POVERTY THE WORLD BANK REDUCTION AND ECONOMIC MANAGEMENT NETWORK (PREM) Economic Premise JUNE 2010 · Number 20 55454 Can Real Exchange Rate Undervaluation Boost Exports and Growth in Developing Countries? Yes, But Not for Long Mona Haddad and Cosimo Pancaro A policy of managed real undervaluation may have been an important factor behind the success of East Asia's export-led growth model. But current discussions over the value of China's currency demonstrate the controversy this kind of policy can generate. Although a managed real undervaluation can enhance domestic competitiveness, it is difficult to sustain--both economically and politically--in the post-crisis environment. We show that a real undervaluation works only for low-income countries, and only in the medium term. In the wake of the global financial crisis, academics and pol- nomic policies rather than by market fluctuations. The real icy makers are questioning the relative merits of export-led exchange rate depends on the balance between savings and growth strategies. Some have argued that many of the con- investment, and on the balance between expenditures and ditions responsible for that model's success are no longer in income. Hence, all policies that produce higher savings rel- place. Many developing countries have relied on an under- ative to investment can lead to real exchange rate deprecia- valued real exchange rate to boost their exports. But global tion. economic prospects are weaker than in the past, and there Governments have a variety of policy instruments avail- is greater uncertainty about the advanced economies' capac- able to achieve a competitive real exchange rate and, poten- ity to continue absorbing developing countries' exports. tially, real undervaluation. Examples include a moderate Moreover, a strategy of export-led growth paired with man- fiscal consolidation in the presence of a low level of private aged undervaluation is likely to incur costs if the real ex- absorption; the introduction of capital controls on capital change rate is kept too low for too long. But how effective inflows and the liberalization of capital outflows; targeted has real devaluation been in boosting exports and growth-- interventions on foreign exchange markets; and a nominal and is it sustainable? depreciation associated with anti-inflationary policies, such as price and wage moderation. The Role of the Real Exchange Rate in Empirical evidence shows that real exchange rate varia- Boosting Economic Growth in Developing tions can affect growth outcomes. Faster economic growth Countries is significantly associated with real exchange rate deprecia- tion (Hausmann, Pritchett, and Rodrik [2005], based on an In most developing countries, the real exchange rate is man- analysis of more than 80 episodes of growth acceleration be- aged to various degrees and is largely determined by eco- tween 1957 and 1992). Real overvaluation hampers exports 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise and leads to a fall in economic growth (Easterly 2005; John- Figure 1. Impact of 50 Percent Real Undervaluation on Growth of son, Ostry, and Subramanian 2007). Annual Per Capita Income Rodrik (2009) argues that real undervaluation promotes real per capita economic growth, increases the profitability of the tradables greater than $6,000 sector, and leads to an expansion of the share of tradables in income $2,500­$6,000 domestic value added. He claims that the tradables sector in developing countries can be too small because it suffers less than $2,500 more than the nontradables sector from institutional weak- nesses and market failures. A real exchange rate undervalu- 0 0.5 1.0 1.5 2.0 ation works as a second-best policy to compensate for the percent negative effects of these distortions by enhancing the sector's Source: Authors' estimates. profitability. Higher profitability promotes investment in the Note: All coefficients are significant at the 1 percent level, but the coefficient for countries with real per capita incomes between $2,500 and tradables sector, which then expands, and promotes eco- $6,000 is not significant. nomic growth. Other researchers showed alternative channels for under- valuation of the real exchange rate to raise growth. Levy- ing Rodrik (2009). Real undervaluation has a positive effect Yeyati and Sturzenegger (2007) claim that an undervalued on economic growth and on export expansion, but this effect real exchange rate boosts output and productivity growth not is significant only for countries with low per capita income. through an expansion of the tradables sector, but through an In developing countries with per capita incomes below increase in savings and capital accumulation. Korinek and $2,500, an increase of 50 percent in real undervaluation is Servén (2010) show that real exchange rate undervaluation associated with an annual 1.7 percent increase in growth (fig- can raise growth through learning-by-doing externalities in ure 1) and an annual 1.8 percent increase in exports over the tradables sector--externalities that are suboptimally pro- GDP (figure 2) in the corresponding five-year period.1 duced in the absence of policy intervention. They suggest that In the long run, the effect of a real exchange rate under- foreign reserves accumulation lowers the real exchange rate valuation on economic growth becomes negative; and on ex- and encourages learning-by-doing externalities of export-led ports, it becomes insignificant. An extended specification of growth without the need for direct subsidies. However, this the model (which includes the lagged effect of undervalua- impact is reduced as more countries embrace such a strategy. tion) shows that in developing countries with per capita in- However, a competing strand of the literature claims that comes below $2,500, real undervaluation has a positive large misalignments of the real exchange rate from its equi- contemporaneous effect on growth but a negative lagged ef- librium hamper economic growth. Its proponents define the fect. In developing countries with per capita incomes lower equilibrium real exchange rate as the rate that guarantees than $6,000 and higher than $2,500, real undervaluation macroeconomic equilibrium in the medium term. They has an insignificant contemporaneous effect and a negative argue that a real overvaluation causes current account lagged effect on growth. Real undervaluation has only a pos- deficits and can lead to currency crises, whereas a real un- itive contemporaneous effect on exports; its lagged effect on dervaluation causes inflation and can lead to an overheating exports is insignificant for all income levels. of the economy. They also assert that the real exchange rate should be determined only by market forces and that large misalignments generate distortions, produce wrong signals Figure 2. Impact of 50 Percent real undervaluation on the Exports- to economic agents, lead to factor misallocation, and cause to-GDP Ratio instability. Consistent with this literature, Aguirre and Calderón (2005) show that both large real exchange rate real per capita greater than $6,000 overvaluation and large real exchange rate devaluations income hamper economic growth. The larger the misalignments, the $2,500­$6,000 larger is the decline in growth. less than $2,500 Evidence of the Positive Impact of Real 0 0.5 1.0 1.5 2.0 Exchange Rate Undervaluation on GDP percent Growth and Export Expansion Source: Authors' estimates. Note: All coefficients are significant either at the 1 percent or at the 5 We provide further evidence on the links among the real ex- percent level, but the coefficient for countries with real per capita incomes change rate, economic growth, and export expansion follow- between $2,500 and $6,000 is not significant. 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Real Exchange Rate Variability and macroeconomic stability (Eichengreen 2008). The effect of Economic Growth this variability on economic growth depends crucially on the level of financial sector development. When financial markets Perhaps more important than the level of the real exchange are sufficiently developed, agents can use sophisticated finan- rate is its variability. An unstable real exchange rate causes cial instruments to hedge against risk (Aghion et al. 2009). more volatile relative prices, creates uncertainty, increases Previous research on the impact of exchange rate volatil- risk, and shortens investment horizons. Large real exchange ity on growth has reached contrasting results. Ghosh et al. rate misalignments do not provide the right incentives for (1997) find that exchange rate variability does not affect investment over time, and they negatively affect the quality economic growth. Bleaney and Greenaway (2001) find that of investment. Frequent shifting of resources between the exchange rate instability negatively affects investment in tradables and nontradables sectors in response to recurrent Sub-Saharan African countries. Bosworth, Collins, and Yu- changes in relative prices leads to higher adjustment costs. chin (1995) provide evidence that in a large sample of in- Frequent movements in exchange rate expectations cause dustrial and developing countries, real exchange rate interest rate volatility and financial instability. volatility hampers economic growth and reduces productiv- A volatile real exchange rate reduces the profitability of ity growth. Aghion et al. (2009) find a similar result, but firms and hampers economic growth. But the final effect of they also show that the negative effect of real exchange rate real exchange rate volatility on economic performance can volatility on economic growth shrinks in countries with also depend on complementary factors, such as political and higher levels of financial development. Servén (2003) shows that real exchange rate volatility negatively affects invest- ment in a large panel of developing countries. This negative Figure 3. Impact of 50 Percent Real Undervaluation on Annual impact is significantly larger in countries with highly open Growth of Per Capita Income economies and less developed financial systems. He also finds evidence of threshold effects, whereby uncertainty only matters when it is relatively high. real per capita greater than $6,000 The literature concerning the effects of real exchange rate income $2,500­$6,000 volatility on export expansion in developing countries also less than $2,500 has not reached a clear-cut consensus. Arize, Osang, and Slottje (2000) find a significant negative relationship be- ­1.5 ­1.0 ­0.5 0 0.5 1.0 1.5 2.0 2.5 tween an increase in exchange rate volatility and exports in percent developing countries. Chit, Rizov, and Willenbockel (2010) lagged effect contemporaneous effect find a similar result for a panel of East Asian countries. Sauer and Bohara (2001) show that volatility has significant neg- Source: Authors' estimates. Note: All coefficients are significant at the 1 percent level, except for the ative effects on exports in Latin America and Africa, but not contemporaneous coefficient for countries with real per capita GDP in Asia. They argue that the effect of real exchange rate between $2,500 and $6,000, which is not significant. volatility on exports depends on the type of goods and the countries involved. Some studies also show that exchange Figure 4. Impact of 50 Percent Real Undervaluation on Annual rate variability has no significant effect in driving export vol- Exports-to-GDP Ratio umes (Klein 1990; Mckenzie 1998). Evidence of the Negative Impact of Real real per capita greater than $6,000 Exchange Rate Volatility on GDP Growth and income $2,500­$6,000 Export Expansion less than $2,500 A stable real exchange rate is a necessary condition for devel- ­2.0 ­1.0 0 1.0 2.0 3.0 4.0 oping countries to achieve sustained economic growth, but percent only large fluctuations matter for exports. Our evidence con- firms the existence of a significant negative relationship be- lagged effect contemporaneous effect tween real exchange rate volatility and real per capita GDP Source: Authors' estimates. growth (figure 5).2 This relationship does not significantly Note: All lagged coefficients are not significant. All contemporaneous change when observations with extremely high real exchange coefficients are significant at the 1 percent level, except for the contemporaneous coefficient for countries with real per capita GDP rate volatility are dropped. Indeed, it becomes even stronger. between $2,500 and $6,000, which is significant at the 10 percent level. Our evidence also confirms the existence of a significant neg- 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Figure 5. Impact of Real Exchange Rate Variability on Growth serves, which is an inefficient outcome. The return produced 10 by foreign reserves is lower than the return produced by the same amount of wealth invested either in infrastructure or in real per capita GDP growth in percent 5 a well-diversified portfolio in international financial markets. Second, a real undervaluation led by nominal deprecia- 0 tion and not associated with anti-inflationary policies (such as price and wage moderation) can cause high and destabi- ­5 lizing liquidity growth and inflation. It also can lead to fi- nancial instability. 0 0.2 0.4 0.6 0.8 Third, an undervalued rate can constrain monetary policy, real exchange rate volatility leaving it no longer free to target domestic objectives. Such Source: Authors' estimates. constraint may cause an artificial process of overlending and Note: The graph presents the cross-country relationship between average overinvestment, mainly in the presence of an open capital real per capita GDP growth and real exchange rate volatility for countries account. The result can be an overheating economy. with per capita incomes of less than $6,000 between 1980 and 2004. Coefficients are significant at the 1 percent level. Fourth, maintaining an undervalued real exchange rate for a long period of time may reduce the incentives to create a more developed financial sector. Figure 6. Impact of Real Exchange Rate Variability on Exports Fifth, an artificial undervaluation is akin to a subsidy for 100 firms that produce tradable goods. The subsidy is paid for by an implicit tax on consumers, who consequently have re- export as percent of GDP 80 duced purchasing power. Finally, it may be difficult to exit a policy of sustained un- 60 dervaluation once it becomes necessary to do so. Governments 40 may be pressured by influential lobbies (that is, tradable goods producers) who derive rents from the status quo. 20 A stable and undervalued real exchange rate is a key step in promoting economic growth, but by no means is it a suf- 0 ficient condition. The adoption of managed undervaluation as a policy aimed at enhancing sustainable growth should be 0 0.2 0.4 0.6 0.8 preceded by an in-depth analysis of its welfare implications. real exchange rate volatility Policy makers must be aware that the costs of such a policy may outweigh the benefits in the medium to long term. To Source: Authors' estimates. Note: The graph presents the cross-country relationship between average be successful, this policy must be accompanied by other nec- exports as a percent of GDP and real exchange rate volatility for countries essary and complementary conditions, such as strong insti- with per capita income less than $6,000 between 1980 and 2004. tutions and macroeconomic stability. A country must be Coefficients are significant at the 1 percent level, but not significant if observations with real exchange rate variability above 0.4 percent are prepared to exit before costs begin outweighing benefits-- dropped. a process that can be done by announcing a clear exit plan when the policy is adopted. To reduce the negative growth ative relationship between real exchange rate variability and effects of possible real overvaluation and of higher real ex- the exports-to-GDP ratio (figure 6). However, this relation- change rate volatility, it is best to move away from this policy ship is no longer significant when observations with extremely strategy when the economy is still strong and confidence is high real exchange rate volatility are dropped. Only large real high. exchange rate variability appears to matter for exports. Notes Maintaining an Undervalued Real Exchange Rate over Time Is Not Sustainable 1. Closely following Rodrik (2009) and using the Penn World Tables 6.3 data (Heston, Summers, and Aten 2009) A stable and undervalued real exchange rate can be a key el- for the period 1950­2004, we first compute a measure of ement in promoting economic growth in low-income coun- real undervaluation for a panel of 187 countries and then tries; but maintaining this policy for too long may have estimate the effects of real undervaluation on per capita in- significant adverse consequences. First, such a policy may come growth for a panel of developing countries. We define cause an excessive accumulation of low-yielding foreign re- "developing countries" as those countries with GDP per 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise capita incomes lower than $6,000. For each time-series, we Arize, A. C., T. Osang, and D. J. Slottje. 2000. "Exchange-Rate Volatility and take a five-year average. This generates a maximum of 11 Foreign Trade: Evidence from Thirteen LDCs." Journal of Business and Economic Statistics 18 (1): 10­17. five-year time periods for each country. The real undervalu- Bleaney, M., and D. Greenaway. 2001. "The Impact of Terms of Trade and ation index is a measure of the deviation of the actual real Real Exchange Rate Volatility on Investment and Growth in Sub-Sa- exchange rate from the purchasing power parity real ex- haran Africa." Journal of Development Economics 65 (2): 491­500. change rate, which takes into account the Balassa-Samuelson Bosworth, B., S. Collins, and C. Yu-chin. 1995. "Accounting for Differences effect--that is, the fact that the relative price of nontradable in Economic Growth." Unpublished manuscript. Brookings Institution, Washington, DC. goods is higher in countries with higher per capita incomes. Chit, M. M., M. Rizov, and D. Willenbockel. 2010. "Exchange Rate Volatil- The real undervaluation index is computed as the residual ity and Exports: New Empirical Evidence from the Emerging East Asian of a panel regression of the real exchange rate on real GDP Economies." The World Economy 33 (2): 239­63. per capita and time-fixed effects. The real exchange rate is Easterly, W. 2005. "National Policies and Economic Growth: A Reappraisal." defined as the ratio of the nominal exchange rate to the pur- In Handbook of Economic Growth, ed. P. Aghion and S. Durlauf, 1015­ chasing power parity conversion factor. To capture the ef- 56. Amsterdam, The Netherlands: Elsevier. Eichengreen, B. 2008. "The Real Exchange Rate and Economic Growth." fects of real undervaluation on growth, we estimate a panel Working Paper No. 4. Commission on Growth and Development, World regression of GDP per capita growth on the real exchange Bank, Washington, Dc. rate undervaluation index, the level of initial income, and Ghosh, A., A.-M. Gulde, J. D. Ostry, and H. Wolf. 1997. "Does the Nominal time- and country-fixed effects. To capture the effects of real Exchange Rate Regime Matter?" Working Paper 5874. National Bureau undervaluation on export expansion, we estimate a panel of Economic Research, Cambridge, MA. Hausmann, R., L. Pritchett, and D. Rodrik. 2005. "Growth Accelerations." regression of export over GDP on the real exchange rate un- Journal of Economic Growth 10 (4): 303­29. dervaluation index, the level of real income per capita, and Heston, A., R. Summers, and B. Aten. 2009. Penn World Table, Version 6.3. time- and country-fixed effects. Technical Report. Center for International Comparisons of Production, 2. This finding is based on estimates of the relationship Income and Prices, University of Pennsylvania, Philadelphia. between real exchange rate volatility--measured as the stan- Johnson, S., J. D. Ostry, and A. Subramanian. 2007. "The Prospects for Sus- dard deviation of RER (defined as the ratio between the tained Growth in Africa: Benchmarking the Constraints." Working Paper 07/52. International Monetary Fund, Washington, DC. nominal exchange rate and purchasing power parity)--and Levy-Yeyati, E., and F. Sturzenegger. 2007. "Fear of Appreciation." Policy average real GDP per capita growth in a panel of countries Research Working Paper 4387. World Bank, Washington, DC. with GDP per capita incomes lower than $6,000 between Klein, M. W. 1990. "Sectoral Effects of Exchange Rate Volatility on United 1980 and 2004. States Exports." Journal of International Money and Finance 9 (3): 299­ 308. About the Authors Korinek A., and L. Servén. 2010. "Undervaluation through Foreign Reserve Accumulation: Static Losses, Dynamic Growth." Policy Research Work- ing Paper 5250. World Bank, Washington, DC. Mona Haddad is sector manager of the Trade Department Mckenzie, M. D. 1998. "The Impact of Exchange Rate Volatility on Aus- (PREM) at the World Bank, Washington, DC. Cosimo Pan- tralian Trade Flows." Journal of International Financial Markets, Institu- caro is a consultant in MNACE at the World Bank Group. tions and Money 8 (1): 21­38. The authors would like to thank Luís Servén and Ha Rodrik, D. 2009. "The Real Exchange Rate and Economic Growth." In Nguyen for valuable comments and Ben Shephard for editorial Brookings Papers on Economic Activity, Fall 2008, ed. D. W. Elmendorf, N. G. Mankiw, and L. H. Summers, 365­412. Washington, DC: Brook- assistance. ings Institution. Sauer, C., and A. K. Bohara. 2001. "Exchange Rate Volatility and Exports: References Regional Differences between Developing and Industrialized Coun- tries." Review of International Economics 9 (1): 133­52. Aghion, P., P. Bacchetta, R. Ranciere, and K. Rogoff. 2009. "Exchange Rate Servén, L. 2003. "Real-Exchange-Rate Uncertainty and Private Investment Volatility and Productivity Growth: The Role of Financial Develop- in LDCs." Review of Economics and Statistics 85: 212­18. ment." Journal of Monetary Economics 56 (4): 494­513. Aguirre, A., and C. Calderón. 2005. "Real Exchange Rate Misalignments and Economic Performance." Working Paper 315. Central Bank of Chile, Santiago. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. It is produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at www.worldbank.org/economicpremise.