90897 MACROECONOMIC DEVELOPMENTS IN LOW-INCOME DEVELOPING COUNTRIES: 2014 REPORT September 18, 2014 OVERVIEW This report examines macroeconomic developments and related vulnerabilities in low- income developing countries (LIDCs)—a group of 60 countries that have markedly different economic features to higher income countries and are eligible for concessional financing from both the IMF and the World Bank. Collectively, they account for about one-fifth of the world’s population. The report examines the strong economic performance achieved by the bulk of LIDCs since 2000 and assesses their short-term economic prospects. It then looks at the economic risks and vulnerabilities that they currently face, against the backdrop of a brittle and uneven global recovery that is vulnerable to important financial and geopolitical risks. The final section of the report examines the evolution of public debt levels in LIDCs in recent years. Key messages in the report include: 1) most LIDCs have recorded strong economic growth for an extended period, but based primarily on factor accumulation rather than productivity growth; 2) about one-half of LIDCs are classified as being at medium/high vulnerability to a growth shock, with weakened fiscal positions forming a key source of vulnerability; 3) fiscal institutions, including debt management capacity, should be strengthened to pre-empt the build-up of potential new imbalances. LIDCs: Macroeconomic Trends and Outlook Economic growth in most LIDCs has been strong over the past 15 years, faster than in previous decades and on par with growth performance in emerging markets. This performance was helped by external factors but domestic factors also played a central role, with sound macroeconomic management and wide-ranging market-oriented reforms providing the building blocks for sustained growth even as the global economy stalled in 2009. The impressive resilience of LIDCs during the global economic crisis was facilitated by the limited direct linkages between domestic financial systems and international financial markets. The positive growth performance for the group as a whole masks a number of more problematic developments. For one, almost one-fifth of LIDCs failed to increase the level of output per capita over the period—mainly countries affected by conflict and weak states, but, in some cases, reflecting flawed economic policies. Second, growth MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT has generally not been very deep or transformative, driven largely by factor accumulation rather than productivity gains. Thirdly, progress in reducing extreme poverty and reaching other Millennium Development Goals has been mixed. Looking ahead, LIDCs may face continued economic headwinds in the form of mediocre growth in many trading partner countries. Nevertheless, growth is projected to remain generally strong, with a number of the larger economies (such as Bangladesh and Kenya) showing significant dynamism and with improved political conditions and/or policy reforms contributing to growth in other cases (such as Myanmar and Democratic Republic of Congo). The Ebola outbreak, if not contained rapidly, could have acute macroeconomic and social consequences on several already fragile economies in West Africa. Over the medium term, maintaining growth at the pace needed to employ fast- growing labor forces will be difficult without achieving structural transformation and associated strong productivity growth. How vulnerable are LIDCs to adverse shocks? Although LIDCs have been resilient in recent years, their still-limited export diversification and weakened policy buffers leave them less well-positioned to handle these shocks than prior to the global crisis. The share of LIDCs that are assessed to be highly vulnerable is easing slightly (to around 10 percent of the total); most of these countries are fragile states. Weak fiscal positions are typically the most important source of vulnerability across countries. Analysis of selected shock scenarios, drawing on the World Economic Outlook, flags the significant adverse impact on LIDCs of a protracted period of slower growth in advanced and emerging market economies. Temporary global oil price shocks have relatively modest output effects on LIDCs, but sizeable fiscal effects on those oil-importing countries that currently subsidize fuel products. Frontier market economies, expanding their links to the global financial system, face new risks; rapid credit growth and the expansion of foreign credit warrant close monitoring in some cases. To enhance resilience, policy actions to rebuild fiscal buffers are a priority in many countries—through a country-specific mix of enhanced revenue mobilization and improved prioritization of public expenditures—as is the strengthening of fiscal institutions including public administration. Foreign reserve levels are insufficient in a sizeable number of LIDCs and need to be given higher priority in framing macroeconomic policies in these cases. The modernization of monetary frameworks underway in many countries will strengthen the effectiveness of monetary and exchange rate policies in responding to shocks. Over the medium term, policies to promote economic diversification would strengthen resilience in the face of shocks, including natural disasters, but will take time to deliver results. 2 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT What do we learn from recent debt trends in LIDCs? Public debt levels are now at relatively low levels in the majority of LIDCs, helped by strong economic growth, low interest rates, and the provision of comprehensive external debt relief to some 34 countries under the Heavily Indebted Poor Countries/Multilateral Debt Relief Initiative (HIPC/MDRI). Some three-quarters of LIDCs are currently assessed as being at low or moderate risk of experiencing external debt distress under the joint Bank-Fund Debt Sustainability Framework. Nevertheless, debt levels are high and/or have increased significantly in recent years in a third of LIDCs. Looking at debt developments since 2007, most countries that had benefited from debt relief prior to that point (“early HIPCs”) have seen public debt levels (as a share of GDP) rise over time, although it is only in a few cases that debt accumulation has become a significant cause for concern. External borrowing (both concessional and non- concessional) has typically accounted for the preponderance of the debt build-up, but domestic debt levels have also risen significantly in a handful of cases (such as Ghana and Malawi). There has been no clear trend in debt levels among non-HIPCs (countries that have not benefited from HIPC/MDRI). However, the supportive conditions that helped stabilize debt ratios in LIDCs since 2007—notably easy global financing conditions—will likely fade away in the period ahead, flagging the need to avoid complacency. The changing external financial landscape has enabled an increasing number of LIDCs to access international financial markets; non-traditional official creditors have also significantly expanded their provision of project finance. New borrowing options open the opportunity to increase development spending but borrowed funds cover their costs only if used to augment development spending on projects that yield appropriately high rates of return. Countries tapping new sources of funding thus need to give due attention to where the incremental funds go and how efficiently they are used. With new risks, such as bunching of repayments and rollover risk, efforts to strengthen public debt management are an imperative, supplemented by the development of a medium-term debt strategy on which new borrowing decisions can be grounded. INTERNATIONAL MONETARY FUND 3 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Approved By Prepared by the Strategy, Policy, and Review Department and Fiscal Siddharth Tiwari and Affairs Department under the overall guidance of Seán Nolan and Vitor Gaspar Sanjeev Gupta. The staff team that prepared the report comprised of: Olumuyiwa Adedeji, Calixte Ahokpossi, Marco Arena, Gilda Fernandez, Jana Gieck, Giang Ho, Alexis Meyer-Cirkel, Nkunde Mwase, Chris Lane, Maxwell Opoku-Afari, Chris Papageorgiou, Andrea Presbitero, Hajime Takizawa, Olaf Unteroberdoerster, Yi Xiong (all SPR), Kerstin Gerling, Marialuz Moreno-Badia, Abdelhak Senhadji, Louis Sears, Priscilla Toffano (all FAD). Research assistance was provided by Mai Anh Bui, Sibabrata Das, Jayendu De, Christian Gonzales, Carla Intal, and Vera Kehayova. Production assistance was provided by Merceditas San Pedro-Pribram and Lucia Hernandez. This document has benefited from comments received from World Bank staff. Contributions to the paper were also provided by Andy Berg, Felipe Zanna (RES), Luc Everaert, Alison Holland, Srobona Mitra (MCM), and LIDC country teams. CONTENTS Acronyms and Abbreviations ___________________________________________________________________ 7 MACROECONOMIC TRENDS AND THE NEAR TERM OUTLOOK _______________________________ 9 A. Introduction ___________________________________________________________________________________ 9 B. Macroeconomic Trends since 2000 ___________________________________________________________ 12 C. Recent Macroeconomic Developments and Outlook _________________________________________ 20 SHIFTING VULNERABILITIES ___________________________________________________________________ 24 A. Introduction __________________________________________________________________________________ 24 B. Trends in Vulnerabilities: The Role of Fundamentals __________________________________________ 25 C. How Vulnerable are LIDCs to Potential Global Shocks? _______________________________________ 27 D. A Closer Look at Financial Vulnerabilities _____________________________________________________ 30 E. Natural Disasters: A Particular Challenge for LIDCs ____________________________________________ 34 F. Building Resilience in LIDCs: Policy Recommendations ________________________________________ 37 DEBT DEVELOPMENTS SINCE DEBT RELIEF ___________________________________________________ 39 A. Stylized Facts _________________________________________________________________________________ 40 B. Risk Diagnostics _______________________________________________________________________________ 47 C. Policy Challenges _____________________________________________________________________________ 51 BOXES 1. Falling Behind _________________________________________________________________________________ 15 4 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT 2. Methodology Underlying the Growth Decline Vulnerability Index_____________________________ 25 3. Frontier FSAPs: Findings from the Basel Core Principles Assessments _________________________ 33 4. The Ebola Outbreak in Guinea, Liberia, and Sierra Leone ______________________________________ 35 5. Public Investment Scaling-Up, Growth, and Debt Sustainability in LIDCs ______________________ 46 6. Risks from International Sovereign Bond Issuance ____________________________________________ 52 FIGURES 1. Map of LIDCs___________________________________________________________________________________ 9 2. LIDC SubGroups by GNI per Capita and Population, 2013 ____________________________________ 11 3. Real GDP Growth______________________________________________________________________________ 12 4. GDP Growth in Past and 2009 Crises __________________________________________________________ 12 5. Growth Heterogeneity Across LIDCs __________________________________________________________ 14 6. Growth Decomposition _______________________________________________________________________ 16 7. Challenges and Potential for Agriculture ______________________________________________________ 16 8. Progress Toward Selected MDGs, by Number of LIDCs _______________________________________ 17 9. Inflation and Commodity Prices _______________________________________________________________ 17 10. Trends in Fiscal and External Sectors _________________________________________________________ 18 11. Capital Flows_________________________________________________________________________________ 19 12. LIDCs’ Export Destinations ___________________________________________________________________ 20 13. Export Product Diversification _______________________________________________________________ 20 14. External Assumptions ________________________________________________________________________ 22 15. GDP Growth and Volatility ___________________________________________________________________ 24 16. Growth Decline Vulnerability Index, 2008–14 ________________________________________________ 25 17. Growth Decline Vulnerability Index by Country Group _______________________________________ 26 18. Growth Decline Vulnerability Index by Sector ________________________________________________ 26 19. LIDCs: Assessment on Budget Planning and Execution_______________________________________ 27 20. Growth Decline Vulnerability Index by Sector and Region ___________________________________ 27 21. Shock Scenarios: Global Growth and Inflation _______________________________________________ 27 22. Impact of Protracted Slowdown _____________________________________________________________ 28 23. Cumulative Change Relative to Baseline _____________________________________________________ 29 24. Energy Subsidies and Impact of Oil Shock ___________________________________________________ 29 25. Bank Return on Assets _______________________________________________________________________ 30 26. Distribution of Z-Scores _____________________________________________________________________ 31 27. First Time Bond Issuances____________________________________________________________________ 31 28. Volatility in Frontier Markets and EMs _______________________________________________________ 32 29. Capital Adequacy Ratios _____________________________________________________________________ 32 30. Natural Disasters and People Affected _______________________________________________________ 34 31. Natural Disasters by Income Distribution ____________________________________________________ 34 32. Natural Disasters in LIDCs: Event Study ______________________________________________________ 36 33. Natural Disasters in LIDCs: Impulse Response Functions _____________________________________ 36 34. Food Supply Crisis in Comparison (1990–2009) ______________________________________________ 37 35. Food Decline Vulnerability Index_____________________________________________________________ 37 36. LIDCs: Public Debt ___________________________________________________________________________ 40 INTERNATIONAL MONETARY FUND 5 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT 37. LIDCs: External Debt by Concessionality _____________________________________________________ 40 38. LIDCs: Changes in Public Debt Ratios, 2007–13 ______________________________________________ 41 39. LIDCs: Changes in PPG External and Domestic Debt-to-GDP Ratios by Country Groups _____ 41 40. Early-HIPCs: Net Lending Flows ______________________________________________________________ 41 41. LIDCs: Debt Service on External Debt ________________________________________________________ 42 42. LIDCs Public Debt Decomposition, 2007–13 _________________________________________________ 42 43. LIDCs: Changes in Expenditures and Revenue, 2007–13 _____________________________________ 44 44. LICDs: Public Investment in LIDCs ____________________________________________________________ 44 45. The Impact of Improving Public Investment Efficiency—Simulation Results__________________ 45 46. LIDCs: Changes in Risk Rating for External Debt Distress_____________________________________ 47 47. LIDCs: Net Lending by Type of Creditors _____________________________________________________ 48 48. Fiscal Trends, 2007–13 _______________________________________________________________________ 49 49. Fiscal Policy Slippages—Frequency and Annual Average of Underperformance _____________ 50 TABLES 1. Selected Macro and Structural Indicators for LIDCs ___________________________________________ 10 2. Selected Macroeconomic Indicators, LIDCs and SubGroups ___________________________________ 21 3. Financial Vulnerability Index: Number and Share of Countries by Vulnerability Rating ________ 32 APPENDICES I. LIDCs and SubGroups__________________________________________________________________________ 54 II. Identifying Frontier Market Economies ________________________________________________________ 55 III. Methodology Underlying the Financial Vulnerability Index ___________________________________ 59 IV. Food Decline Vulnerability Index and Natural Disasters ______________________________________ 61 V. Case Studies: Key Trends ______________________________________________________________________ 63 APPENDIX TABLES 1. Description and Definition of Variables _______________________________________________________ 56 2. Financial Sector Depth Index __________________________________________________________________ 57 3. Financial Vulnerability Index Parameters ______________________________________________________ 60 4. Food Decline Vulnerability Index Estimation Results __________________________________________ 62 References ______________________________________________________________________________________ 64 6 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Acronyms and Abbreviations AMs Advanced Markets BCP Basel Core Principle CPIA Country Policy and Institutional Assessment CRED Centre for Research on the Epidemiology of Disasters DSA Debt Sustainability Analysis DSF Debt Sustainability Framework DIG Debt, Investment, and Growth EMs Emerging Markets EM-DAT Emergency Events Data Base EMDCs Emerging Market and Developing Countries EVD Ebola Virus Disease FAO Food and Agriculture Organization FDI Foreign Direct Investment FDVI Food Decline Vulnerability Index FMs Frontier Market Economies FSAP Financial Sector Assessment Program FVI Financial Vulnerability Index GDVI Growth Decline Vulnerability Index HIPC Heavily-Indebted Poor Countries IFS International Financial Statistics ILO International Labour Organization IRGD Interest-Rate Growth Differential LIDCs Low-Income Developing Countries MDGs Millennium Development Goals MDRI Multilateral Debt Relief Initiative NDPs National Development Plans NPL Non-Performing Loans ODA Official Development Assistance OECD Organization for Economic Cooperation and Development PEFA Public Expenditure and Financial Accountability PFM Public Financial Management PPG Public and Publicly Guaranteed PPP Purchasing Power Parity PPPs Public-Private Partnerships PRGT Poverty Reduction Growth and Trust PRSPs Poverty Reduction Strategy Papers PV Present Value SSA Sub-Saharan Africa TA Technical Assistance TFP Total Factor Productivity UNCTAD United Nations Conference on Trade and Development INTERNATIONAL MONETARY FUND 7 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT VAT Value-Added Tax VIX CBOE Volatility Index WAEMU West African Economic and Monetary Union WDI World Development Indicators WEO World Economic Outlook 8 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT MACROECONOMIC TRENDS AND THE NEAR TERM OUTLOOK A. Introduction 1. This report focuses on macroeconomic developments and policy issues in low-income developing countries (LIDCs). The LIDC group includes all countries that a) fall below a modest per capita income threshold (US$2,500) and b) are not conventionally viewed as emerging market economies (EMs).1 There are 60 countries in this group, accounting for about one-fifth of the world’s population; sub-Saharan Africa (SSA) accounts for some 57 percent of the LIDC population, with a further 28 percent living in Asia (Figure 1). While sharing characteristics common to all countries at low levels of economic development, the LIDC group is strikingly diverse, with countries ranging in size from oil-rich Nigeria (174 million) to tourism-dependent Kiribati (0.1 million), and in 2013 per capita income terms from Mongolia (US$3,770) to Malawi (US$270). The 10 largest economies in the group account for two-thirds of total group output (as measured in PPP terms). Figure 1. Map of LIDCs Source: IMF. 1 The LIDC grouping is a subgroup of the “Emerging Market and Developing Countries” (EMDCs) aggregate used in the IMF’s World Economic Outlook; see Appendix I for a full list of the countries in the group. For further discussion on how the grouping was constructed, see “Proposed New Grouping in WEO Country Classifications: Low-Income Developing Countries” available at http://www.imf.org/external/np/pp/eng/2014/060314.pdf.. INTERNATIONAL MONETARY FUND 9 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT 2. The case for treating LIDCs as a distinct group is that they differ significantly from economies at higher levels of per capita income (Table 1):2  Agriculture has a larger share in economic activity in LIDCs (27 percent of GDP) than in the average emerging market (EM) (8 percent of GDP); the share of the labor force employed in the informal sector is also significantly higher.  LIDCs lag EMs in infrastructure, financial deepening (an average private credit-GDP ratio of 25 percent, compared with 50 percent in EMs), and in quality/capacity of public institutions.  LIDCs rely more heavily on foreign aid, and less on own budgetary revenues, than EMs.  Key development indicators in LIDCs lag the average EM. Poverty and infant mortality rates are much higher, while education levels are significantly lower, but income inequality is similar to EMs. Table 1. Selected Macro and Structural Indicators for LIDCs LIDC Fragile Frontier Commodity EM average states markets exporters average Agricultural sector share (in percent of GDP), 2012 27.4 36.3 21.9 29.8 8.3 Informality (percent employed in informal sector), 2005-10 51.1 57.6 57.6 62.7 35.8 Electric power consumption (kWh per capita, 2011) 517.9 197.2 500.5 536.8 3313.0 Quality of overall infrastructure (score, 7=best), 2013 2.6 2.2 2.8 2.4 4.0 Institutional quality (average rank of Fraser index, 1=highest), 2010 104.3 123.4 90.1 114.4 70.6 Trade openness (exports plus imports, in percent of GDP), 2010 81.0 78.8 76.0 67.0 91.9 Financial development (private credit as in percent of GDP), 2012 24.8 15.7 34.0 18.6 50.0 Net official development assistance (in percent of GDP), 2012 9.3 12.4 5.8 8.3 3.6 Foreign direct investment inflows (in percent of GDP), 2013 2.9 1.8 5.0 3.0 1.2 Revenue (excl. grants, in percent of GDP), 2013 22.0 21.3 20.6 19.9 30.4 Standard deviation of revenue (in percent of GDP), 2000-13 3.6 4.3 2.8 4.3 3.1 Poverty rate (percent of population < $1.25/day), 2005-10 40.6 51.8 44.1 51.1 6.6 Income inequality (Gini coefficient), 2005-10 41.5 42.0 43.8 41.3 41.4 Infant mortality rate (per 1000 live births), 2010 52.7 62.6 44.5 62.2 18.2 Average years of schooling, 2010 5.3 4.4 6.0 4.6 8.7 Sources: WEO, WDI, ILO, and IMF staff estimates. Note: Group aggregates are calculated based on simple (un-weighted) averages. EM refers to the group of non-LIDCs that belong to the WEO category of EMDCs. 3. For analytical purposes, it is useful to divide the LIDC group into subgroups, based on characteristics that are key drivers of economic performance. Four subgroups are identified: 2 In LIDCs, the quality of data on national accounts, employment, and other macroeconomic measures are often of poor quality, and hence need to be interpreted with caution. 10 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT each LIDC appears in at least one subgroup, with some Figure 2. LIDC SubGroups by GNI per featuring in more than one subgroup (Figure 2 and Capita and Population, 2013 Appendix I): 2,000 Commodity Exporters Frontier Markets Population: 579 million 1,800 Population: 666 million No. of countries: 27  Frontier markets are those countries closest to 1,600 No. of countries: 14 resembling EMs in the depth and openness of financial 1,400 GNI per capita (US$) markets and access to international sovereign bond 1,200 All LIDCs Population: 1.3 billion markets.3 There are 14 countries in this grouping, with 1,000 No. of countries: 60 Nigeria, Vietnam, and Bangladesh accounting for about 800 Other LIDCs 600 70 percent of group output. The group contains about half Population: 218 million No. of countries: 15 400 Fragile States of the LIDC population. Population: 416 million 200 No. of countries: 28 0  Commodity exporters have at least 50 percent of export Source: WDL, WEO, and IMF staff estimates. Data for Somalia not available. earnings coming from fuels and primary commodities. There are 27 countries in this group including Nigeria and Uzbekistan that account for about 60 percent of total group output. The group contains over two-fifths of the total LIDC population.  Fragile states are those countries where institutional capacity is especially weak (three-year average of the CPIA score below 3.2) and/or there has been significant internal conflict.4 The group includes 28 countries that contain about one-third of the LIDC population. Myanmar, Sudan, Yemen, and the Democratic Republic of Congo (DRC) are the largest economies in this context, accounting for half of total group output.  Other LIDCs are the 15 countries that do not fall into any of the preceding groupings. These countries collectively contain about 16 percent of the LIDC population, with Ethiopia, Cameroon, Cambodia, and Honduras being the largest economies in the group. 3 See Appendix II for an explanation of the methodology used to construct this group. 4 The CPIA is a diagnostic tool that captures the quality of a country’s policies and institutional arrangements along 16 criteria grouped into four equally-weighted clusters: Economic Management, Structural Policies, Policies for Social Inclusion and Equity, and Public Sector Management and Institutions. Countries are rated on a scale of 1 (low) to 6 (high) for all of the sixteen criteria and are assigned an overall score. INTERNATIONAL MONETARY FUND 11 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT B. Macroeconomic Trends since 2000 Figure 3. Real GDP Growth Economic growth in most LIDCs has been strong… (In percent) 10 4. LIDCs have delivered strong growth performance 8 over the last fifteen years. After an extended period of 6 stagnation, instability, and conflict in most countries, LIDCs 4 entered a period of high and sustained growth from the late- 2 1990s. Over the 2000–13 period, LIDCs recorded average real 0 World GDP growth of 6½ percent, up from 3.6 percent during the -2 Emerging markets LIDCs 1990s and on par with the performance of emerging markets -4 (Figure 3). The growth pick-up was particularly marked for 2000 2002 2004 2006 2008 2010 2012 countries in SSA and the transition economies of Central Asia, Sources: WEO and IMF staff estimates. but also significant in Asia and Latin America. 5. LIDC growth showed notable resilience during the 2009 global financial crisis, providing a marked contrast with the outcome in the wake of previous Figure 4. GDP Growth in Past and 2009 global shocks (Figure 4). The main transmission channels were Crises trade (falling demand for exports), along with a slowing of FDI (In percent) 9 inflows. With LIDC banks relying primarily on a stable deposit 8 LIDC-3 crises LIDC-2009 World-3 crises World-2009 base for funding, the direct impact of the global financial crisis 7 on LIDC financial sectors was very limited—although there 6 were indirect effects on asset quality as exporters dealt with 5 4 falling demand and prices. GDP growth in 2009 remained 3 positive in over 80 percent of LIDCs; average growth was in 2 the order of about 6 percent, 1 point less than the five-year 1 pre-crisis average. The rebound in 2010 was sharp and has 0 -1 -5 -4 -3 -2 -1 0 1 2 3 4 5 been largely sustained since then—a contrast with previous Sources: WEO and IMF staff estimates. downturns, when the decline in growth was prolonged. Note: The chart plots real GDP growth in the world and in LIDCs 5 years before and 5 years after the Growth was supported by countercyclical policy responses— global crises of 1975, 1982 and 1991, and the 2009 facilitated by solid pre-crisis fiscal and external positions—and crisis (0 on the horizontal axis denotes the start of the crisis). by substantial external financial support, including concessional financing from the World Bank and the IMF. 6. The improved growth performance over the past fifteen years reflected favorable external conditions (for most of the period) and better economic policies. LIDCs benefited from robust commodity prices, the emergence of China as an important trade and investment partner (particularly in SSA),5 and increased capital inflows (taking the form of FDI in growing extractive industries in most countries). Countries burdened with high debt levels benefited from international relief initiatives (HIPC/MDRI) that created room to finance development spending. On the domestic 5 See, for example, Drummond and Liu (2013). 12 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT side, improved macroeconomic management contributed to lower inflation and growth volatility compared with the pre-2000 era. Countries also implemented wide-ranging market-oriented reforms in the real and financial sectors, facilitating private sector development (Ostry and others, 2009; Dabla-Norris and others, 2013). 7. Strong performance for the LIDC group as a whole masks considerable heterogeneity of experience both across and within subgroups (Figure 5).  Growth was strongest in frontier market economies, led by Nigeria, Tanzania, and Vietnam. Statistical techniques for identifying structural breaks point to a growth “takeoff”—a period of sustained growth acceleration—for the group as a whole, starting around 2000.6  Commodity exporters experienced both above-average growth7—although not by a large margin—and significantly higher output volatility, the latter linked to export price volatility.  Fragile states experienced below-average growth along with higher output volatility—consistent with several studies of the impact of fragility on economic performance (World Bank, 2011).  Some one-fifth of LIDCs failed to record any growth in output per capita over the period, thereby falling well behind other LIDC peers (see Box 1). 6 A break in growth is identified as the point after which the average growth rate diverges significantly from the previous average growth rate; for frontier LIDCs, annual average growth was 4.1 percent during 1990–99, 7.1 percent during 2000–13. See also Berg, Ostry, and Zettelmeyer (2012). 7 Note that several commodity exporters are also classified as frontier economies, e.g., Nigeria, Mozambique, and Zambia. INTERNATIONAL MONETARY FUND 13 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Figure 5. Growth Heterogeneity Across LIDCs Panel A. Growth and Output Volatility in LIDC Sub-Groups (In percent) 8 Average 2000-13 growth rate 7 Standard deviation of GDP growth over 2000-13 1/ 6 5 4 3 2 1 0 AM EM LIDC Fragile Non- FrontierNon- Comm. Other Fragile Frontier Exp. Panel B. Uneven Growth Across LIDCs (In percent, average 2000-13) 14  Fragile states 12  Frontier markets 10  Commodity exporters Other LIDCs 8 LIDC average 6 EM average 4 AM average 2 0 -2 -4 Madagascar Honduras Nicaragua Malawi Lesotho Myanmar Burundi Rwanda Afghanistan, I.R. of Kenya Ghana Bhutan Nigeria Tanzania Mauritania Mongolia Guinea-Bissau Zimbabwe Djibouti Moldova Benin Burkina Faso Liberia Bolivia Mozambique Ethiopia Papua New Guinea Gambia, The Zambia Guinea Yemen, Republic of São Tomé & Príncipe Bangladesh Haiti Uganda Cameroon Vietnam Niger Kiribati Mali Tajikistan Cambodia Lao People's Dem.Rep Comoros Togo Chad Central African Rep. Eritrea Solomon Islands Nepal Sudan Senegal Uzbekistan Côte d'Ivoire Congo, Dem. Rep. of Congo, Republic of Kyrgyz Republic Sierra Leone Sources: WEO and IMF staff estimates. Note: No data for Somalia and South Sudan. Côte d’Ivoire is both a frontier market and fragile state. 1/ Subgroup averages are GDP-weighted. 14 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Box 1. Falling Behind While most LIDCs have recorded sustained growth since 2000, there is a sizeable group of countries (almost one-fifth of the total) that did not record any increase in output per capita over the period. Among these 11 cases, some countries Average growth Stdev. of experienced significant declines in output per p/c 2000- growth Fragile Frontier Commodity capita (such as Eritrea and Central African Country 13 (%) (%) states markets exporters Eritrea -2.6 5.9 1 0 1 Republic), others effectively stayed put in terms Central African Rep. -2.6 10.4 1 0 1 of income levels (such as Madagascar and Zimbabwe Côte d'Ivoire -2.2 -1.4 9.2 3.8 1 1 0 1 1 0 1 Yemen). Haiti -0.4 3.0 1 0 0 Guinea-Bissau -0.2 2.1 1 0 1 Comoros -0.1 1.5 1 0 0 The weak performance occurred across Madagascar -0.1 5.3 1 0 0 several macro and structural indicators. Over Kiribati -0.1 2.9 1 0 0 Republic of Yemen 0.0 4.6 1 0 1 2000–13, these 11 countries have been less Togo 0.0 2.6 1 0 0 successful in reducing inflation, attracting FDI, Median growth Median LIDC -0.2 1.9 3.8 2.9 developing the financial markets, and improving Source: WEO. social indicators, such as the level of educational attainment. A common feature to all countries in the group is that they are fragile states —countries either with very weak institutions or significantly affected by conflict over the period. The role of fragility in hampering growth is easy to understand in countries affected by sustained internal conflict and political instability over an extended period (such as Côte d’Ivoire, Guinea-Bissau, Comoros, and Yemen). Natural disasters, such as the massive 2010 earthquake in Haiti, can account for sizeable shocks to output levels, but the impact on growth over an extended period of time is likely to be more modest; weak institutions and recurrent political instability play a key role in explaining Haiti’s weak performance as the poorest country in the Western Hemisphere. But a review of the country listing shows that bad policy choices, unlinked to fragility, can also produce income contraction over time, as in Zimbabwe (which experienced hyperinflation) and Eritrea (a tightly regulated/controlled economy). ____________________ 1 In terms of total GDP growth, all 11 countries had average growth rates in the bottom quartile of the LIDC group (less than 3.5 percent). …but not yet deep or transformative 8. LIDC growth has been primarily driven by factor accumulation rather than productivity gains (Figure 6, Panel A). Rapid expansion of the labor force and capital accumulation accounted for the bulk of GDP growth over 2000–10, with little coming from gains in total factor productivity (TFP).8, 9 TFP is estimated to have declined in both fragile states and commodity exporters on average over the decade, although several fast-growing frontier economies have recently experienced acceleration in TFP (e.g., Uganda). Thus, substantial scope exists in LIDCs for moving toward a more intensive pattern of growth, with large potential benefits from economic reforms. 8 The large labor contribution likely reflects employment growth in the service sector; employment growth in the resource sector (even in commodity exporters) is typically limited. 9 This “extensive” pattern of growth has also been observed in East Asia’s newly industrializing countries in the 1990s (see, for example, Young, 1994). INTERNATIONAL MONETARY FUND 15 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Figure 6. Growth Decomposition Panel A. Contribution to Real GDP Growth Panel B. Sectoral Value Added Shares (in percent, average 2000-10) (in percent of GDP) 6.0 50 TFP Human capital Labor Capital Fragile states 5.0 Frontier markets 4.0 40 Commodity exporters 3.0 30 2.0 1.0 20 0.0 fragile Fragile Comm. growing growing frontier Frontier Other All LIDCs Non- -1.0 exp. Non- Slow- 10 exp. Fast- -2.0 0 2000 2012 2000 2012 2000 2012 2000 2012 Agriculture Resources Manufacturing Services Sources: PRMED growth accounting database, WDI, and IMF staff estimates. 9. Growth in most LIDCs has not been accompanied by substantial structural transformation. The relative importance Figure 7. Challenges and Potential for Agriculture of the agricultural sector declined during Panel A. Agricultural Labor Productivity (index, 2000=100) Panel B. Quality Ladders in Agriculture and Manufacturing, 2010 2000–12, but the pace of change has been 160 EMs Panel B. Quality Ladders in modest in most cases and often 150 LIDCs Fragile states 140 accompanied by a decline (rather than 130 Frontier markets Commodity exporters Quality index expansion) in the relative share of 120 manufacturing (Figure 6 Panel B; World 110 100 Bank, 2014a, SSA Regional Economic 90 Outlook, Fall 2012). A significant fraction 80 of the population in LIDCs is employed in 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Agriculture Manufacturing Quality Quality Quality agriculture (particularly in SSA economies), Sources: WDI, Diversification Toolkit, and IMF staff estimates.ofQuality Ladder LIDC of EM ladders where labor productivity on average has reflect the extent of heterogeneity in quality across different varieties of a given product. The length of the quality ladders indicate the potential for quality grown slowly—lagging the corresponding upgrading for each product. growth rate in EMs—over the past decade, notably in fragile states (Figure 7, Panel A). The manufacturing base has remained narrow in the average LIDC, but with important regional differences: the share of manufacturing in GDP was higher in Asia’s LIDCs (12¼ percent average), a number of whom (e.g., Vietnam and Bangladesh) are well- integrated into global manufacturing value chains, but quite limited (and declining) in most SSA economies (7½ percent average), partly reflecting the relative importance of the natural resources sector and its limited positive spillovers to non-resource sectors for these economies (IMF, 2014b). While LIDCs currently occupy a lower position than EMs in estimated export product quality indices for agriculture and manufacturing, the scope for upgrading quality—as indicated by the length of the ladders—is substantial for both agricultural and manufactured products (Figure 7, Panel B; Henn and others, 2013). Services account for close to half of GDP in most countries, albeit reflecting a combination of a high productivity “modern” sector and a low productivity informal sector (see also Dabla-Norris and others, 2013). 16 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT 10. Progress toward meeting the Millennium Development Goals (MDGs) and reducing income inequality has Figure 8. Progress Toward Selected MDGs, by Number of LIDCs been mixed (Figure 8).10 0 While the latest data indicate 11 11 9 that 16 out of 60 LIDCs— 20 16 37 27 including frontier markets 3 5 9 1 8 2 1 such as Ghana, Senegal, 5 8 12 16 20 Uganda, and Vietnam—have 8 9 2 0 already met the target for MDG 1.1 - Extreme poverty (percent pop. MDG 2.0 - Primary MDG 4.1 - Infant mortality completion rate (percent rate (per 1,000 live births) MDG 7.1 Access to an improved water source extreme poverty reduction, living below $1.25 a day) Met of relevant age group) Sufficient progress (percent of population) Insufficient progress 20 countries—17 of them in Moderately off target Seriously off target Insufficient data Sources: WDI; Global Monitoring Report, 2013; and IMF staff estimates. SSA, and 12 are fragile Note: Progress is based on extrapolation of the latest five-year annual growth rates for each country. states—are considered Sufficient progress indicates that the MDG can be attained by 2015. Insufficient progress is defined as being able to meet the MDG between 2016 and 2020. Moderately off target indicates that the MDG “seriously off target,” can be met between 2020 and 2030. Seriously off target indicates that the MDG will not even be met meaning unlikely to meet by 2030. Insufficient data means that not enough data points are available to estimate progress or that the MDG’s starting value is missing. the target even by 2030. Progress has also been slow in regard to key development goals (e.g., primary school completion rate, infant mortality rate, access to an improved water source), with a large number of LIDCs projected to meet the targets only after 2020. This said, measured progress in meeting the MDGs is highly dependent on initial conditions. Many SSA countries have significantly improved their development indicators over the past 15 years—also the period with substantial improvement in the growth performance (SSA Regional Economic Outlook, Spring 2014). In addition, progress in reducing average income inequality in LIDCs has only been modest, but there are several success stories especially in SSA (e.g., Côte d’Ivoire, Mali, Niger, and Sierra Leone). Growth was supported by lower inflation and favorable fiscal and external developments. 11. Inflation has been on a declining trend since 2000, albeit with temporary reversals triggered by spikes in food and fuel prices (Figure 9). Figure 9. Inflation and Commodity Prices Tighter monetary policies, facilitated by reduced fiscal 16 60 dominance, have been central to achieving this trend Fuel price index (in percent change, RHS) Food price index (in percent change, RHS) 14 CPI in LIDCs (in percent change) decline. But the importance of food and fuel in 40 12 consumption patterns in LIDCs is such that surges in 20 10 international price for these products (2008 and 2011) 8 0 inevitably translate into inflation spikes that central 6 -20 banks have to accommodate. Over time, as financial 4 -40 markets develop in the more advanced LIDCs, 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 monetary policy frameworks in these countries (e.g., Sources: WEO and IMF staff estimates. Ghana, Kenya, Uganda, and Rwanda) have been shifting away from monetary-targeting based approaches toward more flexible forward-looking monetary 10 For a current assessment of progress toward meeting the MDGs, see the 2014 Global Monitoring Report. INTERNATIONAL MONETARY FUND 17 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT policy frameworks that give a central role to policy rates as the key policy instrument and the inflation outlook as a central focus of policy-setting. 12. Developments in the fiscal and external sectors were relatively favorable early on but deteriorated somewhat after the global financial crisis (Figure 10). Figure 10. Trends in Fiscal and External Sectors 1/ Panel A. Fiscal Balance Panel B. Public Debt (In percent of GDP) (In percent of GDP) 6 120 Fragile states 4 Frontier markets 100 Commodity exporters 2 All LIDCs 80 0 60 -2 40 -4 20 -6 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Panel C. Current Account Balance + FDI Panel D. Reserve Coverage (In percent of GDP) (months of imports) 20 9 8 15 7 10 6 5 5 4 0 3 2 -5 1 -10 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Sources: WEO and IMF staff estimates. 1/ Fiscal variables refer to the general government subject to data availability. Fiscal variables are GDP fiscal year weighted. External sector variables are PPP GDP weighted.  Fiscal positions improved markedly during the first half of the 2000s (Panel A) as revenue mobilization was stepped up and debt service eased with debt relief. After the global financial crisis, fiscal deficits increased with stimulus measures. Since then, government spending in many LIDCs has remained high while revenue mobilization has yielded relatively little; thus, deficits remain above pre-crisis levels (April 2014, Fiscal Monitor).  Public debt levels have eased significantly over time, reflecting debt relief and strong growth, with average ratios now stabilizing at around 31 percent of GDP (Panel B). Trends in debt are further examined in the final section of this report. 18 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT  Current account balances (augmented by FDI inflows) improved across most countries during the pre-crisis years (Panel C), markedly so in commodity exporters, contributing to reserve accumulation (Panel D). With reduced surpluses in recent years, import coverage levels have fallen again, but remain above three months of import cover target in most countries. Financing and trade structures in LIDCs are changing rapidly. 13. Capital inflows have increased sharply since 2000 (Figure 11) against a backdrop of strong global economic expansion, favorable financing conditions, and benign terms of trade. Starting at less than US$1 billion in 2000, net capital inflows to LIDCs reached some US$54 billion in 2012 (Panel A)—led by inflows into frontier markets—and were interrupted only briefly in the aftermath of the global crisis. Net FDI, the largest component of capital flows to LIDCs, increased sixfold during the period, primarily focused on the extractive sector (UNCTAD, 2014). While most FDI originated from advanced economies, a number of new players have joined from emerging markets, notably China. More recently, private portfolio inflows have also become significant in many frontier LIDCs (Panel B) as average non-FDI inflows to frontier markets increased to 2¼ percent of GDP during 2007–12, from less than 1 percent during 2001–06. Private financing of frontier LIDCs has increased while their net official development assistance (ODA) declined from a peak of 11¼ percent of GDP in 2002 to 5¾ percent of GDP in 2012.11 Figure 11. Capital Flows Panel A. Net Capital Inflows to LIDCs Panel B. Non-FDI Private Inflows to Frontier LIDCs (In US$ billion) (In percent of GDP) 80 FDI 10 Portfolio 60 Average 2001-06 Official flows 8 Bank flows Average 2007-12 40 Net capital flows 6 Net capital flows, frontier markets 20 4 0 2 -20 -40 0 Mozambique Ave. Frontier Zambia Nigeria Mongolia Bangladesh Senegal Kenya Bolivia Tanzania Vietnam Uganda Ghana PNG -60 -2 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 -4 Sources: Financial Flows Analytics database, and IMF staff estimates. Sources: Financial Flows Analytics database and IMF staff Source: Araujo and others (2014). estimates. 14. Trade links have increased steadily to countries other than the traditional advanced country markets (Figure 12). The last decade witnessed a significant shift in LIDCs’ trading partner composition toward emerging and developing countries (“South-South” trade) and away from advanced economies (Panel A). This trend partly reflected the increasingly closer ties between LIDCs and EMs in terms of FDI and development financing. China is emerging as an important export 11 ODA flows to fragile states (including debt relief) rose through 2010, but have declined as a share of recipient GDP since then. INTERNATIONAL MONETARY FUND 19 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT destination for LIDCs; the share of LIDC exports to China tripled from less than 5 percent in 2000 to 15 percent by 2010, with fuels accounting for the bulk of the export basket—but increasing shares of crude materials and manufactured goods (Panel B). Figure 12. LIDCs’ Export Destinations Panel A. Share of LIDC Exports to Partners Panel B. Composition of LIDC Exports to China 0.7 1.0 0.6 0.8 0.5 0.4 0.6 0.3 Advanced economies 0.4 0.2 EMDCs of which, China 0.2 0.1 0.0 0.0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2008 2009 2010 2011 2012 2013 Food and Beverages Crude Materials Fuels Chemicals Sources: COMTRADE, and IMF staff Sources: COMTRADE and IMF staff estimates. 15. Experiences with trade diversification, however, Figure 13. Export Product Diversification are uneven across LIDCs (Figure 13). Export product (Theil index) diversification entails introducing new higher value-added 4.5 Uganda Tanzania Vietnam Nepal products and/or upgrading the quality of the existing export 4.0 LIDCs basket. Diversification has been shown to be conducive to 3.5 faster economic growth in LIDCs in addition to being associated with lower output volatility (IMF, 2014b). Over the 3.0 past decade, most LIDCs have made little progress in 2.5 achieving export diversification, but there are important 2.0 exceptions, including Vietnam and several economies in East 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Africa. Sources: Diversification Toolkit, and IMF staff estimates. Note: Lower values of the Theil index indicate more diversification C. Recent Macroeconomic Developments and Outlook Robust growth and moderate imbalances in 2013… 16. LIDC growth in 2013 continued to be robust (Table 2), recording 6 percent on average (up from about 5¼ percent in 2012), driven primarily by strong domestic demand. While growth remained strong in frontier markets at about 6 percent, it was down compared with previous years, led by recent slowdowns in Ghana, Nigeria, and Vietnam. Meanwhile, growth picked up particularly strongly in fragile states—led by Myanmar and DRC, helped by improved political stability. Softer commodity prices and calibrated monetary policy tightening have helped lower average inflation from over 10 percent during 2010–12 to 8.2 percent in 2013. 20 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT 17. Both fiscal and current account deficits in LIDCs continued to widen in 2013 (Table 2). The deterioration was particularly marked on the fiscal front with the fiscal deficit reaching 3.2 percent of GDP on average. In some cases, the deterioration of the fiscal position reflected large increases in the wage bill and subsidies (e.g., Zambia and Lao P.D.R.) or election-related spending (e.g., Honduras); in others, the main driver was a revenue shortfall (e.g., Chad). Current account balances worsened notably in commodity exporters (e.g., Burundi and Democratic Republic of Congo), partly reflecting weak terms of trade. Meanwhile, reserve cover stood at about 3.7 months of imports in 2013, with fragile states experiencing continued deterioration. Table 2. Selected Macroeconomic Indicators, LIDCs and SubGroups Projections 2010-2011 2012 2013 2014 2015 Growth (percent) Average LIDCs 6.3 5.2 6.0 6.2 6.6 Frontier markets 7.0 5.4 5.8 6.5 6.7 Commodity exporters 6.3 4.3 5.8 6.2 6.9 Fragile states 3.3 3.8 6.1 5.2 6.5 Inflation (percent) Average LIDCs 10.5 10.1 8.2 7.4 7.4 Frontier markets 11.0 9.5 7.4 7.3 7.3 Commodity exporters 11.3 12.7 10.1 8.5 8.6 Fragile states 9.3 11.6 10.9 7.9 7.5 Fiscal Balance (in percent of GDP) Average LIDCs -1.9 -2.2 -3.2 -3.2 -3.0 Frontier markets -2.6 -2.6 -3.7 -3.4 -3.4 Commodity exporters -0.7 -0.4 -2.3 -2.1 -2.0 Fragile states -1.2 -2.7 -2.6 -3.0 -2.4 Current Account Balance (in percent of GDP) Average LIDCs -1.7 -2.5 -2.3 -2.2 -2.4 Frontier markets -0.5 -0.2 0.0 0.0 -0.4 Commodity exporters 0.2 -1.7 -1.3 -1.0 -1.2 Fragile states -3.0 -6.8 -6.5 -6.4 -6.2 Sources: WEO and IMF staff estimates. Note: Aggregates are computed using weighted averages. …with positive outlook despite increased downside risks to the global baseline scenario 18. LIDCs may face greater headwinds in the period ahead. Drawing on baseline projections from the WEO, notwithstanding the ongoing recovery in advanced economies, global growth is expected to reach 3.3 percent in 2014 and 3.9 percent in 2015, still significantly lower than an average of 4½ percent recorded before the global financial crisis (2000–07). Growth in emerging markets, which drove the recent commodities boom, is expected to shift down from over 6½ percent before the crisis (and 6.9 percent during the 2010–11 rebound) to an average of 4.6 percent during 2014–15. LIDCs’ trading partner growth, although picking up, is also set to remain lower than pre- crisis level (Figure 14, Panel A). Coupled with easing commodity prices (Figure 14, Panel B) (which affect net exporters of commodities) and the continued decline of aid flows, external conditions will be less supportive to LIDC growth compared to the period before the crisis. The current Ebola outbreak is expected to have a significant economic toll on the three most-affected economies: Guinea, Liberia, and Sierra Leone. INTERNATIONAL MONETARY FUND 21 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Figure 14. External Assumptions Panel A. Global Growth Projections Panel B. Commodity Price Projections (In percent) (index, 2005=100) 7 200 2000-07 6 2008-13 Fuel prices 2014 190 Food prices 5 2015 4 180 3 170 2 160 1 0 150 World Advanced EMs Trading 2013 2014 2015 2016 2017 2018 2019 par tners 1/ Sources: WEO and IMF staff estimates. 1/ Growth of LIDCs' trading partners, weighted by exports. 19. Nevertheless, the growth outlook in LIDCs is expected to remain strong supported by continuous implementation of structural reforms (Table 2). Real GDP growth in LIDCs as a whole is projected at about 6¼ percent in 2014 and 6.6 percent in 2015, a significant acceleration from 2012–13. The strong performance is set to occur broadly across LIDC subgroups, led by several frontier markets (e.g., Bangladesh, Kenya, Nigeria, and Senegal)—supported by continued efforts to implement critical reforms (e.g., energy sector reform in Nigeria) and improve business environment. Progress in rebuilding peace and stability and implementation of structural reforms (e.g., energy subsidy and civil service reforms) is expected to benefit fragile states, with growth continuing to accelerate in 2014–15, notably in Chad and Myanmar. Robust growth in LIDCs would also be supported by greater macroeconomic stability, with inflation projected to decline by over 1 percentage point between 2013 and 2014–15, with softer commodity prices and prudent monetary policy. 20. In the near term, deficits are expected to remain significantly above pre-crisis levels reflecting uneven efforts to improve fiscal buffers. On current policies, the average deficit in LIDCs is broadly unchanged in 2014 and is set to decline marginally in 2015, stabilizing at around 3 percent of GDP compared to a 0.4 percent of GDP overall surplus in 2008. However, there is large variation across countries. In about half of the LIDCs deficits are expected to decline, mainly in frontier countries (where recent fiscal slippages have resulted in market pressures).12 In some cases, the improvement in the fiscal position reflects delays or cuts in public investment (e.g., Lao P.D.R. and Zambia) and/or wage bill restraint (e.g., Ghana and Lao P.D.R.). In other cases, revenues are projected to increase thanks to higher oil revenues (e.g., Chad) and improvements in tax administration (e.g., Cambodia). In the other half of LIDCs, fiscal deficits are projected to go up mainly on account of higher current spending (e.g., Moldova and Mozambique) but also owing to infrastructure investment (e.g., Djibouti, Liberia, and Mozambique). The average public debt in LIDCs is expected to rise marginally in 2014–15. However, debt increases are likely to be sizable in a handful of countries 12 Despite the consolidation, the fiscal deficit in frontier markets will remain significantly above the LIDC average. 22 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT due to large deficits (e.g., Djibouti and Ghana) and/or significant increases in current spending (e.g., Niger). 21. There are important downside risks to the outlook. Downside risks include a weakening of macroeconomic policies, adverse global spillovers, and natural disasters including a worsening of the current Ebola outbreak. A protracted global growth slowdown, would negatively affect LIDCs through trade, remittances, commodity prices, and financial channels. Sharply higher global oil prices would have differential effects among LIDCs—benefiting oil exporters, but hurting oil importers, especially economies that face energy constraints related to a high cost of electricity. Financial vulnerabilities arising from potential adverse swings in capital flows would be particularly relevant for frontier LIDCs, making vigilant financial sector oversight a policy priority for these economies. Natural disasters can be particularly detrimental to growth in the poorest LIDCs with weak institutions. These risks are discussed further in the next section. INTERNATIONAL MONETARY FUND 23 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT SHIFTING VULNERABILITIES A. Introduction 22. LIDCs have generally performed well in the period since the global financial crisis, but remain vulnerable to shocks, both external and domestic. Figure 15. GDP Growth and Volatility Output volatility has been high for many countries, most 8 Fragile States, 28 Standard Deviation of Real GDP growth countries, notably fragile states (Figure 15). How well are they positioned 7 416 million Commodity Exporter, 27 6 (relative to LIDC average) to cope with shocks? How would different risk scenarios for countries, 5 579 million the global economy affect different LIDCs? These questions 4 3 Frontier are examined here, building on the methods used in the IMF’s 2 LIDCs, 60 countries, Markets, 14 countries, 2013 Low-Income Countries Global Risks and Vulnerabilities 1 1.3 billion 666 million 0 Report. The discussion expands on the analysis of 2 3 4 5 Real GDP Per Capita Growth (PPP GDP weighted, 6 vulnerabilities in previous IMF reports by exploring financial average 2003-2013) Source: WEO. sector vulnerabilities in LIDCs, with a special focus on frontier markets; and by examining the exposure of LIDCs to natural disaster shocks and the ensuing impact on growth and food security, an issue of particular relevance for the poorest LIDCs. 23. Key conclusions of the analysis are:  The share of LIDCs that are highly vulnerable is easing from its crisis peak, but the number of countries in the medium vulnerability group has picked up again since 2012. Weakened fiscal positions remain a key source of vulnerability across most LIDCs.  LIDCs are not immune to domestic financial sector weaknesses and, in frontier markets in particular, to global financial turbulence. While financial sector vulnerabilities have abated since the global recession, rapid credit growth and greater exposure to portfolio inflows warrant close monitoring, especially where modernization of regulatory norms and banking supervision has not kept pace with rapid financial development.  Natural disasters frequently impose large economic and human costs in LIDCs hampering growth but also contributing to weaker fiscal accounts and the likelihood of food crises in the poorest countries. 24. Strengthening macroeconomic policies to rebuild buffers is key to boosting resilience. Enhanced revenue mobilization and spending rationalization are critical to increase fiscal space. Reserve levels need to be rebuilt in a significant number of countries, especially fragile states. In economies where financial deepening is well underway, modernizing monetary frameworks should enhance the transmission mechanism and create room to let exchange rate movements absorb shocks. 24 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT B. Trends in Vulnerabilities: The Role of Fundamentals Some improvement… 25. Amid generally robust growth, the number of LIDCs deemed to be highly vulnerable to an adverse growth shock has Figure 16. Growth Decline Vulnerability Index, 2008–14 (LIDCs with Low, Medium, and High Vulnerabilities; in percent of total, PPP GDP eased since the global financial weighted) crisis. The extent of economic 100 vulnerability is assessed using a 80 growth decline vulnerability 60 index (GDVI), developed in 40 previous IMF board papers (see 20 13 Box 2). Using this metric, some 0 end-2008 end-2009 end-2010 end-2011 end-2012 end-2013 end-2014 10 percent of LIDCs are currently High Medium Low classified as highly vulnerable, Sources: WEO, IFS, DSA and staff reports, World Bank, and EM-DAT. with 43 percent in the medium vulnerability category (Figure 16). After a spike in 2009, the number of countries deemed to be highly vulnerable has eased gradually, but the number in the medium vulnerability grouping has been increasing again since 2012. The latter is due mainly to an increase in vulnerabilities in some key frontier markets (Nigeria, Ghana, and Vietnam). Small and/or poorer countries within the LIDC grouping typically record higher vulnerability scores than the average LIDC. Box 2. Methodology Underlying the Growth Decline Vulnerability Index 1/ The growth decline vulnerability index (GDVI) measures a country’s vulnerability to sudden growth declines in the event of a large exogenous shock. A range of indicators is examined to identify variables and thresholds to separate crisis from non-crisis cases. Thirteen variables are used in calculating three sectoral vulnerability indices: a real sector index based on such variables as GDP growth, Country Policy and Institutional Assessment (CPIA) scores (a broad indicator of political stability and quality of institutions), and natural disaster frequency; an external sector index based on such variables as reserve coverage, real export growth, exchange market pressure and export price changes, and a fiscal sector index based on variables such as the fiscal balance and the level of public debt. The weights assigned to each variable in constructing these indices depend on their ability to distinguish between crisis and non-crisis situations. The three sectoral vulnerability indices are then combined to establish the overall GDVI. ____________________ 1/ See Dabla-Norris and Bal Gündüz (2014). 13 See International Monetary Fund (2011a, 2011b, 2012a, and 2013a). INTERNATIONAL MONETARY FUND 25 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT ...with substantial differences across countries and sectors 26. Vulnerabilities based on the GDVI are lowest among frontier market economies, helped in good part by stronger external sector positions (Figure 17). By contrast, fragile states are most likely to be highly vulnerable, reflecting weaker growth performance, poor institutional quality, and both fiscal and external sector weaknesses. With elevated poverty levels (averaging 42 percent), fragile states are also those where adverse shocks are likely to have the strongest effects on the poor. Figure 17. Growth Decline Vulnerability Index by Country Figure 18. Growth Decline Vulnerability Index by Sector Group (Share of LIDCs with High Vulnerabilities; PPP GDP weighted) (Share of LIDCs with High Vulnerabilities, PPP GDP weighted) 60 60 50 50 40 40 Fragile States 30 Frontier Markets 30 Commodity Exporter 20 20 10 10 0 0 2008 2009 2010 2011 2012 2013 2014 2008 2009 2010 2011 2012 2013 2014 External Sector Fiscal Sector Real Sector Sources: WEO, IFS, DSA, IMF staff reports, World Bank, and EM-DAT. Sources: WEO, IFS, DSA, IMF staff reports, World Bank, and EM-DAT. 27. The fiscal sector is the primary source of vulnerabilities across country groups (Figure 18). The main drivers of fiscal vulnerability are: (1) elevated fiscal deficits (which remain well above pre-crisis levels in most countries); and (2) weak fiscal institutions—notably lack of a well-defined medium-term fiscal anchor and shortcomings in budget planning and execution that often result in spending overruns (Figure 19). By contrast, vulnerabilities in the external sector have receded somewhat relative to the peak crisis levels helped by improvements in current account positions and robust export growth. 26 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Figure 19. LIDCs: Assessment on Budget Planning and Execution (In percent of countries, A=high rating, D=low rating) a. Aggregate expenditure outturn b. Composition of expenditure c. Availability of data for monitoring d. Aggregate revenue outturn compared to original approved outturn compared to original the stock of expenditure payment compared to original approved budget 1/ approved budget1/ arrears2/ budget1/ 50 50 50 50 45 45 45 45 40 40 40 40 35 35 35 35 30 30 30 30 25 25 25 25 20 20 20 20 15 15 15 15 10 10 10 10 5 5 5 5 0 0 0 0 A B C D A B C D A B C D A B C D Sources: Public Expenditure and Financial Accountability Assessment (PEFA). Database and staff estimates. Percentages show the share of countries who received the score in question. The sample comprises of 45 countries and shows latest available data (ranging from 2005 to 2013). 1/ A, B, and C, are scores related to deviation up to 9%, 15% and more than 15% in 1 out of three years. D refers to deviation higher than 15% in 2 or 3 out of three years. 2/ A, B, C, and D are scores related to, respectively, reliable and complete data on the stock of arrears produced at least at the end of each fiscal year, reliable data generated annually but not necessarily complete for a few identified expenditure categories or budget institutions, data generated by at least one comprehensive ad hoc exercise within the last two years, no reliable data from the past two years. 28. Noticeable differences in Figure 20. Growth Decline Vulnerability Index by Sector and vulnerability patterns have emerged Region across regions. About half of LIDCs in the (LIDCs high vulnerabilities; PPP GDP weighted) end-2014 end-2013 Middle East/Central Asia have high 100 Real Sector 100 Fiscal Sector 100 External Sector vulnerabilities as a result of sluggish growth 75 75 75 (some 3 percentage points below the LIDC 50 50 50 25 25 25 average during 2011–14) and weak 0 0 0 institutions (Figure 20). More than one-half LAC MEU ASI SSA LAC LAC MEU ASI MEU ASI SSA SSA of SSA countries are rated at “medium Sources: WEO, IFS, DSA, IMF staff reports, World Bank, and EM-DAT. vulnerability” in 2014—a threefold increase from 2012—due to weaker fiscal indicators. Asian LIDCs have recorded increasing vulnerability scores in recent years, due to weakening external and fiscal indicators—but this group has consistently lower scores than other regions over time, helped by strong growth. C. How Vulnerable are LIDCs to Potential Global Shocks? 29. The exposure of LIDCs to global Figure 21. Shock Scenarios: Global Growth and Inflation shocks has increased significantly in (In percent) recent years. The key spillover channel 6.5 Growth Inflation Protracted slowdown 5.5 from advanced and emerging economies Oil price shock 5.0 5.5 continues to be the trade channel, but there Baseline Asynchronous exit 4.5 4.0 4.5 are also important linkages via investment 3.5 3.5 3.0 flows, remittances, and aid. Drawing on the 2.5 2.5 April 2014 and October 2014 WEO, we look 2.0 1.5 1.5 here at the potential impact on LIDCs of 2014 2015 2016 2017 2018 2014 2015 2016 2017 2018 three specific adverse scenarios (Figure 21): Source: WEO. INTERNATIONAL MONETARY FUND 27 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT  Protracted period of slower growth in advanced and emerging economies through 2018 affecting trading partner import growth and key commodity prices.  An energy price shock stemming from an escalation of geopolitical tensions, with the effects concentrated in 2014–15. Oil prices are assumed to rise relative to the baseline by 10 percent in 2014 and 15 percent in 2015.  Asynchronous normalization of monetary policies in key advanced economies, with a spike in global risk premia that produces some financial turmoil but with only a modest impact on global growth levels. 30. The protracted slowdown in advanced and emerging markets would have a substantial impact on LIDCs relative to baseline WEO projections in Table 2.14 Figure 22. Impact of Protracted Slowdown (Percent deviation from baseline, 2013–18)  Real GDP growth over the medium term (2014– 90 80 70 60 18) would fall short of the baseline by about 70 50 60 50 40 1.4 percentage points on a cumulative basis, with 40 30 30 20 20 the effect being most marked on commodity 10 0.0 0 10 0.0 0 exporters (Figure 22). Trade and investment -0.5 -0.5 -1.0 -1.0 -1.5 -1.5 -2.0 -2.0 linkages with China constitute a significant -2.5 Fragile States Frontier Markets -2.5 -3.0 -3.0 -3.5 Commodity Expor ter -3.5 transmission mechanism in this scenario. -4.0 LIDCs -4.5 -4.0 Other LIDCs -5.0 -4.5 -5.0 GDP growth (in percent) Cumulative Fiscal Cumulative Financing (LHS) Balance(in percent of GDP) Gap (LHS) (in US dollars  The fiscal balance in LIDCs would worsen by billions) (RHS) about 3 percent of GDP on a cumulative basis Source: WEO. relative to baseline, with commodity exporters (dependent on resource revenues) again being the hardest hit with a deterioration of nearly 4 percent of GDP as compared to some 2 percent of GDP for other LIDCs. As a corollary, debt levels would rise by about 3 percent of GDP over the medium term relative to baseline. The scope for discretionary countercyclical policies in the aftermath of the shock would be severely constrained in most countries, given already-elevated fiscal deficits, the likely shrinking of access to external commercial credit, and (for many LIDCs) limited domestic borrowing room in thin financial systems (IMF, 2013b).  Reserves (relative to imports) would fall most among commodity exporters, but other LIDCs, notably fragile states, would still face large financing needs to maintain adequate import coverage. For LIDCs as a group, financing amounting to about US$64 billion would be needed to restore international reserve levels to three months of import cover (or to pre-shock import coverage levels, if this was below three months). 14 The framework to analyze these shocks comprises a system of equations for the growth, fiscal, and external sectors to estimate the impact on LIDCs. For a detailed description of the methodology used for the scenario analysis, see IMF (2013a). 28 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT 31. The energy price shock scenario would have a much less severe impact on LIDCs as a group. While oil exporters would benefit, countries Figure 23. Cumulative Change Relative to Baseline with strong export links to adversely-hit economies 20 would be negatively affected under this shock 18 16 (Figure 23). Specifically: 14 12 10 8  For oil exporters, the primary impact would be 3 6 4 2 on the fiscal balance and reserves, with the 1 0 2 0 aggregates estimated to increase on a -1 -2 -2 -4 -3 Net Oil Exporter -6 cumulative basis by close to 2 percent of GDP -4 -5 Net Oil Importer -8 -10 and 1.8 months of imports respectively, by 2018. GDP growth (in Fiscal balance (in Financing need percent, PPP GDP percent of GDP, (USD Billion) (RHS) weighted) (LHS) GDP FY weighted)  For other LIDCs, output effects would be (LHS) Source: WEO modest, but the fiscal balance would deteriorate by a cumulative 2 percent of GDP by 2018. Given the deterioration in current account positions, the external financing needed to maintain adequate international reserve levels would be around US$17 billion by end-2018.  A key channel through which the price shock hits Figure 24. Energy Subsidies 1/ and Impact of Oil Shock 2/ fiscal positions is via its impact on energy (In percent of GDP) subsidies (Figure 24). Estimates based on 3.5 Post-tax energy subsidies Clements and others (2013) indicate that, on Fiscal impact (via subsidies) 3.0 average, energy subsidies in non-commodity exporters amount to about 2 percent of GDP on 2.5 Increase in subsidies after oil shock a post-tax basis. With a partial pass-through to 2.0 Level of subsidies retail prices (consistent with historical patterns), 1.5 the additional fiscal cost from fuel subsidies is 1.0 estimated at about 1 percent of GDP on average, 0.5 but exceeding 2½ percent of GDP in some cases. 0.0 LIDCs Importer Exporter LIDCs 32. The impact of an asynchronous Sources: Staff estimates, Organization for Economic Cooperation and Development, International Energy Agency, Deutsche normalization of monetary policies in advanced Gesellschaft Internationale Zusammenarbeit, IMF World Economic economies on international financial markets Outlook, and World Bank. 1/ Sample for LIDCs including 53 countries. Data as of 2011. would be significant, but the overall impact on 2/ Sample for LIDCs including 40 countries. Cumulative fiscal LIDCs would be very limited. As can be seen in impact in 2014–2018. Figure 21, the net impact on global output in this scenario is modest, yielding little effect on LIDCs via the trade channel. The impact on most LIDCs via the financial channel is minimal, given few direct links to international financial markets; the impact on frontier markets would be more marked, but is not adequately captured by the modeling methodology used. This issue is explored further below. INTERNATIONAL MONETARY FUND 29 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT D. A Closer Look at Financial Vulnerabilities 33. Financial systems in LIDCs, typically bank-dominated, have traditionally been relatively insulated from international financial developments, given limited access to external funding and the frequent presence of capital controls. With stable domestic funding from resident deposits, the key threat to financial stability has been the erosion of asset quality. Systems where state-owned banks play a lead role have been particularly vulnerable to such erosion, given the potential politicization of lending decisions (as in Vietnam). But asset quality erosion can also easily emerge when there is: a) rapid expansion of credit in an environment of weak internal controls and/or limited supervisory oversight (as in Nigeria, 2008–9); b) excessive exposure to specific sectors or corporates (a common feature in undiversified economies); and c) significant related-party lending by banks that are part of larger financial conglomerates (Cameroon 2010–11). The relatively benign experience of LIDC financial systems during the global financial crisis highlighted the direct insulation from external developments—but also flagged the indirect exposure as weaker exports and exchange rate adjustments took a toll on banks’ corporate borrowers (as in Zambia). 34. There has been significant financial development in many LIDCs in recent years, bringing new risks to financial systems. Financial deepening and broadening has proceeded, foreign investors are investing in domestic capital markets, governments have undertaken sovereign bond issues in international capital markets, and new financial instruments (including mobile banking and salary-backed lending) have taken off. Frontier market economies have seen the most far- reaching changes in these areas. Financial vulnerabilities in LIDCs after the global crisis 35. Financial sector vulnerabilities have abated since the peak of the global financial crisis but some LIDCs show renewed signs of potential Figure 25. Bank Return on Assets pressures. (In percent) 3  Banks’ profitability has been stable and reportedly higher 2 than in more advanced economies (Figure 25). This partly 1 reflects oligopolistic market structures and higher risk 0 levels (e.g., with respect to credit enforcement). In -1 aggregate, non-performing loan (NPL) ratios have also -2 declined steadily (even through the global financial crisis) 2003 2008 2012 -3 from 14 percent in 2003 to about 8 percent in 2013 with -4 provisions stable at above 70 percent throughout the LIDC FM EM AM period. Source: Bankscope. 30 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT  Insights into country-level vulnerabilities can be obtained by comparing trends in key vulnerability indicators with the country-specific Figure 26. Distribution of Z-Scores 15 (In percents) historic experience, using the z-score methodology. 100 Looking at the behavior of six financial variables in 28 90 80 LIDCs, large deviations occur in about 10 percent of 70 cases, a modest pick-up from post-crisis lows in 2011 60 50 (Figure 26; share of major deviation marked red). 40 Signs of stress in segments of the financial system 30 20 across LIDCs in previous years often reflected a 10 0 resumption of rapid credit growth after the global 2010 2011 2012 2013 financial crisis. More recently, pressures have been Green (<0.5 SD from median) Yellow (<0.5 and <2 SD) concentrated in rapid growth of cross-border Red (>2 SD above median) exposures (e.g., Cambodia and Uganda), and in Source: IMF staff estimates. stretched loan-deposit ratios (Haiti, Cameroon, Kenya, and Mongolia). However, z-scores need to be interpreted with caution, to the extent that they may reflect a catch-up from low levels of financial development and/or reflect large transactions of individual multinational investors and banks. A changing financial landscape warranting a closer look at frontier markets 36. Frontier markets pioneered the use of sovereign bond issuance by LIDCs with the number and the size of international bond issuances Figure 27. First Time Bond Issuances having increased markedly in recent years (Figure 27). By tapping non-domestic funding sources, sovereigns provide greater room for domestic bank lending to the private sector, expanding access to finance. In addition, the benchmarking role of sovereign bonds means that their issuance can also pave the way, over time, for the corporate sector to tap external markets, including for critical long-term infrastructure financing. Anecdotal evidence suggests that foreign participation in frontier domestic bond markets is also growing rapidly in a small number of countries. Source: Bloomberg. Given growing cross-border portfolio flows, LIDCs and frontier debt markets in particular, are more exposed to global volatility than in the past. While there appears to be a lower basic level of volatility for frontier markets, when compared to EMs, the volatility of bond spreads in frontier markets increased sharply at the times of more 15 Z-scores are standardized measures of how close an observation is to the historical median of a given variable. Z- scores cannot be interpreted as definitive threshold variables. The indicators used reflect both domestic and foreign channels: return on assets, bank credit to bank deposit ratio, cross border loans to GDP, cross border deposits to GDP, growth rate of M2 to GDP, and the growth rate of private credit to GDP ratio. INTERNATIONAL MONETARY FUND 31 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT extreme increases in the CBOE Volatility Index (VIX) in mid- Figure 28. Volatility in Frontier 2013 (Figure 28).16 In less volatile times, LIDCs thus seem to Markets and EMs (Rolling 20 day st. dev. of average sovereign benefit from a more dedicated investor class, possibly due to spreads) the higher transaction costs (e.g., information gathering) 35 25 involved, but once certain volatility thresholds are breached, 30 20 25 price movements can become more pronounced due to the 20 15 shallowness of markets. 15 10 10 5 37. A model-based assessment suggests that financial 5 0 0 vulnerabilities in some frontier markets are potentially Jan-13 Mar-13 May-13 Apr-13 Dec-13 Aug-13 Jun-13 Jul-13 Feb-13 Nov-13 Sep-13 Oct-13 significant, but data limitations preclude strong FM_Rolling SD EMs_Rolling SD conclusions. VIX (RHS) Source: Bloomberg; IMF staff estimates.  A financial vulnerability index for frontier markets was constructed based on IMF modeling techniques developed for emerging markets (see Appendix III). The analysis suggests that four- fifths of the economies are Table 3. Financial Vulnerability Index: Number and Share of Countries by Vulnerability Rating either “medium” or Number of countries Share of countries (In percent of total) “highly” vulnerable, albeit 2008 2009 2010 2011 2012 2013 2008 2009 2010 2011 2012 2013 with gradual Low 1 1 3 3 4 3 7 7 21 21 29 21 Medium 9 8 4 5 5 6 64 57 29 36 36 43 improvements since 2009 High 4 5 7 6 5 5 29 36 50 43 36 36 (Table 3). For the five Total 14 14 14 14 14 14 100 100 100 100 100 100 Source: IMF staff estimates. countries classified in the “high risk” spectrum in 2012–13, the key drivers of vulnerability are from banks’ balance sheet weaknesses (Bangladesh), rapid cumulative credit growth (Côte d’Ivoire, Mozambique, and Senegal) and/or a sharp rise in foreign liabilities as percent of domestic credit (Côte d’Ivoire and Vietnam). Figure 29. Capital Adequacy Ratios (In percent, as of 2013 Q4) 30  A closer examination of African frontier markets suggests that banking systems have low liquidity risk and are well- 25 capitalized and profitable, but potential vulnerabilities in 20 the area of foreign currency exposures (Tanzania, Uganda, 15 and Zambia) and rising NPLs (Uganda), could be 10 aggravated by the observed weaknesses of supervision in 5 the area of risk assessment and stress testing, and 0 challenges to consolidated supervision posed by the expansion of pan-African banks (Box 3 and Figure 29). Source: IMF FSI database. 16 Based on average spreads for all sovereign bonds of EMs and LIDCs rated “B” to “BB-.“ 32 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT  That said, model-based assessments need to be seen in the broader context of maintaining financial stability while promoting development. While financial deepening is welcome, the speed with which it is happening is a potential cause of concern, given that it involves a move into unchartered territory for regulatory agencies and banks alike. Box 3. Frontier FSAPs: Findings from the Basel Core Principles Assessments An analysis of banking supervision in the most recent FSAPs conducted in frontier countries between 2006 and 2013 suggests that significant progress has been made and many countries continue updating their practices, including through FSAP technical assistance. However, three areas stand out where compliance was low:  Anti-Money laundering supervision (BCP 18): Most frontier countries were found to be non compliant in anti-money laundering supervision. While the overall legal framework was in place, supporting regulations remained lacking; the perimeter of regulation was too narrow and did not capture shadow banks; fines imposed were not punitive; the Financial Integrity Unit lacked operational independence; and regulators still lacked capacity and adequate resources.  Prudential requirements (BCPs 12-15): While capital adequacy requirements were in line with Basel I, other material risks (including market, liquidity, and operational) were typically not properly assessed and regulators lacked the capacity to fully understand them with the result that they did not perceive them as material. Regulators provided limited guidance on these risks, and formal stress testing (or requirements for banks to conduct stress tests) was lacking.  Consolidated and cross border supervision (BCPs 24-25): Consolidated and cross-border supervision was not performed in eight of twelve frontier economies, mainly on account of a lack of legal authority. All countries are host to a number of foreign banks but BCP Non Compliance home regulators have often not been responsive to (In percent of countries) 60 Frontier Countries requests from host frontier countries, as these foreign 50 Emerging and Developing Countries operations are not deemed systemic. One example of recent 40 Advanced Countries progress in this area is in East Africa, where the Kenyan 30 central bank has organized supervisory colleges for all 20 major banks with cross-border operations, and is providing 10 training to host country supervisors in the region. 0 Information Cross Border Objectives, Autonomy, Licensing Enforecement Powers Methods of supevision Prudential Requirements The FSAP findings also highlighted issues that impact Powers supervisors’ “willingness to act.” In particular, independence of the regulator, which, while stipulated in the law, is still lacking in practice given arbitrary hiring and removal of supervisors, Source: Standards and Codes Database, BCP 2006 requirements for regulators to consult with the ministry of Methodology. finance on licensing or the issuance of new regulations, or insufficient legal protection in the carrying out of duties. Most supervisors were found to have adequate enforcement powers but lacked sufficient tools, skills, or proper guidance and procedures on the application of remedial actions, in part reflecting significant human and resource constraints. Finally, there were important organizational gaps, with on- and off-site supervision units often having limited contact, and interagency communications (both domestically and across borders) also being constrained. INTERNATIONAL MONETARY FUND 33 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT E. Natural Disasters: A Particular Challenge for LIDCs 38. LIDCs tend to be more vulnerable to losses from natural disasters than more developed economies.17  Poor infrastructure in LIDCs limits capacity to withstand disasters, buffers/resources to be deployed to affected areas are scarce, and economies are less diversified, reducing post-shock economic resilience (Laframboise and Loko, 2012).  The frequency of natural disasters in LIDCs has increased over time, and is expected to increase further with global warming (World Bank, 2014b), although the percentage of population affected by such natural events has remained largely constant (Figure 30).  LIDCs experience natural disasters more frequently at low per capita income levels within the LIDC sample (Figure 31). Figure 30. Natural Disasters and People Affected Figure 31. Natural Disasters by Income Distribution (Bubble size shows percent share of LIDC population affected) 2,500 average GDP per capita (LHS) 2,369.0 40 share of total natural disasters (in percent) 170 average people affected (in percent of population) Number of Natural Disaster per year 2,000 33.8 150 3.1 2.3 1.0 30 1.4 2.9 2.4 130 5.1 1.5 2.2 1,500 1,340.7 24.1 110 2.9 23.0 20 2.7 4.0 2.4 1,000 815.5 19.0 90 1.6 1.8 1.7 0.8 1.8 70 476.8 10 5.1 2.5 500 50 2.5 2.7 2.6 5.7 3.9 4.5 3.9 30 1.3 0 0 0-25th perc: 25-50th perc: 50-75th perc: 75-100th perc: 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Sources: EM-DAT: The International Disaster Database, CRED and IMF Sources: The International Disaster Database, CRED and IMF staff staff estimates. estimates.  Disasters are likely to have a higher impact on the poorest, given their limited capacity to respond (e.g., via savings and access to credit—see Hallegate and Przyluski, 2010).  Although natural disasters in this section do not specifically focus on epidemic outbreaks, some of its findings also apply to the current Ebola crisis (Box 4). 17 Natural disaster data come from the Emergency Events Database (EM-DAT) managed by the Centre for Research on the Epidemiology of Disasters (CRED). They comprise geophysical (earthquake), metereological (storms), hydrological (floods), climatological (droughts), and biological (epidemics) incidents. A disaster is registered in the database (occurrence of natural disaster) if one the following conditions is met: (i) 10 or more fatalities; (ii) 100 or more people “affected;” (iii) a call for international assistance; and (iv) the declaration of a state of emergency. People “affected” by a disaster include those injured, homeless/displaced, or requiring immediate assistance, but exclude fatalities. 34 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Box 4. The Ebola Outbreak in Guinea, Liberia, and Sierra Leone The Ebola Virus Disease (EVD) epidemic continues to intensify. The outbreak was initially concentrated in an area around the joint border of Guinea, Sierra Leone, and Liberia, but has spread to the capital cities of these countries and, to some extent, crossed other international boundaries. Epidemiologists have suggested that it may take six to nine months to bring the outbreak under control, provided containment efforts are implemented vigorously. The economic impact of the epidemic is expected to be significant. Although the number of infected people is still small as a share of population, the impact of quarantines and declining public confidence is being felt across all sectors of economic activity. Reduced output levels and disruption to both domestic and cross-border trade and services is expected to adversely affect both the fiscal accounts and the balance of payments in all three countries. The economic and social toll of the Ebola outbreak is expected to be most pronounced in Liberia, where both the capacity and resources to cope with the disease are lacking. The Fund is responding to requests for financial assistance from the three countries most affected by Ebola through augmentations of existing Fund-supported programs in Liberia and Sierra Leone, and the Rapid Credit Facility in Guinea. Total new commitments are US$129 million and the affected countries ’ financial needs will be reviewed again later in the year. The World Bank is mobilizing assistance of US$230 million — including a US$117 million emergency financing package—to tackle the outbreak in the three countries hardest hit by the crisis. 18 39. Major natural disasters can lead to a sharp deterioration of fiscal positions. The fiscal impact depends not only on the direct costs of the event (damages and loss of output), but also on the nature of the government’s reaction to the disaster. In general, public expenditure rises as a result of the immediate emergency assistance relief and the ensuing reconstruction efforts, but revenue collection may also fall as the domestic tax base contracts. To assess these effects in LIDCs, two approaches are adopted: 18 There is growing literature that focuses on the economic effects of natural disasters. For a recent analysis, see Melecky and Raddatz (2011); Dahlen and Saxena (2012); and Acevedo (2014). INTERNATIONAL MONETARY FUND 35 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT  Event studies. The EM-DAT database is used to analyze the impact of large natural disasters (defined in terms of the cost of damage) during the period 1980–2013.19 On impact, real GDP growth declines significantly (by about 2 percentage points) but only in the case of severe episodes.20 Following the event, growth picks up—mainly reflecting reconstruction spending— and gradually returns to trend (Figure 32). In parallel, the fiscal deficit widens by about 2 percent of GDP and the effect is persistent, but again only in severe episodes.  Empirical estimates. A panel vector autoregressive model for 54 LIDCs confirms that fiscal accounts mainly deteriorate because of higher spending. Expenditure increases on average by about 2 percent of GDP on impact and an additional 1 percent the year after, only slowly coming back to the baseline (Figure 33); the effect on revenues is not significant. The result is an increase in fiscal deficits with debt remaining some 10 percent of GDP above the baseline 8 years after the event. Figure 32. Natural Disasters in LIDCs: Event Study 1/ Figure 33. Natural Disasters in LIDCs: Impulse Response Real GDP growth Overall public deficit Functions 1/ (In percent) (In percent of GDP) 6 7 Expenditure Public Debt (In percent of GDP) (In percent of GDP) 5 6 6 25 5 5 20 4 4 4 15 3 3 10 3 2 2 5 2 1 0 1 75 percentile 75 percentile 1 90 percentile 90 percentile 0 -5 0 0 -1 -10 t-2 t-1 t0 t+1 t+2 t-2 t-1 t0 t+1 t+2 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 Sources: EM-DAT (International Disaster Database, CRED); and IMF Years (after shock) Years (after shock) staff estimates. 1/ Some 29 episodes were identified as falling above the 75th Sources: EM-DAT (International Disaster Database, CRED); and IMF percentile of the distribution of costs (2.1 percent of GDP), and 13 as staff estimates. falling above the 90th percentile (7.1 percent of GDP). 1/ Following Acevedo (2014), a panel vector autoregressive model with exogenous variables (PVARX) is estimated including the following variables: revenues, expenditures, real GDP growth, debt, terms of trade and debt relief. Confidence bands were obtained through bootstrapping methods based on 1,000 iterations. The red line shows the mean impulse response. The dotted lines show 10th/90th percentiles. 19 Although the frequency of natural disasters in LIDCs since 1980 has been very high, the median cost, at ½ percent of GDP is relatively small. Thus, subject to data constraints, the analysis focuses only on large episodes that are likely to have a significant impact on the fiscal accounts. Nevertheless, results should be interpreted with caution as some LIDCs have low levels of infrastructure and thus the cost of a disaster may not fully reflect its severity. 20 Data constraints do not allow assessing the extent to which the growth impact changes across country groups. However, recent studies have found that smaller economies face a much larger output decline after a disaster (Noy, 2009). 36 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT 40. Natural disasters also play a role in food shortages––a particular challenge in the poorest LIDCs. Figure 34. Food Supply Crisis in Comparison (1990–2009)  Fragile LIDCs and LIDCs in the lowest income 50 Share in sample * 8 quartile are about twice as likely to No. of Food Supply Crisis (overall) 7 40 38.6 experience episodes of a severe decline in 34.9 Probability of food supply crisis (overall) 6 food supply as other LIDCs (Figure 34).21 30 26 26.0 25.7 5 4 20 3.7 16 17 3  Empirical estimates suggest that 2 10 2.3 2.4 macroeconomic buffers (captured by higher 6 1 0.9 reserves and a lower fiscal deficit) limit the 0 0 Fragile State Commodity Frontier LIDCs Other (neither incidence of large food supply declines as Exporter Frag/Com/Front) governments are better able to deploy *Shares do not sum up to 100% since some countries are classified to more than one group. emergency assistance; other structural Sources: Food and Agriculture Organization of the UN (FAO) and Fund features, (e.g., level of education and staff estimates presence of social safety nets) also influence Figure 35. Food Decline Vulnerability Index severeness. For illustrative purposes, the (In percent share of countries, 2014) factors affecting countries’ ability to handle 100 90 natural disasters were aggregated into a risk- 80 weighted index, constructed in a similar 70 60 manner to the GDVI (Appendix IV). The index 50 40 suggests that about 10 percent of LIDCs are 30 currently highly vulnerable to a severe food 20 10 supply decline with the risk being somewhat 0 LIDCs Fragile States Frontier Markets Commodity higher in some fragile states and (some High Medium Low Exporters fragile) commodity exporters (Figure 35). Sources: WEO, IFS, DSAs, IMF staff reports, World Bank, and EM-DAT. F. Building Resilience in LIDCs: Policy Recommendations 41. How should LIDCs respond to the above vulnerabilities and risk exposure and enhance resilience? Macroeconomic policies need to be geared toward restoring buffers over time and enhancing flexibility. These efforts need to be complemented by structural reforms geared at economic diversification–– creating a real hedge to risks–– and improving the ability to cope with shocks (e.g., infrastructure investment for disaster prevention). While appropriate policies will depend on country-specific circumstances, some common priorities are worth highlighting. 42. Rebuilding fiscal policy space is a priority across many LIDCs, given that overall vulnerabilities stem mainly from weakened fiscal positions. As illustrated in the scenario analyses, many LIDCs have little room to operate countercyclical policies in the face of shocks unless fiscal positions are strengthened. This is particularly relevant among countries highly vulnerable to 21 Food decline episodes are identified using the food supply variable provided by the Food and Agriculture Organization (FAO). INTERNATIONAL MONETARY FUND 37 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT natural disasters. Creating fiscal space will require both improvements in revenue mobilization (IMF, 2011c) and spending rationalization. Public financial management (PFM) reforms have also an important role to play in entrenching macro-fiscal stability over time.22 43. Modernizing monetary and exchange rate frameworks as financial systems develop will also help.  Coping better with shocks will require strengthening monetary transmission channels. Financial sector reform and the deepening of markets have generally expanded the scope for active market-based monetary management, increasing the priority for more nuanced policy frameworks and the use of indirect instruments. Frontier markets such as Uganda and Ghana that have deeper financial systems, have made significant progress in this area, although preventing the re-emergence of pressures to accommodate lax fiscal policies may prove to be a challenge.  Greater exchange rate flexibility will facilitate the absorption of external shocks. The process of modernizing monetary frameworks described above is an important pre-requisite in this regard, allowing LIDCs to move away from de-facto exchange rate anchors.  More analysis is needed to assess the value of macroprudential tools in financially developed frontier markets. Early lessons from their adoption in emerging markets suggest that they need to be carefully calibrated against shifting stances in monetary and exchange rate policies and both targeted and communicated effectively. 44. Broader economic reforms can play an important role to support greater resilience over the medium term.  Economic diversification/structural transformation: The available evidence suggests that broad-based policies such as improving education, infrastructure, and the institutional and regulatory environment, and microeconomic reforms (e.g., liberalization of trade and promoting financial deepening) have proved successful in facilitating diversification.23  Natural disasters and food security. Actions are needed to secure supplementary buffers through insurance schemes, such as self-insurance via contingency reserves, and rapid access to disaster support lines from donors, while also freeing up resources for disaster preparedness. Better integration of markets can also play an important role in enhancing food security. LIDCs that are particularly vulnerable to natural disasters and food production shocks, such as in the drylands of SSA, should seek to lower trade barriers and improve road networks to create a larger regional market and allow a faster and more efficient response to localized food shortages.24 22 For details on the sequencing of PFM reforms and some case studies, see Diamond (2012); Alln, Hemming , and Potter (2013); and World Bank (2013). 23 For more details see IMF (2014b). 24 For example, between Burkina Faso, Ghana, and Benin the cost of transporting maize accounts for nearly 60 percent of final market prices due to monopoly and cartel rents, irregular payments, and poor and scarce infrastructure (USAID, 2011). 38 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT DEBT DEVELOPMENTS SINCE DEBT RELIEF 45. Public debt levels have declined markedly in the majority of LIDCs since 2000, but have been on an upward trend in a number of countries in recent years.25 At the same time, the external financing landscape is changing quickly: official concessional lending and grants are declining in relative importance, while a broader range of private financing instruments and lending sources are available for project finance and, more recently, for general budget financing needs. We focus here on the question as to whether recent trends point to threats to debt sustainability in LIDCs over the medium term.26 To explore this issue, this section first takes stock of the evolution of domestic and external public debt in LIDCs since 2007, examining the factors behind the debt dynamics, including the role of debt relief. It also looks at whether the fiscal space created by debt relief was accompanied by rising investment and social spending. Finally, the section reviews the risks and opportunities from the expanded access to new financing sources, using seven case studies to illustrate the major institutional challenges faced by these countries. 46. The main findings of this section are:  Debt sustainability is not a pressing concern for a majority of LIDCs. Public debt has stabilized at relatively low levels in most countries, helped by strong growth performance and by debt relief. That said, 12 LIDCs are classified as being either at high risk of, or in external debt distress.  Weaknesses in fiscal institutions persist that could endanger macroeconomic stability over time. Despite institutional improvements, many countries lack a well defined medium-term fiscal strategy; reforms to strengthen budget planning and execution are still a work-in-progress; and revenue mobilization efforts have stalled in many cases.  The lack of a comprehensive debt management strategy also poses risks—particularly for LIDCs now gaining external market access. Monetary policy normalization in advanced economies could create rollover and re-pricing risks, underscoring the importance of developing buffers and a medium-term debt strategy. 25 The Heavily Indebted Poor Countries (HIPCs) initiative was launched in 1996 as the coordinated debt relief effort of multilateral, bilateral, and commercial creditors. The initiative was enhanced in 1999 and complemented in 2005 by the Multilateral Debt Relief Initiative (MDRI). 34 LIDCs have benefited from the HIPCs Initiative, with four eligible countries yet to reach the completion point (IMF, 2013c). 26 For a recent discussion on debt sustainability in LIDCs, see Merotto, Thomas, and Stucka (2012); and Prizzon and Mustapha (2014). INTERNATIONAL MONETARY FUND 39 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT A. Stylized Facts Public debt in most LIDCs has stabilized at relatively low levels… 47. Over the last decade, public debt ratios in LIDCs have, on average, declined substantially (Figure 36).27 Following the early wave of debt relief that restored debt sustainability in a third of LIDCs (henceforth called “early HIPCs),28 public debt ratios declined considerably, falling below an average level of 35 percent of GDP in these countries by 2007; public debt also declined in non-HIPCs, albeit at a much slower pace. Since then, average debt levels have remained broadly stable while the dispersion across countries has been reduced as 14 “late-HIPCs” received debt relief. Debt-to-export and debt-to-revenue ratios also have shown similar trends. Meanwhile, the public and publicly guaranteed (PPG) external debt-to-GDP ratio has continued to decline, while domestic debt has been on the rise, taking the place of external debt in recent years. The bulk of PPG external debt remains concessional in nature (Figure 37),29 although the average level of concessionality (or grant element) of PPG external debt contracted since 2009 has trended down, as non-traditional creditors (e.g., emerging bilateral creditors, private investors)—increased their role. Figure 36. LIDCs: Public Debt 1/ Figure 37. LIDCs: External Debt by (In percent of GDP) Concessionality 1/ 2/ (In percent of GDP) 120 120 50 0.50 Concessional Non Concessional Domestic Debt 0.45 100 PPG External Debt 100 40 0.40 non-HIPCs 0.35 80 80 Early-HIPCs 30 0.30 60 Late-HIPCs 60 0.25 20 0.20 40 40 0.15 10 0.10 20 20 0.05 0 0 0 0.00 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2004 2005 2006 2007 2008 2009 2010 2011 2012 Sources: IMF, WEO and staff estimates. Source: World Bank, IDS. 1/ Afghanistan, Bolivia, Mongolia, Nicaragua, Papua 1/ Kiribati, Somalia, South Sudan, and Zimbabwe are excluded. New Guinea, Somalia, South Sudan, and Zimbabwe are 2/ Concessional debt is defined as debt that has a grant element excluded because of missing data. of at least 25 percent, calculated against a discount rate of 10 percent per annum. Loans from major regional development banks and the World Bank are classified as concessional according to each institution’s classification. 27 Throughout this section, public debt ratios refer to nominal general government gross debt in percent of GDP unless otherwise indicated. Data to calculate averages are weighted by annual nominal fiscal year GDP converted to U.S. dollars at average market exchange rates as a share of the group GDP. 28 “Early HIPCs” are countries that had reached the HIPC completion point before 2007; “late HIPCs” are LIDCs that have reached HIPC completion point since 2007; and “non-HIPCs” are those either not eligible for HIPC initiative or else eligible but yet to reach the HIPC completion point. 29 Caution is needed in interpreting the data on the evolution of concessional loans in Figure 37, as the definition of concessionality that is used to classify part of external debt is taken from the OECD-DAC and differs from that used by the World Bank and the IMF. 40 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT 48. The aggregate data mask emerging divergences across LIDC groupings (Figure 38). Some LIDCs have enjoyed sharp declines in public debt levels since 2007, most obviously “late HIPCs,” who saw debt-to-GDP ratios drop by Figure 38. LIDCs: Changes in Public Debt Ratios, 2007–13 1/ 51 percentage points on average. By contrast, (In percent of GDP) debt levels rose by about 7½ points of GDP in 20 Top percentile early HIPCs over the period, with countries 10 Middle 0 taking advantage of both external and domestic -10 Quartiles range markets to finance higher deficits. Public debt -20 Low percentile ratios in non-HIPCs showed no clear trends over -30 -40 the period, with debt ratios rising in some cases, -50 Weighted average falling in others. -60 -120 -70 Frontier Fragile Commodity Late HIPC Non HIPC Early HIPC All LIDCs exporters 49. There also have been changes in the composition of debt: Sources: IMF, WEO and staff estimates. 1/ Madagascar, Somalia, and South Sudan are excluded because of  In early HIPCs, external borrowing accounts missing data. for most of the debt increase (Figure 39), with non-concessional external borrowing increasing in importance (Figure 40). But domestic borrowing rose sharply in a number of countries, including Ghana and Malawi.  In non-HIPCs, domestic debt has been on the rise, in step with expanding local currency bond markets, in many cases substituting for external debt. Thus, Nigeria and Kenya have developed active local currency bond markets, supported by regular issuance and with well-defined yield curves. Figure 39. LIDCs: Changes in PPG External and Domestic Figure 40. Early-HIPCs: Net Lending Flows 1/ Debt-to-GDP Ratios by Country Groups 1/ 2/ (In percent of GDP) (In percentage points) 15 15 5 5 10 10 Non-Concessional Concessional 5 5 4 4 0 0 -5 -5 3 3 Domestic Debt -10 -10 PPG External Debt 2 2 -15 -15 -20 -20 1 1 -25 -25 -30 -30 0 0 Non HIPC Early HIPC All LIDCs Late HIPC Fragile 2/ Frontier 2/ Exporters 2/ Commodity -1 -1 2000 2002 2004 2006 2008 2010 2012 Source: IMF, WEO. Source: World Bank, IDS. 1/ Afghanistan, Bolivia, Mongolia, Nicaragua, Papua New Guinea, 1/ Concessional debt is defined as debt that has a grant element of at Somalia, South Sudan, and Zimbabwe are excluded because of least 25 percent, calculated against a discount rate of 10 percent per missing data values. annum. Loans from major regional development banks and the World 2/ Excludes late HIPCs. Bank are classified as concessional according to each institution’s classification. INTERNATIONAL MONETARY FUND 41 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT 50. The external debt service burden has eased, Figure 41. LIDCs: Debt Service on External Debt 1/ reflecting both debt relief and, more recently, favorable (In percent of GDP) 4.0 4.0 global financial conditions. The average interest rate on HIPCs non-HIPCs 3.5 3.5 PPG external debt has hovered around 2-2½ percent of 3.0 3.0 GDP over the period, while the average debt service 2.5 2.5 ratio has fallen below 2 percent in recent years (Figure 2.0 2.0 41). Average interest rates on domestic debt, at 1.5 1.5 8½ percent, have been broadly stable in line with the 1.0 1.0 inflation differential with the U.S. 0.5 0.5 0.0 0.0 2004 2005 2006 2007 2008 2009 2010 2011 2012 51. However, in one third of countries debt levels Source: World Bank, IDS. are high and/or increasing quite significantly. Nine of 1/ Kiribati, Somalia, South Sudan, and Zimbabwe are excluded. the 60 LIDCs currently have public debt levels that exceed the risk threshold levels identified in the Bank-Fund Debt Sustainability Framework (DSF); a further seven countries are deemed to be either at high risk of, or already in external debt distress; and a further three countries have seen public debt levels increase by at least twenty points of GDP since 2007. …but stable debt ratios have been underpinned by Figure 42. LIDCs Public Debt Decomposition, 2007–13 1/ favorable macroeconomic conditions. (In percent of 2013 GDP, cumulative) 40 52. Debt dynamics in the post-HIPC/MDRI period 20 Primary deficit have been largely driven by strong economic growth 0 r–g differential rather than improvements in the primary balance. -20 The change in the debt-to-GDP ratio over time can be Stock-flow adjustment accounted for by four variables: (1) the primary fiscal -40 balance (in percent of GDP); (2) the interest rate-growth Debt relief -60 rate differential (IRGD);30 (3) debt relief; and (4) stock flow -80 Change in Commodity Fragile Frontier Late HIPC All LICs Early HIPC Non-HIPC adjustments, which usually reflect accumulation of debt-to-GDP ratio exporters arrears, valuation changes from exchange rate Sources: World Bank, IDS and IMF, WEO staff estimates. movements, and the realization of contingent liabilities Note: Madagascar, Somalia, and South Sudan are excluded because of missing data. (see Weber, 2012). For LIDCs as a whole, this decomposition reveals that debt accumulation during 2007–13 has been largely contained by a) strong growth—which, helped by concessional interest rates,31 resulted in a negative interest rate- growth differential (IRGD); and b) for late HIPCs, debt relief (Figure 42). Over the same period, the 30 For a given debt stock, higher interest rates mean that a larger share of public resources needs to go toward paying interest, leaving fewer resources to pay down the debt. In contrast, faster growth brings down debt ratios more quickly, by increasing the denominator in the debt-to-GDP ratio and by making it easier to run larger primary balances. 31 In many LIDCs, real interest rates have been negative (particularly in fragile states), in some cases offsetting the impact of subdued economic growth (Eritrea and Guinea). 42 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT average primary balance in LIDCs deteriorated by 2.1 percent of GDP, switching from a surplus into a deficit position. The contribution of the primary deficit to debt accumulation was the highest among early HIPCs—notably in Ghana, Honduras, Senegal, and Tanzania. Stock-flow adjustments contributed significantly to debt accumulation across all country groups. 53. Relative to the pre-crisis period, the scope for expansionary fiscal policy (“fiscal policy space”) is now more constrained. On average, the fiscal deficit in LIDCs has widened by 3.6 percent of GDP compared to 2008, limiting the room for maneuver in many cases (IMF, 2014d). The severity of this constraint depends on the debt stock and debt service levels and, ultimately, on the access to finance. The ability to finance higher deficits is tightly constrained in most LIDCs as they have shallow financial systems and limited access to international financial markets (IMF, 2013b). 54. Tighter financing conditions and lower aid flows are expected to put pressure on fiscal positions:  Increase in financing costs. To illustrate the sensitivity of the fiscal space to changes in financing conditions stemming from higher global interest rates and increased risk aversion, a model of the IRGD is estimated for a panel of 51 LIDCs over 2000–13.32 The results suggest that for every 100 bp increase in the U.S. long-term bond yields, the IRGD is expected to increase by 60 bp in LIDCs as a whole. The impact varies markedly with the prevailing global risk aversion, as measured by the VIX.33 Under a high risk aversion scenario, the primary balance gap would deteriorate by about ½ percent of GDP, with a somewhat larger effect among early HIPCs and fragile countries.  Decline in net official development assistance (ODA). Aid data from donors indicate that disbursements to LIDCs will likely continue their downward trend (as a share of recipient GDP) during 2014–15, with sizeable effects on many recipients: in a quarter of LIDCs, ODA could decrease at least 2 percentage points of GDP. A shock of this magnitude to budgetary resources, whether to revenues or concessional loan resources, would place a severe strain on budgetary finances, either forcing significant fiscal adjustment or worsening debt dynamics unless offset by revenue mobilization efforts.34 32 The model specification follows closely Escolano, Kolerus, and Lonkeng Ngouana (2014), controlling for U.S. interest rates, global risk aversion (measured by the VIX), inflation, current account deficit, openness, public debt, and concessional borrowing. The sensitivity of the IRGD to monetary tightening is somewhat lower than in EMs, which likely reflects the higher reliance on concessional lending among LIDCs. 33 The impact of a 100 bp increase in the U.S. yields rises to [179] bp under a high risk aversion scenario whereby the VIX is at the [90] percentile of historical values. 34 The literature generally finds a negative association between net ODA and domestic tax revenues (Benedek and others, 2014). INTERNATIONAL MONETARY FUND 43 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Rising public spending in recent years partly reflects higher social and capital spending. 55. Expenditure ratios in LIDCs, and in particular HIPCs, Figure 43. LIDCs: Changes in Expenditures and Revenue, 2007–13 /1 have been on the rise, not always matched by revenue (In percent of GDP) increases (Figure 43). The trend on expenditures reflects 7 Current several factors: (i) a conscious effort to use the newly created spending 5 fiscal space to respond to pressing developmental needs; Capital 3 spending (ii) fiscal stimuli in the aftermath of the global recession; and 1 (iii) expenditure overruns relative to original budgets—in some Revenues -1 cases related to exogenous shocks such as natural disasters -3 Revenues w/o (see case studies below). Revenue mobilization efforts have grants -5 often fallen short of what is needed to offset the impact of the Fragile Commodity Frontier Late HIPC Non HIPC Early HIPC All LIDCs exporters ongoing decline in grants. Sources: IMF, WEO and staff estimates. 1/ Afghanistan, Bhutan, Bolivia, Kiribati, 56. In more than half of countries, spending increases Madagascar, Mongolia, Papua New Guinea, Solomon Islands, Somalia, and South Sudan are are associated with the scaling-up of investment: excluded because of missing data.  In late HIPCs, capital spending has gone up by Figure 44. LICDs: Public Investment in 5 percentage points of GDP since 2007 (Figure 44). This LIDCs 1/ (In percent of GDP) sharp increase partly reflects investment related to the 12 12 exploitation of natural resources (the Republic of Congo) Late-HIPCs LIDCs Non-HIPCs Early-HIPCs 10 10 or reconstruction spending in the aftermath of natural disasters and conflict (Haiti and Liberia). But higher capital 8 8 spending to build infrastructure is not limited to these 6 6 cases, with increases also being recorded in Democratic 4 4 Republic of Congo, The Gambia, and Côte d'Ivoire. 2 2  In early HIPCs, the increase in public investment has 0 0 generally been more muted in recent years, as most of 2000 2002 2004 2006 2008 2010 2012 Source: IMF, WEO. them achieved high levels of investment immediately after 1/ Afghanistan, Kiribati, Papua New Guinea, São receiving debt relief (prior to 2007). Nevertheless, as in late Tomé and Príncipe, Solomon Islands, Somalia, South Sudan, and Zimbabwe are excluded because of HIPCs, some countries have recently undertaken large missing data values. investment projects aiming to improve their infrastructure (Cameroon, Sierra Leone, and Rwanda). 57. In about half the countries, the bulk of spending increases has been on current outlays. The data suggest that the public sector wage bill has gone up by ¾ percent of GDP on average since 2007, reaching 6¼ percent of GDP or about 40 percent of current spending; some of the increase in the wage bill has benefited health and education, but much has gone elsewhere in the public sector. Increases have been particularly large among some early HIPCs (Ghana, Malawi, and Mozambique) and non-HIPCs (Laos, Lesotho, and Zimbabwe). 44 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT 58. Concerns remain about the effectiveness of public spending and extent to which it is supporting growth.  Public investment. The efficiency with which LIDCs convert public investment into public capital is generally much lower than in advanced and emerging economies (Gupta and others, 2014). The efficiency of public investment plays a central role in determining the size of the growth dividend from investment scaling up, and indeed the wisdom of undertaking such strategy in the first place (Figure 45 and Box 5). It is estimated that reducing all inefficiencies by 2030 would provide the same boost to capital stock as increasing government investment by 14 percentage points of GDP (April 2014, Fiscal Monitor). Boosting efficiency will require improving project implementation (via competitive bidding and internal audit) and project selection,35 as well as ensuring maximum standards of fiscal transparency and performance accountability when projects are executed through public-private partnerships (PPPs).36 Figure 45. The Impact of Improving Public Investment Efficiency—Simulation Results 1/ Public capital Growth Total public debt (percent deviation from the steady state) (In percent) (In percent of GDP) 70 6 90 Decreasing Efficiency 60 Keeping Efficiency Constant 80 5 Increasing Efficiency 70 50 4 60 40 50 3 30 40 20 2 30 Decreasing Efficiency Decreasing Efficiency 10 Keeping Efficiency Constant 20 1 Keeping Efficiency Constant Increasing Efficiency 10 Increasing Efficiency 0 0 1 3 5 7 9 11 13 15 17 19 21 23 25 0 1 3 5 7 9 11 13 15 17 19 21 23 25 1 3 5 7 9 11 13 15 17 19 21 23 25 Source: IMF staff estimates. 1/ For illustrative purposes, the DIG (Debt, Investment and Growth) model is calibrated to an average LIDC, following Buffie and others (2012), and simulated as a stylized version of the experiments that were run in the country applications. The average LIDCs efficiency is set at 0.5, in line with Pritchett (2001) and Gupta and others (2014).  Current spending. Energy subsidies in LIDCs, amounting to about 3¼ percent of GDP when measured on a post-tax basis, crowd out private investment and distort resource allocation by encouraging excessive energy consumption (Clements and others, 2013). There is also significant scope for improving the efficiency of health and education spending in LIDCs (Kapsoli and Grigoli, 2013; Grigoli, 2014; and April 2014, Fiscal Monitor). 35 Institutional reforms take time, so policymakers need to balance the implications of capital scarcity (and associated high rates of return to capital) against the costs of scaling up too rapidly when the institutions supporting the efficiency of spending are not fully in place. 36 As of 2012, PPP commitments in LIDCs amounted to 8 percent of GDP on average. INTERNATIONAL MONETARY FUND 45 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Box 5. Public Investment Scaling-Up, Growth, and Debt Sustainability in LIDCs A model-based framework is used to analyze the effects of public investment scaling-ups on growth and debt sustainability in 10 developing countries. The framework, developed by Buffie and others (2012), incorporates the public investment-growth nexus and allows for different financing strategies (external concessional, external commercial, and domestic). The simulations reflect ambitious, front-loaded investment plans in line with countries’ National Development Plans (NDPs) or PRSPs. The plans are not fully funded by concessional loans and grants, resulting in a financing gap that could be potentially covered by fiscal adjustment, external commercial or domestic borrowing. The analyses reveal that, despite long-run growth benefits macroeconomic challenges and trade-offs accompany the different financing strategies, raising the stakes for policymakers. Covering the financing gap associated with the scaling-up program with tax increases or spending rationalization requires sharp and unrealistic macroeconomic adjustments, crowding out private investment and consumption in the short to medium term—e.g., in Rwanda, an increase of 5 percentage points of public investment to GDP financed with taxes would lead to a 10 percent decline in consumption relative to its initial level. Financing the program with more domestic borrowing can still crowd out the private sector. On the other hand, external commercial borrowing can ease the fiscal and macroeconomic adjustment. But these gains have to be balanced against the build-up of debt distress risks—e.g., in a simulation for Togo, more external borrowing comes at the cost of increasing the debt-to GDP ratio by 10 percentage points. Securing more aid, as well as more gradual public investment scaling-ups, can help counterbalance these risks. The applications also show that improving investment efficiency and revenue mobilization are key to reaping growth benefits of scaling-ups while minimizing the associated debt sustainability risks in LIDCs. For the average LIDC, a 10-year average scaling-up of 4.5 percentage points of public investment to GDP coupled with an increase of efficiency from 0.5 to 0.75 could deliver a gain of 1.4 percentage points in per-capita GDP growth over a decade.1 By contrast, with declining efficiency (from 0.5 to 0.25), scaling-up plans would generate much smaller growth dividends; and if these plans were to be financed with more non-concessional borrowing, public debt could become unsustainable (see Figure 45). To address these risks, additional reforms to improve efficiency (e.g., by strengthening the appraisal, selection, implementation, and evaluation of projects) coupled with revenue mobilization and spending rationalization would help to strengthen debt sustainability, even with external commercial borrowing, while still securing growth dividends. ____________________ 1 For illustrative purposes, the DIG (Debt, Investment, and Growth) model is calibrated to an average LIDC, following Buffie and others (2012), and simulated as a stylized version of the experiments that were run in the country applications. The average LIDCs efficiency is set at 0.5, in line with Pritchett (2001) and Gupta and others (2014). 46 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT B. Risk Diagnostics The risks of experiencing external debt distress remain contained in most LIDCs… 59. Based on the joint analysis of IMF and World Bank staffs, the risk to LIDCs of experiencing debt distress over the medium term has improved somewhat since 2007 (Figure 46).37 Risk ratings for 44 countries out of a sample of 56 countries are currently low or moderate—including all frontier markets and early HIPCs. Overall, risk ratings have improved, even when post-2007 debt relief recipients are excluded. Some 12 other countries have recorded an improvement in their risk rating since 2007, while eight countries have received a downgrade. Downgrades of the rating of debt distress risk have typically resulted from at least one of the following: a) a steep increase in non-concessional debt (Chad, Mongolia, Mozambique, and Niger), b) downward revisions to the macroeconomic outlook and/or weaker-than-expected fiscal adjustment (Honduras, Mongolia, and Sao Tomé and Principe); and c) a deterioration in the outlook for natural resource revenues and exports (Mali and Sao Tomé and Principe). Nevertheless, stress tests simulated under the Bank-Fund DSA indicate that the debt outlook is particularly sensitive to adverse shocks to the exchange rate and external financing terms. Figure 46. LIDCs: Changes in Risk Rating for External Debt Distress 1/ In Debt Distress Zimbabwe, Sudan Comoros, Congo, Dem. Rep., Central High Chad, Sao Tome and Principe Haiti, Djibouti, Afghanistan African Republic, Burundi Sierra Leone, Nicaragua, Niger, Malawi, 2013/14 Mozambique, Mongolia, Mali, Mauritania, Lesotho, Kyrgyz Republic, Yemen, Tajikistan, Solomon Togo, Guinea-Bissau, Guinea, Cote Moderate Honduras, Cameroon Ghana, Bhutan, Burkina Faso Islands, Lao PDR, Gambia d'Ivoire Zambia, Vietnam, Uganda, Tanzania, Senegal, Nigeria, Madagascar, Moldova, Kenya, Rwanda, Papua New Guinea, Cambodia, Low Bolivia, Bangladesh Ethiopia, Benin Nepal, Congo, Rep. Myanmar, Liberia Low Moderate High In Debt Distress 2007/08 Source: IMF, DSAs. 1/ Bolded italicized letters indicate late HIPCs. Ratings in 2007/08 are based on DSA issued in 2007 (or 2008 if no DSA was issued in 2007) and ratings in 2013/14 are based on the latest DSA issued in 2013 and 2014. For Chad, Honduras, and Zimbabwe, the rating in 2007/08 is that in 2009 as no DSAs were issued in 2007/08. Eritrea, Kiribati, South Sudan, Somalia, and Uzbekistan are excluded because of missing DSAs. Downgrade of Niger (2011) did not lead to a change in rating from 2007/08 to 2013/14 as it was a reversal of the previous rating. 37 These assessments are conducted using the Bank-Fund Low-Income Country Debt Sustainability Framework. The risk ratings are dependent on both medium- to long-term macroeconomic projections and the impact of standardized stress tests upon the baseline projections. INTERNATIONAL MONETARY FUND 47 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT …but the composition of the creditor base is changing rapidly, bringing new opportunities and challenges… 60. Bilateral creditors are taking on a more prominent role. Bilateral net lending was on a declining trend through 2007, as some non-HIPCs (Nigeria) Figure 47. LIDCs: Net Lending by Type of made principal repayments. In the wake of the debt relief Creditors (In percent of GDP) 1/ initiatives, bilateral creditors expanded their lending to now- 1.6 1.6 credit worthy HIPCs (Figure 47). Net lending by multilateral 1.4 Private 1.4 Bilateral creditors rose temporarily in 2009–10, as LIDCs sought 1.2 Multilateral 1.2 1.0 1.0 external financing to help cushion the impact of the economic 0.8 0.8 slowdown, but contracted in the ensuing years. By 2012, net 0.6 0.6 lending from bilateral creditors was close to the level provided 0.4 0.4 by multilateral creditors—a large increase from pre-crisis 0.2 0.2 0.0 0.0 years—reflecting the expanding role of new creditors in 2007 2008 2009 2010 2011 2012 Sources: World Bank, IDS and IMF, WEO. infrastructure finance. 1/ Kiribati, Somalia, South Sudan, and Zimbabwe are excluded. 61. Private creditors have also expanded their lending recently. In particular, frontier markets have begun to issue international sovereign bonds with the stated objective of increasing public investment (discussed in vulnerabilities section. However, the correlation between bond issuance and public investment has been weak, with the funds being used instead for both capital and current spending, including on wages and subsidies (IMF, 2013b and 2014c). Some countries (Nigeria) issued bonds to provide a benchmark for sub-national and corporate issuers, others (Côte d’Ivoire) for finance debt restructuring. 62. The new sources of funding can provide major benefits but also create risks. Issuance abroad can provide funding for large projects, creates domestic financing space for the private sector, can be less costly than local issuance, and can contribute to greater policy discipline and transparency through increased market scrutiny. However, the new debt also comes with refinancing risk, re-pricing risk, and exposure to exchange rate depreciation. If debt management capacity is limited and public investment capacity weak, use of international sovereign bond issuance, even if used solely for public investment, may still be inadvisable (Box 6). …and case studies point to the need for better fiscal institutions. 63. Seven case studies highlight different aspects of debt dynamics in LIDCs. The sample included countries where the debt-to-GDP ratio has declined since 2007 (Djibouti and Haiti), remained broadly stable (Kenya and Mozambique), and increased (Ghana, Honduras, and Vietnam)— (Appendix V presents selected indicators).  Debt accumulation was not always reflected in higher public investment. In four cases (Djibouti, Haiti, Kenya and Mozambique), the increase in the primary deficits was associated with capital spending (Figure 48); in two cases (Ghana and Honduras), the public sector wage bill was a major source of pressure. Energy subsidies have been pervasive in all countries except in Kenya, with an estimated cost of about 3 percent of GDP on average. 48 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Figure 48. Fiscal Trends, 2007–13 a. Decomposition of Public Debt b. Changes in Expenditures and Revenues (In percent of 2013GDP, cumulative) (In percent of GDP) Primary deficit r–g differential Stock-flow adjustment Debt relief Current spending Capital spending Revenues Revenues w/o grants Change in debt-to-GDP ratio 60 12 10 40 8 20 6 4 0 2 -20 0 -2 -40 -4 -60 -6 LIDCs DJI GHA HTI HND KEN MOZ VNM LIDCs DJI GHA HTI HND KEN MOZ VNM Sources: World Bank, IDS; and IMF, WEO and staff estimates.  Debt increases also arose from other factors, including quasi-fiscal operations of state-owned enterprises (Mozambique), clearance of arrears (Honduras), and the materialization of contingent liabilities, such as bank recapitalization or called guarantees (Vietnam).  In general, poverty-reducing social spending has not increased with debt accumulation, undermining the argument that higher borrowing levels are being used to accelerate progress toward meeting the MDGs.  Budget overruns on current expenditure contributed significantly to elevated deficits (Figure 49), including in Djibouti, Honduras, Kenya, and Ghana even though capital spending was under-executed because of weak institutional capacity (Haiti) or to create some room for additional current outlays (Honduras and Ghana).  In some cases, exogenous shocks (including natural disasters) have contributed to weakened fiscal performance and subsequent debt accumulation. The Haiti earthquake in 2010 impaired tax-collection capacity while emergency relief and reconstruction spending went up. Similarly, in the face of the drought in the Horn of Africa in 2011 and a rise in food prices, fiscal deficits widened in Djibouti and Kenya due to the associated fall in tax revenues and emergency spending. In Ghana, revenue underperformance in 2013 partly reflected a slowdown in economic activity and weaker corporate income taxes due to the fall in gold prices. INTERNATIONAL MONETARY FUND 49 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Figure 49. Fiscal Policy Slippages—Frequency and Annual Average of Underperformance 1/ (Deviation from targets, 2007–13) Overall Balance Capital Expenditure 2 4 Annual average (pct of GDP) Annual average (pct of GDP) HTI MOZ MOZ 1 VNM 2 VNM DJI 0 KEN KEN 0 HND HND -1 DJI -2 -2 GHA GHA HTI -3 -4 0 50 100 0 50 100 Current Expenditure Underperformance (pct of years) Public Wage Bill Underperformance (pct of years) 2 1 Annual average (pct of GDP) GHA GHA Annual average (pct of GDP) KEN, DJI MOZ 1 0 DJI HND KEN HTI 0 HTI HND -1 -1 -2 -2 MOZ VNM VNM -3 -3 0 20 40 60 80 0 20 40 60 80 Underperformance (pct of years) Underperformance (pct of years) Sources: IMF, WEO and Staff Reports. 1/ Shaded areas highlight areas of policy slippages. The latter are estimated as the difference between actual outcomes and targets. The latter is captured by within-year projections taken from that year’s IMF Spring WEO vintage complemented with data from Staff Reports. The horizontal axis shows the frequency of missed targets on the under-performing side; the vertical axis shows the annual average deviation from the target in a symmetrical way, i.e., on the under- and over-performing side. 64. Countries can contain or reduce emerging debt vulnerabilities by addressing existing weaknesses in fiscal institutions. Based on the case studies, policy reform priorities would include:  Establishing a fiscal anchor. Except for Kenya, none of these countries have a fiscal rule or a clearly defined medium-term objective/target. This poses particular challenges in countries that have recently discovered natural resources (e.g., Mozambique) where the fiscal framework should be taking into account the volatility of revenues and exhaustibility of resources to avoid procyclicality and ensure fiscal sustainability (IMF, 2012c).  Formulation of more realistic budgets. Deviations from approved budgets often stem from overly-optimistic macroeconomic or revenue projections and under-provisioning for compulsory expenditure. The latter often leads to expenditure arrears. This was the case in Ghana where the government accumulated net arrears of 2.8 percent of GDP in 2012, nearly half of which was to state-owned enterprises (mainly linked to energy subsidies). 50 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT  Strengthening budgetary controls. In some cases, cash management is an issue while in others within-year spending controls are deficient, with line ministries revising budget allocations without the involvement of the finance ministry and bypassing commitment controls (Honduras). Reforms in these areas are a priority in fragile states with severe capacity constraints where budgets may not always be implemented as intended (Gupta and Ylaoutinen, 2014).  Monitoring of fiscal risks. There is generally no recognition of fiscal risks in the budget nor is there a strategy to manage them. Risks arise from reliance on donor support for external financing (Djibouti, Haiti, and Mozambique), activities of financial and non-financial public enterprises (Vietnam and Mozambique), lower levels of government (Kenya), large public- financed infrastructure projects (Mozambique, Kenya, and Djibouti), and PPPs (Mozambique and Honduras).  Tackling weaknesses in revenue mobilization. Revenue mobilization efforts have been impaired by inadequate tax administration (Haiti, Honduras, Djibouti, and Ghana); while tax incentives and exemptions have grown, thereby eroding the tax base (Vietnam and Djibouti). In addition, a broad-based and easy-to-administer VAT system is missing in some cases (e.g. Haiti). Also, there is a need to design appropriately calibrated tax regimes for natural resource sectors in those countries with recent discoveries. C. Policy Challenges 65. Although debt sustainability is, for the majority of countries, not a pressing concern, LIDCs are confronted with the challenge of maintaining sound macroeconomic policies while implementing growth enhancing structural reforms to take full advantage of the space afforded by low debt levels. On the institutional front, there are two priorities:  Strong fiscal institutions are needed to control the accumulation of debt. Budgetary institutions need to have the capacity to control spending and budget overruns; a medium-term fiscal framework is needed to provide an anchor for deficits and borrowing needs; and timely reporting and monitoring of contingent liabilities (e.g., PPPs) are key as such commitments expand in many LIDCs. Improving the efficiency of spending including by strengthening public investment management capacity is also a key requirement if growth potential is to be realized.  Comprehensive medium-term debt-management strategies (underpinned by adequate institutional capacity) need to be developed. This is particularly pressing for countries resorting to substantial external borrowing on market terms, notably sovereign bond issues, to finance large infrastructure projects. Measures are needed to manage risks arising from sovereign bond issuance and to build institutional capacity. INTERNATIONAL MONETARY FUND 51 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Box 6. Risks from International Sovereign Bond Issuance 1/ International sovereign bonds could bring significant benefits to LIDCs, but also entail risks. More diversified sources of financing can alleviate funding constraints for much needed investment in infrastructure but also can exacerbate vulnerabilities:  Exchange rate risk. International debt issuance may worsen the currency mismatch between debt service and fiscal revenues (Das, Papaioannou, and Polan, 2008). This risk is smaller for countries that are commodity exporters trading in U.S. dollars, and can be mitigated through currency swaps, although such markets are not prevalent or deep in LIDCs.  Refinancing risk. Because the majority of first-time issuers are issuing bullet bonds, the projected amortization increases dramatically at the maturity date. This could lead to pressures on the balance of payments or, rollover may happen at punitive interest rates and shorter maturities.  Interest rate risk. Since the majority of recent issuances took the form of fixed rate instruments (with Tanzania a notable exception), a spike of global interest rates does not lead to an immediate increase in borrowing costs. However, these countries might need to refinance their Eurobonds in the future in an environment of higher interest rates, declining growth, and lower investors’ interest.  Rating risk. LIDCs that have recently issued sovereign bonds are paying relatively high interest rates relative to rates on their existing debt portfolios. To make market issuance sustainable, yields must come down over time along with a steady improvement in credit ratings. Limited debt management capacity in LIDCs poses an additional challenge. Often the required skills to undertake a thorough and comprehensive analysis of international debt issuance are absent and high staff turnover is problematic. This increases the risks of a poorly executed transaction and could undermine the fiscal and external sustainability of a country. Also, institutional arrangements are needed to ensure that the proceeds from the issuance are used for intended purposes. Addressing these challenges requires comprehensive medium-term debt-management strategies. Measures to manage refinancing risks include adopting a structured issuance strategy, and building cash buffers. Currency risk can be handled through the development of swap markets. Similarly, fixed rate issuance or interest rate swaps could mitigate interest rate risk, although, given thin markets, efforts in this area will take time. Also, the institutional capacity should be strengthened, ensuring the operational independence of the debt management office and close coordination with fiscal and monetary policy. These measures should be complemented with a communication strategy to keep investors abreast of economic and policy developments. ____________________ 1/ The box also draws on IMF (2014c). 52 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Box 6. Risks from International Sovereign Bond Issuance (concluded) First-Time Issuers: Rollover Risk 1/ Changes in the Cost of Borrowing 1/ 2/ (In percent of GDP) (In percent) 1.6 max =1.6 12 12 average interest rates on current debt portfolio 1/ 1.4 Maturity profile average= 1.3 10 global bond interest rate 2/ 10 without new bond First-time bonds 1.2 (period average) falling due 8 8 1 (period average) average= 0.8 6 6 0.8 average= 0.7 4 4 0.6 2 2 0.4 0.2 0 0 0 2007 2009 2011 2013 2015 2017 2019 2021 2023 2025 2027 2029 2031 BBB BB NR B Source: IMF country desks and staff calculations. Source: Bloomberg; and IMF, WEO and staff calculations. 1/ Country sample includes Armenia, Bolivia, Honduras, Mongolia, 1/ Interest cost over previous year’s debt stock. Namibia, Paraguay, Rwanda, Tanzania, and Zambia. 2/ Yield as of July 2nd. INTERNATIONAL MONETARY FUND 53 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Appendix I. LIDCs and SubGroups Frontier Markets Commodity Fragile States Other Exporters Bangladesh Afghanistan # Afghanistan # Benin* Bolivia* Bolivia* Burundi # Bhutan Cote d’Ivoire # Burkina Faso* Central African Rep. # Cambodia Ghana* Burundi # Chad Cameroon* Kenya* Central African Rep. # Comoros # Ethiopia* Mongolia Chad Congo, Dem. Rep. # Gambia, The # Mozambique* Congo, Dem. Rep. # Congo, Rep. # Honduras* Nigeria* Congo, Rep. # Cote d’Ivoire # Kyrgyz Republic Papua New Guinea* Eritrea Djibouti Lao PDR Senegal* Guinea # Eritrea Lesotho Tanzania* Guinea-Bissau # Guinea # Moldova Uganda* Malawi* Guinea-Bissau # Nepal Vietnam Mali* Haiti # Nicaragua* Zambia* Mauritania* Kiribati Rwanda* Mongolia Liberia # Tajikistan Mozambique* Madagascar* Niger* Malawi* Nigeria* Mali* Papua New Guinea* Myanmar Sierra Leone* Sao Tome and Principe # Solomon Islands Sierra Leone* South Sudan Solomon Islands Sudan Somalia Uzbekistan South Sudan Yemen, Rep. Sudan Zambia* Togo # Zimbabwe Yemen, Rep. Zimbabwe Note: *Early HIPC: Completion point before 2007. #Late HIPC: Completion point after 2007. 54 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Appendix II. Identifying Frontier Market Economies This note identifies economies within the wider group of LIDCs that can be viewed as “frontier market economies” (FMs). The specific countries identified as FMs are Bangladesh, Bolivia, Côte d’Ivoire, Ghana, Kenya, Mongolia, Mozambique, Nigeria, Papua New Guinea, Senegal, Tanzania, Uganda, Vietnam, and Zambia. 1. There is no generally accepted definition of frontier markets. Common dimensions used in identifying FMs have included (i) a financial sector that is starting to share some similar characteristics of emerging market economies in terms of financial market depth; (ii) access to international capital markets; (iii) fast sustained growth; and (iv) sound institutions.1 2. Whether a country is considered to be a frontier market or not could depend on the issue at hand. If the objective is to assess the risks associated with the issuance of sovereign bonds, it is important to focus on countries that have issued sovereign bonds. Analyzing possible macroeconomic impacts of capital flow reversals would entail a focus on countries with significant portfolio inflows. An assessment of financial vulnerabilities would tend to emphasize the depth of financial sectors as well as their integration. Accordingly, there has been varying coverage of FMs in analytical pieces and flagships at the Fund. 3. This note proposes a framework focusing on a set of financial indicators for identifying FMs. The starting point is the recently introduced WEO list of 60 LIDCs. The selection criteria focus on the depth and openness of the financial system and the issuance of sovereign bonds. Appendix Table 1 (Section A) provides information on the variables used for the analysis and data sources. Each LIDC is benchmarked against EMs as follows: (i) LIDCs that are within one standard deviation below 3 the EM average for the following variables: M2 to GDP;2 cross border loans/deposits, stock market 4 capitalization, and portfolio inflows; and LIDCs that have accessed (or have the potential to access: 5 proxied by sovereign ratings similar to those that have issued sovereign bonds) sovereign bond markets, putting them on the radar screen of international fund managers. Figure 1 provides a good visualization of the benchmarking exercise between EMs and FMs, focusing on two of the key variables: M2 and cross-border loans and deposits over GDP. 1 IMF (2011d) adopts the following criteria to come up with a list of F-LICs: (i) recent growth dynamics and perspectives; (ii) financial market development; (iii) general institutional conditions and evolution; (iv) natural resource richness; and (v) political conditions and perspectives. 2 Between two closely correlated summary indicators of financial depth, M2/GDP and credit/GDP, the former is preferred as a selection criterion because it turns out to be more widely and frequently reported. 3 According to BIS, these capture asset and liability positions of banks located in a country vis-à-vis nonresidents. 4 This indicator focuses on countries that have attracted portfolio inflows into the domestic markets by looking at the actual portfolio inflows (capturing nonresident involvement in the stock and bond markets). 5 A country with a sovereign rating of BB-, consistent with the most common rating for LIDCs that have issued, is considered to have the potential to issue. INTERNATIONAL MONETARY FUND 55 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Appendix Table 1. Description and Definition of Variables Variable Variable Definition Source A. Financial Sector Depth M2/GDP Sum of money and quasimoney WDI Cross border loans or deposits External loans and deposits of reporting banks vis-à-vis all sectors. BIS Stock market capitalization (in percent of GDP) Market capitalization (also known as market value) is the share price times the number of shares outstanding. Portfolio inflows Portfolio inflows (debt securities) CPIS Survey Sovereign bond issuance Sovereign bond issuance or potential to issue (similar sovereign Bloomberg L.P. ratings with other LIDCs that have issued). B. Institutions CPIA Country Policy and Institutional Assessment World Bank Heritage Foundation overall score Score based on 10 broad factors of economic freedom. Heritage Foundation C. Sustained real GDP growth Real GDP growth WEO 4. An LIDC is identified as an FM if it meets four out of the five conditions indicated above. The resulting list comprises the following (14) countries: Bangladesh, Bolivia, Côte d’Ivoire, Ghana, Kenya, Mongolia, Mozambique, Nigeria, Papua New Guinea, Senegal, Tanzania, Uganda, Vietnam, and Zambia (Appendix Table 2). 5. Are FMs more advanced than peers in a broader sense? By way of validating the focus on financial selection criteria, FMs turn out to have had stronger institutions and sustained growth 6 performance than peers. Strength of institutions is measured by a composite index assigning equal weight to the Heritage Foundation overall score (2004–12), and the overall CPIA score (2010–12). Growth performance is assessed by real GDP growth during the period 2003–13. With the exception of Côte d’Ivoire, all FMs have relatively stronger institutions and sustained growth performance, well above the respective LIDC averages. 6 Each of the variables relating to institutions is standardized from 0 to 1 into an index Z’ using the Min and Max approach: Z '  X  X min  , where Xmin and Xmax stand for the minimum and maximum of X, respectively. X max  X min  56 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Appendix Table 2. Financial Sector Depth Index 1/ 2/ 3 Cross Border Loans Stock Market M2 (in percent of Sovereign Bond Country or Deposits (in Capitalization (in Portfolio Inflows1/ Index 3/ GDP) 1/ Issuance2/ percent of GDP) 1/ percent of GDP) 1/ Ghana 1 1 1 1 1 1.0 Kenya 1 1 1 1 1 1.0 Mongolia 1 1 1 1 1 1.0 Papua New Guinea 1 1 1 1 1 1.0 Vietnam 1 1 1 1 1 1.0 Nigeria 1 1 1 1 1 1.0 Bangladesh 1 1 1 0 1 0.8 Bolivia 1 1 1 0 1 0.8 Cote D'Ivoire 1 1 1 1 1 0.8 Senegal 1 1 … 1 1 0.8 Tanzania 1 1 1 0 1 0.8 Uganda 1 1 1 0 1 0.8 Zambia 0 1 1 1 1 0.8 Mozambique 1 1 … 1 1 0.8 Burkina Faso 1 1 … 0 1 0.6 Cambodia 1 1 … 0 1 0.6 Honduras 1 1 … 0 1 0.6 Malawi 1 1 1 0 0 0.6 Nepal 1 1 1 0 0 0.6 Zimbabwe 1 1 1 0 0 0.6 Eritrea 1 1 … 1 0 0.6 Solomon Islands 1 1 … 1 0 0.6 Burundi 1 1 … 1 0 0.6 Congo, Rep. 1 1 … 1 0 0.6 Rwanda 0 1 … 1 1 0.6 Afghanistan 1 1 … 0 0 0.4 Benin 1 1 … 0 0 0.4 Bhutan 1 1 … 0 0 0.4 Cameroon 0 1 … 0 1 0.4 Comoros 1 1 … 0 0 0.4 Djibouti 1 1 … 0 0 0.4 Ethiopia 1 1 … 0 0 0.4 Gambia, The 1 1 … 0 0 0.4 Guinea 1 1 … 0 0 0.4 Guinea-Bissau 1 1 … 0 0 0.4 Haiti 1 1 … 0 0 0.4 Kyrgyz Republic 1 1 0 0 0 0.4 Laos 1 1 … 0 0 0.4 Lesotho 1 1 … 0 0 0.4 Mali 1 1 … 0 0 0.4 Mauritania 1 1 … 0 0 0.4 Moldova 1 1 … 0 0 0.4 Nicaragua 1 1 … 0 0 0.4 São Tomé and Príncipe 1 1 … 0 0 0.4 Sudan 1 1 … 0 0 0.4 Togo 1 1 … 0 0 0.4 Yemen 1 1 … 0 0 0.4 Sierra Leone 0 1 … 1 0 0.4 Liberia 1 … … 0 0 0.2 C.A.R. 0 1 … 0 0 0.2 Congo, Dem. Rep. 0 1 … 0 0 0.2 Kiribati … 1 … 0 0 0.2 Madagascar 0 1 … 0 0 0.2 Myanmar 1 0 … 0 0 0.2 Uzbekistan 0 1 … 0 0 0.2 South Sudan 0 … … 0 0 0.0 Chad 0 0 … 0 0 0.0 Niger 0 0 … 0 0 0.0 Somalia … … … 0 0 0.0 Tajikistan 0 0 … 0 0 0.0 1/ 1 if a country is three-year average exceeds one standard deviation from EM average and 0 otherwise in 2010-2012; "…" indicates missing data. Sample of countries includes 60 LIDCs for which there is data available. Armenia and Georgia are excluded because they graduated from PRGT. EM includes 49 countries based on VEE sample of countries. Stock market capitalization criteria met for Côte d’Ivoire, which hosts the regional (WAEMU) stock market and accounts for 31 of its 36 listings.” 2/ Indicates whether a country has issued, or has the potential to issue a sovereign bond; Yes=1, No=0. Congo, Rep. is excluded due to issuance in the context of debt restructuring. 3/ Measures the proportion of indicators that meet the qualification threshold. INTERNATIONAL MONETARY FUND 57 58 100 120 20 40 60 80 0 100 120 20 40 60 80 0 Liberia Cape Verde Vietnam Dominica Eritrea St. Lucia Grenada Djibouti Djibouti Guyana Dominica São Tomé and Príncipe Cape Verde 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 28.38342 Mauritania St. Lucia Gambia, The Nepal Solomon Islands Bolivia Maldives St. Vincent and the … Kenya Maldives Mozambique Bangladesh Moldova Uzbekistan Gambia, The Togo Bhutan Lesotho Mongolia Honduras Papua New Guinea Bhutan Guyana Sources: BIS and World Bank, WDI. Nicaragua Honduras Ghana Kenya Guinea Cambodia Bolivia Laos EM_VEE Average Afghanistan Benin Moldova Togo EM_VEE Average Grenada Haiti INTERNATIONAL MONETARY FUND Papua New Guinea Solomon Islands Sierra Leone Senegal Madagascar Guinea-Bissau Zambia Myanmar Senegal Liberia Vietnam São Tomé and… Malawi Comoros Malawi Cote D'Ivoire Nigeria Haiti Yemen Kyrgyz Republic Mauritania Nigeria Afghanistan Uganda Lesotho Cambodia Nicaragua M2 (In percent of GDP) in 2011-2013 Zimbabwe Tanzania MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Laos Mozambique Yemen Mali Nepal Congo, Republic of Tanzania Kyrgyz Republic Eritrea Burkina Faso Average Burundi Ghana Congo, Dem. Rep. of Guinea Total Cross-border Loans and Deposits (In percent of GDP) in 2011-2013 Cameroon Zimbabwe Ethiopia EM_VEE Average Congo, Republic of Cameroon Rwanda Zambia Mongolia Uzbekistan Bangladesh Sudan Guinea-Bissau Niger Burkina Faso Madagascar Mali Uganda One standard deviation within C.A.R. Sierra Leone Benin Burundi Sudan Rwanda Ethiopia Tajikistan One standard deviation within EM_VEE Niger C.A.R. Chad Cote D'Ivoire Tajikistan Chad Myanmar Congo, Dem. Rep. of Top 3 Top 3 Latvia Jamaica Georgia Algeria Uruguay Malaysia Pakistan Armenia Bottom 3 Lebanon Bottom 3 deviation. Panama 1/ Lebanon 1/ China, Mainland standard deviation. Appendix Figure 1. Selected Indicators of Financial Sector Depths and Openness EM_VEE average and GDP) in 2012 1/ Panama exluded from 1/ Lebanon exluded from EM_VEE average and standard EM_VEE Total Cross-border Loans and Deposits in 2012 4.17 7.41 8.18 19.55 135.65 604.25 EM_VEE M2 (in percent of 2.23 28.64 29.28 137.58 182.83 304.22 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Appendix III. Methodology Underlying the Financial Vulnerability Index 1. For frontier markets, whose financial system share similar characteristics with emerging markets, the VE-E financial vulnerability index (FVI) is calculated as an indicator of underlying financial vulnerability. The VE-E exercise (IMF, 2007) distinguishes between vulnerability and crisis risk assessments. This is because countries with severe vulnerabilities may still face low risks in a benign global environment. The risks envisaged relate to capital account crises (connected to domestic and external financial imbalances), which have been marked by large swings in cross-border capital flows. These have often prompted currency crises and frequently lead to sovereign debt crises, financial and corporate sector crises, as well as both. The vulnerability index is constructed as a summary measure of underlying vulnerability, combining information from a wide range of economic and financial indicators. 2. Five variables make up the financial vulnerability index of the VE-E: capital adequacy ratio, return on assets, bank credit over bank deposits, cumulative three-year growth of credit over GDP, and foreign liabilities measured as cross border loans and deposits over domestic credit.1 Other potential risk indicators, such as non-performing loans (NPLs) of the banking sector or provisioning thereof, have been excluded as they are often backward looking and/or not standardized across countries. 2 3. The financial vulnerability of frontier markets is assessed using the thresholds and weights of the VE-E (Appendix Table 3). The FVI ranges from zero (low vulnerability) to one (high vulnerability). It is a weighted average of zero/one scores for individual indicators (one if the indicators breach their thresholds). The thresholds are determined by searching for the values of each indicator that separates crisis from non-crisis cases while minimizing the combined percentages of missed crisis (Type I error) and false alarms (Type II error).3 The aggregation weights depend on the individual indicator’s ability to discriminate between crisis and non-crisis cases. A country is considered low-risk if the index is lower than 0.33, high-risk, if it exceeds 0.45, and medium-risk otherwise.4 1 Based on the 2013 Vulnerability Exercise for Emerging Markets (VE-E) methodology. There is a concern that the VE-E financial vulnerability index applied here may not reflect the full extent of vulnerabilities in EMs. As a result, a newer methodology (including variables such as real estate prices) is now applied to a subset of more advanced EMs for which data is available. 2 For example, NPLs tend to be understated when credit is growing too fast and underwriting standards weaken, such as in a credit bubble, thus failing to point to underlying vulnerabilities. 3 A crisis event is identified by combining quantitative indicators of sudden stop and analysis of narrative records (to eliminate spurious crisis events). 4 Thresholds for the overall index is derived by minimizing an asymmetrically weighted loss function which penalizes missing crisis (Type I error) more. INTERNATIONAL MONETARY FUND 59 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Appendix Table 3. Financial Vulnerability Index Parameters Threshold Weight Capital adequacy ratio (percent) * 12.3 0.22 Return on assets (percent) * 0.2 0.11 Loan-to-deposit ratio (percent) 110 0.22 3-year cumulative credit-to-GDP growth (percent) 9.4 0.19 Foreign liabilities (percent domestic credit) 14.7 0.26 Sources: BankScope; BIS; World Bank; and IMF staff estimates. * Larger values of this variable are associated with a lower risk of crisis. 60 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Appendix IV. Food Decline Vulnerability Index and Natural Disasters This note develops a food decline vulnerability index (FDVI) to assess the vulnerability of the least- developed countries among the LIDCs to sudden food declines.1 1. Understanding the interplay between the impact of natural disasters on access to food and the available coping mechanisms is important. The analysis requires operational definitions of food supply and food decline episodes. Based on the database of the Food and Agriculture Organization (FAO) of the United Nations, food supply measures the “total quantity of foodstuffs produced in a country added to the total quantity imported and adjusted to any change in stocks that may have occurred since the beginning of the reference period.” Food decline episodes are matched with natural disaster shocks which are identified if the number of people affected or the damage caused by natural disasters falls above the 50th percentile of the country-specific distribution.2 2. Drawing on the literature,3 a general-to-specific approach was used in selecting variables for the univariate signaling approach. The analysis uses 12 variables, based on results from probit regressions and a number of variables considered in the literature as main determinants of food supply decline (Appendix Table 4): food supply growth (measured in kilograms per capita for the period 1990–2009);4 The World Bank’s CPIA index; Gini coefficient; life expectancy; share of agriculture in total GDP; natural disasters; fiscal and external indicators; and growth in trading partners. 3. Results show that underlying structural and institutional aspects best explains food decline vulnerabilities. For each indicator, the 2nd and 3rd columns in Appendix Table 4 show the food decline and non-food decline observations. A maximum of 22 food decline episodes were found to match with the occurrence of natural disasters.5 The fourth column presents the estimated thresholds followed by the corresponding Type I (i.e., ability of an indicator to predict a food supply 1 The basic approach draws on the methodology for the GDVI. For details on the GDVI, see the following: IMF (2013a) and “Dabla-Noris and Bal Gündüz (2014). 2 Conditional on these natural disaster shocks, food decline episodes are identified if the post two-year average of food supply per capita falls below its pre three-year average ((fst + fst+1 )/2 < (fst + fst-1 +fst-2 )/3) and if food supply growth was negative. 3 See the following; (i) Burg (2008); (ii) Lucas and Hilderink (2004); (iii) Capaldo and others (2010); and (iv) The Economist Intelligence Unit (2013). 4 To take into account food supply growth beyond 2009 in the FDVI indicator, we use a simple regression based on lagged food supply growth, lagged natural disaster shocks and a constant to forecast food supply growth in subsequent years. 5 In the overall sample, 22 food decline episodes were found. However, Appendix Table 4 only reports the crisis episodes which coincide with the data availability of each variable. Therefore, the number of crisis episodes varies for each variable between 2 and 20 observations. INTERNATIONAL MONETARY FUND 61 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT decline) and Type II errors (i.e., occurrence of false alarms). The weight of each indicator in the overall FDVI is shown in the last column and represents the importance of each indicator in the overall index. The top predictor of food supply decline is the overall economy and institutional sector (responsible for 55 percent of the index weight), with the Gini coefficient and food supply growth being the main predictors within this cluster. The second most important sector with a weight of 25 percent is the fiscal sector followed by the external sector with a weight of 20 percent. Appendix Table 4. Food Decline Vulnerability Index Estimation Results /1 Direction to be Crisis Non-crisis Variables 1/ Thresholds Type I error Type II error Index weight safe observations observations Overall economy and institutions Food supply growth (kgs) > 20 259 -0.9 0.1 0.3 0.18 CPIA (t-1) > 20 261 3.1 0.5 0.4 0.01 Gini coefficient (t-1) < 2 15 56.4 0.0 0.3 0.24 Food supply growth average (kgs) > 20 263 0.6 0.3 0.4 0.04 Life expectancy (t-1) > 20 257 65.0 0.0 0.8 0.02 Agriculture value added (in percent of GDP) (t-1) > 13 193 25.6 0.5 0.3 0.02 Natural disaster (in percent population) < 13 249 3.3 0.2 0.6 0.01 GDP growth (t-1) > 15 201 2.0 0.7 0.2 0.02 0.55 External Sector Reserve coverage (months of imports) (t-1) > 19 226 1.0 0.8 0.1 0.10 Growth in trading partners weighted by lagged exports to GDP > 19 262 2.0 0.1 0.9 0.10 0.20 Fiscal Sector Government balance (in percent of GDP) (t-1) > 20 277 -8.3 0.6 0.1 0.23 Government tax revenue (in percent of GDP) (t-1) > 12 181 11.9 0.4 0.6 0.02 0.25 Sources: FAO; World Bank; EM-DAT; and IMF, WEO and staff estimates. 1/ Sample includes 63 PRGT-eligible countries for the time period 1990–2008. Due to the identification of food supply crisis episodes (where the post two year average is necessary) the estimation period has to stop in 2008 despite the availability of food supply data until 2009. 62 INTERNATIONAL MONETARY FUND Appendix V. Case Studies: Key Trends (2013 and in percent of GDP, unless otherwise indicated) DJI GHA HTI HND KEN MOZ VNM Population (Mio)* 0.9 24.9 10.2 7.9 40.7 25.2 88.8 GNI per capita' Atlas method (current USD)** 1,030(2005) 1,760 810 2,180 930 590 1,730 Public debt*** 53.46 57.39 16.91 43.39 43.46 53.13 50.55 External public debt*** 47.57 25.65 15.03 27.88 23.30 44.16 28.83 o/w concessional (share of total, 2012)** 75.83 47.10 85.35 55.09 65.73 78.07 50.87 Domestic currency denom. domestic public debt*** 5.89 31.75 1.88 15.51 20.15 8.98 21.72 Debt vulnerability trend (rating in past 3 DSAs)*** 1/ → (H,H,H) → (M,M,M) → (H,H,H) ↓ (L,M,M) → (L,L,L) ↓ (L,M,M) → (L,L,L) HIPC relief (PV, decision point year)**** × 29.7 2.9 7.8 × 51.2 × MDRI relief**** × 8.81 11.37 14.47 × 13.28 × Lowest public debt-to-GDP ratio since 2001 (year)*** 53.46/2013 26.13/2006 9.73/2011 23.36/2008 39.16/2007 21.85/2007 22.88/2001 CPIA overall score: trend, rating, score 2/** →W (3.17) ↓↓W (3) →W (3.17) ↓↓W (3) →S (4.5) →S (4.17) →S (4.17) Macroeconomic management →M (3.5) ↓↓W (3) ↑S (4) ↓↓W (3) →S (4.5) →S (4.5) →S (4) Fiscal policy →W (3) ↓↓W (2.5) ↓W (3) ↓W (2.5) →S (4.5) →S (4) →S (4.5) Debt policy →W (3) ↓M (3.5) →W (2.5) ↓M (3.5) →S (4.5) →S (4) →S (4) Social protection →W (3) ↑S (4) →W (2) ↓W (3) ↑M (3.5) ↑M (3.5) →M (3.5) Public sector mngt. & institutions cluster →W (2.7) ↓M (3.7) →W (2.4) ↓W (3) →M (3.4) →M (3.3) →M (3.5) Quality of budgetary and financial mngt. →W (2.5) ↓↓W (3) →W (2.5) ↓M (3.5) →M (3.5) ↑S (4) ↓M (3.5) Efficiency of revenue mobilization →M (3.5) →S (4) →W (2.5) ↓M (3.5) →S (4) →S (4) ↑S (4) Quality of public administration →W (2.5) →M (3.5) →W (2.5) →W (2.5) →M (3.5) →W (3) →M (3.5) Trade openness *3/ 95.6 80.9 70.4 89.0 52.8 120.2 163.0 Financial openness (2011) *****4/ 237.0 84.9 57.1 119.9 73.6 165.7 108.0 Aid-dependency (2012) *5/ n.a. 4.69 16.12 3.34 7.00 14.76 2.76 Market access year ******6/ No 2007, 13 No 2013 2014 2013 2005, 09, 10 Fund program ******* ECF(99,08), SMP(05) ECF(09) ECF(06,10) ECF(04),SBA(08,10),SCF(10 ECF(00,03,11) ECF(99,04),PSI(07,10,13),E ECF(01) ) SF(09) MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Macro stability (average in last 5 years)* Real p.c. growth 1.7 5.9 0.7 0.3 2.9 4.4 4.6 CPI inflation 4.0 16.8 6.6 5.4 6.0 5.6 5.2 Int. reserves (months of next year's imports) -4.3 -2.9 -3.4 -3.2 -3.4 -3.3 0.0 Fiscal balance -3.3 -8.7 -3.7 -4.4 -4.7 -4.3 -4.4 Current account balance -14.1 -9.3 -4.0 -6.7 -7.3 -26.6 0.3 Progress towards MDGs**** #1: Eradicate extreme hunger and poverty Met Met Seriously off target Met Met Seriously off target Met #2: Achieve universal primary education on track Met Insufficient data Met Met Seriously off target Met #3: Promote gender equality and empower women off track Seriously off target Insufficient data Met moderate off target Insufficient data Sufficient progress INTERNATIONAL MONETARY FUND 63 #4: Reduce child mortality on track Moderate off target Moderate off target Sufficient progress seriousely off target Sufficient progress On track #5: Improve maternal health on track Moderate off target Seriously off target Moderate off target seriousely off target Moderate off target Sufficient progress #6: Combat HIV/AIDS, malaria and other diseases off track Seriously off target Seriously off target Insufficient data insufficient data Seriously off target Met #7: Ensure access to improved water sources off track Met Seriously off target Met moderate off target Seriously off target Insufficient progress #8: Global partnership for development off track Seriously off target Insufficient data Seriously off target insufficient data Seriously off target Sufficient progress Sources: * IMF, WEO. **World Bank, WDI. ***IMF, SPR/DSA database. **** IMF HIPC and MDRI statistical update 2013, IMF PRSP Djibouti 2012, Kenya MDGs progress Leo and Barmeier (2010), UN MDGs report on Vietnam 2011. ***** IMF, the External Wealth of Nations Mark II database. ****** IMF, Bonds, Equities and Loans (BEL) database. ******* IMF press release. Note: 1/ L, M and H mean low, medium and high risk of debt distress. 2/ The CPIA score ranges from 1 (=low) to 6 (=high). A s trong CPIA performer (labeled“S”) is defined as one whose CPIA score is 3.75 or above, whilst a medium performer (labeled“M”) has a score within the range of 3.25 to 3.75, and a weak performer (labeled “W”) has a score of less than 3.25. ↑ [resp. ↓] indicates an upgrade [resp. downgrade] of one rank (e.g., from S to M [resp. M to S]) and ↑↑ a change of two ranks (e.g., from W to S [resp. S to W]) between 2007 and the latest available observation in 2013. 3/ Calculated as (Imp+Exp/GDP). 4/ Calculated as (FA+FL/GDP). 5/ Defined as ODA as a proportion of total revenue. 6/ Defined as issuance of sovereign bond. MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT References Abiad, Abdul, Enrica Detragiache, and Thierry Tressel, 2010, "A New Database of Financial Reforms," IMF Staff Papers 57, pp. 281–302 (Washington: International Monetary Fund). Acevedo, Sebastian, 2014, “Debt, Growth and Natural Disasters: A Caribbean Trilogy,” IMF Working Paper 125 (Washington: International Monetary Fund). African Center for Economic Transformation (ACET), 2014, “African Transformation Report: Growth with Depth” (Accra, Ghana: African Center for Economic Transformation). AfDB, OECD, and UNDP, 2014, “African Economic Outlook: Global Value Chains and Africa’s Industrialization.” Allen, Richard, Richard Hemming, and Barry H. Potter (eds.), 2013, The International Handbook of Public Financial Management (London: Palgrave MacMillan). Araujo, Juliana, Antonio David, Sarwat Jahan, Svitlana Maslova, and Chris Papageorgiou, 2014, “Non- FDI Private Capital Flows in Low-Income Countries: Catching the Wave?” IMF Staff Discussion Note, forthcoming (Washington: International Monetary Fund). Bai, Jushan, and Pierre Perron, 1998, “Estimating and Testing Linear Models with Multiple Structural Changes,” Econometrica, Vol. 66 (1), pp. 47–78. Berg, Andrew, Jonathan Ostry, and Jeromin Zettelmeyer, 2012, “What Makes Growth Sustained,” Journal of Development Economics, Vol. 98 (2), pp 149–166. Benedek, Dora, Ernesto Crivelli, Sanjeev Gupta, and Priscilla Muthoora, 2014, “Foreign Aid and Revenue: Still a Crowding-Out Effect?” FinanzArchiv: Public Finance Analysis, Mohr Siebeck, Tübingen, Vol. 70 (1), pp. 67–96. Buffie, Edward E., Andrew Berg, Catherine Pattillo, Rafael Portillo, Luis-Felipe Zanna, 2012, "Public Investment, Growth, and Debt Sustainability: Putting Together the Pieces," IMF Working Paper 12/144 (Washington: International Monetary Fund). Burg, Jericho, 2008, “Measuring Populations’ Vulnerabilities for Famine and Food Security Interventions: The Case of Ethiopia’s Chronic Vulnerability Index,” Disasters, 32 (4), pp. 609- 630. Capaldo, Jeronim, Panagiotis Karfakis, Marco Knowles, and Mark Smulders, 2010, “A Model of Vulnerability to Food Insecurity,” Food and Agriculture Organization, ESA Working Paper 10- 03. 64 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Cavallo, Eduardo, Sebastian Galiani, Ilan Noy, and Juan Pantano, 2013, “Catastrophic Natural Disasters and Economic Growth,” Review of Economics and Statistics, MIT Press, Vol. 95 (5), pp 1549–1561. Clements, Benedict, David Coady, Stefania Fabrizio, Sanjeev Gupta, Trevor Alleyne, and Carlo Sdralevich, 2013, “Energy Subsidy Reform: Lessons and Implications,” (Washington: International Monetary Fund). Dabla-Norris, Era, and Yasemin Bal Gündüz, 2014 “Exogenous Shocks and Growth Crises in Low- Income Countries: A Vulnerability Index,” World Development, Vol. 59 (C), pp. 360–378. Dabla-Norris, Era, Giang Ho, Kalpana Kochhar, Annette Kyobe, and Robert Tchaidze, 2013, “Anchoring Growth: The Importance of Productivity-Enhancing Reforms in Emerging Market and Developing Economies,” IMF Staff Discussion Note 13/08 (Washington: International Monetary Fund). Das, Udaibir S.; Michael G. Papaioannou, and Magdalena Polan, 2008,“ Strategic Considerations for First-Time Sovereign Bond Issuers,” IMF Working Paper 08/261 (Washington: International Monetary Fund). Diamond, Jack, 2013, “Good Practice Note on Sequencing PFM Reforms,” (Washington: PEFA Secretariat). Drummond, Paulo, and Estelle Xue Liu, 2013, “Africa’s Rising Exposure to China: How Large Are Spillovers Through Trade?” IMF Working Paper 13/250 (Washington: International Monetary Fund). Economist Intelligence Unit, The, 2013, “Global Food Security Index: An Annual Measure of the State of Global Food Security”. EM-DAT: The OFDA/CRED International Disaster Database–www.emdat.be, Université Catholique de Louvain, Brussels (Belgium). Escolano, J., Kolerus, C., and C. Lonkeng Ngouana, 2014, “Global Monetary Tightening: Emerging Markets Debt Dynamics and Fiscal Crises,” IMF Working Paper (forthcoming). Grigoli, Francesco, 2014, “A Hybrid Approach to Estimating the Efficiency of Public Spending on Education in Emerging and Developing Economies,” IMF Working Paper 14/19 (Washington: International Monetary Fund). ———, and Javier Kapsoli, 2013, “Waste Not, Want Not: The Efficiency of Health Expenditure in Emerging and Developing Economies,” IMF Working Paper 13/187 (Washington: International Monetary Fund). INTERNATIONAL MONETARY FUND 65 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Gupta, Sanjeev, Alvar Kangur, Chris Papageorgiou, and Abdoul Wane, 2014, “Efficiency-Adjusted Public Capital and Growth,” World Development Vol. 57, pp. 164–178. Gupta, Sanjeev, and Sami Ylaoutinen, 2014, “Budget Institutions in Low-Income Countries: Lessons from G-20,” IMF Working Paper 14/164 (Washington: International Monetary Fund)). Hallegate, Stéphane, and Valentin Przyluski, 2010, “The Economics of Natural Disasters: Concepts and Methods,” Policy Research Working Paper 5507 (Washington: World Bank). Henn, Christian, Chris Papageorgiou, and Nikola Spatafora, 2013, “Export Quality in Developing Countries,” IMF Working Paper 13/108 (Washington: International Monetary Fund). Infrastructure Consortium for Africa (ICA), 2008, Infrastructure Consortium for Africa–Annual Report. www.icafrica.org/fr. International Monetary Fund (IMF), 2007, “Assessing Underlying Vulnerabilities and Crisis Risks in Emerging Market Countries-A New Approach,” (Washington). ———, 2010, “Regional Economic Outlook: Sub-Saharan Africa,” (Washington). ———, 2011a, “Managing Volatility: A Vulnerability Exercise for Low-Income Countries” (Washington). March ———, 2011b, “Managing Global Growth Risks and Commodity Price Shocks—Vulnerabilities and Policy Challenges for Low-Income Countries” (Washington). September ———, 2011c, “Revenue Mobilization in Developing Countries” (Washington). ———, 2011d, “Regional Economic Outlook: Sub-Saharan Africa” (Washington). ———, 2012a, “Global Risks, Vulnerabilities, and Policy Challenges Facing Low-Income Countries” (Washington). October ———, 2012b,”Revisiting the Debt Sustainability Framework for Low-Income Countries” (Washington). ———, 2012c, “Macroeconomic Policy Frameworks for Resource-Rich Developing Countries” (Washington). ———, 2012d, “Regional Economic Outlook: Sub-Saharan Africa” (Washington). ———, 2013a, “Low-Income Countries Global Risks and Vulnerabilities Report,” IMF Policy Paper (Washington). ———, 2013b, “Regional Economic Outlook: Sub-Saharan Africa” (Washington, May). 66 INTERNATIONAL MONETARY FUND MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT ———, 2013c, “Heavily Indebted Poor Countries (HIPC) Initiative and Multilateral Debt Relief Initiative (MDRI)—Statistical Update” (Washington). ———, 2014a, “Regional Economic Outlook for Sub-Saharan Africa” (Washington, May). https://www.imf.org/external/pubs/ft/reo/2014/afr/eng/sreo0414.htm ———, 2014b, “Sustaining Long-Run Growth and Macroeconomic Stability in Low-Income Countries—The Role of Structural Transformation and Diversification,” IMF Policy Paper (Washington). ———, 2014c,”Managing Risks in Sovereign Bond Issuance in Low-Income Developing Countries,” forthcoming IMF Staff Discussion Note (Washington). ———, 2014d, “Regional Economic Outlook for Sub-Saharan Africa,” forthcoming, October (Washington). Laframboise, Nicole, and Boileau, 2012, “Natural Disasters: Mitigating Impact, Managing Risks”, IMF Working Paper 12/245 (Washington: International Monetary Fund). Leo, Benjamin, and Julia Barmeier, 2010, “Who Are the MDG Trailblazers? A New MDG Progress Index,” CGD Working Paper 222 (Washington: Center for Global Development). Lucas, Paul L., and Henk B. M. Hilderink, 2004. “The Vulnerability Concept and Its Application to Food Security,” National Institute for Public Health and the Environment (RIVM Report 550015004). Melecky, Martin and Claudio Raddatz, 2011, “How Do Governments Respond After Catastrophes? Natural-Disaster Shocks and the Fiscal Stance,” (Washington: The World Bank). Merotto, Dino, Mark Roland Thomas, and Tihomir Stucka, 2012, “How Clean is the Slate? African Debt Since Debt Relief,” unpublished manuscript (Washington: The World Bank). Noy, Ilan, 2009, “The Macroeconomic Consequences of Disasters,” Journal of Development Economics, Vol. 88, pp. 221-231. Ostry, Jonathan, Alessandro Prati, and Antonio Spilimbergo, 2009, “Structural Reforms and Economic Performance in Advanced and Developing Countries,” IMF Occasional Paper No. 268 (Washington: International Monetary Fund). Peter, Goetz von., Sebastian von Dahlen, and Sweta Saxena, 2012, “Unmitigated Disasters? New Evidence on the Macroeconomic Cost of Natural Catastrophes,” BIS Working Papers, No. 394 (Basel: Bank of International Settlements). Pritchett, Lant, 2000, “The Tyranny of Concepts: CUDIE (Cumulated, Depreciated, Investment Effort) is Not Capital,” Journal of Economic Growth, Vol. 5 (December), pp. 361–384. INTERNATIONAL MONETARY FUND 67 MACROECONOMIC DEVELOPMENTS IN LIDCS: 2014 REPORT Prizzon, Annalisa, and Shakira Mustapha, 2014, “Debt Sustainability in HIPCs in a New Age of Choice,” ODI Working Paper 397 (London: Overseas Development Institute). UNCTAD, 2014, “Investing in the SDGs: An Action Plan,” World Investment Report, United Nations Conference on Trade and Development (New York: United Nations). USAID, 2011, “Regional Agricultural Transport and Trade Policy Study,” West Africa Trade Hub Technical Report No. 41 (Washington: U.S. Agency for International Development). Weber, Anke, 2012, “Stock-Flow Adjustments and Fiscal Transparency: A Cross-Country Comparison,” IMF Working Paper 12/39 (Washington: International Monetary Fund). World Bank, 2011, “Conflict, Security, and Development,“ World Development Report” (Washington: The World Bank). ———, 2013, “Beyond the Annual Budget: Global Experience with Medium-Term Expenditure Frameworks,” Public Finance (Washington: The World Bank). ———, 2014a, “Africa’s Pulse,” April 2014, Vol. 9, (Washington: The World Bank). ———, 2014b, “Turn Down the Heat: Climate Extremes, Regional Impacts, and the Case for Resilience” (Washington: The World Bank). Young, Alwyn, 1994, “Lessons from the East Asian NICs: A Contrarian View,” European Economic Review, Vol. 38, No. 3–4, pp. 964-973. 68 INTERNATIONAL MONETARY FUND