.1 1.8 1996­2002 4.6 4.6 3.9 7.8 3.1 3.0 4.2 5.8 3.4 2003­08 6.2 6.2 5.4 8.7 5.9 3.5 5.0 7.4 5.6 2009­15 6.3 6.3 5.8 8.1 5.7 3.8 5.4 7.4 6.5 2016­50 4.5 4.5 5.2 5.5 4.5 1.6 2.9 4.9 5.1 Higher base rates and return to "normal" spreads 1980­95 2.6 2.6 2.6 8.3 0.6 2.4 2.6 5.1 1.8 1996­2002 4.6 4.6 3.9 7.8 3.1 3.0 4.2 5.8 3.4 2003­08 6.2 6.2 5.4 8.7 5.9 3.5 5.0 7.4 5.6 2009­15 5.9 5.9 5.4 7.8 5.2 3.3 4.9 7.0 6.3 2016­50 4.5 4.4 5.2 5.4 4.4 1.5 2.8 4.8 5.1 Higher base rates, initially "normal" spreads that fall subsequently due to improved policies 1980­95 2.6 2.6 2.6 8.3 0.6 2.4 2.6 5.1 1.8 1996­2002 4.6 4.6 3.9 7.8 3.1 3.0 4.2 5.8 3.4 2003­08 6.2 6.2 5.4 8.7 5.9 3.5 5.0 7.4 5.6 2009­15 6.0 6.0 5.5 8.1 5.3 3.4 5.1 7.1 6.3 2016­50 4.9 4.8 6.0 5.6 5.1 1.9 3.3 5.5 5.5 (Percentage point difference in average potential output growth rates) Difference compared with baseline Higher base rates and return to normal spreads 2003­08 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2009­15 0.4 0.4 0.4 0.3 0.5 0.5 0.4 0.4 0.3 2016­50 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.0 Higher base rates, initially "normal" spreads that fall subsequently due to improved policies 2003­08 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2009­15 0.3 0.3 0.3 0.0 0.4 0.4 0.3 0.3 0.2 2016­50 0.3 0.3 0.7 0.1 0.5 0.3 0.4 0.6 0.3 (Cumulative change in potential output) Change in final potential output compared with baseline Higher rates scenario 4.8 4.8 4.9 4.3 6.0 6.3 5.7 5.4 2.8 Higher rates plus improved policy scenario 7.4 7.0 20.4 1.5 14.8 7.9 8.8 17.2 8.1 Source: World Bank. 106 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H through a mixture of regulatory reform and savings and investments could be substantial improved macroeconomic management. It because middle-income countries have become compares the outcomes from the third scenario the source of more than half of global invest- in table 3.6 (the scenario with higher base rates ment growth. Moreover, lower investment and a return to "normal" interest rate spreads) rates will likely have an impact on regional and a scenario that builds upon the same in- imbalances. In Europe and Central Asia lower crease in base rates and interest rate spreads rates will contribute to a reduction in current but assumes that developing countries con- account deficits, in a way similar to what hap- tinue to make strides in reducing borrowing pened in East Asia after the 1998 financial cri- costs (by 25 basis points a year until spreads sis. In East Asia, where annual growth rates reach 150 basis points). In this simulation, de- during the next five years are expected to be veloping countries still experience a period of almost 1 percentage point lower than during significantly slower growth in the five to seven the boom, lower investment rates will increase years following the crisis, but the assumed cu- current account surpluses--unless savings mulative improvements in fundamentals result rates decline. in a gradual increase in growth rates relative In low-income countries, capital needs re- to both the higher base rates and baseline sce- main substantial. The impact of higher bor- narios. Overall, these improvements mean that rowing costs could be counteracted by total potential output is actually some 7.4 percent factor productivity gains that are already in higher than in the original baseline scenario the pipeline and by sustained population and 12.2 percent higher than in the higher growth. rates scenario, with the largest gains accruing The increase in borrowing costs described to those countries and regions currently facing in the preceding pages reflects both an in- the highest spreads. crease in the cost of intermediation (stemming from stricter regulation of financial markets and consolidation in the banking sector) and an increase in the price of a given quantity Implications for the global of risk (caused by the disappearance of some balance between savings and risk-management instruments and a general- investment ized increase in risk aversion). The risk-free T he preceding analysis focused on the im- pact of increasing costs of capital on po- tential output. However, GDP growth may fall interest rate that is relevant for savers is not affected directly by these factors. It is affected indirectly, however, because higher interme- short of historical levels for reasons other than diation costs and risk premiums will increase a decline in the capital-output ratio during the the wedge between borrowers' costs and the transition period. The adjustment process, dur- risk-free interest rate that motivates savers. If ing which investment rates are suppressed, may savings are interest-rate elastic, this will pro- also have an impact on effective demand, and voke a rightward movement along both the thus on utilization rates, although that impact demand and supply of savings curves, reduc- is more speculative than the analysis of poten- ing the risk-free rate of interest (from rs1 to tial growth. rs2 in figure 3.14). This decline will reduce Especially in middle-income countries, in- savings and increase domestic demand--at vestment rates are likely to remain substan- least partially compensating for the decline tially lower during the next five years than in investment and working to eliminate cur- during the boom period as the capital stock rent account surpluses. The overall impact on adjusts to higher borrowing costs and slowing global imbalances will depend on the balance population growth. The impact of lower in- of these affects across current account surplus vestment rates on the global balance between and deficit countries. 107 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 financial environment, but also one character- Figure 3.14 Higher intermediation costs will ized by less liquid and more expensive financing reduce the risk-free interest rate relevant to savings decisions conditions. The combination of a more constrained Increased intermediation costs raise borrower's interest rates and lower those of savers and more expensive financing environment is Percent likely to have important real-side implications 10 for developing countries. For low-income countries with relatively weak domestic finan- rb(2) rb(1) cial sectors and binding capital constraints, weaker bank finance and FDI flows will be rs(1) rs(2) particularly problematic. In some of these countries FDI inflows represent more than 40 0 percent of total investment, so the projected 0 100 30 percent cut in flows from pre-crisis levels Volumes could have particularly strong repercussions. Source: World Bank. For middle-income countries with access to international financial markets and better de- veloped domestic markets, the main impact is Apart from increasing consumption shares likely to be through the increased cost of bor- in some of the emerging economies, measures rowing. Here, estimates suggest that the ad- to counteract the potential decline in invest- justment to higher capital costs could reduce ments may be needed too. The possibility potential growth rates in developing countries of ineffective demand provides another rea- by between 0.2 and 0.7 percentage points over son to accelerate productivity growth, either the next five to seven years, and that, longer by eliminating bottlenecks or by opening up term, potential output in these countries could new areas of growth. Only with such growth be lower by between 3.4 and 8.0 percent. impulses can investment rates in developing The extent to which domestic finance will countries return to the levels seen during the be able to step in to supplement a weaker in- boom. In that sense the most successful strat- ternational financial system is very uncertain. egy to create sufficient effective demand in the Countries with strong policies and institu- medium run is consistent with policies that tions have the most to lose from a weakening aim to maximize potential growth. of the international financial system because they benefit most from it. However, in gen- eral they also have the strongest domestic Conclusion intermediation systems and are well placed D eveloping countries are likely to face a much more constrained financial envi- ronment over the next decade or so than they to compensate for a weaker international system by expanding this sector. Increased domestic intermediation could ultimately did during the pre-crisis boom. The discredit- generate larger benefits than the cost that ing of many of the financial innovations that reduced access to international capital might led investors to believe that the ultimate cost impose. For many low-income countries the of holding a risky asset had declined sharply, domestic financial sector may be too weak a tighter regulatory environment, more risk to respond effectively. Moreover, while the aversion on the part of investors, and the ne- immediate costs from reduced international cessity of both the high-income banking and flows may be smaller--because these coun- household sectors to consolidate and rebuild tries have little access and therefore little to their balance sheets is expected to result in a lose--the longer-terms costs could be signifi- more stable and ultimately more robust global cant. In particular, if international finance 108 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H and the expertise and spillover effects that of intermediation suggests that policies that can accompany it are not available or are succeed in increasing the efficiency of the do- available to a lesser degree, in the future mestic banking sector--through improved this could hamper the ability of low-income, enforcement of property rights, enhanced weakly intermediated countries to make the competition, or better regulation--could have transition to middle-income countries with significant impacts on domestic intermedia- more robust financial systems. tion, investment levels, and growth potential. Developing countries themselves can be ex- Borrowing costs in many regions could be re- pected to react to the crisis in ways that are duced by more than 300 basis points, a reduc- not necessarily beneficial. The crisis has made tion that could be associated with a long-term clear once again the lesson that financial flows increase in potential output of 8 percent or can be very volatile and that this volatility more. Of course, implementing the reforms to can generate extremely costly real-side adjust- take advantage of this potential will be a slow ments. With GDP growth off by as much as 4 and difficult process--especially if they run percent, it is a small consolation to a country afoul of domestic interests. that has pursued prudent macroeconomic and structural policies that the growth recession Notes that it has experienced was much smaller than 1. It was possible to retain the risk of off-balance-sheet that of countries that went into the crisis with assets but avoid capital requirements by abusing excep- large current account and budget deficits. tions to the "true sale" rule governing securitization; that Throughout the world public pressure and rule states that to avoid capital requirements the seller of prudent policy making will force a rethink- loans must not retain any responsibility for subsequent loan performance or collateral. U.S. regulators were ing of the costs and benefits of both financial aware of this practice but found it difficult to control and trade openness. The evidence supports a (Calomiris and Mason 2004). cautious approach to capital account liberal- 2. Calomiris and Mason (2004) argued that the ization, while supporting the view that rela- use of securitization with implicit recourse to evade tively flexible exchange rate regimes are less capital requirements was socially beneficial, essen- susceptible to the kind of crises that have been tially because regulatory capital requirements were observed in the past. Countries with weak do- too high. 3. Efforts by regulators to close insolvent banks and mestic institutions and limited intermediation impose capital requirements that are commonly based on can find a too rapid capital account liberaliza- the risk of an individual bank can fail to mitigate sys- tion to be destabilizing, while those with more temic risk (Acharya 2009). mature domestic systems can benefit from 4. For a broad view of potential changes in financial the additional competition and in some cases sector regulation, see the G-20 statement on "Strength- funding that a more liberalized external ac- ening the Financial System," issued at the London count can provide. Similarly, while a buildup Summit on April 2, 2009; "A New Foundation: Re- building Financial Supervision and Regulation" (issued of reserves gives a country additional room for by the Obama administration in June 2009); and "The maneuver, in the event of a crisis these reserves Turner Review: A Regulatory Response to the Global can be very expensive to maintain and, of Banking Crisis" (issued by the United Kingdom's Fi- course, cannot prevent exchange rate revalu- nancial Services Authority in March 2009). One exam- ations in the presence of overwhelming funda- ple of legislative progress is passage by the US House mental forces. of Representatives of a bill to strengthen consumer pro- Although individual developing countries tection, extend regulation to some over the counter de- rivatives, and create a process for addressing troubled may be powerless to influence global develop- firms that may pose systemic risks. The bill may yet ments, much can done at the domestic level undergo significant modification before a compatible to mitigate these costs. The sensitivity of version is passed by the U.S. senate and it becomes law. developing-country GDP and growth pros- 5. See cross-country studies in Claessens, Demirgüç- pects to borrowing costs and the high cost Kunt, and Huizinga (1998); Barth, Caprio, and Levine 109 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 (2001); and Claessens and Laeven (2003). Bayraktar at low cost may also reduce incentives for monitoring and Wang (2004) find that domestic banks' efficiency managers, leading to weaker corporate governance and gains from foreign entry are greatest in countries that less productive investments (Levine and Zervos 1998). liberalized their stock markets before domestic finan- 13. Levine and Zervos (1998) find that stock market cial markets. liquidity (along with banking sector development) is pos- 6. The data on local claims on foreign-owned banks itively associated with growth, capital accumulation, and include only banks reporting to the BIS and thus ex- productivity improvements in a cross-country regression. clude claims by some foreign-owned banks (particu- Demirgüç-Kunt and Maksimovic (1996) find that active larly from many developing countries). Also, these data stock markets are associated with higher-than-predicted on the stock of claims are not adjusted for changes in rates of firm growth. Kassimkatis and Spyrous (2001) exchange rates. Thus the appreciation of the dollar dur- find that equity markets boost growth only in relatively ing the second half of 2008 contributed to reducing the liberalized economies. In addition, some authors have level of claims. It is unclear, however, whether the cur- found that the lifting of restrictions on stock market de- rency denomination of claims differed greatly between velopment is positively associated with growth, either foreign and locally owned banks. temporarily or over the long term (Fuchs-Schundeln and 7. Up to 20 percent of the $15.8 trillion of world Funke 2003). merchandise trade in 2008 involved secured documen- 14. Foreign investors may spur investment in in- tary transactions (such as letters of credit), and other frastructure services such as clearing and settlement forms of trade finance play an important role (either in systems, as countries compete for the limited pool financing inventories or accounts receivables) in help- of foreign investors willing to devote resources to ing bridge the gap between the time when costs are in- developing-country markets, and provide information curred in producing a product and the time when final on practices in more developed markets. In a cross- payment is made upon receipt. country regression, Chinn and Ito (2006) find that 8. World Bank calculations using data from openness to external financial flows contributes to the Dealogic. development of equity markets. 9. Mutual funds based in developed countries began 15. While the database does not distinguish whether investing in the 1980s in the form of closed-end funds the acquisition is cross-border or domestic, we assume (whose shares cannot be redeemed), to limit turnover that multinational firms borrow internationally mostly in the relatively illiquid markets in many developing for cross-border acquisitions. countries (Kaminsky, Lyons, and Schmuckler 2001). 16. Outward FDI from developing countries actu- As liquidity increased, open-end funds became more ally increased in 2008 to $164 billion (from $138 billion common. Other major foreign investors in emerging in 2007), in contrast to the 25 percent fall in outflows stock markets (many of whom invest through mutual from developed countries, and is expected to slow only funds) include pension funds, insurance companies, slightly in 2009. Chinese companies have spent more hedge funds, and individual investors. than $20 billion on oil assets overseas since December 10. This includes evidence that stock market liberal- 2008, including in Kazakhstan, Nigeria, and the Syrian izations reduce the cost of capital (Henry 2000, Bekaert Arab Republic. and Harvey 2000). Kose, Prasad, and Terrones (2008) 17. The superior technology (and marketing and find a positive relationship between portfolio equity lia- management practices) often used by foreign firms is bilities and total factor productivity growth in a sample transmitted to domestic firms either through observa- of industrial and developing countries. tion or because domestic firms hire workers trained by 11. For example, Gupta and Yuan (2005) find that foreign firms (Fosfuri, Motta, and Ronde 2001). Foreign following equity market liberalizations, industries that firm entry also can increase the intensity of competition depend on external finance grow faster than industries in an industry, potentially forcing domestic firms to im- dependent on finance internal to the firm. Similarly, prove their efficiency (Blomström, Kokko, and Zejan Vanassche (2004) finds that financial openness has a 2000; Javorcik, Keller, and Tybout 2006). The extent positive effect on growth of industrial sectors generally, of efficiency gains from FDI depends on the quality of and that this impact is greatest in industries that rely domestic policies and institutions, including policies to- more on external finance. Eichengreen, Gullapalli, and ward FDI (Beamish 1998); policies promoting the diffu- Panizza (2009) find that capital account openness has a sion of technology (Lall 2003); the level of human capital positive impact on the growth of financially dependent (Borensztein, DeGregorio, and Lee 1998); the level of industries only in high-income countries and that this technology in the recipient country and how close it is effect disappears during periods of crisis. to technology used by foreign firms (Saggi 2002); trade 12. There remains some theoretical argument over policy (Moran 2002); and financial development (Alfaro this effect, however, because the opportunity to exit and others 2002). 110 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H 18. See World Bank (2004) for a discussion of in- what might have been expected if high-income policy frastructure financing in developing countries, and rates increased by some 600 basis points. In fact during Dailami and Hauswald (2007) for an example of how the period, the effective Federal Reserve rate actually complex financing arrangements helped ensure the fell by close to 180 basis points, although corporate success and lower the costs of a major project finance bond defaults (another important fundamental) did rise transaction in Qatar. significantly. 19. For example, the share in dollar terms of 30. Prediction errors are largest for Brazil, developing-country institutions in project finance trans- Hungary, and to a lesser extent Bulgaria and Malaysia; actions increased from about 0.5 percent in 1997 to 9 for more see Kennedy and Palerm 2009. percent in 2008 (Project Finance International, http:// 31. The interest rate spread rose by more than 30 www.pfie.com). Moreover, participation by firms in percent more than expected by the model in Argentina high-income countries that recently were viewed as and Brazil and by 30 percent less than expected in the developing--for example the Republic of Korea; Taiwan, case of Poland. China; and Singapore--also increased markedly. 32. The basic OECD model has the form 20. The high levels of volatility in consumption SynRiskIndex 1.24 0.41*BaseRate Other in developing countries imply large welfare benefits variables. In Kennedy and Palerm (2009), develop- to consumption smoothing (Pallage and Robe 2003). ing-country spreads (EMBI spreads) are a function Consumption smoothing may also facilitate specializa- of the SynRiskIndex (and indirectly, the BaseRate). tion and hence development by reducing the impact on Two specifications were retained: one for coun- welfare of higher volatility that may be associated with tries with good credit histories, and one for those specialization. with less good histories: EMBI(poor risk) 140* 21. The correlation between ratios of short-term SynRiskIndex Other variables; and EMBI(better debt to reserves and the change in GDP growth is 0.2. risk) 27* SynRiskIndex Other variables. And using our forecasts of GDP growth in 2009, the Combining the two equations results in a reduced correlation between reserves-to-import ratios in 2007 form equation for EMBI spreads as a function of the and the fall in GDP growth in 2008­09 is zero. BaseRate: EMBI(poor risk) 140* 0.41*BaseRate 22. This last term reflects the difference between 57.4*BaseRate and EMBI(better risk) 27* the interest on domestic-currency securities issued in 0.41*BaseRate 11.1*BaseRate. Dailami, Masson, the course of past efforts to sterilize capital inflows and Padou (2008) modeled the relationship between and the domestic-currency return on the international fluctuations in high-income policy interest rates and reserves. developing-country risk premiums directly, in contrast 23. Hauner's country groups do not correspond to with the less direct approach here, which first estimates the World Bank's distinction between developing and the impact of policy rates on the global price of risk high-income countries. and then the impact of the price of risk on specific 24. This figure, which refers to reserves holdings developing-country risk premiums. where the currency composition is known, is taken from 33. Empirically, depreciation rates range from 3 the International Monetary Fund's COFER database. to 13 percent in the manufacturing sector (Nadiri and 25. World Bank calculations based on table 7A of Prucha 1993) for manufacturing in the United States the Locational Banking Statistics from the Bank for to as high as 18 to 36 percent in the high-tech research International Settlements. and development sector. Economy-wide depreciation 26. The Chinn-Ito index aggregates information rates represent a weighted average of very long depreci- on restrictions on financial transactions reported in ation rates on physical capital such as roads and build- the IMF's Annual Report on Exchange Arrangements ing and other much shorter ones on high technology. to produce a single indicator of financial openness for The aggregate depreciation rate varies from country to most of the world's countries. The data can be found at country with these values and the weight of different www. http://web.pdx.edu/~ito/. activities in overall output. While the 7 percent assump- 27. Information asymmetries may be smaller at re- tion used here is a rough approximation, it corresponds gional levels than at the global level (Ocampo 2006). to the assumption used in IMF (2005) in a slightly dif- 28. As reported in the People's Daily Online, ferent modeling exercise and is broadly consistent with May 4, 2008 (http://english.people.com.cn/90001/ implicit values used in the OECD interlink model. 90780/91421/6650574.html). ASEAN 3 refers to the 34. In the model of potential output described in ASEAN countries plus China, Japan, and the Republic chapter 2, a 100 basis point fall in the cost of capital for of Korea. a country with a cost of capital of 1,000 basis points 29. The observed increase in the synthetic price of should result in a 4 percent increase in its capital-output risk in the fourth quarter of 2008 was equivalent to ratio over the long term (assuming capital's share in 111 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 value added of 40 percent, a real interest rate of 3 per- Blomström, M., A. Kokko, and M. Zejan. 2000. For- cent, and an average depreciation rate of 7 percent). eign Direct Investment: Firm and Host Country 35. The simulations are run out to 2050 to allow Strategies. London: Macmillan. medium-term dynamics to resolve themselves. Borensztein, E., J. De Gregorio, and J. W. Lee. 1998. 36. The dynamic interaction between investment "How Does Foreign Direct Investment Affect rates and the desired capital stock actually causes the Economic Growth?" Journal of International ratio to overshoot its final level about by 3 percent of Economics 45: 115­35. the baseline capital output ratio. Brunnermeier, Markus K, Andrew Crockett, Charles 37. The IMF analysis is based on an ex post analy- A Goodhart, Avinash Persaud, and Hyun Song sis of the factors that explain past post-crisis declines Shin. 2009. "The Fundamental Principles of in potential output, among which is the same fall in Financial Regulation." Geneva Reports on the capital-output ratios described here. World Economy. Geneva: International Center 38. See World Bank (2007) for a more in-depth for Monetary and Banking Studies. discussion of likely long-term scenarios and their long- Calomiris, Charles W., and Joseph R. Mason. 2004. term implications for poverty. "Credit Card Securitization and Regulatory Ar- 39. Based on an assumed import elasticity of 4. bitrage." Journal of Financial Services Research 26 (1): 5­27. Calvo, Guillermo. 1996. "Varieties of Capital Market Crises." In The Debt Burden and its Conse- quences for Monetary Policy, ed. Calvo Guillermo References and Mervyn King. London: Macmillan in associa- Acharya, Viral V. 2009. "A Theory of Systemic Risk tion with the International Economic Association. and Design of Prudential Bank Regulation." Jour- Cetorelli, Nicola, and Linda Goldberg. 2009. "Global- nal of Financial Stability 5 (3): 224­55. ized Banks: Lending to Emerging Markets in the Alfaro, Laura, Areendam Chanda, Sebnem Kalemli- Crisis." Staff Report 377. Federal Reserve Bank Ozcan, and Selin Sayek. 2004. "FDI and Eco- of New York, New York. nomic Growth: The Role of Local Financial Chinn, Menzie D., and Hiro Ito. 2006. "What Matters Markets." Journal of International Economics 64 for Financial Development? Capital Controls, (October): 89­112. Institutions, and Interactions." Journal of Devel- Al-Yousif, Khalifa Yousif. 2002. "Financial Develop- opment Economics 81: 163­92. ment and Economic Growth: Another Look Claessens, Stijn, and Luc Laeven. 2003. "What at the Evidence from Developing Countries." Drives Bank Competition? Some International Review of Financial Economics 11 (2): 131­50. Evidence." Journal of Money, Credit and Bank- Arena, Marco, Carmen Reinhart, and Francisco ing 36 (3): 563­83. Vazquez. 2007. "The Lending Channel in Emerg- Corm, George. 2006. "The Arab Experience." In ing Economies: Are Foreign Banks Different?" Ocampo, Regional Financial Cooperation. IMF Staff Working Paper 07/48. International Cortavarria, Luis, Simon Gray, Barry Johnston, Laura Monetary Fund, Washington, DC. Kodres, Aditya Narain, Mahmood Pradhan, and Barth, James R., Gerard Caprio, and Ross Levine. Ian Tower. 2009. Lessons of the Financial Crisis 2001. "Bank Regulation and Supervision: What for Future Regulation of Financial Institutions Works Best?" Policy Research Working Paper and Markets and for Liquidity Management. 2775. World Bank, Washington, DC (August). Washington, DC: International Monetary Fund. Bayraktar, Nihal, and Yan Wang. 2004. "Foreign Crystal, Jennifer S., B. Gerard Dages, and Linda S. Bank Entry, Performance of Domestic Banks and Goldberg. 2002. "Has Foreign Bank Entry Led to the Sequence of Financial Liberalization." Policy Sounder Banks in Latin America?" Current Issues Research Working Paper 3416. World Bank, in Economics and Finance 8 (1):1­6. Washington, DC. Cull, Robert, and Maria Soledad Martinez Peria. 2007. Beamish, P. W. 1988. Multinational Joint Ventures in "Foreign Bank Participation and Crises in Devel- Developing Countries. London: Routledge. oping Countries." World Bank, Washington, DC. Bekaert, Geert, and Campbell R. Harvey. 2000. "For- Dailami, Mansoor, and Robert Hauswald. 2007. eign Speculators and Emerging Equity Markets." "Credit-spread Determinants and Interlocking Journal of Finance 55 (2): 565­613. Contracts: A Study of the Ras Gas Project." Jour- Blanchard, Olivier, Hamid Faruqee, and Vladimir nal of Financial Economics 86: 248­78. Klyuev. 2009. "Did Foreign Reserves Help Weather Dailami, Mansoor, Paul R. Masson, and Jean Jose the Crisis?" IMF Survey Magazine (October 8). Padou. 2008. "Global Monetary Conditions versus 112 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H Country-Specific Factors in the Determination of a Result of the Global Financial Crisis?" Research Emerging Market Debt Spreads." Journal of Inter- Paper 60. World Bank, Development Research national Money and Finance 27: 1325­36. Group, Washington, DC. De Haas, Ralph, and Iman van Lelyveld. 2006. "For- Fuchs-Schundeln, Nicola, and Norbert Funke. 2003. eign Banks and Credit Stability in Central and "Stock Market Liberalizations: Financial and Eastren Europe: A Panel Data Analysis." Journal Macroeconomic Implications." Review of World of Banking and Finance 30: 1927­52. Economics 139 (4): 730­61. Demirgüç-Kunt, Asli, and Vojislav Maksimovic. 1996. Fung, Laurence Kang-por, Chi-sang Tam, and Ip-wing "Financial Constraints, Uses of Funds and Firm Yu. 2008. "Assessing the Integration of Asia's Growth: An International Comparison." Policy Equity and Bond Markets." In Regional Financial Research Working Paper 1671. World Bank, Integration in Asia: Present and Future. BIS Pa- Washington, DC. pers 42. Bank for International Settlements, Basel. Demirgüç-Kunt, Asli, Harry Huizinga, and Stijn Furman, Jason, and Joseph Stiglitz. 1998. "Economic Claessens. 1998. "How Does Foreign Entry Crises: Evidence and Insights from East Asia." Affect the Domestic Banking Sector?" Policy In Brookings Papers on Economic Activity, ed. Research Working Paper 1918. World Bank, William C. Brainard and George Perry. Washing- Washington, DC. ton, DC: Brookings Institution. Detragiache, Enrica, and Poonam Gupta. 2004. "For- Glick, Reuven, and Michael Hutchison. 2001. "Bank- eign Banks in Emerging Market Crises: Evidence ing and Currency Crises: How Common Are from Malaysia." IMF Working Paper 04/129. Twins?" In Financial Crises in Emerging Markets, International Monetary Fund, Washington, DC. ed. Reuven Glick, Ramon Moreno, and Mark M. Di Giovanni, Julian. 2005. "What Drives Capital Spiegel. New York: Cambridge University Press. Flows? The Case Of Cross-Border M&A Activity Goldberg, Linda, B. Gerard Dages, and Daniel Kinney. and Financial Deepening." Journal of Interna- 2000. "Foreign and Domestic Bank Participation tional Economics 65 (1): 127­49. in Emerging Markets: Lessons from Mexico and Dorrucci, Ettore, Alexis Meyer-Cirkel, and Daniel Argentina." NBER Working Paper 7714. National Santabarbara. 2009. "Domestic Financial Devel- Bureau of Economic Research, Cambridge, MA. opment in Emerging Economies: Evidence and Gupta, Nandini, and Kathy Yuan. 2005. "On the Implications." Occasional Paper Series 102, Eu- Growth Effects of Liberalizations." Available at ropean Central Bank, Frankfurt (April). SSRN: http://ssrn.com/abstract=774605. Edwards, Sebastian. 2005. "Capital Controls, Sudden Hauner, David. 2005. "A Fiscal Price Tag for Inter- Stops, and Current Account Reversals." NBER national Reserves." IMF Working Paper 05/81. Working Paper 11170. National Bureau of Eco- International Monetary Fund, Washington, DC. nomic Research, Cambridge, MA. Hellman, T., K. Murdock, and J. Stiglitz. 2000. "Liber- Eichengreen, Barry, Rachita Gullapalli, and Ugo alization, Moral Hazard in Banking and Prudential Panizza. 2009. "Capital Account Liberalization, Regulation: Are Capital Requirements Enough?" Financial Development and Industry Growth: American Economic Review 90(1): 147­65. A Synthetic View." University of California at Henry, Peter B. 2000. "Stock Market Liberalization, Berkeley, Berkeley, CA. Economic Reform, and Emerging Market Equity Eichengreen, Barry, and David Leblang. 2002. "Capi- Prices." Journal of Finance 55 (2): 529­64. tal Account Liberalization and Growth: Was ------. 2007. "Capital Account Liberalization: The- Mr. Mahathir Right?" NBER Working Paper ory, Evidence and Speculation." Journal of Eco- 9427. National Bureau of Economic Research, nomic Literature 45 (4): 887­935. Cambridge, MA. IMF (International Monetary Fund). 2005. World Ellis, Nan S., Lisa M. Fairchild, and Harold D. Economic Outlook. Washington, DC. Flether. 2008. "The NYSE Response to Specialist ------. 2009. World Economic Outlook: Sustaining Misconduct: An Example of the Failure of Self- the Recovery. Washington, DC (October). Regulation." www.works.bepres.com\cgi. IMF/BAFT (International Monetary Fund­Bankers' Fimetrix. 2009. www.fimetrix.com. Association for Finance and Trade). 2009. "IMF- Fosfuri, Andrea, Massimo Motta, and Thomas Ronde. BAFT Trade Finance Survey: A Survey among 2001. "Foreign Direct Investment and Spillovers Banks Assessing the Current Trade Finance Envi- through Workers' Mobility." Journal of Interna- ronment" (http://baft.org/content_folders/Issues/ tional Economics 53 (1): 205­22. IMFBAFTSurveyResults20090331.ppt). Friedman, Jed, and Norbert Schady. 2009. "How Javorcik, Beata Smarzynska, Wolfgang Keller, Many More Infants Are Likely to Die in Africa as and James R. Tybout. 2006. "Openness and 113 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Industrial Responses in a Wal-Mart World: A Financial Integration." Working Paper 14916. Case Study of Mexican Soaps, Detergents and National Bureau of Economic Research, Cam- Surfactant Producers." NBER Working Paper bridge, MA. 12457. National Bureau of Economic Research, Kose, M. Ayhan, Eswar Prasad, and Marco E. Cambridge, MA. Terrones. 2005. "Growth Volatility in an Era Jeanne, Olivier. 2007. "International Reserves in of Globalization." IMF Staff Papers 52. In- Emerging Market Economies: Too Much of a ternational Monetary Fund, Washington, DC Good Thing?" Brookings Papers on Economic: (September). 1, ed. William C. Brainard and George Perry. ------. 2008. "Does Openness to International Finan- Washington, DC: Brookings Institution. cial Flows Raise Productivity Growth?" NBER Jeon, Bang Nam. 2002. "Progress and Prospects of Re- Working Paper 14558. National Bureau of Eco- gional Financial Arrangements and Cooperation: nomic Research, Cambridge, MA. A Critical Suvey." Paper prepared for the 2002 Lall, Sanjaya. 2003. "The Technological Structure and International Finance Seminar ADBI/ADB in Pen- Performance of Developing Country Manufac- ang, Malaysia. http://ssrn.com/abstract=1305688. turing Exports, 1985­98." QEH Working Paper Jickling, Mark. 2009. "Causes of the Financial Crisis." 44. Queen Elizabeth House, Oxford University, Congressional Research Service, Washington, Oxford, U.K. DC. http://digitalcommons.ilr.cornell.edu/key Lau, Lawrence J. 2000. "Research on the Cost of workplace/600. Capital: Past, Present and Future." Chapter in Jorgenson, Dale W. 1963. "Capital Theory and Invest- Econometrics and the Cost of Capital: Essays in ment Behavior." American Economic Review 53: Honor of Dale Jorgenson. Cambridge, MA: Mas- 247­57. sachusetts Institute of Technology Press. Kaminsky, Graciela A., Richard K. Lyons, and Sergio Levine, Ross. 1996. "Foreign Banks, Financial Devel- L. Schmukler. 2001. "Mutual Fund Investment in opment and Economic Growth." In International Emerging Markets: An Overview." World Bank Financial Markets, ed. Claude E. Barfied. Wash- Economic Review 15 (2): 315­40. ington, DC.: AEI Press. Kaminsky, Graciela L. and Sergio L. Schmukler. 2002. Levine, Ross, and Sara Zervos. 1998. "Stock Markets, "Short-Run Pain, Long-Run Gain: The Effects of Banks, and Economic Growth." American Eco- Financial Liberalization." NBER Working Paper nomic Review 88 (3): 537­58. 9787. National Bureau of Economic Research, Machinea, Jose Luis, and Guillermo Rozenwurcel. Cambridge, MA. 2006. "Macroeconomic Coordination in Latin Kassimatis, Konstantinos, and Spyros I. Spyrou. 2001. America: Does It Have a Future?" In Ocampo, "Stock and Credit Market Expansion and Eco- Regional Financial Cooperation. nomic Development in Emerging Markets: Fur- Magud, Nicolas, and Carmen M. Reinhart. 2006. ther Evidence Utilizing Cointegration Analysis." "Capital Controls: An Evaluation." NBER Work- Applied Economics 33: 1057­64. ing Paper 11973. National Bureau of Economic Kennedy, Michael, and Angel Palerm. Forthcoming. Research, Cambridge, MA. "Emerging Market Bond Spreads: The Role of Malouche, Mariem. 2009. Trade and Trade Finance World Financial Market Conditions and Country- Developments in 14 Developing Countries post Specific Factors." World Bank Prospects Working September 2008: A World Bank Survey." Policy Paper, Washington, DC. Research Working Paper 5138. World Bank, Klein, Michael W., and Giovanni P. Olivei. 2008. Washington, DC. "Capital Account Liberalization, Financial Depth Martinez Peria, Maria Soledad, Andrew Powell, and and Economic Growth." Journal of International Ivanna Vladkova-Hollar. 2002. "Banking on Money and Finance 27: 861­75. Foreigners: The Behavior of International Bank Klein, M. W., J. Peek, and E. S. Rosengren. 2002. Lending to Latin America, 1985­2000." IMF "Troubled Banks, Impaired Foreign Direct In- Staff Papers 52. International Monetary Fund, vestment: The Role of Relative Access to Credit." Washington, DC. American Economic Review 92: 8. McKibbin, Warwick J., and Andrew Stoeckel. 2009. Kose, M. Ayhan, Eswar Prasad, Kenneth Rogoff, and "The Potential Impact of the Global Financial Shang-Jin Wei. 2006. "Financial Globalization: Crisis on World Trade." Mimeo. A Reappraisal." IMF Working Paper 06/189. Miyachi, Masato. 2007. "Regional Cooperation in De- International Monetary Fund, Washington, DC. veloping Asian Bond Market." Presentation for Kose, M. Ayhan, Eswar Prasad, and Ashley D. Taylor. the Asian Development Bank, given in Shanghai 2009. "Thresholds in the Process of International September 25. 114 M E D I U M - T E R M I M P A C T S O F T H E C R I S I S O N F I N A N C E A N D G R O W T H Moran Theodore. 2002. How to Investigate the Impact Survey." World Bank Research Observer 17 (2): of Foreign Direct Investment on Development 191­235. and Use the Results to Guide Policy. Washington, SEC (Security and Exchange Commission). 2008. D.C.: Brookings Trade Forum. "Summary Report of Issues Identified in the Mukerji, Purba. 2008. "Ready for Capital Account Commission's Staff Examinations of Select Credit Convertibility?" Journal of International Money Rating Agencies." Washington, DC (July). and Finance. Shakya, Malika. 2009. "Integration of Developing Nadiri, Ishaq, and Ingmar Prucha. 2007. "Sources of Countries into Global Supply Chains: A Global Growth of Output and Convergence of Produc- Buyers' and Producers' Perspective." Interna- tivity in Major OECD Countries." International tional Trade Department, World Bank, unpub- Journal of Production Economics 52 (1997): lished working paper. 133­46. Sløk, Torsten, and Michael Kennedy. 2005. "Factors Ocampo, Jose Antonio. 2006. "Regional Financial Driving Risk Premia." OECD Economics Depart- Cooperation: Experiences and Challenges." In ment Working Papers, No. 385. Paris, OECD. Regional Financial Cooperation, ed. Jose Anto- UNCTAD (United Nations Conference on Trade and nio Ocampo. Washington, DC: The Brookings Development). 2009. World Investment Report. Institution. New York: Unied Nations. OECD. 2004. OECD Economic Outlook No. 75. Paris. Vanassche, Ellen. 2004. "The Impact of International ------. 2009. OECD Economic Outlook No. 85. Financial Integration on Industry Growth." Cen- Paris. ter for Economic Studies Discussion Paper 04.12. Pallage, Stephane, and Michael A. Robe. 2003. "On Catholic University of Leuven, Leuven, Belgium. the Welfare Cost of Economic Fluctuations in World Bank. 2004. Global Development Finance: Developing Countries." International Economic Harnessing Cyclical Gains for Development. Review 44 (2): 677­98. Washington, DC: World Bank. Peek, Joe, and Eric S. Rosengreen. 2000. "Implications ------. 2006. Global Development Finance: The of the Globalization of the Banking Sector: The Development Potential of Surging Capital Flows. Latin American Experience." Proceedings. Fed- Washington, DC: World Bank. eral Reserve Bank of Boston Conference Series ------. 2007. Global Development Finance: The Glo- (June): 145­70. balization of Corporate Development Finance in Rodrik, Dani. 2006. "The Social Cost of Foreign Ex- Developing Countries. Washington, DC: World change Reserves." CEPR Discussion Paper 5483. Bank. Centre for Economic Policy Research, London. ------. 2008. Global Economic Prospects: Techology Saggi, Kamal. 2002. "Trade, Foreign Direct Invest- Diffusion in the Developing World. Washington, ment and International Technology Transfer: A DC: World Bank. 115 1 Appendix Regional Economic Prospects East Asia and the Pacific East Asia's rebound from the global down- turn over the course of 2009 was quicker and Recent developments more robust than in other parts of the world. A s chapter 1 outlined, economies in the East Asia and Pacific region were particu- larly hard hit by the collapse of global business China led the global recovery in industrial pro- duction, with contributions to growth from the high-income OECD countries emerging only investment in the fall of 2008. The crisis cur- later in the year. The recovery in East Asia was tailed financing flows to private firms world- underpinned by domestic stimulus programs wide, and together with depressed growth put in place by a number of economies, most expectations, investment plans were marked notably by China; a shift from large inventory down sharply. Household wealth, incomes, reduction to restocking by firms; and a return and demand for consumer durables were af- to positive growth in exports and production fected just as adversely. Outside of China, by the second quarter of 2009.3 Against this investment in the East Asia region was hit background, GDP losses for East Asia were exceptionally hard. Local equity markets fell limited in 2009, with current growth estimates by almost 60 percent from January to Octo- placed at 6.8 percent, down from 8 percent ber 2008; currencies tumbled between 5 and in 2008. China grew by an estimated 8.4 per- 25 percent against the dollar through the first cent during the year, while performance in quarter of 2009. And bond spreads increased Indonesia (4.5 percent) and Vietnam (5.5 per- by 515 basis points from January 2007 to cent) was strong. Output contractions were reach 645 points by November 2008.1 limited to Cambodia ( 2.2 percent), Malaysia As one of the key producing regions for ( 2.3 percent), Thailand ( 2.7 percent), and durable and capital goods--highly integrated several Pacific islands. However, when China into global production networks--East Asian is excluded from the 2009 growth estimates, economies experienced dramatic declines in GDP numbers for the remainder of the region trade and production between September offer a better reflection of the crisis, with an 2008 and March 2009. Dollar-based exports advance of just 1.3 percent following 4.8 per- dropped 25 percent, while production (ex- cent growth in 2008 (table A1). cluding China) plummeted 15­30 percent over As the global downturn took hold across the period. High-income economies within the East Asia in late 2008, many developing coun- region--including Japan; Republic of Korea; tries, together with major developed econo- Singapore; Taiwan, China; and Hong Kong, mies, began to implement large-scale fiscal and China--were equally or more severely hit by monetary stimulus measures to support do- these developments (figure A1).2 mestic demand and to counter the drag from the 117 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Figure A1 East Asian exports and production hard hit by downturn in capital goods demand Exports in U.S. dollars Industrial production percent change (saar) 60 60 percent change (saar) NIEs 40 40 China Indonesia 20 20 0 0 ­20 ­20 Malaysia Thailand ­40 ­40 Korea, Rep. of Other East Asian ­60 and Pacific countries ­60 ­80 ­80 Jan. May Sep. Jan. May Sep. Jan. May Sep. Jan. May Sep. 2008 2008 2008 2009 2009 2009 2008 2008 2008 2009 2009 2009 Sources: Haver Analytics and World Bank. Table A1 East Asia and Pacific forecast summary (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009e 2010f 2011f GDP at market prices (2005 US$)b 7.4 10.1 11.4 8.0 6.8 8.1 8.2 GDP per capita (units in US$) 6.3 9.2 10.5 7.2 6.0 7.2 7.3 PPP GDPc 7.3 10.1 11.3 8.0 6.8 8.0 8.2 Private consumption 5.7 8.1 8.7 6.7 5.9 7.3 7.5 Public consumption 8.1 8.2 9.8 7.8 11.1 8.4 7.3 Fixed investment 8.1 12.4 8.7 5.3 14.3 9.3 9.3 Exports, GNFSd 12.5 18.8 15.4 7.4 13.5 6.6 8.8 Imports, GNFSd 9.7 12.7 11.0 4.9 12.1 6.2 8.5 Net exports, contribution to growth 0.7 3.4 3.0 1.7 2.0 0.7 0.8 Current account balance/GDP (%) 2.2 8.4 9.9 8.8 7.1 6.5 6.4 GDP deflator (median, LCU) 5.9 4.4 3.5 4.3 3.2 3.3 3.4 Fiscal balance/GDP (%) 1.8 0.7 0.0 0.6 3.3 3.7 3.1 Memo items: GDP East Asia excluding China 3.5 5.7 6.2 4.8 1.3 4.7 5.1 China 9.1 11.6 13.0 9.0 8.4 9.0 9.0 Indonesia 2.7 5.5 6.3 6.1 4.5 5.6 5.8 Thailand 2.7 5.3 4.9 2.6 2.7 3.5 4.0 Source: World Bank. a. Growth rates over intervals are compound average; growth contributions, ratios, and the GDP deflator are averages. b. GDP measured in constant 2005 U.S. dollars. c. GDP measured at PPP exchange rates. d. Exports and imports of goods and non-factor services (GNFS). e. Estimate. f. Forecast. collapse of export markets. The Chinese stim- in part by a surge in credit expansion, with ulus package is of special note; it entails some total new lending equivalent to 30 percent of $575 billion to be spread proportionately over GDP in 2009. Elsewhere in the region, gov- time from late 2008 through 2010, financed ernment deficits (as a share of GDP) increased 118 R E G I O N A L E C O N O M I C P R O S P E C T S Figure A2 Malaysia: A profile of recovery Figure A3 China's stimulus program supports exports from regional partners Percentage change (saar) in real GDP and components 40 Export values, US$, seasonally adjusted annual percentage change 30 80 20 10 60 Developing East Asian exports 0 40 Chinese imports 10 20 20 0 30 40 20 50 40 East Asian 60 NIE exports GDP Government Consumer Investment Exports 60 spending spending 80 2008Q4 2009Q1 2009Q2 2009Q3 08 08 08 09 09 09 20 20 20 20 20 20 n. ay p. n. ay p. Ja Se Ja Se M M Sources: Haver Analytics and World Bank. Sources: Haver Analytics and World Bank. significantly, reflecting both automatic stabi- lizers and countercyclical measures. importantly) with existing excess capacity, and Over the course of 2009, East Asia's stimu- expectations among Malaysian business that lus measures began to bear fruit, supporting conditions in the main developed markets for incomes and helping to boost household spend- electronics and other equipment would be in ing, underpinning infrastructure development decline for an extended period (see figure A2). though public investment outlays, and provid- Exports fell abruptly, by 45 percent (saar), as ing support for the financial sector (figure A2 the synchronous global shutdown of demand shows the recovery for Malaysia). China's for capital and related goods took hold. GDP stimulus had regionwide impacts, by boosting declined by a sharp 9.1 percent during the demand for East Asian exports. China's infra- quarter. However, fiscal stimulus measures structure outlays also underpinned demand were implemented that helped to shore up for regional commodities and raw materials confidence and provide direct support to the used in construction, from countries such as construction sector. The overall impact on the Indonesia, Papua New Guinea, and Lao Peo- economy during the first two quarters of 2009 ple's Democratic Republic (as well as Austra- was limited, as decision and implementation lia). As replenishment of inventories got under lags affected the speed and rate of disburse- way, firms in China began to restock parts and ment. While the steep decline in exports deeply components from regional suppliers, notably affected the manufacturing sector, the impact electronic goods from countries such as Ma- on the economy as a whole was mitigated by laysia and Thailand (as well as NIEs). Also, the compression in processing imports. rising Chinese consumer expenditure stimu- Still, GDP declined by 18.3 percent (saar) lated demand for a variety of consumer du- in the first quarter of 2009, as exports con- rable goods across the region (figure A3). tinued in sharp decline, even though the In Malaysia, fixed investment declined by falloff in investment was mitigated by a 35 percent during the fourth quarter of 2008 second fiscal measure (RM15 billion, or (saar) as an indirect result of the surge in 2.3 percent of GDP). By the second and third international capital costs, combined (more quarters, Malaysia had emerged to recovery, 119 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 after additional government spending and effects of earlier stimulus helped to turn Figure A4 Equity markets have nearly recovered from earlier declines around investment and consumer spend- Local currency­based equity prices ing, while exports rebounded sharply (in part because of Chinese demand). The Index, Jan. 1, 2008 = 100 annualized pace of GDP growth re- 110 bounded strongly to 14.0 and 12.2 percent, 100 Thailand respectively, in the second and third quarters. 90 Capital flows are returning to East Asia, 80 Indonesia with a notable pickup during the fall of 70 2009. East Asian bond and initial public offerings (IPO) increased as conditions in 60 China international markets became more hospi- 50 table, with spreads much reduced from cri- 40 Singapore sis levels and underlying demand conditions 30 firming. Some $12.8 billion in bonds were Jan. 1, Apr. 1, Jul. 1, Oct. 1, Jan. 1, Apr. 1, Jul. 1, Jul. 1, Jan. 1, 2008 2008 2008 2008 2009 2009 2009 2009 2010 issued in the year to October, representing a doubling of issuance from the like period of Sources: Morgan-Stanley through Thomson/Datastream. 2008. IPO issuance increased to $38 billion, largely from China, but also from Malaysia and Indonesia, up 65 percent from the same capital for firms, restore a significant por- period in 2008. But cross-border bank lend- tion of earlier wealth losses, and lift overall ing disappointed, largely reflecting risk aver- confidence. The return of foreign capital also sion on the part of international commercial helped to reverse some of the earlier sharp de- banks, where deleveraging continues to be clines in local currencies (figure A5). Under the order of the day. Syndicated loans to East Asian entities amounted to $12.8 billion Figure A5 East Asian currencies continue during the year through October, down sub- to recoup against U.S. dollar stantially from the $37 billion raised in the Local currency unit, percentage change same period of 2008. Total gross flows to 45 developing East Asia amounted to $63 bil- 40 lion over the year through October--a 35 Stronger local currency 4.5 percent decline from the $66 billion ac- 30 crued in the same period of 2008. 25 Local financial market developments have 20 provided further impetus to the recovery. A return of capital from the United States, where 15 funds had earlier fled to safe-haven securities 10 has underpinned a rapid rebound in regional 5 equity markets following steep declines in 0 2008 and early 2009 (figure A4). Bourses in In wo an C pia ia r gg n Ph olla re pe nes na K t ol g, in e C on bah lla rin sia nm es ru es , d on d o re do n hi h ng it ilip r Th so C ar bi do ap pi re hin n ay Ko l H i a Indonesia and Thailand are close to regaining na al hi g M Si levels that prevailed in January 2008, while Percentage change (US$ per local currency unit) Indonesia's market has more than doubled from October 2008 troughs, as have equity Cumulative depreciation, Sep. 2008­March 2009 markets in Thailand and Singapore. The re- Cumulative appreciation, March 2009­Dec. 2009 bound in equity markets and falling interest Source: Thomson/Datastream. rate spreads have helped reduce the cost of 120 R E G I O N A L E C O N O M I C P R O S P E C T S these circumstances the potential formation of include Vietnam at 9.4 percent of GDP, a "new" financial market bubble in the region Malaysia at 7.8 percent, Thailand (4.2 per- is an increasing cause for concern. cent) and the Philippines (3.8 percent). The Inflation has eased broadly in East Asia, unwinding of these fiscal support measures given the slowing in activity and lower food will play an important role in shaping the and fuel prices, although conditions vary economic recovery over the forecast period. widely across countries. Increases in the con- Although trade conditions have improved sumer price index for 2009 range between over the course of 2009 as Chinese imports highs of 20 and 12 percent in Cambodia and recovered, regional export volumes (goods Vietnam, to 3 and zero percent (or slightly and services) dropped 13.5 percent during negative) in China, the Philippines, and the year, while imports fell 12.1 percent, Thailand. leading to a narrowing of the aggregate In step with the cyclical downturn (a sharp current account position from a surplus of drop in government revenues) and with large 8.8 percent of GDP in 2008 to 7.1 percent discretionary stimulus packages, fiscal defi- for 2009. This was aided in particular by a cits have widened across both middle- and sharp decline in China's current account sur- low-income countries in the region--this plus, which fell from 9.8 percent of GDP in even as fiscal space for the latter countries 2008 to 6.4 percent of GDP during the first such as Cambodia and Lao PDR appears lim- six months of 2009. ited. The World Bank's East Asia and Pacific Department in a recent "East Asia Update" (November 2009) estimates that fiscal stimu- Medium-term outlook lus in the regions' middle-income countries Momentum underlying economic activity in amounted to 2.1 percent of GDP in 2009, the region should be sustained, as a gradual up from an earlier estimate of 1.7 percent. decline in the effects of domestic stimulus China's fiscal shortfall is projected to have measures is countered over the course of reached a record 3.3 percent of GDP during 2010 by the return to growth (albeit moder- 2009, but a number of countries exceeded ate) in East Asia's main OECD export mar- this deficit when the deficit is expressed as kets. But contrasted with earlier episodes of a proportion of GDP (figure A6). Examples global downturns (for example the 2001­03 "dot-com" bust), the rebound and recov- ery path of GDP in East Asia is expected to be more muted, reflecting weaker global Figure A6 Stimulus measures yield widening demand and less buoyant financial condi- fiscal shortfalls across the region tions. Continued strong advances in China's Overall government fiscal balance, percent of GDP domestic demand, and associated imports, 2 should play an important role in underpin- 0 ning a second export-led revival phase for 2 the remainder of the region. At the same time, world trade growth is anticipated to 4 revive from the estimated 14.4 percent de- 6 cline in 2009 to a gain of 6.2 percent by 2011. 8 Against this background, East Asian export 10 volumes are forecast to advance by 6.6 per- Vietnam Malaysia Thailand China Indonesia cent in 2010 and 8.8 percent in 2011, picking 2007 2008 2009 up additional market share. The regional cur- Sources: National authorities and Bank staff projections. rent account surplus position is anticipated to moderate from 7.1 percent of GDP in 2009 to 121 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 6.4 percent by 2011 (see table A1), reflecting an increased contribution to overall growth Figure A7 East Asia will enjoy a moderate rebound in 2010­11 from domestic demand. The recovery in business investment is ex- Growth of real GDP, percent 10 pected to be gradual (by historic standards), as excess capacity will first have to be 8 worked down. Growth of public sector out- 6 lays should ease from 2009 peaks of 11.1 per- 4 cent to 7.3 percent by 2011. Recognizing 2 that prospects for an export-led recovery 0 are less favorable than in the past, policy is likely to shift further toward fostering 2 growth in household demand, helping, in 4 Developing China East Asia NIEs turn, to offset the profile of weaker govern- East Asia excluding ment spending. China On balance, regional GDP growth is ex- 2008 2009 2010 2011 pected to increase to 8.2 percent by 2011 from the 6.8 percent registered in 2009. Source: World Bank. This is a modest recovery by historic stan- dards, but, at the end of the first year of financial crisis, the regional downturn has been equally moderate, compared, for ex- ample, with the East Asian crisis of the late and in particular the surge in liquidity over 1990s. Domestic demand will be the key the course of 2009, raises uncertainties growth driver, with more modest net trade regarding the future growth path. Prospects contributions. China will lead the regional for low-income countries (Cambodia, Lao advance with GDP growth of 9 percent by PDR, and Vietnam) will depend heavily on 2010 (figure A7; table A2). Growth ex- improvements in the international environ- cluding China is anticipated to pick up to ment for bank lending. Recent developments 5.1 percent by 2011, from an estimated in Dubai and credit rating downgrades for 1.3 percent in 2009. In particular, shifts Greece are indicative of continued uncertain- from negative to positive growth in Malay- ties, which, should they become widespread, sia and Thailand and a solid acceleration may have serious implications for bank lend- in activity in Indonesia and Vietnam should ing and growth around the globe, particu- underpin the turnaround. larly in developing countries. Consequently, banking flows may remain sluggish for an extended period of time as commercial banks Risks remain cautious and rebuild balance sheets. Downside risks facing the region have di- Furthermore, for middle-income countries minished owing to improvements in the currently benefiting from the return of large- global financial environment and positive scale inflows, driven by international inves- growth developments within East Asia and tors' search for yields above those available the Pacific. The possibility of a "double- in mature markets, there is a risk of yet dip" global recession remains, particularly another round of "asset bubbles," this time as mature economies will be unwinding both in emerging markets, the bursting of which monetary and fiscal stimulus. Also within could carry adverse effects over the short to the region, the Chinese stimulus program, medium term. 122 R E G I O N A L E C O N O M I C P R O S P E C T S Table A2 East Asia and Pacific country forecasts (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009c 2010d 2011d Cambodia GDP at market prices (2005 US$)b 8.3 10.8 10.2 6.7 2.2 4.2 6.0 Current account balance/GDP (%) 4.4 3.6 6.3 10.2 3.5 4.0 4.0 China GDP at market prices (2005 US$)b 9.1 11.6 13.0 9.0 8.4 9.0 9.0 Current account balance/GDP (%) 2.6 9.7 11.0 9.8 5.6 4.1 4.0 Fiji GDP at market prices (2005 US$)b 2.3 3.6 6.6 0.2 0.3 1.8 2.2 Current account balance/GDP (%) 4.8 22.5 14.4 19.6 25.4 24.8 27.7 Indonesia GDP at market prices (2005 US$)b 2.7 5.5 6.3 6.1 4.5 5.6 5.8 Current account balance/GDP (%) 1.5 3.0 2.4 0.1 1.4 0.5 0.3 Lao PDR GDP at market prices (2005 US$)b 6.2 8.4 7.6 7.3 6.4 7.5 7.3 Current account balance/GDP (%) 9.8 1.4 2.5 12.5 8.1 6.0 7.0 Malaysia GDP at market prices (2005 US$)b 4.8 5.8 6.2 4.6 2.3 4.1 4.8 Current account balance/GDP (%) 6.5 16.3 15.5 17.5 15.3 15.5 15.0 Papua New Guinea GDP at market prices (2005 US$)b 0.7 2.6 6.5 6.6 3.9 3.7 3.3 Current account balance/GDP (%) 3.0 2.2 1.8 2.8 6.7 4.7 4.3 Philippines GDP at market prices (2005 US$)b 4.2 5.3 7.1 3.8 1.0 3.5 3.8 Current account balance/GDP (%) 1.4 4.5 4.9 2.5 3.4 2.8 2.3 Thailand GDP at market prices (2005 US$)b 2.7 5.3 4.9 2.6 2.7 3.5 4.0 Current account balance/GDP (%) 1.9 1.1 5.7 0.1 5.5 3.5 3.0 Vanuatu GDP at market prices (2005 US$)b 1.5 7.4 6.8 6.6 4.2 4.5 5.5 Current account balance/GDP (%) 9.8 4.1 5.3 5.9 4.7 4.4 3.4 Vietnam GDP at market prices (2005 US$)b 7.2 8.2 8.5 6.2 5.5 6.5 7.0 Current account balance/GDP (%) 2.5 0.3 9.8 11.9 5.1 4.5 4.4 Source: World Bank. Note: World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries' pros- pects do not significantly differ at any given moment in time. American Samoa; Micronesia, Fed. Sts.; Kiribati; Marshall Islands; Myanmar; Mongolia; N. Mariana Islands; Palau; Korea, Dem. Rep.; Solomon Islands; Timor-Leste; and Tonga are not forecast owing to data limitations. a. Growth rates over intervals are compound average; growth contributions, ratios, and the GDP deflator are averages. b. GDP measured in constant 2005 U.S. dollars. c. Estimate. d. Forecast. Europe and Central Asia region in the degree of impact. Aggregate GDP is estimated to have contracted 6.2 percent in Recent developments 2009, nearly twice as much as the 3.3 percent A mong developing regions, the Europe and Central Asia region4 has been the most negatively affected by the global financial estimated decline in high-income countries, and sharply more negative than the (2.2 percent) contraction for the remaining developing coun- crisis, albeit with large variations across the tries excluding China and India (table A3). 123 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Table A3 Europe and Central Asia forecast summary (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009e 2010f 2011f GDP at market prices (2005 US$)b 4.1 7.6 7.1 4.2 6.2 2.7 3.6 GDP per capita (units in US$) 4.0 7.5 7.1 4.2 6.2 2.6 3.5 PPP GDPc 4.0 7.7 7.4 4.5 6.5 2.7 3.6 Private consumption 4.8 7.5 9.2 6.4 4.6 2.2 3.3 Public consumption 2.0 6.0 5.2 4.1 2.3 2.1 2.6 Fixed investment 4.7 16.5 14.2 8.7 16.5 4.1 4.7 Exports, GNFSd 7.9 8.1 7.1 3.9 13.2 4.3 6.6 Imports, GNFSd 8.7 13.9 17.9 9.0 12.9 3.7 6.0 Net exports, contribution to growth 0.1 1.5 3.4 1.9 0.3 0.1 0.0 Current account balance/GDP (%) 0.9 1.1 0.6 0.3 0.5 0.4 0.2 GDP deflator (median, LCU) 18.8 9.3 7.7 9.5 3.5 6.7 4.0 Fiscal balance/GDP (%) 5.5 3.0 2.4 0.7 6.2 4.5 3.4 Memo items: GDP Transition countries 4.0 6.9 5.7 3.0 4.1 2.2 3.8 Central and Eastern Europe 3.8 6.8 6.8 5.0 2.5 1.3 3.5 Commonwealth of Independent States 4.1 8.3 8.4 5.4 8.1 3.1 3.3 Russian Federation 3.9 7.7 8.1 5.6 8.7 3.2 3.0 Turkey 4.3 6.9 4.7 0.9 5.8 3.3 4.2 Poland 4.3 6.2 6.7 4.9 1.6 2.2 3.4 Source: World Bank. a. Growth rates over intervals are compound average; growth contributions, ratios, and the GDP deflator are averages. b. GDP measured in constant 2005 U.S. dollars. c. GDP measured at PPP exchange rates. d. Exports and imports of goods and nonfactor services. e. Estimate. f. Forecast. The severity of the impact of the crisis and corporations held large foreign cur- in the region reflects significant preexisting rency obligations (Armenia, Bulgaria, Croa- vulnerabilities in many countries. Many econ- tia, Latvia, Lithuania, Romania, Turkey, omies were heavily reliant on foreign finance and Ukraine), and where pre-crisis growth (a result of excessive credit expansion that had relied heavily on foreign capital inflows been enabled by foreign banks, large current (Bulgaria, Georgia, Latvia, Lithuania, the account deficits, elevated external debt levels, former Yugoslav Republic of Macedonia, and considerable currency mismatches in both Moldova, Montenegro, and Romania are corporate and household debt). As a result, among the largest, with current account defi- this region was particularly vulnerable to the cits equivalent to 10 percent or more of GDP reversal in capital flows that accompanied the in 2008). At the same time, petroleum export- initial phases of the financial crisis. ers (Kazakhstan, the Russian Federation) were Sharply reduced external demand for ex- also hit hard by the plunge in international ports, a halving of foreign direct investment commodity prices. inflows, and falling remittances exacerbated Sharp declines in international financ- the collapse in investor confidence and credit ing have forced large adjustments in domestic tightening, forcing a sharp contraction of demand. Gross capital inflows to the region fell 4.6 percent in regional private consump- 54 percent during 2009, versus the 19 percent tion, and a decline in gross fixed investment increase posted by other developing countries of 16.5 percent in 2009--down from expan- (figure A8). This decline in inflows primarily sions of 6.4 percent and 8.7 percent, respec- reflects the drying up of syndicated bank lend- tively, in 2008. The impact of the crisis was ing, which represented 60 percent of total flows most negative in countries where households to the region in 2007, before the crisis. Partly 124 R E G I O N A L E C O N O M I C P R O S P E C T S the region's ex post financing needs declined, Figure A8 Gross capital flows to Europe and while at the same time external assistance Central Asia picked up in mid-2009 and moral suasion helped prevent access to US$ (billions) external finance from declining as sharply as 70 60 had been initially expected. 50 Reflecting these developments, financing 40 conditions have improved. Spreads on sov- 30 ereign debt, which rose sharply in the third 20 quarter of 2008 and into the first quarter 10 0 of 2009, have since narrowed. In the case of Ukraine, for example, spreads over U.S. 4 1 2 3 4 1 2 3 4 Q Q Q Q Q Q Q Q Q Treasuries jumped by as much as 3,100­3,660 07 08 08 08 08 09 09 09 09 20 20 20 20 20 20 20 20 20 basis points in March 2009 but have since re- Bank Bond Equity versed to a spread of 768 points, as of early Sources: Dealogic: Bondware, Loanware and World Bank. January. These improved market conditions have also been supported by an easing of infla- tionary pressures, which has enabled monetary policies to focus on cushioning the downturn. Figure A9 Many European and Central Asian Many governments also implemented coun- economies post sharp current account tercyclical fiscal policies to support domestic adjustments demand. Reflecting these measures, as well as Change in current account balances (2008 to 2009) the depth of the recession and much weaker Percent of GDP commodity prices, government deficits have 18 increased by about 7 percent of GDP, mov- 16 14 ing from a surplus of 0.7 percent in 2008 to a 12 10 6.2 percent of GDP deficit in 2009. 8 Although economic activity in Europe and 6 4 Central Asia remains depressed, the pace of 2 0 contraction is moderating. Thus, although industrial production in the region began a ia ia va ia ne ia nd ia ey ni tv ar an rg at do rk ai la ua La ro lg eo om kr Po Tu ol expanding at a 4.8 percent annualized pace in th Bu C U G M Li R Source: World Bank. the second quarter of 2009, output in October 2009 remained 6.0 percent below its pre-crisis level in October 2008 (figure A10). In Russia, the 2009 recession is estimated to have been much sharper than was the one reflecting substantial support from international following the 1998 crisis. During the 2009 financial institutions, bond and equity flows to recession, GDP is estimated to have fallen countries in the region began recovering in the 8.7 percent, compared with 5.3 percent in third quarter of 2009, although bank lending the 1998 crisis, and represents the largest de- remains very weak. cline in growth since the breakup of the Soviet Reflecting the cut in capital inflows (and Union.5 The contraction reflects both external the associated cuts in domestic demand), re- factors (import demand among Russia's main gional current account deficits have narrowed, trading partners decreased by an estimated with Bulgaria, Latvia, and Lithuania post- 15 percent in 2009) and domestic factors (an ing double-digit improvements in current ac- 18 percent decline in investment and a 4.7 per- count positions measured as a share of GDP cent contraction in private consumption). Re- (figure A9). As a result of the cuts in spending, flecting widespread economic slack, inflation 125 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 1.9 percent of GDP in 2009 from an estimated Figure A10 Pace of contraction in industrial 5.8 percent of GDP in 2008. To support do- production moderates in Europe and Central Asia mestic demand, the Central Bank of Turkey Annual percentage change has more than halved its key policy interest 20 rate, cutting it by a cumulative 1,025 basis 15 points from 16.75 percent in October 2008 to 10 6.5 percent as of January 2010. 5 0 Economic conditions remained difficult over 5 the first three quarters of 2009 for the five new 10 members of the European Union that are devel- 15 oping countries, although the pace of contrac- 20 tion in real output moderated from 12.6 percent 08 08 08 8 09 09 09 9 00 00 20 20 20 20 20 20 year-over-year in the first quarter to 4.4 in the .2 .2 n. r. l. n. r. l. ct ct Ju Ju Ap Ap Ja Ja O O third quarter. Partly in response to strengthening Latin America and Caribbean South Asia demand in high-income Europe, industrial pro- Middle East and North Africa East Asia and Pacific Sub-Saharan Africa Europe and Central Asia duction grew at a 6.5 percent annualized rate in High-income countries the three months ending in October. Sources: IMF and World Bank. Among the five European Union accession members that are developing countries,6 Poland has best weathered the economic storm and is one of a handful of the 24 developing countries has fallen below the 10 percent level, and the in Europe and Central Asia not to witness a con- Russian central bank has repeatedly lowered its traction in output. Poland's good performance refinancing rate, so that it is nearly zero in real reflects comparatively resilient service and agri- terms. The government has also put in place a cultural sectors, compared with industrial out- large fiscal stimulus program, and as a result the put, which fell by 9 percent in the first half of fiscal budget is projected to move from a surplus 2009 over the first half of 2008. Exports were of 4.3 percent of GDP in 2008 to a deficit equiv- also relatively resilient, and as a result, net ex- alent to 7 percent of GDP in 2009. ports contributed positively to growth. Turkey's economy is projected to have con- Outside of Poland, the other four develop- tracted by 5.8 percent in 2009, nearly on par ing European Union accession economies posted with the 5.7 percent decline posted during the marked contractions in output during 2009, 2001 economic crisis and its largest contraction given the bursting of the credit boom and con- on record since 1969. The economy has been traction in demand from Western Europe. The hit by an investor pullback and sharp decline economies of Latvia and Lithuania were under in demand from export markets, notably from significant stress before the onset of the acute western and eastern Europe, where economies phase of the crisis--a situation that was exacer- have posted some of the sharpest slowdowns bated by heightened international risk aversion globally. The pace of contraction in growth (and concerns about the sustainability of their hit the trough in the first quarter of 2009, at pegs to the euro given huge accumulated imbal- 14.7 percent compared with the previous year, ances). GDP in both countries is estimated to but has eased significantly to a 3.3 percent rate have declined by well over 10 percent in 2009. of contraction in the third quarter--a relatively All four countries entered the crisis with very rapid turnaround. Unemployment has surged, large current account deficits. While the collapse contributing to the marked decline in private in domestic demand has improved their external consumption and fixed investment. With import positions, substantial external debt obligations volumes contracting even faster than export vol- remain, further undermining the business and umes, the current account deficit improved to investment environment. 126 R E G I O N A L E C O N O M I C P R O S P E C T S Among the Commonwealth of Independent Asian and Caucasus countries, weaker eco- States, Ukraine is projected to post the deepest nomic conditions--notably a sharp reduction contraction in GDP of 15 percent in 2009-- in trade demand from Russia, lower oil and indeed one of the sharpest contractions in the commodity prices, and significant reductions world. The plunge in metal prices in 2008 in investment and remittance flows--have been took a toll on the economy, where nonpre- partially offset by sustained economic assis- cious metals represent over 40 percent of goods tance from Russia. exports. Further, political strains in the lead- Overall, the number of poor or vulnerable up to the January 2010 presidential elections people in the Europe and Central Asia region have delayed the government from meeting the is estimated to have increased by some 10 mil- full set of IMF policy measures (such as rais- lion in 2009--compared with what might ing household gas prices) under its $11 billion have been had the crisis not arisen (based on a November 2008 stand-by facility. Thus, while $5-a-day poverty line). The contraction in eco- the government has made some progress in nomic activity has led to a 2.5 percentage point meeting its commitments to the IMF, it appears jump in the median unemployment rate of the that the release of the latest $3.7 billion tranche 10 countries reporting data (compared with will be postponed until after the elections. This August 2008). Unemployment is expected to uncertainty--along with ongoing political remain high for some time, curtailing house- instability--has undermined confidence and hold expenditures and contributing to higher contributed to the depreciation and heightened poverty rates. Partly as a result of higher un- volatility in the hryvnya, which depreciated by employment in destination countries (notably 50 percent against the U.S. dollar in 2009. the European Union and Russia) for migrants, Economic growth in the five Central Asian remittances are projected to decline by 15 per- countries has been relatively more robust than in cent in 2009--placing additional pressure on the rest of the region.7 However, this aggregate poor households. The macroeconomic impact picture masks wide differences in economic per- from the decline in remittances will be larg- formances at the country level. Turkmenistan and est in countries such as Albania, Armenia, Uzbekistan--among the least open economies in Moldova, and Tajikistan, where remittances the Commonwealth of Independent States and represent between 9 percent of GDP (Ar- exporters of natural gas--were only modestly menia) and as much as 50 percent of GDP affected by the global crisis. In addition, these (Tajikistan). In Tajikistan an estimated 30 per- economies benefited from the implementation cent contraction in remittances may cause an of fiscal stimulus measures and are estimated additional 5 percent of the population to move to have posted the strongest GDP growth out- into poverty.9 comes in the Europe and Central Asia region, with 8 percent and 5.5 percent, respectively, Medium-term outlook in 2009. Growth in Tajikistan and Kyrgyz The recovery in economic growth in the region Republic was buoyed by an upswing in agri- is expected to be slow and marked by a rise cultural output stemming from good harvests. in poverty. GDP is projected to rise a modest In contrast, GDP in Kazakhstan is estimated 2.7 percent and 3.6 percent in 2010 and 2011, to have contracted, led by the negative fiscal respectively. This growth path contrasts sharply effects from the collapse in oil prices. with the average growth rate for the region of 7 Among the three Caucasus countries,8 the percent from 2003 through 2007, and with the global crisis has had a particularly pronounced aggregate growth of 5.6 percent and 6.1 per- impact on Armenia and to a lesser extent cent projected for other developing countries Georgia--with economic conditions in the lat- in 2010 and 2011, respectively. While resur- ter also negatively affected by the conflict with gent demand in parts of Europe and Asia-- Russia in 2008. In most of the other Central combined with stable and/or modestly rising 127 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 commodity prices--should support a turn- a projected firming of growth, which should around in the region's exports, the projected support stronger revenues, fiscal consolida- weak recovery for developed Europe will result tion is projected to progressively reduce the in relatively muted overall export growth. Simi- regional fiscal deficit from 6.2 percent of GDP larly, foreign direct investment--which corre- in 2009 to 4.5 percent in 2010 and to 3.4 per- lates strongly with trade activity--and credit cent in 2011. However, the adjustment in an inflows are expected to remain significantly environment characterized by large negative lower than the levels observed before the crisis. output gaps and low growth will be difficult, Given the region's overleveraged private sec- particularly as the recovery in tax revenues tor, weakness in the banking sector, and house- may initially underperform, and that recovery hold indebtedness, the recovery in domestic will be exacerbated by additional pressure in demand is expected to be muted. Higher tax the medium term emanating from extensive rates, cuts in public spending, higher unem- social assistance and pension regimes to sup- ployment, and lower wages will curb private port the aging population. consumption, which is projected to firm to Monetary policy is expected to remain ac- 2.2 percent in 2010 and 3.3 percent in 2011-- commodative in most regional economies over half the unsustainable 8.4 percent pace recorded much of the forecast horizon. Inflationary in 2006­07. Excess capacity and crowding-out pressures should remain subdued, given large from increased government borrowing will excess capacity, weak domestic demand, and constrain investment, which is projected to a relatively open economy. In countries fac- grow by 4.1 percent in 2010 and by 4.7 percent ing continued adjustment in demand to reduce in 2011--well below the double-digit growth external and internal imbalances, monetary rates recorded in the pre-crisis years. policy is expected to remain relatively restric- Because of weak domestic demand and tive to help dampen activity. Monetary policy relatively tight financial conditions, the re- in countries with IMF programs will be guided gional current account balance is forecast to by the framework defined by the ongoing remain close to zero over the forecast horizon. Stand-by Arrangements. For those economies Across the region, however, there is greater that are moving toward adoption of the euro, variety (table A4). For instance, hydrocar- or whose currencies are pegged to the euro bon exporting economies are expected to re- (Bosnia and Herzegovina, Bulgaria, Latvia, cord rising surpluses or reductions in deficits and Lithuania), monetary policy will be influ- resulting from somewhat higher petroleum enced by the European Central Bank's policy prices and increased production (Azerbaijan, stance--which is expected to remain support- Kazakhstan, Russia, Turkmenistan, and ive of growth over the forecast horizon, but Uzbekistan). This improvement is projected to incrementally withdraw stimulus measures be offset by an expansion in the current account (including reversal of policy interest rate cuts) deficits--given a more rapid recovery in do- as demand conditions permit. A moderate up- mestic demand leading to import volumes tick in median regional headline inflation is that exceed exports--in Moldova, Poland, projected for 2010, as the downward pressure Romania, Ukraine, and Turkey. For developing from the fall in oil prices in the second half of European countries with important automotive 2008 ceases and the recent uptick in commod- industries, sales are expected to decelerate as ity prices starts to work through the system. cash-for-clunkers programs in high-income These pressures should be partly neutralized European countries unwind. by the strong appreciation in currencies since Most countries have little room for further March 2009 (particularly the Russian ruble fiscal expansion. Indeed, government spending and the Turkish lira), which will help to re- is projected to moderate as a result of planned duce import costs. However, core inflation structural fiscal consolidation. Combined with will continue to be subject to disinflationary 128 R E G I O N A L E C O N O M I C P R O S P E C T S Table A4 Europe and Central Asia country forecasts (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009c 2010d 2011d Albania GDP at market prices (2005 US$)b 5.4 5.0 6.0 6.5 2.2 3.0 4.5 Current account balance/GDP (%) 5.5 5.9 8.6 13.4 12.8 7.6 6.7 Armenia GDP at market prices (2005 US$)b 8.6 13.2 13.8 6.8 13.0 1.5 3.5 Current account balance/GDP (%) 11.7 1.8 2.6 4.9 2.8 0.6 3.7 Azerbaijan GDP at market prices (2005 US$)b 10.2 34.5 25.0 10.8 3.1 5.2 8.5 Current account balance/GDP (%) 16.6 17.7 28.5 37.6 19.5 27.2 26.2 Belarus GDP at market prices (2005 US$)b 6.9 10.0 8.6 10.0 1.0 2.0 4.0 Current account balance/GDP (%) 3.2 4.0 6.8 8.4 9.2 6.3 5.1 Bulgaria GDP at market prices (2005 US$)b 2.2 6.7 6.2 6.0 6.5 2.0 3.6 Current account balance/GDP (%) 3.6 18.4 25.2 25.4 9.8 5.2 4.9 Georgia GDP at market prices (2005 US$)b 6.6 9.4 12.3 2.2 4.0 2.0 3.5 Current account balance/GDP (%) 10.0 16.2 16.9 22.8 18.2 15.8 16.7 Kazakhstan GDP at market prices (2005 US$)b 6.4 10.7 8.2 3.0 1.9 1.8 3.5 Current account balance/GDP (%) 2.3 2.5 7.0 9.5 1.3 2.2 1.4 Kyrgyz Republic GDP at market prices (2005 US$)b 4.7 3.1 8.5 7.6 0.6 2.4 2.8 Current account balance/GDP (%) 10.2 10.6 0.6 4.6 5.2 2.4 4.9 Latvia GDP at market prices (2005 US$)b 6.9 12.2 10.3 4.6 17.5 3.9 2.4 Current account balance/GDP (%) 7.5 22.7 21.5 11.3 5.3 6.0 7.0 Lithuania GDP at market prices (2005 US$)b 6.0 7.8 8.9 3.0 17.5 3.5 2.2 Current account balance/GDP (%) 7.9 10.7 14.6 16.1 0.5 0.3 0.5 Macedonia, FYR GDP at market prices (2005 US$)b 2.2 4.0 5.9 5.0 1.3 1.9 3.8 Current account balance/GDP (%) 5.9 0.5 4.4 12.5 9.4 8.3 7.3 Moldova GDP at market prices (2005 US$)b 2.3 4.8 3.0 7.2 9.0 1.4 2.8 Current account balance/GDP (%) 7.9 11.3 16.5 17.4 9.0 10.2 11.1 Poland GDP at market prices (2005 US$)b 4.3 6.2 6.7 4.9 1.6 2.2 3.4 Current account balance/GDP (%) 3.3 2.7 4.7 5.5 0.9 2.6 2.5 Romania GDP at market prices (2005 US$)b 2.2 7.9 6.2 7.1 7.8 0.5 4.2 Current account balance/GDP (%) 5.8 10.4 13.5 12.4 4.2 4.9 5.5 Russian Federation GDP at market prices (2005 US$)b 3.9 7.7 8.1 5.6 8.7 3.2 3.0 Current account balance/GDP (%) 7.6 9.6 5.9 6.2 3.1 2.5 1.7 Tajikistan GDP at market prices (2005 US$)b 4.6 7.0 7.8 7.9 2.0 5.0 5.0 Current account balance/GDP (%) 4.5 2.8 8.6 7.9 10.9 11.1 10.2 Turkey GDP at market prices (2005 US$)b 4.3 6.9 4.7 0.9 5.8 3.3 4.2 Current account balance/GDP (%) 1.5 6.0 6.1 5.8 1.9 2.5 2.8 Ukraine GDP at market prices (2005 US$)b 2.7 7.3 7.9 2.1 15.0 2.2 3.0 Current account balance/GDP (%) 2.7 1.5 3.7 7.2 0.6 0.1 2.1 Uzbekistan GDP at market prices (2005 US$)b 4.6 7.3 9.5 9.0 5.5 6.5 6.5 Current account balance/GDP (%) 3.3 14.4 19.5 26.3 16.9 20.4 19.2 Source: World Bank. Note: World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries' pros- pects do not significantly differ at any given moment in time. Bosnia and Herzegovina, Turkmenistan, and Yugoslavia, FR (Serbia/Montenegro) are not forecast owing to data limitations. a. Growth rates over intervals are compound average; growth contributions, ratios, and the GDP deflator are averages. b. GDP measured in constant 2005 U.S. dollars. c. Estimate. d. Forecast. 129 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 tendencies and headline inflation is expected Figure A11 Nonperforming loans rise across to ease into 2011. much of developing Europe and Central Asia % share of total Risks Belarus Despite the weak baseline forecasts for the re- Bulgaria 2007 2009 gion, risks remain tilted toward the downside, Bosnia and Herzegovina a result of financing constraints, the limited Turkey scope for supportive fiscal policy, large and Poland rising banking sector vulnerabilities, and a lack Macedonia, FYR of economic diversification. If the domestic Russian Federation recovery is slow and subdued with continued Moldova high interest rates stifling investment growth, Montenegro potential output could suffer--leading to a Serbia rise in structural unemployment. A more pro- Latvia tracted and deeper-than-projected recession Lithuania Ukraine could place further pressure on banking sys- tems and on currencies in those countries with 0 5 10 15 20 25 30 relatively inflexible exchange rate regimes. Percent share of total Balance-sheet consolidation by parent banks Sources: IMF Global Stability Report, October 2009; World Bank. of foreign subsidiaries may manifest as fur- Note: Ranked by 2009 median for all reporting countries. Median was 2.7 percent for 2007; 4.4 percent for 2009. ther cuts in financial flows to the region in the months ahead. Rising domestic nonperform- ing loans and inadequate provisioning thereof Reinvigorating the reform programs that pose significant risks to regional growth by have stalled with the global crisis could help de- restricting capital availability or, in a worst liver stronger growth outturns than projected.11 case scenario, leading to a freezing of banking Regional governments have space to introduce systems (figure A11). This already somber sce- institutional reforms to improve the regulatory nario may be further darkened if it coincides framework and reduce red tape, tighten legal with a global double-dip scenario, particularly standards and further adopt international con- if the region's major export markets (such as tract and property rights norms, and clamp Germany) are severely affected. down on corruption to improve competition A related and enduring risk for the re- and efficiency, among other reforms. Failure to gion derives from the high level of household reform the pension systems poses a long-term and corporate foreign-currency-denominated threat to growth, given high social security fi- debt. Exposure to foreign exchange loans ex- nancing burdens. Successful implementation of ceeds 50 percent of total lending in Hungary, these reforms may lower precautionary savings, Kazakhstan, Latvia, Lithuania, Romania, and with positive spin-offs for private consumption Ukraine for both corporate and household bor- and growth. Higher private consumption in the rowers. For households in particular, high levels region is indeed identified as a possible upside risk of foreign exchange debt post significant risks, and incorporated in the global "more buoyant because unlike corporations, households are private-sector reaction" scenario (see chapter 1). unlikely to have hedged against exchange rate Finally, given the degree of dislocation en- movement.10 For countries with relatively in- gendered by the crisis, black market activity is flexible exchange rate regimes, outturns could expected to rise, posing challenges for policy find these regimes under assault, which in turn makers and undermining greater fiscal con- would limit the ability of regional central banks solidation. In the Commonwealth of Indepen- to conduct accommodative monetary policy. dent States, a lack of economic diversification 130 R E G I O N A L E C O N O M I C P R O S P E C T S outside of mineral-export-led activities is a the contraction in domestic output. Central common structural weakness and remains a American economies (including Mexico) key vulnerability. were the worst affected, with output contract- ing a sharp 6.4 percent, while growth in the Caribbean economies stagnated. Latin America and the Caribbean In the immediate aftermath of the crisis, the region was hit by a sharp slowdown in Recent developments private capital inflows, while increased uncer- T hanks to sound macroeconomic fundamentals in place before the onset of the crisis, the Latin America and tainty and credit tightening led to a marked contraction in private consumption and pri- vate investment. The capital outflows induced Caribbean region has been able to weather a sharp depreciation of currencies in the re- the global financial crisis much better than gion, a decline in equity markets, and much previous external shocks. Nevertheless, higher borrowing costs. Nevertheless, the re- economic activity in the region deceler- gion managed to avoid falling into a balance ated sharply in the aftermath of the crisis. of payments and/or financial crisis. For the 2009 calendar year, GDP is esti- Private consumption contracted by nearly mated to have fallen 2.6 percent, following 2 percent, while fixed investment declined an expansion of 3.9 percent in 2008 (table A5). sharply by 13.6 percent, after growing at This aggregate result masks a high degree of double-digit rates in the previous years. The heterogeneity among countries in the region region was also affected by the collapse in ex- with respect to the timing and magnitude of ternal demand for commodity exports, falling Table A5 Latin America and the Caribbean forecast summary (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009e 2010f 2011f GDP at market prices (2005 US$)b 2.9 5.4 5.5 3.9 2.6 3.1 3.6 GDP per capita (units in US$) 1.4 4.0 4.1 2.6 3.8 1.8 2.3 PPP GDPc 2.9 5.5 5.7 4.2 2.3 3.0 3.5 Private consumption 3.4 6.1 3.5 4.2 1.9 3.2 3.4 Public consumption 2.2 2.8 2.9 4.1 2.9 2.8 2.6 Fixed investment 3.3 13.4 20.7 11.7 13.6 6.1 5.8 Exports, GNFSd 6.0 6.7 4.9 1.6 11.2 7.8 5.0 Imports, GNFSd 6.2 14.0 11.9 9.2 15.8 10.3 5.6 Net exports, contribution to growth 0.2 1.5 1.7 2.0 1.6 0.7 0.3 Current account balance/GDP (%) 1.6 1.4 0.4 0.6 0.9 1.0 1.0 GDP deflator (median, LCU) 7.1 7.2 5.4 8.4 7.2 3.0 4.0 Fiscal balance/GDP (%) 3.5 1.1 1.1 0.9 3.3 2.8 2.5 Memo items: GDP Latin America excluding Argentina 3.0 5.2 5.2 3.7 2.6 3.2 3.7 Central America 3.6 5.0 3.7 1.7 6.4 3.3 3.6 Caribbean 4.2 9.0 6.1 3.6 0.1 2.3 3.3 Brazil 2.4 4.0 5.7 5.1 0.1 3.6 3.9 Mexico 3.6 4.8 3.3 1.4 7.1 3.5 3.6 Argentina 2.3 8.5 8.7 6.8 2.2 2.3 2.4 Source: World Bank. a. Growth rates over intervals are compound average; growth contributions, ratios, and the GDP deflator are averages. b. GDP measured in constant 2005 U.S. dollars. c. GDP measured at PPP exchange rates. d. Exports and imports of goods and nonfactor services. e. Estimate. f. Forecast. 131 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Exports collapsed in the first half of the year, Figure A12 Industrial production and trade dragging output down 7 percentage points, volumes are gaining momentum in Latin America and the Caribbean while the collapse in import volumes boosted growth by close to 9 percentage points. In Industrial production growth (3 month saar) 2009 private consumption is estimated to 50 Export volume Industrial 40 production have contracted by 6.9 percent, as the labor 30 market was severely affected by the economic 20 10 slowdown, with formal unemployment almost 0 doubling to 6.4 percent by September, and as 10 remittances fell 13.4 percent in the first nine 20 30 months of the year. 40 In Argentina, the global recession in con- 50 junction with policy-related uncertainty took Jan. Jul. Jan. Jul. Jan. Jul. Jan. 2000 2001 2003 2004 2006 2007 2009 a toll on investment and trade. Collapsing Source: World Bank. imports and declining fiscal revenues point to weak domestic demand and relatively poor output performance in the first half of 2009, while a severe drought added to the economy's weak performance. In República Bolivariana de Venezuela, commodity prices, lower remittance inflows, GDP is estimated to have declined 2.4 per- and declining tourism activity. The decline cent as a result of the collapse in external in domestic demand translated into a sharp demand, weak private consumption, and 15.8 percent contraction in import volumes. lower investment spending. Manufacturing and As a result, and despite an 11.2 percent con- retail sales plunged owing to weak domestic traction in export volumes, net trade contrib- demand, and output fell at a 4.5 percent an- uted 1.6 percent to growth. Reflecting these nualized pace in the third quarter of 2009. The developments industrial activity fell rapidly, oil sector is becoming increasingly dominant plunging at a 20 percent saar rate in the last in the economy. Supply bottlenecks, a difficult quarter of 2008 and at 16 percent in the first business environment, and a lack of private quarter of 2009 (figure A12). investor confidence are undermining new in- Countries that rely heavily on trade with vestment, impairing much-needed economic the United States were especially hard hit by diversification. the crisis. Mexico's economy suffered the Strong retrenchment in private investment steepest contraction in the region (7.1 percent) spending and a steep drawdown in stocks and its worst economic performance in seven (close to 1 percent of GDP) caused Chile's out- and a half decades, both because of its close put to decline 4.7 percent year-on-year in the economic ties with the United States in the sec- second quarter of 2009. Marked weakness in tors most affected by the crisis (construction, domestic demand resulted in a sharp contrac- automotive, and electrical appliances) and be- tion in imports that exceeded the plunge in cause of the AH1N1 flu outbreak in the sec- exports. ond quarter of 2009. The flu outbreak hit the Peru's economic growth decelerated from tourism sector especially hard and is estimated a double-digit pace in the first half of 2008 to have reduced overall GDP by 0.5 percent. to a standstill in the first half of 2009, with Furthermore Mexican firms suffered foreign the sharp contraction in investment spending derivatives losses in December 2008 after the leading to a 5.4 percent contraction in domes- global crisis drove the peso to record lows. tic demand. Weak external demand resulted 132 R E G I O N A L E C O N O M I C P R O S P E C T S in a 6.3 percent decline in exports, although respectively, during the first half of 2009. In imports contracted more sharply on account the first quarter of 2009, FDI inflows to Costa of weak domestic demand. Rica fell by 19 percent (year-on-year), and by Countries in Central America and the 41 percent to the Dominican Republic. Caribbean were afflicted by the recession in the In response to the crisis, many governments in United States and major economic partners in the region implemented countercyclical macro- the European Union, particularly Spain, which economic policies in an effort to support domes- has resulted in a contraction in trade, tourism, tic demand, with government spending being the FDI, and remittances. The Caribbean econo- only demand component that registered growth mies contracted only 0.1 percent in 2009, during 2009. The aggressiveness of the fiscal down from the 3.6 percent growth recorded policies implemented depended on the fiscal in 2008. Jamaica recorded one of the sharpest space available in each country and the extent declines in GDP in the subregion, attributable to which they had access to financial markets. to its heavy dependence on the U.S. economy That said, the region entered the crisis much (remittances declined 17 percent in the first better prepared with respect to both the fiscal half of 2009), and to sharp cuts in mining and external accounts. In Mexico, the declining production. In the Dominican Republic, eco- oil revenues constrained the countercyclical re- nomic performance deteriorated sharply, with sponse. In Chile, fiscal stimulus has helped limit output down by 0.1 percent after 5 percent the output contraction, and the government also growth in 2008, reflecting developments in the provided credit support to SMEs through the de- U.S. economy that affected remittances, FDI, velopment bank Banco Estado. The implementa- and tourism. The improvement in the terms tion of the fiscal stimulus in Peru was to some of trade, as the oil price declined, has had a extent hindered as budget appropriation and positive impact on economic performance, how- distribution rules limited the increase in govern- ever. Caribbean economies benefited somewhat ment spending, even as procurement rules have from the AH1N1 outbreak in Mexico as visitors become more lax. Furthermore the government shifted holiday destinations from Mexico to the provided credit support to SMEs through the de- Caribbean islands, and consequently in the early velopment bank Banco de la Nacion to help ease stages of the crisis, tourism and offshore finan- the impact of the credit crunch. cial services proved somewhat resilient. To support domestic demand at the time The Central American economies, exclud- that external demand was collapsing, coun- ing Mexico, contracted by 1.0 percent in 2009. tries more integrated in the global economy External demand for their exports was hit by lowered interest rates aggressively and allowed the global economic crisis, while remittances real exchange rates to depreciate (figure A13, and tourism revenues also declined. Costa figure A14). During the monetary-easing cycle, Rica's economy was afflicted by a 10.3 per- the central bank of Colombia cut rates by a cent decline in U.S. tourist arrivals in the cumulative 6.5 percentage points. Chile cut first nine months of 2009, but investment in rates by 7.75 percentage points since the begin- the services sector continued and back-office ning of 2009, while Peru also eased monetary services were resilient. The decline in tourist policy substantially. Brazil cut the SELIC12 arrivals has prompted large price cuts for tour- rate by an unprecedented 500 basis points to ism packages as countries competed for a de- 8.75 percent. clining number of tourists. Remittances have As elsewhere, many economies in the re- also suffered because of weak labor markets gion showed signs that the recession bottomed in high-income countries. Compared with a out in the second half of 2009, with external year earlier, remittances to Guatemala and demand rebounding faster and more strongly El Salvador were down by 9.5 and 10.3 percent, than initially anticipated (figure A15). 133 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Figure A13 Central banks across Latin Figure A15 Quarterly GDP points to output America eased monetary policy aggressively stabilization in Latin America Percent Percent change 15 6 4 2 10 0 2 4 5 6 8 le o a ru a il ic bi in az hi Pe ex om nt C Br 0 ge M ol Ar C Aug. Nov. Feb. May. Aug. Aug. 2008 2008 2009 2009 2009 2009 2008Q1 2008Q2 2008Q3 2008Q4 Mexico Brazil Chile 2009Q1 2009Q2 2009Q3 Colombia Peru Source: Haver Analytics. Sources: Central banks; Datastream; Thomson Financial. Note: Data represent the seasonally adjusted change quarter over quarter except for Mexico, which are seasonally adjusted annual rates. Figure A14 Real effective exchange rates depreciated in more integrated economies for a sharp recovery in industrial production, Jan 2006 100 which increased by an annualized 17.6 percent 150 in the second quarter and by 20.5 percent in 140 the third quarter. However, because of base 130 effects and a rather moderate recovery in 120 external demand, as Chinese restocking tailed 110 off, output is estimated to have remained rel- 100 atively flat in 2009, implying the worst eco- 90 nomic performance since the early 1990s. 80 In Mexico, the rate of contraction moder- ated in the second and third quarters, supported 70 Jan. Jul. Jan. Jul. Jan. Jul. Jan. Jul. by less dramatic output declines in the manu- 2006 2006 2007 2007 2008 2008 2009 2009 facturing and service industries. In Argentina, Brazil Chile Colombia an improved external environment has ignited Mexico Peru a modest recovery and led to improvements in external balances, as commodity prices firmed Source: JPMorgan. and demand for exports increased, in partic- ular from its main trading partner Brazil. In Chile significant fiscal and monetary stimu- In Brazil, a swift rebound in domestic de- lus contributed to the moderation in output mand was boosted by expansionary monetary contraction to 1.7 percent year-on-year in the policy and countercyclical fiscal policy. These third quarter, bringing the decline in GDP over steps pulled the economy out of recession in the first three quarters of the year to 2.7 per- the second quarter of 2009. Brazil's economy cent. In Colombia the improved external en- is also benefiting from the shift in the inven- vironment and the lagged effect of aggressive tory cycle. This in conjunction with the stimu- monetary easing helped the economy recover lus for the automotive sector has set the stage in early 2009. Output growth in the first two 134 R E G I O N A L E C O N O M I C P R O S P E C T S quarters of 2009 was also boosted by strong growth in public investment spending, even Figure A16 Capital flows return to Latin America though private consumption and investment US$ (billions) remained weak. In Peru a significant rebuild- ing of inadequate stocks is projected to con- 70 60 tribute to growth in the second half of 2009. 50 Uruguay's economy expanded by 0.5 percent 40 in the second quarter of 2009 relative to the 30 previous quarter, bolstered by growth in con- 20 struction and transportation, reflecting the 10 impact of several megaprojects, which offset 0 output declines elsewhere, particularly in en- 08 08 08 08 09 09 09 09 20 20 20 20 20 20 20 20 1 2 3 4 1 2 3 4 ergy, agriculture, and manufacturing. Q Q Q Q Q Q Q Q Corporate and sovereign spreads have re- Bank Equity Bond treated to pre-crisis levels in countries more integrated into the global financial system-- Sources: Haver Analytics; World Bank. demonstrative of a return of investors' confi- dence, while access to the international debt market has also improved. Lower-rated coun- tries in the region continue to be perceived as risky by investors and this is reflected in weaker investment growth. The shape of the spreads remaining above pre-crisis levels. recovery will, to a large extent, be determined Overall, capital inflows to the region have by the growth path of the United States and returned, especially in economies that proved other major economic partners of the region. resilient to the crisis, such as Brazil, with total Growth is expected to remain strong for the capital inflows rising to $57.4 billion in the next couple of quarters but to weaken in the fourth quarter of 2009, up from $15.7 billion second half of 2010, as the impact of stimu- in the second quarter. Bond issuance increased lus measures and the rebuilding of depleted almost sixfold, nearing $30 billion, while eq- inventories cease to bolster growth. A double uity inflows more than tripled to $14.8 billion. dip or a more buoyant growth scenario is also Bank lending recovered modestly, totaling possible as a result of close linkages with high- $13.2 billion, down 33 percent compared with income countries. the second quarter of 2008 (figure A16). Economies more integrated through trade and financial linkages with the global econ- Medium-term outlook omy, which have been the worst affected by Fiscal stimulus, lagged impacts of accom- the global downturn are expected to benefit modative monetary policy, the shift in the most from the global economic recovery. The inventory cycle, improvements in the terms region's exports are projected to rebound of trade, rising consumer and business con- strongly, expanding by 7.8 percent in 2010 as fidence, stronger demand from high-income demand from major trading partners recov- countries, and an easing of external financing ers. Higher commodity prices will also ben- conditions are all expected to support growth efit commodity exporters in the region, easing in the region over the next few quarters. GDP pressures on external balances and in some growth in the region is projected to accelerate to cases fiscal balances. A weakening of growth 3.1 percent in 2010, following an estimated momentum or even a double dip in high- 2.6 percent contraction in 2009, but growth income countries (see chapter 1) could lead to will not regain the growth rates recorded a deceleration in export growth in the second during the boom years, in part because of half of 2010 and into 2011. 135 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Private consumption in the region is pro- private consumption in many countries in the jected to bounce back strongly, rising 3.2 per- Caribbean and Central America. cent in 2010, partly because of a low base The recovery in the United States will help effect (it contracted by an estimated 1.9 per- Mexico exit the deep recession it entered fol- cent in 2009) but also owing to improvements lowing the collapse in U.S. demand. Mexico's in labor markets throughout the region and economy is forecast to expand by 3.5 percent in migrant-destination countries. Domestic in 2010 and growth will accelerate marginally demand growth may be supported by a pro- to 3.6 percent in 2011 (table A6). Govern- nounced bounce back in fixed investment as ment spending is not expected to grow as it confidence returns and financing constraints implements fiscal adjustments to compensate ease (see chapter 1 regarding a more buoyant for lower oil revenues associated with declin- private sector reaction scenario). Less restric- ing oil production. Both exports and imports tive financing conditions compared with the are projected to rebound strongly in 2010, as crisis period and a return of investor confi- external and domestic demand strengthen, but dence together with resumptions in delayed net trade will be a drag to growth, as the ac- investment, are projected to boost fixed in- celeration in imports due to stronger domestic vestment by 6.1 percent in 2010. However, demand will outpace export growth. A strong investment growth will remain below the rebound in the service sector is projected, after double-digit pace recorded in the boom years, a subdued performance in 2009 on account as excess capacity lingers. Large output gaps, of the negative impact of the AH1N1 flu. weak international financing conditions, and Mexico's growth outlook is clouded, however, weak public sector investment will all weigh by concerns about the long-term sustainability on prospects. The lagged impact of the sub- of fiscal accounts. The fiscal shortfall over the stantial monetary easing in some countries, 2009­10 period is estimated at a cumulative along with stronger fiscal stimulus, and a one- 6.6 percent of GDP, with almost half of the off benefit from inventories accumulation will deterioration related to lower oil prices and bolster growth into 2010. In other countries, production. The expected fiscal reform should such as Chile, there will be fiscal consolidation reduce government discretionary spending, in 2010, which will moderate the contribution which may have a negative impact on growth of government spending to growth. over the short(er) term. The tourism sectors in many countries Domestic demand in Brazil should benefit in the region are expected to stage a recov- from strong fiscal and monetary stimuli, while ery after a sharp decline in tourist arrivals in exports are projected to rise in response to 2009, although a recovery in Mexico's tour- strong external demand from China. Overall, ism sector may weaken the recovery in some the economy is projected to stage a comeback of the Caribbean countries that had seen a in 2010, with growth accelerating to 3.6 per- lower-than-expected decline in tourist arrivals cent. Economic growth will be largely driven by in 2009, as they managed to attract tourists by the recovery in private consumption and invest- offering discounted packages. ment, as well as stronger external demand. Remittances are expected to recover only Recovery in external demand will help modestly in the 2010­11 period, undermined Argentina's economic recovery strengthen into by weak labor market conditions in the 2010 as job creation in export-oriented indus- United States and other high-income coun- tries will underpin a mild recovery in private tries, although the bottoming out of the hous- consumption. The expected recovery in the ag- ing sector in the United States bodes well for riculture sector will boost economic activity, countries receiving remittances from the con- as will less restrictive external financing condi- struction sector. The weak recovery in remit- tions. The recovery will be fragile, however, tances will limit the strength of the recovery in with investment remaining a drag on growth 136 R E G I O N A L E C O N O M I C P R O S P E C T S Table A6 Latin America and the Caribbean country forecasts (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009c 2010d 2011d Argentina GDP at market prices (2005 US$)b 2.3 8.5 8.7 6.8 2.2 2.3 2.4 Current account balance/GDP (%) 0.2 3.6 2.8 2.2 2.3 2.2 2.2 Belize GDP at market prices (2005 US$)b 5.6 4.7 1.2 3.8 0.1 1.7 2.3 Current account balance/GDP (%) 12.1 2.1 4.0 10.8 7.7 7.7 7.6 Bolivia GDP at market prices (2005 US$)b 3.8 4.6 4.6 6.1 2.6 3.2 3.8 Current account balance/GDP (%) 3.0 11.5 12.5 12.0 2.6 1.8 2.9 Brazil GDP at market prices (2005 US$)b 2.4 4.0 5.7 5.1 0.1 3.6 3.9 Current account balance/GDP (%) 2.0 1.3 0.1 1.7 1.1 1.6 1.8 Chile GDP at market prices (2005 US$)b 4.2 4.6 4.7 3.2 1.8 4.7 4.5 Current account balance/GDP (%) 1.5 4.9 4.4 2.0 1.5 1.1 1.4 Colombia GDP at market prices (2005 US$)b 2.4 6.9 7.5 2.5 0.1 2.6 3.9 Current account balance/GDP (%) 2.2 1.8 2.9 2.7 2.9 2.6 2.3 Costa Rica GDP at market prices (2005 US$)b 4.5 8.8 7.8 2.6 1.8 2.1 2.9 Current account balance/GDP (%) 4.0 4.5 6.3 9.2 4.2 5.1 6.3 Dominica GDP at market prices (2005 US$)b 1.4 3.2 0.9 3.1 1.7 1.4 3.0 Current account balance/GDP (%) 19.8 17.3 28.5 36.5 24.2 24.1 23.8 Dominican Republic GDP at market prices (2005 US$)b 5.2 10.7 8.5 5.0 0.1 2.4 2.6 Current account balance/GDP (%) 0.8 3.6 5.1 10.1 6.8 7.2 6.7 Ecuador GDP at market prices (2005 US$)b 3.2 3.9 2.5 6.5 2.2 1.7 3.0 Current account balance/GDP (%) 1.4 3.9 3.6 2.2 3.0 3.3 3.4 El Salvador GDP at market prices (2005 US$)b 2.7 4.2 4.7 2.5 2.1 0.8 2.3 Current account balance/GDP (%) 2.5 3.6 5.4 7.2 2.6 3.5 4.7 Guatemala GDP at market prices (2005 US$)b 3.5 5.4 6.3 3.8 0.4 1.6 3.0 Current account balance/GDP (%) 4.9 5.2 5.4 4.8 2.8 4.1 4.4 Guyana GDP at market prices (2005 US$)b 1.7 2.4 5.4 3.2 1.1 2.5 3.0 Current account balance/GDP (%) 9.4 19.8 17.8 20.2 12.6 18.1 18.2 Haiti GDP at market prices (2005 US$)b 0.9 2.3 3.2 1.4 0.3 1.9 2.1 Current account balance/GDP (%) 4.0 9.0 5.7 8.2 7.9 9.1 10.6 Honduras GDP at market prices (2005 US$)b 3.8 6.3 6.3 4.0 2.5 1.8 2.8 Current account balance/GDP (%) 6.7 4.7 9.8 14.3 8.7 10.9 9.3 Jamaica GDP at market prices (2005 US$)b 0.8 2.7 1.5 1.0 3.7 0.3 2.2 Current account balance/GDP (%) 5.5 9.9 15.3 19.8 14.3 12.6 9.9 Mexico GDP at market prices (2005 US$)b 3.6 4.8 3.3 1.4 7.1 3.5 3.6 Current account balance/GDP (%) 1.9 0.5 0.8 1.5 1.4 1.7 1.9 (continued) 137 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Table A6 (continued) (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009c 2010d 2011d Nicaragua GDP at market prices (2005 US$)b 4.1 3.7 3.2 3.2 2.5 1.7 1.7 Current account balance/GDP (%) 20.2 13.4 17.6 23.8 15.2 19.8 21.7 Panama GDP at market prices (2005 US$)b 4.5 8.5 11.5 9.2 1.2 2.7 3.8 Current account balance/GDP (%) 5.3 3.1 7.3 12.3 7.4 10.1 10.0 Paraguay GDP at market prices (2005 US$)b 1.2 4.3 6.8 5.8 3.8 2.6 3.7 Current account balance/GDP (%) 1.5 1.4 0.8 2.1 0.3 1.5 1.9 Peru GDP at market prices (2005 US$)b 3.3 7.7 9.0 9.8 1.2 3.9 5.2 Current account balance/GDP (%) 3.3 3.0 1.6 3.4 2.4 2.5 2.3 St. Lucia GDP at market prices (2005 US$)b 2.9 2.2 1.7 0.7 1.4 1.5 2.7 Current account balance/GDP (%) 13.8 33.1 32.6 33.6 26.5 27.9 29.0 St. Vincent and the Grenadines GDP at market prices (2005 US$)b 4.2 10.8 6.7 2.3 1.0 1.2 1.8 Current account balance/GDP (%) 18.3 24.1 26.3 27.8 20.1 21.4 21.8 Uruguay GDP at market prices (2005 US$)b 1.5 4.6 7.6 8.9 1.3 3.2 3.4 Current account balance/GDP (%) 0.9 2.0 0.9 3.8 1.4 2.4 2.5 Venezuela, R. B. de GDP at market prices (2005 US$)b 1.6 10.3 8.4 4.8 2.4 0.2 1.4 Current account balance/GDP (%) 7.5 14.3 8.7 12.4 2.2 3.5 2.5 Source: World Bank. Note: World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries' prospects do not significantly differ at any given moment in time. Barbados, Cuba, Grenada, and Suriname are not forecast owing to data limitations. a. Growth rates over intervals are compound average; growth contributions, ratios, and the GDP deflator are averages. b. GDP measured in constant 2000 U.S. dollars. c. Estimate. d. Forecast. owing to policy uncertainty. Furthermore, funded by an increasing public debt issuance as the unsustainable fiscal stimulus implemented well as price and exchange controls, is under- ahead of the presidential elections will likely mining growth. Inflationary pressures are likely fade in the second half of the year, weakening to continue to be fueled by currency mismanage- one of the growth engines. ment as well as rising import costs, partly stem- República Bolivariana de Venezuela is ex- ming from the government's decision to import pected to buck the regional trend of economic through Argentina instead of Colombia. Fur- recovery and is likely to continue to contract for thermore, inadequate investments will exacer- a second consecutive year in 2010, as private bate domestic shortages, thereby exerting further consumption, investment, and exports continue upward pressure on prices. to shrink. Macroeconomic imbalances--the Small open economies like Chile are likely to result of inadequate macroeconomic policies benefit most from the global economic recovery and high inflation (notwithstanding economic as their business cycles are highly correlated with contraction)--will undermine investment. Also, the global economy. Chile's recovery will also be the strong growth of government spending, supported by domestic factors because aggressive 138 R E G I O N A L E C O N O M I C P R O S P E C T S and front-loaded countercyclical policies are faster than anticipated, leading to an inflationary boosting domestic demand. Improved terms of environment. In particular, the risks for Brazil trade as well as rising consumer and business con- have shifted to the upside as domestic demand is fidence should also bolster the recovery, bringing rebounding strongly, while the effects of already growth closer to potential. enacted monetary loosening and countercyclical Peru's recovery will benefit from stronger fiscal policy easing have not yet run their course. demand for commodity exports, particularly Another upside risk emanates from commodity from Asia. Furthermore, the Free Trade Agree- prices, should the world economy (particularly ment with China, which comes into operation in resource-intensive economies such as China) in January 2010, will further boost exports, stage a stronger-than-expected rebound. in particular those of fishmeal and minerals. The recent run-up in equity markets and Government consumption and investment stronger capital inflows in general, stemming in should be firm in 2010 as the government part from still large interest rate differentials, maintains efforts to support economic growth have put upward pressure on real effective ex- through new spending on public works and change rates in some countries. The surge in social programs, and it should remain a high capital inflows to the region, which reached priority ahead of the April 2011 presidential $87.2 billion in the second half of 2009 (of and congressional elections. which $34.3 billion came in December), com- Growth in Central America is expected to pared with $36.8 billion in the first half of the bounce back in 2010 in line with developments year, has prompted the Brazilian government to in the United States and other major economic impose a 2 percent financial transaction tax on partners. Recovery in the region is highly depen- foreign portfolio inflows. However, this mea- dent on workers' remittances from the United sure has been ineffective in preventing capital States and Europe (El Salvador, Guatemala, inflows and real currency appreciation. Should Honduras, and Nicaragua), and is projected to be such flows persist, this may lead to renewed more gradual, as the expected jobless recovery in asset price bubbles. Also, some economies may high-income countries will put pressure on remit- lose external competitiveness because of real tances, thereby delaying the recovery in private currency appreciation at a time when external consumption in these countries. Similarly, tour- demand recovery remains fragile. ism in the region (of particular importance for the Caribbean) is expected to recover only mod- erately as labor markets in client countries recover Middle East and North Africa only gradually. FDI, which was a major source of growth over the 2003­08 period, is unlikely Recent developments to return to pre-crisis levels while excess capacity lingers. The recovery in most countries in Central America will thus be anemic at best. In Jamaica, T he impact of the global financial crisis for the developing economies of the Middle East and North Africa region varied across oil low alumina and bauxite production and export exporters and importers of the region.13 Initially, prices will constrain the recovery. Growth in these the decline in regional equity markets was regions will continue to be undermined further by sharper than the average for emerging markets crime, corruption, weak democratic institutions, (figure A17). Since then, recovery in these mar- and a lack of competitiveness. kets has been hesitant owing to the unfolding of the Dubai World debt problems in the United Risks Arab Emirates as well as concerns regarding In countries where domestic demand is strength- growth prospects for the broader region. ening rapidly, delays in withdrawing policy Conditions at the outset of the financial stimulus represent an upside risk to growth and crisis were less than propitious for the Middle inflation. In such cases, output gaps could close East and North Africa. The "food-fuel" crisis 139 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 economies, but stimulus measures and stronger Figure A17 Equity markets in the Middle East dropped quickly, and GCC recovery has been non-oil developments helped to maintain posi- muted tive overall growth. For the more diversified U.S.$-based equity prices (Jan. 1, 2008 = 100) economies (Egypt, Jordan, Lebanon, Morocco, 120 and Tunisia) steep declines in external demand 100 (notably from the dominant Euro Area) had a negative effect on merchandise exports, com- 80 pounded by falling tourism volumes, lower 60 worker remittances, and declining FDI inflows, 40 notably those from the GCC economies. The decline in FDI was at first linked to falling oil 20 Jan. 1, Apr. 1, Jul. 1, Oct. 1, Jan. 1, Apr. 1, Jul. 1, Oct. 1, Jan. 1, revenues among the GCC and softer conditions 2008 2008 2008 2008 2009 2009 2009 2009 2010 in host markets. This came to be further clouded MSCI-All GCC Egypt, Arab. Rep. UAE by the question of sovereign debt sustainability on the part of Dubai, owing to past overinvest- Source: Morgan-Stanley through Thomson/Datastream. ment in real estate and tourism ventures within Note: MSCI-AII is all emerging markets; GCC is Gulf Cooperation Council; UAE is United Arab Emirates. the United Arab Emirates. Against this background, GDP growth in 2009 for the developing countries of the region is estimated to have eased to 2.9 per- cent, from 4.3 percent in 2008. For develop- of 2007­08 was a challenge for the region, ing oil exporters, growth almost halved, to the largest net exporter of oil and the larg- 1.6 percent from 2.9 percent in 2008. GDP est net importer of food. Oil exporters were gains for the oil importers (diversified econo- less adversely affected, but food import bills mies) faltered by almost 2 percentage points rose sharply. Hardest hit were countries in in the year, from a strong 6.6 percent outturn the Maghreb, as well as Jordan and Lebanon, in 2008 (powered by growth of more than which are large importers of both food and 7 percent in Egypt) to 4.7 percent in 2009. fuel; and the Arab Republic of Egypt (high And for the high-income GCC economies cov- food-import dependence). The policy environ- ered in this report, GDP is estimated to have ment had to shift quickly from mitigating the contracted by 0.6 percent in 2009 following a effects of higher commodity prices to shoring firm 4.6 percent growth in the preceding year, up banking systems and applying fiscal stimu- as the sharp slide in oil production and rev- lus to bolster domestic demand. enues dampened output (table A7). Over the course of 2009, net terms-of-trade movements for the developing oil export- Developments among regional oil exporters. The ers (Algeria, Islamic Republic of Iran, Syrian global economic crisis ended the oil boom that Arab Republic, and Republic of Yemen) and saw oil prices peak at more than $150 a barrel the Gulf Cooperation Council (GCC) were fa- on an intra-day basis in mid-2008 (figure A18), vorable, as oil prices increased and food prices and prices have settled into a range of $65­$80 declined. But high oil prices have been main- a barrel, supported by OPEC (Organization tained at the expense of much reduced output. of Petroleum-Exporting Countries) production Because of falling oil production, key GCC cuts. As part of this effort, regional oil export- oil exporters suffered modest GDP declines ers scaled back production by nearly 10 percent during the year, only partially offset by fiscal (11 percent among high-income producers and stimulus programs and more buoyant non-oil 7.3 percent among the developing exporters of sectors. Developing oil exporters in contrast the region). The combination of much lower saw a marked downturn in oil sectors of their prices and reduced output caused oil and gas 140 R E G I O N A L E C O N O M I C P R O S P E C T S Table A7 Middle East and North Africa forecast summary (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009g 2010h 2011h GDP at market prices (2005 US$)b 4.4 5.2 5.9 4.3 2.9 3.7 4.4 GDP per capita (units in US$) 2.7 3.5 4.1 2.6 1.2 2.0 2.8 PPP GDPc 4.5 5.4 6.2 4.3 2.7 3.6 4.4 Private consumption 4.2 4.8 6.3 1.2 4.6 4.5 5.1 Public consumption 3.3 5.7 2.2 11.3 10.6 7.5 6.4 Fixed investment 6.5 5.9 18.7 19.4 8.0 4.0 4.6 Exports, GNFSd 4.9 5.9 6.5 0.2 8.8 2.3 5.2 Imports, GNFSd 5.7 7.0 12.4 9.0 1.1 5.2 6.7 Net exports, contribution to growth 0.2 0.1 1.6 3.0 3.5 1.1 0.8 Current account balance/GDP (%) 2.9 11.6 10.1 10.5 0.1 1.5 1.0 GDP deflator (median, LCU) 5.2 8.3 6.1 16.0 6.7 6.2 3.9 Fiscal balance/GDP (%) 2.2 0.9 0.4 1.9 6.1 4.1 3.7 Memo items: GDP Middle East and North Africa geographic 4.0 4.6 4.9 4.4 1.4 3.5 4.3 regione Selected GCC Countriesf 3.6 3.8 3.6 4.6 0.6 3.2 4.1 Egypt, Arab Rep. 4.4 6.8 7.1 7.2 4.7 5.2 6.0 Iran, Islamic Rep. of 4.8 5.9 7.8 2.5 1.0 2.2 3.2 Algeria 4.0 2.0 3.0 3.0 2.1 3.9 4.0 Source: World Bank. a. Growth rates over intervals are compound average; growth contributions, ratios, and the GDP deflator are averages. b. GDP measured in constant 2005 U.S. dollars. c. GDP measured at PPP exchange rates. d. Exports and imports of goods and non-factor services. e. Geographic region includes high-income countries: Bahrain, Kuwait, Oman, and Saudi Arabia. f. Selected GCC countries: Bahrain, Kuwait, Oman and Saudi Arabia. g. Estimate. h. Forecast. Figure A18 Oil prices, 2004­09 Figure A19 Lower oil prices and lower Middle Eastern output yield sharp decline in oil US$ per barrel revenues in 2009 140 Jul. 2008: $133/bbl Revenues, Middle East oil exporters, US$ (billions) 120 800 100 Dec. 2009: Other $75/bbl 600 Middle East 80 60 400 Other GCC Jan. 2007: $53/bbl 40 Jan. 2009: $43/bbl 20 200 Jan. Jan. Jan. Jan. Jan. Jan. Jan. Saudi Arabia 2004 2005 2006 2007 2008 2009 2010 Sources: IEA and DECPG Commodities Group. 0 2004 2005 2006 2007 2008 2009 Sources: World Bank and International Energy Agency. revenues for all exporters to drop from $755 billion in 2008 to $485 billion in 2009--a de- cline equivalent to 30 percent of the group's nonetheless a substantial 12.5 percent of GDP. GDP (figure A19). For the developing export- Current account surplus positions fell sharply ers, the decline in revenues was less severe, but across the region for all oil exporters, from 141 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 revenues plummeted 40 percent. This placed Figure A20 Reduced oil revenues cut current account surplus positions in 2009 substantial pressures on budget revenues, which normally support domestic demand. In- Current account balance for Middle East oil exporters flation continues at rates near 20 percent, and Percent of GDP the current account surplus fell from 22 per- 45 40 cent of GDP in 2008 to 7.5 percent in 2009. 35 Iraq is facing a major short-term financing gap 30 in the year ahead owing to the global slowdown, 25 with a fiscal deficit of 26 percent of GDP ac- 20 crued in 2009 and a current account that moved 15 10 from substantial surplus in 2008 (13.3 percent of 5 GDP) to major deficit in 2009 (31 percent). 0 The diversified economies. The Euro Area is the po st le C a t Ira rter g .o c ria ai ep i bi o in C R lam ex Ea idd w ge rte oil n, s f ra xp lop G Ku lM iA Al Is rs le e destination for more than 70 percent of export oi ev ud Al D Sa goods from the diversified economies of the 2008 2009 Middle East and North Africa region. Moreover, the Euro Area is also the host for overseas work- Sources: Word Bank data and estimates. ers from the Maghreb and Mashreq and an im- portant source of remittance flows and tourism arrivals to the developing region. As investment 25 percent to 7.3 percent of GDP between and trade plummeted in key Euro Area econo- 2008 and 2009, and from 19.7 percent to mies, GDP for the zone declined to 0.5 percent 3.3 percent for developing oil exporters (fig- growth in 2008, and it is anticipated to contract ure A20). With public expenditure growing by a sharp 3.9 percent in 2009, the deepest reces- at a rapid pace, fiscal deficits for developing sion since WWII. exporters increased sharply during 2009, to The effects of the European downturn on 11 percent of GDP in Algeria (though well cov- exports from the region have been dramatic, ered by reserves of some $150 billion), 5.5 per- with Egypt's merchandise exports declining cent in Syria, 3.8 percent in the Islamic Republic from growth of 33 percent in 2008 to minus of Iran, and 2 percent in the Republic of Yemen. 15 percent by July 2009 (year-on-year). Simi- GDP growth in Algeria slowed to 2.1 percent lar patterns of export decline were registered in 2009 from 3 percent in 2008. A 2 percent de- in Jordan, Morocco, and Tunisia (figure A21). cline in the oil sector was partly offset by non- Together with only modest declines in imports oil activity, which increased by 5.7 percent, (supported by stimulus measures), the current supported by construction and services tied account position for the group deteriorated to a long-running infrastructure development from a deficit of 2.1 percent of GDP in 2007 plan (PIP). The program has continued to be to 5.2 percent by 2009. Deficits during 2009 implemented in part as a stimulus measure, varied between 2.5 percent of GDP in Egypt and in early 2009 the government announced and Tunisia, 5.8 percent in Morocco, and it would put about $60 billion from its oil- 7.0 percent in Jordan. linked fiscal surplus toward the investment Slackening economic activity and wors- program. Even though partial national ac- ening labor conditions in Europe, as well as counts data for 2009 are not available for the across the GCC economies over the course of Islamic Republic of Iran, growth is estimated 2009 caused worker remittances flows into the to have slowed to 1 percent in 2009, from 2.5 developing region to decline by 6.3 percent percent during 2008, as crude oil production for the year--in contrast to the strong gains contracted 7.3 percent and oil and gas export of 23.0 and 11.3 percent in 2007 and 2008, 142 R E G I O N A L E C O N O M I C P R O S P E C T S Figure A21 Middle Eastern exports declined Figure A23 Tourism receipts fall from record sharply as European demand fell 2008 performance, but modest recovery likely Import, export values (US$), percentage change, three-month Tourism receipts, US$ (billions) average, year over year 30 125 25 Egypt exports 100 20 15 75 10 50 Morocco exports 5 25 0 Euro area imports 2004 2005 2006 2007 2008 2009 0 Tunisia exports Egypt, Arab Rep. of Morocco 25 Tunisia Jordan 50 Sources: World Bank, IMF, World Tourism Organization. 09 09 08 09 08 08 20 20 20 20 20 20 p. ay ay n. p. n. Se Ja Se Ja M M Sources: Haver Analytics; World Bank. Tourism receipts are a key source of foreign currency (equivalent to 14 percent of GDP for the diversified economies of the region). respectively (figure A22). Among the larger re- With Europe suffering increasing unemploy- cipient countries, Egypt appears to have been ment rates, faltering wage growth, and efforts most adversely affected, with flows declining by households to repair balance sheets badly 9 percent, while Morocco experienced an damaged by the financial market meltdown of 8 percent drop in receipts. Jordan, Lebanon, 2008, tourism receipts are estimated to have and Tunisia experienced lesser declines, vary- declined by 5 percent during 2009, following ing between 1 and 3 percent. strong gains in the 20 percent range since 2006 (figure A23). Tunisia appears to have bucked the downtrend with a gain of 4 percent. But Figure A22 Worker remittances fell by a declines elsewhere range from 8 percent in moderate 6.3 percent in 2009 Morocco to 3 percent in Egypt. Workers' remittance receipts, US$ (billions) Foreign direct investment (FDI) inflows to the 35 diversified group, which is increasingly sourced 30 from the GCC economies, fell to 4.3 percent of GDP in FY09 from 8.1 percent a year earlier. 25 Morocco and Tunisia registered a 35 percent de- 20 cline in inflows during calendar year 2009, while 15 FDI in Jordan dropped by 80 percent during the 10 first quarter of 2009. These declines reflect the 5 substantial deterioration of financial conditions in the wake of the Dubai World debt-payment 0 2004 2005 2006 2007 2008 2009 standstill, inducing GCC economies to scale back on current investment projects and putting Morocco Egypt, Arab Rep. of Jordon earlier planned FDI endeavors on hold. Tunisia Lebanon Algeria In addition to pressures on FDI, the Dubai Yemen, Rep. of Syrian Arab Rep. financial crisis may have adverse consequences Source: World Bank; IMF; national agencies. for the countries of the Mashreq (Jordan, Leb- anon, and Syria), which hold particularly close 143 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 ties to the GCC. Lower investment within the GCC portends fewer job opportunities for Figure A24 Oil price and export revival hold key to recovery in the Middle East and workers from these countries, lower remit- North Africa tances, and consumption in home markets. Growth of real GDP, percent Growth in Egypt slowed to 4.7 percent in 8 FY09, from 7 percent during the three previous years. The slowdown was driven by lower exter- 6 nal demand with exports of goods and services declining by 25 percent; growth was negative 4 in economic sectors with a strong exposure 2 to external markets such as the Suez Canal (down by 7.2 percent, compared with 18 per- 0 cent growth in FY08) and hotels, restaurants, and related activities linked to tourism (down 2 Developing Oil exporters Diversified GCC by 1.3 percent in real terms compared with Middle East economies 30 percent growth). Declining fixed investment 2008 2009 2010 2011 (down 10 percent compared with 14.8 per- cent growth a year ago) has moved in tandem Source: World Bank. with increases in unemployment, which rose to 9.4 percent from 8.4 percent a year earlier. In response, the government implemented a the regional profile masks both the diversity of crisis stimulus plan featuring fiscal, monetary, performance across countries and the driving and direct support measures in the form of forces for growth. LE 15 billion in additional spending, includ- Oil prices are expected to remain broadly ing higher subsidies and social benefits. On the stable over the forecast period, at around monetary side, the Central Bank of Egypt cut $75 a barrel. Stronger global activity should policy rates six times between February and allow for crude oil and gas production to re- September 2009, reducing overnight deposit turn to positive growth, implying moderate and lending policy rates by 325 and 275 basis revenue gains. As a result, current account points, respectively. positions for developing oil exporters are projected to stabilize near 5 percent of GDP by 2011. GDP growth for developing oil ex- Medium-term outlook porters should reach 3.1 and 3.7 percent, re- Following the tortuous conditions of 2009, spectively, in 2010 and 2011 (figure A24). prospects for both the developing and high- By 2011 growth will vary from 3 percent in income economies of the Middle East and the Islamic Republic of Iran to 5.5 percent in North Africa should improve through 2011. Syria, grounded in developments in non-oil Growth is projected to increase to 4.4 percent sectors and in investment in hydrocarbons by that year, the same pace registered on aver- capacity (table A8). age between 1995 and 2005. Though domes- GDP for the high-income GCC economies tic absorption will be a continuing source of is anticipated to increase by 3.2 percent in strength, the forecast for regional recovery is 2010 and 4.1 percent in 2011, as oil produc- premised on a revival in global oil demand tion firms and a higher average oil price help and a rebound in key export markets. Despite to restore revenues, albeit in more moderate the gradual withdrawal of fiscal stimulus mea- increments. Current account surplus posi- sures, moderate advances in consumer and tions for the group are expected to rebound capital spending are expected to underpin the from 11 percent of GDP in 2009 to 14.5 per- strengthening of growth (see table A7). But cent by 2011, providing a means to support 144 R E G I O N A L E C O N O M I C P R O S P E C T S Table A8 Middle East and North Africa country forecasts (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009c 2010d 2011d Algeria GDP at market prices (2005 US$)b 4.0 2.0 3.0 3.0 2.1 3.9 4.0 Current account balance/GDP (%) 8.2 25.0 22.4 20.8 3.4 2.7 5.6 Egypt, Arab Rep. of GDP at market prices (2005 US$)b 4.4 6.8 7.1 7.2 4.7 5.2 6.0 Current account balance/GDP (%) 0.4 2.4 0.3 0.9 3.2 3.5 3.2 Iran, Islamic Rep. of GDP at market prices (2005 US$)b 4.8 5.9 7.8 2.5 1.0 2.2 3.2 Current account balance/GDP (%) 7.3 9.2 12.0 22.2 7.5 3.6 3.2 Jordan GDP at market prices (2005 US$)b 4.7 8.0 8.9 7.9 3.2 3.9 4.5 Current account balance/GDP (%) 0.0 10.8 16.7 11.4 10.1 9.7 9.2 Lebanon GDP at market prices (2005 US$)b 3.2 0.6 7.5 8.5 7.0 7.0 7.0 Current account balance/GDP (%) 20.0 11.3 11.1 20.5 14.5 15.2 14.2 Morocco GDP at market prices (2005 US$)b 4.4 7.8 2.7 5.6 5.0 3.0 4.4 Current account balance/GDP (%) 0.7 2.0 0.3 5.4 5.9 5.7 5.2 Syrian Arab Republic GDP at market prices (2005 US$)b 3.2 5.1 4.2 5.2 3.0 4.0 5.5 Current account balance/GDP (%) 2.9 2.8 3.3 4.0 3.2 4.3 4.0 Tunisia GDP at market prices (2005 US$)b 5.0 5.7 6.3 4.5 3.3 3.8 5.0 Current account balance/GDP (%) 3.0 2.0 2.6 4.2 3.5 2.6 2.0 Yemen, Rep. of GDP at market prices (2005 US$)b 4.9 3.2 3.3 3.6 4.2 7.3 4.5 Current account balance/GDP (%) 3.1 1.1 7.0 5.6 5.2 2.3 2.5 Source: World Bank. Note: World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries' pros- pects do not significantly differ at any given moment in time. Djibouti, Iraq, Libya, and West Bank and Gaza are not forecast owing to data limitations. a. Growth rates over intervals are compound average; growth contributions, ratios, and the GDP deflator are averages. b. GDP measured in constant 2005 U.S. dollars. c. Estimate. d. Forecast. domestic growth while once more accumu- be driven by domestic demand, with the help lating international reserves. A rekindling of fiscal and monetary stimulus measures, as of interest in regional FDI may emerge as external contributions fade. The anticipated financial and economic conditions begin to normalization of agriculture in Morocco (fol- normalize. lowing the post-drought boom of 2009) will Economic recovery in Europe and among be a drag on growth in 2010, and gains for the the GCC countries will be supportive of a re- diversified group are projected to pick up to vival for the diversified economies, suggesting 4.5 percent in 2010 and 5.4 percent in 2011, a resumption of export growth, a rebound in respectively. remittances and various services receipts, and improvement in business expectations, lead- Risks ing to a revival in capital spending. GDP gains The broadly favorable outlook for the Mid- in Jordan, Morocco, and Tunisia are likely to dle East and North Africa over 2010­11 145 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 remains subject to substantial downside 6.0 percent in 2009, which was largely driven risks, which would pose additional chal- by a pronounced decline in investment growth lenges to policy makers already grappling and, to a lesser extent, private consumption. with the current crisis. A deeper and more While exports contracted sharply with exter- protracted global recession (the deeper nal demand, the decline in imports was steeper, growth recession discussed in chapter 1) and net trade actually supported growth on cannot be ruled out. Within the region, po- the regional level. As the crisis took hold, eq- litical tensions remain a constant, tending uity markets and exchange rates plunged in to restrain international capital flows that most countries in the region. Sovereign bond might otherwise contribute to a deepening of spreads spiked with the contraction in capital capital markets and private investment. Fur- flows, as both domestic and international inves- ther, needed reform efforts, some initiated tors sought safe-haven assets outside the region. during the crisis period, could receive less Although the global financial crisis had a attention and commitment once economic sharp negative impact on South Asia, the slow- conditions start to normalize. down in regional GDP growth was the least The recent difficulties of Dubai World hold- pronounced among all developing regions. ing company--an entity of the Government This partly reflects the relatively closed na- of Dubai, United Arab Emirates--in asking ture of the region's economies. Private capi- its creditors for a six-month standstill on all tal inflows--a key transmission channel of scheduled debt payments, indicates that finan- the crisis--are less significant as a share of cial institutions in the region were not entirely South Asia's GDP (particularly foreign direct unaffected by the global financial crisis. Given investment), compared with most other re- the very high investment levels of the past sev- gions. Economic activity in South Asia is also eral years, as well as asset inflation (property less specialized in manufacturing and natural prices increased particularly sharply in Egypt resources--sectors that have been particularly and Morocco), there may be additional large- negatively affected by the crisis. Correspond- scale financial losses that have yet to be real- ingly, the region's greater reliance on services ized. Though a systemic crisis in Dubai will trade--roughly double the 7.7 percent aver- likely be averted thanks to the diversified age share of GDP for developing countries in holdings of the Dubai government and emer- 2008--also provided a buffer to the crisis, as gency support from the emirate of Abu Dhabi services tend to be more resilient during down- (both bilaterally and through the federal au- turns (although smaller countries with impor- thorities), it may have an adverse impact on tant tourism sectors, such as the Maldives, the balance sheets of local and regional banks were hit hard). Domestic demand in the region holding Dubai World debt. The financial was relatively resilient, having been cushioned problems facing Dubai, along with previous by countercyclical macroeconomic policies. defaults by two large Saudi private companies, Interest rates were rapidly cut across most will continue to raise concern amidst the need economies. Although fiscal space in most econ- for comprehensive corporate governance and omies was limited, substantial fiscal stimulus debt restructuring reforms in the region. measures were introduced in India (including pre-election spending), Bangladesh and Sri Lanka (in the form of incentives and safety South Asia net expenditures). Relatively robust, albeit moderating, regional remittance inflows have Recent developments been supportive, particularly in Bangladesh, T he global financial crisis contributed to a marked deceleration in real GDP growth in South Asia, from 8.7 percent in 2007 to Nepal, and Sri Lanka, where they continue to represent over 5 percent of GDP. Real incomes were also boosted by the collapse in global 146 R E G I O N A L E C O N O M I C P R O S P E C T S commodity prices--particularly for food and Figure A25 Gross capital inflows to South fuel, which represent a large share of regional Asia are recovering but remain below household outlays. pre-crisis levels The extent of the downturn in the indi- US$ (billions) per quarter US$ (billions) per quarter vidual economies has been mixed and reflects 140 16 initial conditions. Growth has been weakest 120 14 in countries that entered the crisis with large 100 12 internal and external imbalances and that 80 10 were forced to severely constrain domestic de- 60 8 6 mand, such as the Maldives, Pakistan, and Sri 40 4 Lanka. Countries that entered the crisis with 20 2 stronger fundamentals, such as Bangladesh 0 0 Bhutan, and India, weathered the crisis better. 1 2 3 4 1 2 3 4 Q Q Q Q Q Q Q Q 08 08 08 08 09 09 09 09 A number of regional economies also faced 20 20 20 20 20 20 20 20 ongoing internal conflicts that continued to dis- Total gross flows to emerging markets excluding rupt economic activity, notably Afghanistan, South Asia (left axis) Pakistan, Sri Lanka (which ended a decades- Total gross flows to South Asia (right axis) old civil war in mid-2009), and to a lesser ex- Sources: Dealogic and World Bank. tent Nepal (where warring factions reached a peace accord in late 2006, but are still vying for political control). The stabilization and progressive thaw- trends across developing countries (figures A25 ing of global financial markets in early 2009 and A26). This process has been supported and the rebound of world trade and output by improved investor sentiment on com- growth beginning in the second half of 2009 paratively strong growth outturns (India and have contributed to improving conditions in Bangladesh), ongoing or new International South Asia. Since the second quarter of 2009, Monetary Fund (IMF) stabilization programs local equity markets and capital inflows to the (Pakistan, Sri Lanka, and most recently the region began to recover--largely in line with Maldives), steep reductions in interest rates, Figure A26 Local equity markets have generally returned to pre-crisis levels MSCI indexes (US$), Jan. 1, 2009 = 100 500 450 400 350 300 250 200 150 100 50 0 Jul. Sep. Nov. Jan. Mar. May Jul. Sep. Nov. Jan. Mar. May Jul. Sep. Nov. Jan. 2007 2007 2007 2008 2008 2008 2008 2008 2008 2009 2009 2009 2009 2009 2009 2010 India Emerging markets Sri Lanka Bangladesh Pakistan Sources: Bloomberg and World Bank. Note: MSCI = Morgan Stanley Capital International. 147 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 and improved political stability. While the region experienced a sharp decline in gross Figure A27 Recovery in industrial produc- tion has taken hold in South Asia ahead of capital inflows during the first half of 2009, most other regions portfolio inflows surged in the fourth quarter Annual percent change and bond issuance and syndicated bank lend- 20 ing jumped in the fourth quarter, such that 15 total gross inflows firmed in 2009 to an esti- 10 mated $31 billion. As inflows to other regions 5 shrank (particularly in Europe and Central 0 Asia), South Asia's share of total capital in- 5 flows to developing countries rose to 8.9 per- 10 15 cent in 2009 from 6.7 percent in 2008. While 20 most local stock exchanges have recovered to Jan. Apr. Jul. Oct. Jan. Apr. Jul. Oct. pre-crisis levels, the majority remain well below 2008 2008 2008 2008 2009 2009 2009 2009 peak levels posted in late 2007 and early 2008 Latin America and Caribbean South Asia (in both local currency and U.S. dollar terms). Middle East and North Africa East Asia In Bangladesh, where capitalization of listed Sub-Saharan Africa and Pacific Europe and companies (relative to GDP) is lower than in High-income countries Central Asia its neighbors and where foreign participation Sources: Thomson Datastream, IMF and World Bank. is limited, the equity market remained stable during the crisis and posted strong growth in recent months. Sri Lanka's equity market is also an exception, with a recovery to pre-crisis However, a poor monsoon season in India in highs of 2007 supported by the improvement 2009 suggests that agricultural growth will in sentiment following the end of the civil war be modest in the current 2009­10 crop year and the formal standby arrangement reached (which began in late 2009). Regional services with the IMF in mid-2009. activity decelerated with the decline in global Regional industrial activity, which did not tourism, hitting the Maldives, Nepal, and Sri contract as much as in most other develop- Lanka, in particular, where tourism is a key ing regions, has shifted into positive growth, sector. In contrast, in India, services activity led by India, Bangladesh, and more recently was supported by resilient outsourcing. Pakistan. Fiscal stimulus measures have sup- Merchandise trade growth remains below ported the rebound in output by helping to previous-year levels for the region, with im- revive consumer demand. Further, continued ports down much more sharply than exports, robust remittance inflows boosted construc- given the sharp compression of demand in tion activity, especially in Bangladesh and Maldives, Pakistan, and Sri Lanka in partic- Nepal. The recovery in regional output is ular. Indeed, the 32 percent decline in South ahead of most other developing regions--with Asia's import volumes through July 2009 the exception of East Asia and the Pacific-- compared with the previous year is the second and of high-income countries (figure A27). steepest among developing regions after that of Regional agricultural output was buoyed by Europe and Central Asia (39 percent). In con- a good monsoon in 2008 that contributed trast, the decline in the region's merchandise to a good harvest in 2009 across much of the export volumes was less severe than in most region. One exception is Afghanistan, where ag- other developing regions, with the exceptions ricultural output contracted sharply (16.5 percent of East Asia and the Pacific and Latin America in FY2008/09). In Sri Lanka the agricultural and the Caribbean. This partly reflects the low sector benefited from the end of fighting and manufacturing and commodity content (sec- from acreage brought back into production. tors particularly hard hit by the recession) of 148 R E G I O N A L E C O N O M I C P R O S P E C T S the region's exports. Some sectors also dem- upswing in government outlays and a surge in onstrated marked resilience during the crisis, imports for resort-related construction mate- such as ready-made garments in Bangladesh, rials contributed to the sharp deterioration in where competitive pricing has enabled produc- the current account balance. ers to build market shares (i.e., the "Wal-Mart While the adjustment was less stark, India effect") and in Sri Lanka, where long-term also posted a shrinking current account defi- strategic partnerships with mid- to high-end cit in 2009, as imports fell faster than exports. retailers in the United States and the European Bangladesh and Nepal recorded rising current Union, (such as Victoria's Secret, Diesel, and account surpluses, as the moderation of export Nike) created a buffer, and in India, where growth was less pronounced than the decline information technology software also proved in imports, supported by continued firm remit- relatively resilient. tances inflows. In contrast, Bhutan's current Overall, the combination of a sharp fall in the account deficit is estimated to have grown from value of imports, a somewhat less steep decline in 10 percent of GDP in 2008 to 12.3 percent in exports (both reflecting favorable terms-of-trade 2009, partly reflecting the start of interest pay- developments), and resilient remittance inflows ments for the Tala hydropower scheme (figure meant that current account balances generally A28). Afghanistan's current account deficit, improved in 2009 (figure A28). Regional ex- including official transfers (equivalent to some ternal positions had come increasingly under 50 percent of official GDP) is estimated to have strain from the multiyear boom in food and fuel shifted from a surplus of 0.9 percent of GDP in prices before mid-2008. During 2009, the Mal- 2008 to a deficit of 1.6 percent in 2009. dives, Pakistan, and Sri Lanka posted the largest Remittance inflows--a key source of for- adjustments in their current account deficits. eign exchange for the region--declined in 2009, Domestic demand was sharply compressed in pushed down by the decline in economic activ- the three economies, where large fiscal deficits ity and the rise in unemployment in migrant- had contributed to the buildup of consider- host countries. However, remittance inflows able external imbalances before the crisis. The remained relatively strong compared with other Maldives is an extreme case, where a massive sources of foreign exchange, and indeed are above their 2007 levels (figure A29). Remittance inflows to South Asia contracted by a modest Figure A28 Current account balances improve across much of South Asia in 2009 Percent of GDP Figure A29 International flows to South Asia 6 4 Percentage point change between 2007 and 2009 (as a share 4 of GDP) 0 1.0 2 0.5 4 6 0 8 10 0.5 12 1.0 14 an n ka a sh al 1.5 ta di ep an de ut In is N Bh k iL la Pa ng Sr Ba 2.0 Equities Bank Bonds FDI Remittances 2007 2008 2009e lending Source: World Bank. Source: World Bank. 149 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 1.8 percent in 2009, compared with a 7.5 per- Figure A30 Budget deficits in most South cent decline for developing countries excluding Asian countries outstrip developing- South Asia (World Bank 2009). Growth in the country average Arabian Gulf and East Asian economies, which host a significant share of South Asia's migrant Maldives workers, has not been as adversely affected as India growth in other key host economies, such as the High-income countries United States, the European Union, and Rus- Bhutan sia. Among South Asia's economies, India--the largest recipient of remittances in the world in Sri Lanka dollar terms--posted a contraction in remittance Pakistan inflows in 2009, while Bangladesh, Nepal, Paki- Developing stan, and Sri Lanka, experienced a slower pace countries of growth of remittances inflows. Bangladesh With the moderation in demand and collapse Nepal in energy prices, inflationary pressures across 0 5 10 15 20 25 30 35 the region subsided following the onset of the Percent share of GDP, ranked by 2009 estimates crisis, particularly in the first half of 2009. This helped reverse the buildup of inflationary 2008 2009e pressures that became increasingly evident in Source: World Bank. 2007 and 2008, as fuel and food prices spiked-- despite efforts by authorities to contain the price increases. Lower oil prices have eased nearly all of the regional economies. Even be- pressures on fiscal deficits stemming from fuel fore the crisis, sizable fiscal deficits were already price subsidies. The moderation in inflationary a problem for many South Asian economies, pressures and falling international commodity where weak tax administration and structure prices also provided scope for regional central resulted in low domestic resource mobilization banks to introduce expansionary measures to (figure A30). support domestic demand in response to the crisis. Bangladesh, India, Pakistan, and Sri Lanka cut policy interest rates. Activity in Medium-term outlook Bhutan and Nepal, where the currencies are South Asia's GDP growth is projected to tied to the Indian rupee, was supported by firm from an estimated 6 percent in 2009 to India's expansionary monetary policy stance. 7.0 percent in 2010 and 7.4 percent in 2011. Regional fiscal positions deteriorated in 2009 External demand for goods and services is in response to a combination of reduced tax anticipated to recover, while improving con- receipts resulting from the decline in economic sumer and business confidence, combined with activity and higher outlays. Corresponding to the lagged effects of expansionary monetary the introduction of more accommodative mon- and fiscal policy measures and a positive turn etary policies, expansionary fiscal policy mea- in the inventory cycle, should contribute to sures were introduced in Bangladesh, India, and strengthening domestic demand. A projected Sri Lanka to support domestic demand through firming of capital inflows will also support various expenditure and incentive programs. regional economic activity. The regional cur- Pakistan also sought to stimulate its economy rent account deficit is projected to rise mod- through an increase in its public sector develop- estly, from 1.2 percent in 2009 to 2.2 percent ment program. While these stimulus measures in 2010 and 2.4 percent in 2011, a result of helped offset the negative effects of the global firming domestic demand that is expected to crisis, they also led to higher fiscal deficits in drive import growth ahead of export growth. 150 R E G I O N A L E C O N O M I C P R O S P E C T S Table A9 South Asia forecast summary (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009g 2010h 2011h GDP at market prices (2005 US$)b,f 6.0 9.0 8.5 5.7 5.7 6.9 7.4 GDP in calendar year basisc 6.1 9.3 8.7 6.9 6.0 7.0 7.4 GDP per capita (units in US$) 4.1 7.3 6.8 4.2 4.3 5.5 6.1 PPP GDPd 6.0 9.0 8.5 5.7 5.7 6.9 7.4 Private consumption 4.7 6.0 7.0 2.7 4.2 6.0 6.7 Public consumption 5.0 9.9 5.6 21.1 7.1 7.3 7.2 Fixed investment 8.0 14.6 13.6 7.6 4.1 9.7 10.2 Exports, GNFSe 11.3 17.7 3.5 10.5 4.6 10.3 12.1 Imports, GNFSe 10.6 22.7 6.8 14.9 6.9 10.6 12.3 Net exports, contribution to growth 0.2 1.7 1.0 1.7 0.9 0.6 0.7 Current account balance/GDP (%) 0.6 1.5 1.3 3.3 2.3 3.2 3.4 GDP deflator (median, LCU) 5.9 5.2 7.4 7.2 13.8 6.9 6.5 Fiscal balance/GDP (%) 8.1 5.1 5.7 8.9 9.5 8.6 7.8 Memo items: GDPf South Asia excluding India 4.5 6.4 6.0 3.9 4.4 4.1 4.8 India 6.4 9.7 9.1 6.1 6.0 7.5 8.0 Pakistan 4.1 6.2 5.7 2.0 3.7 3.0 4.0 Bangladesh 5.3 6.6 6.4 6.2 5.9 5.5 5.8 Source: World Bank. a. Growth rates over intervals are compound average; growth contributions, ratios, and the GDP deflator are averages. b. GDP measured in constant 2005 U.S. dollars. c. GDP figures are presented in calendar years (CY) based on quarterly history for India. For Bangladesh, Nepal, and Pakistan, CY data is calculated taking the average growth over the two fiscal year periods to provide an approximation of CY activity. d. GDP measured at PPP exchange rates. e. Exports and imports of goods and nonfactor services. f. National income and product account data refer to fiscal years (FY) for the South Asian countries with the exception of Sri Lanka, which reports in calendar year (CY). The fiscal year runs from July 1 through June 30 in Bangladesh and Pakistan, from July 16 through July 15 in Nepal, and from April 1 through March 31 in India. Because of reporting practices, Bangladesh, Nepal, and Pakistan report FY2007/08 data in CY2008, while India reports FY2007/08 in CY2007. g. Estimate. h. Forecast. Although regional GDP growth is projected contribute to an easing of inflationary pres- to accelerate, a return to boom-period growth sures by 2011 across the region. Further, given rates is not anticipated over the forecast hori- strong aversion to food price inflation within zon, as investment growth is expected to con- the region, monetary authorities are particu- tinue to be constrained by supply bottlenecks larly responsive to signs of inflationary pres- and higher capital costs in the wake of the sures building. crisis. (table A9). External demand is expected The recovery path for the individual econo- to firm, but it too will expand less quickly mies will vary substantially (table A10). India, than during the boom years. The regional fis- Bangladesh, and Bhutan are expected to emerge cal deficit is projected to narrow on planned from the global crisis with stronger growth per- structural fiscal consolidation and cyclical fac- formances, backed by generally sound economic tors, as well as a reversal of stimulus measures policies and greater resilience of trade, invest- introduced to support demand during the cri- ment, and remittances. Sri Lanka is also forecast sis. Nevertheless, the aggregate regional fiscal to post a relatively firm recovery, supported by deficit is projected to continue to exceed its the recent surge in capital inflows and improve- pre-crisis 2007 deficit of 5.7 percent. ment in investor confidence following the cessa- Expected progressive tightening of mon- tion of fighting after nearly three decades of civil etary conditions over the forecast horizon will war. Elsewhere in the region, conflict-affected 151 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Table A10 South Asia country forecasts (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009c 2010d 2011d Calendar year basis Bangladesh GDP at market prices (2005 USD)b 5.0 6.3 6.5 6.3 6.1 5.7 5.7 Current account balance/GDP (%) 0.6 2.0 1.3 1.4 3.1 1.8 1.5 India GDP at market prices (2005 USD)b 6.7 9.9 9.3 7.3 6.4 7.6 8.0 Current account balance/GDP (%) 0.4 1.1 0.7 2.6 2.4 3.5 3.6 Nepal GDP at market prices (2005 USD)b 3.9 3.4 3.5 4.3 5.0 4.3 4.3 Current account balance/GDP (%) 3.1 0.1 1.5 3.7 4.4 1.0 1.4 Pakistan GDP at market prices (2005 USD)b 3.7 6.9 5.9 3.8 2.9 3.3 3.5 Current account balance/GDP (%) 1.0 5.7 6.3 10.2 5.2 4.1 4.6 Sri Lanka GDP at market prices (2005 USD)b 4.5 7.7 6.8 6.0 3.6 5.0 6.0 Current account balance/GDP (%) 3.2 5.8 4.7 9.1 1.7 2.7 2.9 Fiscal year basis Bangladesh Real GDP at market prices 5.3 6.6 6.4 6.2 5.9 5.5 5.8 Current account balance/GDP (%) 0.8 1.3 1.4 0.9 2.8 2.2 1.5 India Real GDP at market prices 6.4 9.7 9.1 6.1 6.0 7.5 8.0 Current account balance/GDP (%) 0.2 1.2 1.0 1.4 2.6 3.1 3.3 Nepal Real GDP at market prices 4.1 3.7 3.3 5.3 4.7 4.0 4.5 Current account balance/GDP (%) 0.2 2.3 0.1 3.2 4.6 2.4 1.1 Pakistan Real GDP at market prices 4.1 6.2 5.7 2.0 3.7 3.0 4.0 Current account balance/GDP (%) 0.8 4.0 4.9 8.6 5.9 4.5 4.1 Source: World Bank. Note: World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries' pros- pects do not significantly differ at any given moment in time. Afghanistan, Bhutan, and the Maldives are not forecast owing to data limitations. National income and product account data refer to fiscal years (FY) for the South Asian countries with the exception of Sri Lanka, which reports in calendar year (CY). The fiscal year runs from July 1 through June 30 in Bangladesh and Pakistan, from July 16 through July 15 in Nepal, and April 1 through March 31 in India. Because of reporting practices, Bangladesh, Nepal, and Pakistan report FY2007/08 data in CY2008, while India reports FY2007/08 in CY2007. GDP figures are presented in calendar years (CY) based on quarterly history for India. For Bangladesh, Nepal, and Pakistan, CY data is calculated taking the average growth over the two fiscal year periods to provide an approximation of CY activity. a. Growth rates over intervals are compound average; growth contributions, ratios, and the GDP deflators are averages. b. GDP measured in constant 2000 U.S. dollars. c. Estimate. d. Forecast. countries--Afghanistan, Pakistan, and to a boost private consumption and support lesser extent, Nepal--are expected to face more growth--particularly in Bangladesh, Nepal, moderate growth outturns, as political uncer- Pakistan, and Sri Lanka. However, the re- tainty and fighting continue to disrupt economic covery in remittance growth is anticipated activity. not to take hold immediately as job growth Regional economies are projected to ben- typically lags output growth in high-income efit from stronger remittance inflows over markets--a lag that could be more extended the forecast horizon, which in turn should than usual given the synchronicity of the global 152 R E G I O N A L E C O N O M I C P R O S P E C T S downturn. However, the slowdown in growth in the Arabian Gulf and East Asia--South Figure A31 Interest payments represent a significant burden in South Asian economies Asia's key migrant destination countries--was Percent share of fiscal expenses in 2007 a generally less pronounced than in other labor- 30 importing countries, which is expected to allow a relatively rapid recovery in remittances inflows 25 to South Asia. 20 Risks 15 As the global economic recovery begins to take 10 hold in the second half of 2009, risks to the GDP growth forecast for South Asia have lessened. 5 Nevertheless, downside risks remain and center 0 on the extent of the upswing and durability of a sh a an d d d di nk st an ia an a an n In de La ki the global recovery. gla i Pa pe As st fric ica ea Ea h A er ribb an Sr ro ral Downside risks to the forecast are repre- B Eu nt e dl rt Am Ca C e id No tin M sented by the region's large fiscal imbalances La and its relatively high reliance on trade taxes. Source: World Bank. An extended period of weak external demand a. Pakistan data are from 2008. would likely erode these revenues and increase pressures on government coffers. The region's large fiscal imbalances also represent a poten- tial drag on long-term growth by crowding out private investment through the public sector's surprise to the downside, pointing to down- large financing requirement and higher inter- side risks for South Asian migrants working est rates. Interest payments in South Asia rep- in the Gulf and reduced remittances flows to resented 21.7 percent of central government their home countries. Correspondingly, should expenditures in 2007, more than double the a significant portion of the stock of expatriate share represented in other developing regions workers return home with accumulated sav- (figure A31). By reducing the large fiscal defi- ings due to the downturn in the Gulf, near- cits and payment obligations, regional gov- term remittances inflows might rise. ernments could free up resources to devote to Overheating is also a risk. Should the recent development spending. The region has a very surge in capital inflows to developing countries low tax base compared with other developing (see chapter 1) be sustained, they could lead regions, so improving tax collection would to ballooning asset markets and appreciation help alleviate fiscal pressures. Similarly, re- of currencies (with the latter hindering export vamping the tax structure (including introduc- prospects), creating challenges for monetary tion of value-added taxes in some countries) authorities. Failure to mop up excess liquidity could help boost revenue mobilization. in banking systems or to bring down the re- Remittances inflows--which provided a gion's large fiscal deficits could lead to higher cushion for the region--could fail to recover inflationary pressures. Separately, while global in the event of a prolonged global recession or rice markets appear well-supplied, and stock- a jobless economic recovery (potentially cou- to-use ratios have returned to more normal pled with tighter immigration controls). The levels (along with maize and wheat stocks), a debt payment problems of Dubai World in serious weather event or policy action could the United Arab Emirates that erupted in late- also cause prices to rise significantly, as only November 2009 suggest that economic ac- seven percent of global rice production is tivity in the Arabian Gulf economies could traded. 153 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Table A11 Sub-Saharan Africa forecast summary (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009e 2010f 2011f GDP at market prices (2005 US$)b 4.0 6.4 6.5 5.1 1.1 3.8 4.6 GDP per capita (units in US$) 1.4 3.9 4.0 3.1 0.8 1.9 2.7 PPP GDPc 4.0 6.4 6.5 5.2 1.6 4.1 4.9 Private consumption 2.0 7.0 8.1 3.5 0.4 3.2 4.5 Public consumption 5.2 5.8 5.8 5.6 5.6 5.2 5.2 Fixed investment 6.5 16.9 19.5 12.2 0.3 6.3 5.7 Exports, GNFSd 4.9 4.8 3.8 4.6 5.2 6.6 5.9 Imports, GNFSd 6.1 13.2 11.8 6.7 5.2 7.5 6.6 Net exports, contribution to growth 0.1 2.7 2.9 0.9 0.2 0.6 0.5 Current account balance/GDP (%) 1.7 0.7 0.1 0.1 3.4 2.5 2.4 GDP deflator (median, LCU) 7.3 7.3 7.6 9.7 6.2 6.1 4.1 Fiscal balance/GDP (%) 2.3 4.3 0.4 0.9 4.2 2.1 1.7 Memo items: GDP Sub-Saharan Africa excluding South Africa 4.5 6.9 7.1 5.9 2.8 4.8 5.6 Oil exporters 4.6 7.5 7.9 6.3 2.8 4.9 5.3 CFA countries 4.4 2.7 4.5 4.0 1.6 3.4 3.8 South Africa 3.3 5.6 5.5 3.7 1.8 2.0 2.7 Nigeria 4.6 6.2 6.3 5.3 4.3 4.8 5.1 Kenya 2.9 6.4 7.1 1.7 2.8 3.7 4.8 Source: World Bank. a. Growth rates over intervals are compound average; growth contributions, ratios, and the GDP deflator are averages. b. GDP measured in constant 2005 U.S. dollars. c GDP measured at PPP exchange rates. d. Exports and imports of goods and nonfactor services. e. Estimate. f. Forecast. Sub-Saharan Africa Figure A32 Middle-income and oil-exporting Recent developments countries of Sub-Saharan Africa hit hardest by global crisis T he global financial crisis has had a marked, negative impact on economic performance in Sub-Saharan Africa, affecting trade, foreign Percent growth 10 9 Oil exporters Low-income 8 direct investment, tourism, remittances, and 7 countries Middle-income countries official assistance. GDP is estimated to have 6 5 grown only 1.1 percent for the region as a 4 3 whole in 2009 (table A11). Notwithstanding 2 1 0.27 the severity of the shock, the improved mac- 0 1 roeconomic fundamentals in place in many 2 countries of the region as they entered the cri- 3 4 Fragile countries sis meant that the impact was less pronounced 5 than in other regions and relative to previous 90 92 94 96 98 00 02 04 06 08 19 19 19 19 19 20 20 20 20 20 external shocks. The growth slowdown has Source: World Bank. varied across countries in Sub-Saharan Africa, with oil exporters and middle-income coun- tries affected more severely, at least initially, than low-income, fragile, and less globally likely to have long-term consequences for Sub- integrated countries (figure A32). Per capita Saharan African countries, as more people fall GDP has declined by an estimated 0.8 percent into poverty in the region, according to Chen in 2009, the first decline in a decade. This is and Ravallion (2009); 30,000 to 50,000 more 154 R E G I O N A L E C O N O M I C P R O S P E C T S infants are likely to die of malnutrition in Figure A34 Current account balances of 2009, with a larger impact among infant girls middle-income and oil-exporting countries (Friedman and Schady 2009). deteriorated the most, while low-income As Sub-Saharan Africa is a major commod- countries remained aid-dependent ity exporting region, lower commodity prices, Percent of GDP declining export volumes, as well as lower tour- 15 ism revenues, and declining remittances have 10 all undermined income and private consump- tion, which decelerated to 0.4 percent growth 5 in 2009, down from 3.5 percent the previous 0 year. Weak external demand for commodities, excess capacity, scarce credit, and tight liquid- 5 ity all led to delays and scaling back of invest- 10 ment spending. Although FDI declined by 19 percent in 2009 the decline was more muted 15 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 than in other regions except South Asia, mainly because of sustained investment in the extrac- Low-income countries Middle-income countries tive sectors. Weak private consumption and Oil exporters investment resulted in lower imports, partially offsetting the negative growth impact emanating Source: World Bank. from the sharp contraction in export volumes (figure A33). As expected, the contraction in both ex- integrated into the global economy. Because port and import volumes was more severe of the marked declines in oil prices, the de- in middle-income countries, which are more terioration in current account balances was most pronounced in oil-exporting countries, where it fell from 9.7 to 1.4 percent of GDP (figure A34). Meanwhile lower tourism rev- Figure A33 Private consumption and trade contracted markedly in South Africa enues, remittances, and private current net transfers brought the current account bal- Percent 20 ances in middle-income countries to a deficit 10 of 1.2 percent of GDP--down from a surplus of 3.3 percent of GDP in 2008. Low-income 0 countries remain dependent on foreign assis- 10 tance to finance deficits of nearly 10 percent 20 of GDP. For most of these countries the terms- 30 of-trade boost from lower oil prices was offset 40 by lower export prices or volumes (primarily 50 metals and minerals, agricultural products), or 60 both. Indeed, the marked decline in oil imports 70 merely offset the decline in current account in- pt ate di nt m d rts rts flows, with current account balances improv- st fixe en e po po n ng t um riv sp nm en io Ex Im ve s P er in ros ing by less than 1 percent of GDP. ov G G ns co Sovereign spreads rose sharply in the 2008Q3 2008Q4 2009Q1 wake of the financial crisis but have declined 2009Q2 2009Q3 significantly since the first quarter of 2009. Source: South African Reserve Bank. In many cases, however, these spreads re- main above the pre-crisis level (figure A35). 155 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Seychelles, which were among the small number Figure A35 EMBI-stripped spreads retreat as investor confidence returns of countries that were able to implement signifi- cant countercyclical fiscal policies. A large part of Basis points 2,000 the deterioration in fiscal balances in commodity 1,800 exporters is linked to both lower volumes and 1,600 lower export prices for commodities. For the re- 1,400 gion as a whole, the fiscal balance deteriorated 1,200 1,000 from a surplus of 0.9 percent of GDP in 2008 800 to a deficit of 4.2 percent of GDP in 2009. Fur- 600 thermore, monetary policy remains ineffective in 400 bolstering domestic demand in many countries 200 0 owing to a lack of depth in the financial systems and weak transmission mechanisms. O . 20 8 N . 20 8 D . 20 8 Ja . 20 8 Fe . 20 8 M . 20 9 Ap . 20 9 M . 20 9 Ju . 20 9 n. 09 Au . 20 9 Se . 20 9 O . 20 9 N . 20 9 D . 20 9 .2 9 9 p 0 ct 0 ov 0 ec 0 n 0 b 0 ar 0 r 0 ay 0 Ju 200 g 0 p 0 ct 0 ov 0 ec 0 00 Se . 20 Lower food and oil prices since mid-2008 g l Au contributed to a sharp decline in headline infla- Ghana Gabon South Africa tion, which has eased to low single-digit levels Source: JPMorgan. in many countries (figure A36). By September 2009 several countries in the region were re- porting falling headline prices. However, in East Africa, as a result of high food inflation related Most countries in the region have been to recurrent droughts, inflation has remained spared from the most abrupt financial tur- stubbornly high. Subdued inflation in many bulences experienced by other regions, but countries in the Sub-Saharan Africa region has in countries that enjoyed rapid credit expan- created room for lowering interest rates. More- sion in the boom years (like Cape Verde, the over, lower food and energy prices have relieved Democratic Republic of Congo, Ethiopia, some of the pressure on fiscal balances, although Nigeria, Rwanda, Tanzania, Uganda, and sharply lower trade volumes and lower export Zambia), nonperforming loans will mount in the quarters ahead, putting strains on the shallow financial systems. Expectations about an imminent recovery Figure A36 Inflation in Sub-Saharan Africa in the global economy triggered a return of down to low single digits on lower food investors to stock markets across the world, prices boosting share prices. In South Africa share Median inflation, percent prices rose by 56 percent in dollar terms since 14 January 2009. The capital inflows have also 12 supported the rand, which gained 28.1 per- 10 cent against the U.S. dollar during the course of 2009. Furthermore, a more positive inves- 8 tor attitude toward risk taking in emerging- 6 market economies boosted inflows of direct 4 and portfolio investment during the second quarter. 2 On the policy front many countries in the 0 region had only limited space for countercycli- 06 06 07 07 08 08 09 09 20 20 20 20 20 20 20 20 cal measures, notwithstanding more prudent n. l. n. l. n. l. n. l. Ju Ju Ju Ju Ja Ja Ja Ja fiscal stances during the boom period. Auto- Source: World Bank. matic stabilizers worked in South Africa and the 156 R E G I O N A L E C O N O M I C P R O S P E C T S disruptions in oil production, GDP expanded Figure A37 Quarterly GDP readings point to 4.5 percent and 7.2 percent in the first two output stabilization in Sub-Saharan Africa quarters of the year, and growth remained Percentage change (saar) strong in the third quarter, largely on ac- 20 count of the non-oil sectors. Agriculture and 15 wholesale and retail trade made positive con- tributions, suggesting continued strength in 10 domestic demand. However poor performance 5 in the manufacturing, mining, and electricity- 0 generation sectors led to a decline in industrial output. The central bank's recent bailouts of 5 Nigeria's top five commercial banks under- 10 score the weakness of Nigeria's financial South Africa Kenya system. Together these banks account for 2008Q1 2008Q2 2008Q3 2008Q4 40 percent of all loans, 30 percent of deposits, 2009Q1 20092Q2 2009Q3 and 31.5 percent of total assets. These banks had a very large exposure to capital markets Source: Haver Analytics. and the gas and oil sectors, as well as a high level of nonperforming loans, thanks to poor corporate governance practices, lax credit ad- and import prices have slashed trade-related ministration, and nonadherence to the banks' government revenues, which has been par- credit risk management practices. The stock ticularly acute in the Southern Africa Customs market has been severely battered since the Union countries. onset of the global crisis, with share prices down In South Africa output contracted for 57.6 percent and market capitalization down three consecutive quarters, starting with the 50.7 percent in the first nine months of 2009. final quarter of 2008, then posted 0.9 percent Kenya's economic growth has been con- growth (saar) in the third quarter of 2009, strained by recurrent drought and ensuing elec- thereby ending the recession (Figure A37). tricity shortages while also suffering the effects Domestic demand remains weak, undermined of the global economic crisis. When compared by declining disposable income, higher unem- with a year earlier, growth decelerated from ployment, and high levels of debt. Growth in 4.3 percent in the first quarter, to 2.1 percent in both government consumption and fixed in- the second quarter, to ­ 0.1 percent in the third vestment deteriorated in the second quarter, quarter, in contrast with growth rates of above the latter partly attributable to more conser- 6 percent in 2007, before the election-related vative lending practices. Weaker domestic de- tensions. The drought-generated general crop mand, in particular postponements of capital failure affected food security, leading to higher expenditure by the private sector, led to a imports of basic foods. Major export crops sharp contraction in imports during the first fell, with tea output down 11.6 percent and half of 2009. This in conjunction with less horticulture output down 7.4 percent in the rapidly falling exports brought the second first eight months of 2009. Power shortages, quarter trade balance into surplus, and helped higher power costs, and weak global demand bring the current account deficit down to caused mining and quarrying, manufacturing, 3.2 percent of GDP in the third quarter from construction, and wholesale and retail trade to 7 percent of GDP in the first quarter. contract in the second quarter. On the bright Nigeria's economy, the second largest in side, transport and communications, as well as the region, seems to have weathered the cri- the hotel and restaurant sector rebounded as sis well. Notwithstanding lower oil prices and the effects of the postelection violence in early 157 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 2008 dissipated. Tourist arrivals rose 42 per- Growth performance has been stronger in cent in the first eight months of 2009. West and Eastern Africa: major economies In Ethiopia economic activity has been sup- in the regions have recovered and reform- ported by growth in the agricultural sector, oriented economies such as Burkina Faso, underpinned by an expansion of roads and Mali, Senegal, and Tanzania have turned better market access that have enabled sub- in relatively strong performances. In Côte sistence farmers to enter the commercial sec- d'Ivoire, which has been enjoying a peace tor. The economy was faced with significant dividend following the easing of political external shocks, however. The global reces- tensions, growth accelerated to above 3 per- sion caused remittances to fall by 6 percent in cent in 2009, as agricultural, mining, and the first half of 2009 relative to a year earlier, hydrocarbon output increased. In Central while merchandise exports fell 11 percent. In Africa, growth remained plagued by weak manufacturing, capacity utilization has been performances in the oil sectors of Cameroon affected by weak demand, shortages of water and Gabon. and electricity, insufficient raw materials and other inputs, and a shortage of capital. For- Medium-term outlook eign direct investment has also been affected, The recovery in growth is projected to be mod- making it more difficult to finance the large est and fragile, with output in Sub-Saharan current account deficit. Economic growth is Africa expected to accelerate to below-trend estimated to have decelerated to 7.2 percent in growth rates of 3.8 percent in 2010 and 2009, as remittances, investment, and export 4.6 percent in 2011. The growth pace will be growth weaken. Two new hydroelectric dams, well below the 6 percent growth rate recorded one commissioned in November 2009 and the during the boom years, as a result of lower real other to become operational in the next few commodity prices and slower global growth. months, will help ease power shortages and Excluding South Africa, the region is projected remove some of the growth constraints. to enjoy a modest acceleration in growth, from Lower demand for minerals is estimated 2.8 percent in 2009 to 4.8 and 5.6 percent in to have weakened performance in southern 2010 and 2011, respectively, as global growth Africa, and the region was also negatively af- recovers; however, this is still below the aver- fected by the recession in South Africa, with age 6.6 percent experienced during the boom which it has close trade, investment, and fi- years. The South African economy is ex- nancial links. Angola's economy also per- pected to recover modestly in 2010, growing formed poorly: oil output declined to below by 2.0 percent, before accelerating further to 1.8 million barrels a day, while falling oil rev- 2.7 percent in 2011. In per capita terms, GDP enues forced the government to cut back on in Sub-Saharan Africa is projected to grow investment spending, and private consumer 1.9 percent in 2010 and 2.7 percent in 2011. spending contracted. Lower demand for min- The rebound in economic activity will ing and hydrocarbon products pushed the primarily be fueled by a recovery in private Democratic Republic of Congo into recession demand, exports, and investment, with the in the last quarter of 2008, and this contraction largest contribution expected to come from was extended into the first half of 2009, when exports. However, the overall strength of the output dropped a cumulative 5.8 percent. Strong recovery will depend on the growth perfor- growth in the agriculture sector helped Mozam- mance in key export markets and investment bique's economic growth to stay at 5.9 percent partners, particularly the United States, the in the first quarter of 2009 notwithstanding a European Union, and China. The projected re- deceleration in growth in the services sector. bound in growth in these economies, fueled by Subsequently, growth accelerated to above the inventory cycle and impressive countercy- 6 percent in the second and third quarters. clical policies, is expected to result in stronger 158 R E G I O N A L E C O N O M I C P R O S P E C T S external demand for Sub-Saharan African ex- Sub-Saharan Africa have very limited social ports and should trigger a modest recovery in safety nets, which means that recovery of pri- investment flows. However growth in external vate consumption will be weaker than in other demand is expected to wane in the second half regions. Indeed private demand is projected to of 2010, as the growth impact of the inventory grow by 3.2 percent, partly fueled by higher restocking cycle and fiscal stimulus wanes. incomes in export-oriented sectors that benefit Stronger domestic demand will cause import from stronger external demand. growth to accelerate, with net exports con- Middle-income countries such as Botswana, tributing negatively ( 0.6 percent) to overall Seychelles, South Africa, and oil-exporting coun- growth. Furthermore, given that recovery in tries like Angola are likely to register the most global labor markets will lag, the recovery in dramatic turnaround from low bases owing tourism revenues and remittances is expected to weak performance in 2009 (table A12; fig- to be modest in 2010. Many countries in ure A38). Growth in middle-income countries Table A12 Sub-Saharan Africa country forecasts (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009c 2010d 2011d Angola GDP at market prices (2005 US$)b 8.3 18.6 20.3 13.2 0.9 6.5 8.0 Current account balance/GDP (%) 2.2 25.1 15.6 8.5 4.2 5.9 6.2 Benin GDP at market prices (2005 US$)b 4.6 3.8 4.6 5.0 3.1 3.3 4.8 Current account balance/GDP (%) 7.2 7.1 12.1 8.7 8.5 7.4 7.3 Botswana GDP at market prices (2005 US$)b 6.8 3.0 4.4 2.9 8.3 4.8 5.6 Current account balance/GDP (%) 8.1 17.6 15.6 7.8 7.3 7.2 7.7 Burkina Faso GDP at market prices (2005 US$)b 6.4 5.5 3.6 4.9 3.6 4.6 5.2 Current account balance/GDP (%) 10.1 11.5 8.7 10.4 9.8 9.5 10.2 Burundi GDP at market prices (2005 US$)b 0.4 5.1 3.6 4.4 2.6 3.7 5.1 Current account balance/GDP (%) 13.7 35.3 26.7 28.4 23.2 21.8 21.8 Cameroon GDP at market prices (2005 US$)b 4.2 3.2 3.3 3.1 1.4 2.6 3.2 Current account balance/GDP (%) 3.5 0.8 2.7 1.0 6.0 5.0 5.5 Cape Verde GDP at market prices (2005 US$)b 5.2 10.8 7.8 5.9 3.3 4.4 5.4 Current account balance/GDP (%) 10.1 6.9 13.5 18.3 23.1 22.3 19.9 Central African Republic GDP at market prices (2005 US$)b 0.7 4.0 3.7 2.2 2.4 2.8 2.7 Current account balance/GDP (%) 4.4 7.6 5.9 8.5 7.2 7.3 7.6 Chad GDP at market prices (2005 US$)b 8.6 0.2 0.2 0.2 0.8 2.7 3.0 Current account balance/GDP (%) 24.2 7.5 10.7 12.2 20.7 14.8 14.3 Comoros GDP at market prices (2005 US$)b 2.1 1.2 1.0 0.6 0.5 1.7 2.3 Current account balance/GDP (%) 6.3 5.5 6.8 11.8 8.2 8.2 8.5 Congo, Dem. Rep. of GDP at market prices (2005 US$)b 0.1 5.6 6.3 7.1 3.0 5.2 6.9 Current account balance/GDP (%) 1.7 4.0 2.7 14.5 13.6 12.8 12.0 (continued) 159 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 Table A12 (continued) (annual percent change unless indicated otherwise) 1995­2005a 2006 2007 2008 2009c 2010d 2011d Congo, Rep. of GDP at market prices (2005 US$)b 3.4 6.2 1.6 5.8 6.8 11.0 2.9 Current account balance/GDP (%) 2.2 1.6 9.3 2.6 9.3 3.2 0.9 Côte d'Ivoire GDP at market prices (2005 US$)b 1.6 0.7 1.6 2.3 3.2 4.0 4.1 Current account balance/GDP (%) 0.2 2.8 0.7 2.6 23.6 2.6 0.8 Eritrea GDP at market prices (2005 US$)b 1.7 1.0 1.3 1.2 1.5 4.2 4.3 Current account balance/GDP (%) 15.3 20.8 15.7 16.3 8.7 9.3 10.1 Ethiopia GDP at market prices (2005 US$)b 5.5 11.5 11.5 11.6 7.2 7.0 7.5 Current account balance/GDP (%) 3.3 9.2 4.5 5.6 5.8 8.1 6.5 Gabon GDP at market prices (2005 US$)b 1.0 1.2 5.6 2.3 1.2 2.3 3.4 Current account balance/GDP (%) 10.6 15.7 13.6 17.1 1.7 6.7 7.8 Gambia, The GDP at market prices (2005 US$)b 4.4 6.6 6.3 6.1 4.6 5.0 5.1 Current account balance/GDP (%) 5.3 13.9 13.1 15.6 18.3 16.8 16.3 Ghana GDP at market prices (2005 US$)b 4.7 6.4 5.7 7.3 4.1 4.6 17.5 Current account balance/GDP (%) 5.4 8.2 12.9 18.2 12.6 15.5 12.7 Guinea GDP at market prices (2005 US$)b 3.7 2.2 1.8 3.0 2.0 2.6 4.1 Current account balance/GDP (%) 5.2 11.4 10.5 15.6 11.5 11.1 11.6 Guinea-Bissau GDP at m arket prices (2005 US$)b 1.4 3.5 2.7 2.9 2.1 3.4 3.4 Current account balance/GDP (%) 13.5 18.5 9.7 12.5 16.4 15.4 15.3 Kenya GDP at market prices (2005 US$)b 2.9 6.4 7.1 1.7 2.8 3.7 4.8 Current account balance/GDP (%) 7.5 2.3 3.5 6.9 8.3 7.6 6.9 Lesotho GDP at market prices (2005 US$)b 2.8 6.5 2.4 4.5 0.6 2.3 2.8 Current account balance/GDP (%) 22.0 4.4 13.7 9.8 3.1 18.5 19.6 Madagascar GDP at market prices (2005 US$)b 3.1 5.0 6.2 6.9 0.9 3.1 3.6 Current account balance/GDP (%) 8.6 9.5 14.7 21.6 17.1 15.8 15.2 Malawi GDP at market prices (2005 US$)b 2.4 8.2 8.6 9.7 6.5 5.4 4.6 Current account balance/GDP (%) 5.7 4.2 1.6 6.3 3.4 4.8 4.6 Mali GDP at market prices (2005 US$)b 5.8 5.3 4.3 5.1 3.9 4.7 4.8 Current account balance/GDP (%) 8.7 3.9 7.4 8.5 6.8 7.9 8.5 Mauritania GDP at market prices (2005 US$)b 3.3 11.7 1.9 2.2 2.5 4.1 5.0 Current account balance/GDP (%) 3.2 3.4 10.9 16.4 15.0 16.3 17.6 Mauritius GDP at market prices (2005 US$)b 4.8 3.6 5.5 4.5 1.9 3.5 4.4 Current account balance/GDP (%) 0.1 9.4 6.4 9.0 8.2 8.1 8.8 Mozambique GDP at market prices (2005 US$)b 8.0 8.7 7.0 6.8 5.0 5.5 5.7 Current account balance/GDP (%) 15.1 10.9 12.9 11.5 10.6 10.1 9.1 (continued) 160 R E G I O N A L E C O N O M I C P R O S P E C T S 1995­2005a 2006 2007 2008 2009c 2010d 2011d Namibia GDP at market prices (2005 US$)b 4.2 7.1 5.5 2.9 1.9 3.0 3.3 Current account balance/GDP (%) 3.0 12.7 9.1 1.9 1.3 2.0 1.2 Niger GDP at market prices (2005 US$)b 3.5 5.8 3.3 9.4 1.6 4.9 5.3 Current account balance/GDP (%) 7.1 8.6 7.8 12.8 17.9 16.3 17.7 Nigeria GDP at market prices (2005 US$)b 4.6 6.2 6.3 5.3 4.3 4.8 5.1 Current account balance/GDP (%) 5.4 15.7 17.7 19.1 8.0 10.6 10.3 Rwanda GDP at market prices (2005 US$)b 8.3 7.3 7.9 11.2 5.1 5.5 5.8 Current account balance/GDP (%) 4.6 6.4 3.8 7.0 6.9 7.0 7.1 Senegal GDP at market prices (2005 US$)b 4.4 2.4 4.7 2.5 2.1 3.4 4.2 Current account balance/GDP (%) 5.7 9.2 11.2 12.4 11.4 10.5 10.6 Seychelles GDP at market prices (2005 US$)b 2.8 8.3 7.3 0.1 10.1 2.7 3.7 Current account balance/GDP (%) 13.4 15.0 29.9 47.5 8.8 16.0 17.5 Sierra Leone GDP at market prices (2005 US$)b 4.6 7.3 6.4 5.3 4.0 4.7 6.5 Current account balance/GDP (%) 12.4 9.5 14.3 16.1 17.0 16.6 16.6 South Africa GDP at market prices (2005 US$)b 3.3 5.6 5.5 3.7 1.8 2.0 2.7 Current account balance/GDP (%) 1.3 6.2 7.3 7.4 5.0 5.7 6.0 Sudan GDP at market prices (2005 US$)b 6.2 11.3 10.2 6.8 3.8 4.9 5.1 Current account balance/GDP (%) 6.3 15.2 12.5 9.1 11.5 9.6 8.8 Swaziland GDP at market prices (2005 US$)b 3.5 2.9 3.5 2.4 0.2 1.1 2.4 Current account balance/GDP (%) 0.8 7.3 0.8 4.3 8.6 10.5 11.5 Tanzania GDP at market prices (2005 US$)b 5.4 6.7 7.1 7.5 4.6 5.5 6.2 Current account balance/GDP (%) 6.3 8.0 9.4 9.9 7.9 8.5 8.4 Togo GDP at market prices (2005 USS)b 3.2 3.9 1.9 1.0 1.7 2.0 3.2 Current account balance/GDP (%) 9.6 7.8 7.7 11.2 7.0 7.6 7.6 Uganda GDP at market prices (2005 US$)b 6.4 10.8 8.4 9.0 5.1 5.6 5.9 Current account balance/GDP (%) 7.1 4.4 3.7 3.5 4.8 6.0 6.4 Zambia GDP at market prices (2005 US$)b 3.8 6.2 6.2 5.7 5.2 5.4 5.9 Current account balance/GDP (%) 11.8 1.2 6.1 7.6 4.2 4.3 4.4 Zimbabwe GDP at market prices (2005 US$)b 2.4 6.3 6.9 14.1 4.7 7.1 6.3 Current account balance/GDP (%) 11.5 16.6 10.3 26.6 20.8 23.6 21.3 Source: World Bank. Note: World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries' pros- pects do not significantly differ at any given moment in time. Liberia, Somalia, and, São Tomé and Principe are not forecast owing to data limitations. a. Growth rates over intervals are compound average; growth contributions, ratios and the GDP deflator are averages. b. GDP measured in constant 2000 U.S. dollars. c. Estimate. d. Forecast. 161 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 demand, and in particular investment (with Figure A38 Growth in middle-income and high import content), will boost import de- oil-exporting countries in Sub-Saharan Africa will accelerate faster mand, causing imports to grow faster than exports. Furthermore the recovery in tourism Growth, percent 9 revenues and remittances will also be moder- 8 ate because of a weak (even jobless) recovery 7 in high-income labor markets during the fore- 6 cast period. Current account balances in these 5 middle-income countries are projected to im- 4 prove by less than 1.5 percent of GDP between 3 2009 and 2011. Current account balances in 2 oil-exporting countries should improve by 1 2.7 percent of GDP, with stronger oil revenues 0 partially offset by stronger import demand 2007 2008 2009 2010 2011 and higher profit repatriations. Low-income countries Higher commodity prices and increased Middle-income countries trade volumes, along with more robust levels Oil exporters of domestic economic activity, should reduce Source: World Bank. fiscal deficits in the region. After fiscal bal- ances deteriorated to 4.2 percent of GDP in 2009, fiscal balances are expected to amelio- is projected to accelerate from 0.3 percent in 2009 rate in 2010, narrowing to 2.1 percent of to 3.5 percent in 2010 and to accelerate further to GDP, before improving further to 1.7 per- 4 percent in 2011, boosted by stronger external cent of GDP by 2011 as accelerating output demand and a moderate recovery in tourism and growth raises tax revenues. However, most remittances. Meanwhile growth in oil-exporting low-income countries will continue to ex- countries will almost double, reaching 4.9 per- perience fiscal gaps in excess of 3 percent of cent in 2010, and accelerate marginally, to GDP. Fiscal balances in oil-exporting coun- 5.3 percent in 2011, helped by stronger demand tries are expected to improve from a deficit of for energy. For low-income countries, a more 5.2 percent of GDP to marginal surpluses over moderate acceleration in growth of slightly less the forecast horizon. than 1 percentage point is forecast, as remit- tances, tourism, and private capital flows recover Risks more slowly. Indeed, remittances to Sub-Saharan The major risk facing the Sub-Saharan African are projected to rise only modestly, by economies is that the world economy could 1.8 percent in 2010, after having declined by experience a double dip or economic stagna- 2.9 percent in 2009, as weak labor markets in tion. This would undermine the recovery in destination countries undermine migrants' in- external demand for the Sub-Saharan econo- comes. In contrast, remittances grew 40 percent mies and would put pressure on commodity over the 2006­2008 period (World Bank Group prices, undermining government revenues 2009). Growth in fragile states will accelerate by and possibly pushing debt to unsustainable slightly more than 1 percentage point to above levels. This could in turn force governments 4.2 percent having weathered the global crisis to implement procyclical fiscal cuts, increase remarkably, benefiting in some cases from the taxation, or both, with adverse implications peace dividend. for poverty, health, education, and long- Current account balances in middle-income term growth prospects. Tourism, remit- countries will improve only marginally in the tances, and private capital flows may also next couple of years, as recovery in domestic decline further, thereby negatively affecting 162 R E G I O N A L E C O N O M I C P R O S P E C T S growth and incomes and ultimately causing affecting the financial sector, which in turn more people to fall into poverty. Further- would have an adverse impact on the real more, safety nets in many countries in the sector. Countries like Cape Verde, the Demo- region are very limited, which means that cratic Republic of Congo, Ethiopia, Nigeria, the impact on the poor cannot be cushioned. Rwanda, Tanzania, Uganda, and Zambia have A jobless recovery in high-income countries registered rapid increases in nonfinancial pri- would have similar negative consequences vate sector claims (as a share of broad money) for tourism and remittances to Sub-Saharan and therefore run larger risks. For countries African countries, some of which depend with low capital adequacy, the effect of dete- heavily on these revenues. riorating balance sheets on performance will The Sub-Saharan Africa region will face be even more severe. a large external financing burden in 2010, equivalent to close to 12 percent of GDP, and growth could fall short of the baseline fore- Notes cast if unmet financing requirements lead to 1. Fixed investment plummeted across most devel- oping and high-income countries of the East Asia re- lower investment and growth prospects. For gion from the final quarter of 2008 through the second countries with external financing needs, ma- quarter of 2009. For example, investment in Thailand turing foreign debt will amount to close to tumbled 40 percent in the first quarter of 2009 (saar), 6.5 percent of GDP in 2010, while the cur- Malaysia experienced sequential falloffs of 35 and rent account deficit including grants is fore- 14 percent over the final quarter of 2008 and the first cast at 5.2 percent of GDP. There is thus a of 2009, while several newly industrialized economies risk in many Sub-Saharan economies, and in (NIEs) were much more severely affected, with Taiwan, China, suffering four successive quarterly declines, two particular in low-income countries, that con- of which were in excess of 40 percent. cessional lending will fall short of the need 2. The developing East Asia region as referenced to finance a swift return to growth. In some in this report comprises the larger countries of China, cases, this shortfall may be exacerbated by Indonesia, Malaysia, the Philippines, and Thailand, institutional capacity constraints, which also as well as Fiji, Cambodia, Lao People's Democratic limit effectiveness. Given the role that foreign Republic, Papua New Guinea, Vanuatu, and Vietnam. investment flows play in the region, a reversal Smaller Pacific island nations generally carry insuf- ficient economic and financial data for inclusion in in these flows not only would directly affect the database and projections. The importance of high- external financing needs, but also would have income East Asian countries--those noted in the a severe impact on investment and growth. text--as well as Australia, should be underscored in Given the increased global growth uncertain- the current context of crisis and recovery, because the ties, investment flows to the region may be strong trade relationships among all countries in East adversely affected. Asia tend to amplify the down-phase of recession, but The fiscal position in some of the smaller should come to support the rebound and recovery in a similar fashion as recovery evolves over coming months members of the Southern Africa Customs Union and years. (particularly Lesotho and Swaziland) may come 3. Dollar-based exports picked up to growth of under severe pressure over the next two to three 52 percent for China by November 2009 (saar) from years, because one of the major revenue sources declines of 54 percent in March; to 41 percent for the of the union's revenue pool is related to taxes remainder of the developing region by October, and to on South African imports, which have dete- 17 percent for the NIEs, also by October of the year. riorated rapidly in the aftermath of the global At the same time industrial production for most econo- mies rebounded sharply, for example, to 25 percent for financial crisis. Thailand in September (saar) from trough declines of In countries that experienced rapid credit 48 percent in December 2008. growth during the boom years, there is a 4. The countries covered in the Europe and Central marked risk that nonperforming loans will Asia section of the appendix are those that fall into the rise sharply during the economic downturn World Bank's definition of low- and middle-income 163 G L O B A L E C O N O M I C P R O S P E C T S 2 0 1 0 classifications (with 2008 per capita Gross National are covered in this report under the category of "other Income equal to or below $3,855). These 24 coun- high-income countries." But as this group has become tries are Albania, Bosnia and Herzegovina, Bulgaria, increasingly more integrated with the developing Kosovo, Latvia, Lithuania, Former Yugoslav Republic economies of the region, discussion of economic and of Macedonia, Montenegro, Poland, Romania, and financial developments for the group is a feature of this Serbia (in the Central European subregion); Armenia, appendix. Among the GCC, insufficient data exists for Azerbaijan, Belarus, Georgia, Kazakhstan, Kyrgyz inclusion of Qatar and the United Arab Emirates. Republic, Moldova, Russian Federation, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan (in the Com- monwealth of Independent States subregion); and References Turkey. Transition countries include all 24 countries, Chen, S., and M. Ravallion. "The Impact of the Global with the exception of Turkey. Among these develop- Financial Crisis on the World's Poorest." Vox: ing countries, Bulgaria, Latvia, Lithuania, Poland, and Research-Based Policy Analysis and Commen- Romania are new European Union members. Owing to tary from Leading Economists, www.voxeu.org/ data limitations, forecasts are not available for Bosnia index.php?q=node/3520 [accessed Dec. 9, 2009]. and Herzegovina, Kosovo, Montenegro, Serbia, Tajiki- Friedman, J., and N. Schady. 2009. "How Many More stan, and Turkmenistan. Infants Are Likely to Die in Africa as a Result of 5. See World Bank (2009b). the Global Financial Crisis?" Research Paper 60 6. Bulgaria, Latvia, Lithuania, Poland, and Roma- (August). World Bank, Washington, DC. nia. See World Bank, 2009a. World Bank. 2009a. "EU10 Regular Economic 7. Kazakhstan, Kyrgyz Republic, Tajikistan, Turk- Report: From Stabilization to Recovery." menistan, and Uzbekistan. Washington, DC (October). http://go.worldbank 8. Armenia, Azerbaijan, and Georgia. .org/UNCGIKEPH0. 9. World Bank Group (2009). ------. 2009b. "Russian Economic Report #20: 10. World Bank (2010a). From Rebound to Recovery?" Washington, 11. World Bank (2010b). DC (November10). http://go.worldbank.org/ 12. SELIC stands for Sistema Especial de Liquida- FKGLSQ4NF0. ção e Custodia, or Special System of Clearance and ------. 2010a. "The Crisis Hits Home: Stress-Testing Custody, which is Banco Central do Brasil's overnight Households in Europe and Central Asia." Pre- lending rate. pared by Erwin R. Tiongson, Naotaka Sugawara, 13. The low- and middle income countries of the Victor Sulla, Ashley Taylor, Anna I. Gueorguieva, Middle East and North Africa region as presented Victoria Levin, and Kalanidhi Subbarao. http:// in this report include Algeria, the Arab Republic of go.worldbank.org/H92ZH3CK20. Egypt, the Islamic Republic of Iran, Jordan, Lebanon, ------. 2010b. "Turmoil at Twenty." Washington, Morocco, the Syrian Arab Republic, Tunisia, and the DC. http://go.worldbank.org/ZQTRLRED70. Republic of Yemen. Several developing economies are World Bank Group. 2009. "Migration and Remit- not covered owing to data insufficiencies, including tance Trends 2009." Migration and Development Djibouti, Iraq, Libya, and the West Bank and Gaza. Brief 11. Washington, DC (November 3). http:// High-income economies of the broader geographic siteresources.worldbank.org/INTPROSPECTS/ region, including Gulf Cooperation Council (GCC) Resources/334934-1110315015165/ members Bahrain, Kuwait, Oman and Saudi Arabia MigrationAndDevelopmentBrief11.pdf. 164 Eco-Audit Environmental Benefits Statement The World Bank is committed to preserv- Saved: ing endangered forests and natural re- · 23 trees sources. The Office of the Publisher · 644 pounds of has chosen to print Global Economic solid waste Prospects 2010 on recycled paper with · 10,609 gallons 30 percent post-consumer waste, in ac- of wastewater cordance with the recommended stan- · 2,203 pounds of dards for paper usage set by the Green net greenhouse Press Initiative, a nonprofit program sup- gases porting publishers in using fiber that · 7 million British is not sourced from endangered for- thermal units of ests. For more information, visit www. total energy greenpressinitiative.org. "The crisis has deeply impacted virtually every economy in the world, and although growth has returned, much progress G lobal Economic Prospects 2010: Crisis, Finance, and Growth explores both the short- and medium-term impacts of the financial crisis on developing countries. Although global growth has resumed, the recovery is fragile, and unless business and consumer demand strengthen, the world economy in the fight against poverty could slow down again. has been lost. More difficult Even if, as appears likely, a double-dip recession is avoided, the international conditions in the recovery is expected to be slow. High unemployment and widespread years to come will mean that restructuring will continue to characterize the global economy for the developing countries will have next several years. Already, the crisis has provoked large-scale human to place even more emphasis on suffering. Some 64 million more people around the world are expected to be living on less than a $1.25 per day by the end of 2010, and improving domestic economic between 30,000 and 50,000 more infants may have died of malnutrition conditions to achieve the kind in 2009 in Sub-Saharan Africa, than would have been the case if the of growth that can durably crisis had not occurred. eradicate poverty." Over the medium term, economic growth is expected to recover. But --Justin Yifu Lin increased risk aversion, a necessary and desirable tightening of financial Chief Economist and regulations in high-income countries, and measures to reduce the Senior Vice President exposure of developing economies to external shocks are likely to make finance scarcer and more costly than it was during the boom period. The World Bank As a result, just as the ample liquidity of the early 2000s prompted an investment boom and an acceleration in developing-country potential output, higher costs will likely yield a slowing in developing-country potential growth rates of between 0.2 and 0.7 percentage points, and as much as an 8 percent decline in potential output over the medium term. In the longer term, however, developing countries can more than offset the implications of more expensive international finance by reducing the cost of capital channeled through their domestic financial markets. For more information, please visit www.worldbank.org/gep2010. To access Prospects for the Global Economy, an online companion publication, please visit www.worldbank.org/globaloutlook. ISBN 978-0-8213-8226-4 THE WORLD BANK SKU 18226