International Monetary Fund World Economic Outlook October 212 Resilience in Emerging Market and Developing Economies: Will It Last? Abdul Abiad, John Bluedorn, Jaime Guajardo, and Petia Topalova with support from Angela Espiritu and Katherine Pan
Percent EMDEs have done well over the past decade, and through the global crisis Percent 8 6 4 2 Growth of Real GDP per Capita Contribution to Growth in World Real GDP per Capita 5 4 3 2 1-2 Emerging Market and Developing Economies Emerging Market and Developing Economies -1-4 Advanced Economies Advanced Economies World Growth -2-6 199 1993 1996 1999 22 25 28 211 199 1993 1996 1999 22 25 28 211-3 Source: World Economic Outlook database. 1
EMDE resilience has improved since the 197s and 198s. Dynamics of Output per Capita Following Peaks 5s and 6s 7s and 8s 199s 2 6 Great Recession 14 Advanced Economies Emerging Market and Developing Economies 14 13 13 12 12 11 11 1 1 9 9 8 Peak t + 1 t + 2 t + 3 t + 4 t + 5 t + 6 t + 7 t + 8 t + 9 t + 1 8 Peak t + 1 t + 2 t + 3 t + 4 t + 5 t + 6 t + 7 t + 8 t + 9 t + 1 2
Percent In fact, from the 2s, EMDEs spent more time in expansion and had smaller downturns than AEs Percent 5s and 6s 7s and 8s 199s 2s 1 Time Spent in Expansion Median Downturn (Peak-to-Trough Amplitude) 5 8 6-5 -1 4-15 2-2 Advanced Economies Source: IMF staff calculations. Emerging Market and Developing Economies Advanced Economies Emerging Market and Developing Economies -25 3
Percent While AE growth in expansion has fallen over time, EMDE growth in expansion has been more stable. 5s and 6s 7s and 8s 199s 2s 5 Median Growth in Real GDP per Capita during Expansion 4 3 2 1 Advanced Economies Source: IMF staff calculations. Emerging Market and Developing Economies Emerging Market Economies Low-income Countries 4
Unpacking further, what factors are associated with greater resilience? We look at three broad areas: External and domestic Policy frameworks and policy space Structural characteristics Analytical approach: Look at how these factors affect the length of expansions and the speed of recoveries using standard tools of duration analysis Bivariate consider factors one-by-one Multivariate consider multiple factors simultaneously 5
Shocks, both external and domestic, tend to bring EMDE expansions to an end Average Probability of Expansion Coming to an End EXTERNAL SHOCKS Spike in global uncertainty* Large rise in U.S. real interest rates* Recessions in AEs* Sudden stop in capital flows* Terms-of-trade bust* DOMESTIC SHOCKS Banking crisis* Credit boom* Without shock With shock 5 1 15 2 25 3 35 Percent Source: IMF staff calculations. Note: Statistically significant differences at the 1 percent level are denoted by starred labels. 6
but improved policy frameworks and enhanced policy space help prolong expansions, and hasten recoveries Effects of Policies on Expansion Duration Effects of Policies on Speed of Recovery Without characteristic With characteristic Without characteristic With characteristic POLICY FRAMEWORKS POLICY FRAMEWORKS Inflation targeting* Inflation targeting* Countercyclical fiscal policy* Countercyclical fiscal policy* Nonpegged exchange rate* Nonpegged exchange rate POLICY SPACE POLICY SPACE Low inflation* Low inflation* Fiscal surplus* Fiscal surplus Low public debt Low public debt* Current account surplus* Current account surplus* Low external debt* Low external debt* High reserves* High reserves* 2 4 6 8 1 12 14 16 2 4 6 8 1 12 14 16 Source: IMF staff calculations. Note: Statistically significant differences at the 1 percent level are denoted by starred labels. Recovery is defined as the number of years required to get back to pre-downturn levels of per capita output. 7
while structural characteristics are more of a mixed bag, with fewer robust relationships. Effects of Structural Characteristics on Expansion Duration Without characteristic With characteristic Effects of Structural Characteristics on Recovery Duration Without characteristic With characteristic Trade openness Trade openness* Trade liberalization Trade liberalization High intra-emde exports High intra-emde exports* High financial integration High financial integration* High capital account openness High capital account openness* High FDI flows* High FDI flows* Low income inequality* Low income inequality 2 4 6 8 1 12 14 16 2 4 6 8 1 12 14 16 Source: IMF staff calculations. Note: Statistically significant differences at the 1 percent level are denoted by starred labels. Recovery is defined as the number of years required to get back to pre-downturn levels of per capita output. 8
Percent Improved performance from the 198s to the 2s has been mainly due to policies and policy space Contribution to Change in Expected Mean Duration of Expansions from 198s to 2 7 5 Expected Mean Duration of Expansions 18 4 3 15 12 2 1 9 6-1 3-2 External Domestic Policies Structure Total 198s 199s 2 7 28 9 21 11 AE Crisis Scenario Source: IMF staff calculations. 9
Concluding remarks EMDE resilience is not a recent phenomenon; it has been building over many years. Not just in EMs, but low-income countries as well Better policymaking has a lot to do with it But resilience cannot be taken for granted: These economies remain vulnerable to external and domestic Policy space, partly used up in response to the 28-9 crisis, needs to be rebuilt Improvements in policymaking (e.g., greater ER flexibility, more countercyclical policies, inflation targeting) should be maintained 1
It is not just a commodity story 13 Commodity Exporters Dynamics of Output per Capita Following Peaks 5s and 6s 7s and 8s 199s 2 6 Great Recession Noncommodity Exporters 13 12 12 11 11 1 1 9 9 8 Peak t + 1 t + 2 t + 3 t + 4 t + 5 t + 6 t + 7 t + 8 t + 9 t + 1 8 Peak t + 1 t + 2 t + 3 t + 4 t + 5 t + 6 t + 7 t + 8 t + 9 t + 1 12
13 Largest EMDEs 1 nor is it only for the largest EMDEs. Dynamics of Output per Capita Following Peaks 5s and 6s 7s and 8s 199s 2 6 Great Recession Other EMDEs (excluding the largest) 13 12 12 11 11 1 1 9 9 8 8 Peak t + 1 t + 2 t + 3 t + 4 t + 5 t + 6 t + 7 t + 8 t + 9 t + 1 Peak t + 1 t + 2 t + 3 t + 4 t + 5 t + 6 t + 7 t + 8 t + 9 t + 1 1 Refers to the 3 largest emerging market and developing economies based on their average real GDP over the sample period. 13
14 Among EMDEs, EMs did better from the 199s, while LICs improved most from the 2s. Emerging Market Economies Dynamics of Output per Capita Following Peaks 5s and 6s 7s and 8s 199s 2 6 Great Recession Low-income Countries 14 13 13 12 12 11 11 1 1 9 9 8 Peak t + 1 t + 2 t + 3 t + 4 t + 5 t + 6 t + 7 t + 8 t + 9 t + 1 8 Peak t + 1 t + 2 t + 3 t + 4 t + 5 t + 6 t + 7 t + 8 t + 9 t + 1 14
Among regions, EMD Asia and Latin America have seen the most dramatic improvements. Dynamics of Output per Capita Following Peaks 14 5s and 6s 7s and 8s 199s 2 6 Great Recession Emerging and Developing Asia Latin America 14 13 13 12 12 11 11 1 1 9 9 8 Peak t + 1 t + 2 t + 3 t + 4 t + 5 t + 6 t + 7 t + 8 t + 9 t + 1 8 Peak t + 1 t + 2 t + 3 t + 4 t + 5 t + 6 t + 7 t + 8 t + 9 t + 1 15
Sub-Saharan Africa also improved, but mostly in the 2s. CIS-EM Europe was derailed by the crisis. Dynamics of Output per Capita Following Peaks 5s and 6s 7s and 8s 199s 2 6 Great Recession 14 Sub-Saharan Africa Commonwealth of Independent States and Emerging Europe 14 13 13 12 12 11 11 1 1 9 9 8 Peak t + 1 t + 2 t + 3 t + 4 t + 5 t + 6 t + 7 t + 8 t + 9 t + 1 8 Peak t + 1 t + 2 t + 3 t + 4 t + 5 t + 6 t + 7 t + 8 t + 9 t + 1 16
Percent What is behind the gains in EMDE resilience? Higher steady-state growth and lower variability Percent 5s and 6s 7s and 8s 9s and 2s 4. Median Steady-State Growth Median Growth Variability 5. 3.5 4.5 3. 4. 3.5 2.5 3. 2. 2.5 1.5 2. 1. 1.5 1..5.5. Advanced Economies Emerging Market and Developing Economies Advanced Economies Emerging Market and Developing Economies. Source: IMF staff calculations. 17
Why has performance improved? Some have become less frequent, others more frequent Relative frequency of shock in EMDEs (percent) 5 Banking Crises 4 3 2 1 15 Credit Booms 12 9 6 3 12 Spikes in Global Uncertainty 1 8 6 4 2 12 Sudden Stops in Capital Flows 1 8 6 4 2 12 Recessions in AEs 1 8 6 4 2 18
but EMDE policy frameworks have improved, and policy space has increased. Share of EMDEs with characteristic (percent) 6 Pegged Exchange Rate 5 4 3 2 1 5 Countercyclical Fiscal Policy 4 3 2 1 18 Inflation Targeting 15 12 9 6 3 1 High International Reserves 8 6 4 2 1 Low Public Debt 8 6 4 2 9 Single-Digit Inflation 6 3 19
Structural factors are mixed some are more supportive, others less so. Share of EMDEs with characteristic (percent) 12 High Trade Openness 1 8 6 4 2 12 High Intra-Emerging Market and Developing Economies Trade 1 8 6 4 2 12 High Financial Integration 1 8 6 4 2 9 High Net Foreign Direct Investment 6 3 7 Low Income Inequality 6 5 4 3 2 1 2
Mixed patterns in other factors 35 Terms-of-Trade Busts 3 25 2 15 1 5 6 4 2 Relative frequency of shock in EMDEs (percent) 8 Spikes in U.S. Real Short-Term Interest Rate 5 Current Account Surplus 4 3 2 1 Share of EMDEs with characteristic (percent) 8 Low External Debt 3 Fiscal Surplus 6 25 2 4 15 2 1 5 21
What Ends Expansions? Expansions Explanatory Variable All Z statistic Pre-199 Z statistic Post-1989 Z statistic Implied S&P 1 Volatility (VXO) 1.951*** [-4.179].981 [-.985].943*** [-4.565] U.S. Ex Ante Real Interest Rate.956 [-1.461].993 [-.158].835*** [-3.479] Terms-of-Trade-Bust Indicator.968 [-.214].82 [-1.34] 1.134 [.74] Sudden Stop (capital inflows) Indicator.59*** [-2.927].497* [-1.885].841 [-1.254] Advanced Economy Recession Indicator.642*** [-4.74].668** [-2.42].68* [-1.911] Credit Boom during Past Three.616*** [-3.913].591*** [-2.621].75*** [-2.61] Banking Crisis Indicator.55*** [-3.376].54*** [-3.32].538*** [-2.83] Single-Digit Inflation Indicator 1.473*** [3.185] 1.574** [2.474] 1.276** [2.12] Low Public Debt to GDP Indicator 1.9 [.713].998 [-.117] 1.19 [.132] International Reserves to GDP 1.9*** [2.866] 1.6 [1.289] 1.4 [.93] Income Inequality (Gini coefficient).986** [-2.144].976*** [-2.833].997 [-.459] Trade Openness (exports plus imports to GDP).999 [-.451] 1.1 [.373] 1. [-.17] Financial Openness (external assets plus liabilities to GDP).999*** [-3.121].999*** [-4.84] 1. [-.549] Observations 1,264 Number of Episodes 188 Number of Exits 126 Number of Economies 75 Weibull Shape Parameter 1.516 1.48 2.277 Z statistic of Shape Parameter 6.829 3.258 2.928 Log Likelihood -13. -88.1 Model Chi-Squared p Value.. Source: IMF staff calculations. Note: Exponentiated coefficients shown are time ratios, which indicate whether the variable tends to shorten (less than 1) or lengthen (greater than 1) the expected time-in-episode. Z statistics are given in brackets underneath the coefficient estimates. A negative z statistic indicates that the associated variable tends to shorten an episode; if the z statistic is positive, it tends to lengthen an episode. *, **, and *** denote significance at the 1 percent, 5 percent, and 1 percent levels, respectively. 1 VXO = Chicago Board of Exchange S&P 1 volatility index.
External Domestic Policies Structural characteristics Total External Domestic Policies Structural characteristics Total Percent Percent Across regions and groups, some differences in drivers of improved resilience, but overall similar. Contribution of Shocks, Policies, and Structure to the Length of Expansions 5 Emerging and Developing Asia Latin America 5 4 4 3 3 2 2 1 1-1 -1-2 -2 Source: IMF staff calculations. 23
External Domestic Policies Structural characteristics Total External Domestic Policies Structural characteristics Total Percent Percent Across regions and groups, some differences in drivers of improved resilience, but overall similar. Contribution of Shocks, Policies, and Structure to the Length of Expansions 5 Sub-Saharan Africa Heavily Indebted Poor Countries 5 4 4 3 3 2 2 1 1-1 -1-2 -2 Source: IMF staff calculations. 24
External Domestic Policies Structural characteristics Total External Domestic Policies Structural characteristics Total Percent Percent And it s not just commodities Contribution of Shocks, Policies, and Structure to the Length of Expansions 5 Commodity Exporters Noncommodity Exporters 5 4 4 3 3 2 2 1 1-1 -1-2 -2 Source: IMF staff calculations. 25