Anticipating Capital Flow Reversals

International Monetary Fund. External Relations Dept.
Published Date:
January 1995
  • ShareShare
Show Summary Details

A review of the experience of developing countries that received major private capital inflows in the first half of the 1990s suggests that some of them were better able to manage these flows than others. Countries that have largely avoided sudden reversals of inflows have a number of common characteristics: stable macroeconomic policies; avoidance of protracted real exchange rate appreciation; high domestic savings; significant investment in industries in the traded goods sector; relatively restrained current account deficits; high export growth; and relatively little reliance on short-term capital inflows. In contrast, countries that have proven to be more vulnerable to shifts in market sentiment have tended to exhibit these characteristics to a far lesser degree.

A number of lessons emerge from the experiences of both sets of countries that can help policymakers avoid sudden reversals of capital flows.

Lessons of experience

Large real exchange rate appreciation. Some of the countries that experienced significant private capital flow reversals in 1995—such as Argentina, Hungary, and Mexico—had also undergone large and protracted real exchange rate appreciation in the first half of the 1990s (see chart). A surge in capital inflows attracted by high returns in the wake of a comprehensive economic reform may contribute to some appreciation of the real exchange rate. But prolonged appreciation of the real exchange rate that threatens external competitiveness and puts the external sector under pressure may create the conditions for a subsequent loss of confidence by foreign investors.

The experience of a number of countries shows that this pattern need not always prevail. For one thing, some key reform measures, such as fiscal consolidation and trade liberalization, are apt to lead to real depreciation rather than appreciation. Indeed, many of the most successful reformers and recipients of capital inflows, particularly those in Asia, have seen only modest real appreciation or even significant real depreciation of their currencies (see table). Increases in private savings in these countries have tended to offset the excess demand and inflationary pressures generated by large capital inflows, pressures that would otherwise result in real appreciation. For example, Malaysia, which was among the countries receiving the highest volume of private capital inflows relative to GDP, experienced a small depreciation of its real exchange rate. In general, the correlation between private capital inflows and real appreciation was not significant for the 16 major private capital recipients shown in the table.

In contrast, significant real appreciation is often the undesirable side effect of using the exchange rate as a nominal anchor in stabilization programs in high-inflation countries. In a number of countries, high real appreciation of the exchange rate in the 1990s was associated with the use of nominal exchange rate anchors to curb the momentum of high inflation built up in earlier years. Of the 16 countries shown in the chart, the 8 countries with the largest real appreciation had a median inflation rate of 53 percent in the 1980s while the countries with the smallest real appreciation (mostly in Asia) had a median inflation rate of only 9 percent. Stabilization programs based on a nominal exchange rate anchor were in effect at one time or another during the early 1990s in several of the countries showing high real appreciation, such as Argentina, Hungary, Mexico, and Poland. Persisting in this policy much beyond the initial phase of such programs, instead of switching to greater reliance on tight fiscal and monetary policies, risks undermining external competitiveness.

Uri Dadusha French national, is Chief of the International Economic Analysis and Prospects Division of the World Bank’s International Economics Department.

Milan Brahmbhatta UK national is an Economist in the International Economic Analysis and Prospects Division of the World Bank’s International Economics Department.

Declining private saving rates. Some decline in private saving rates could, in principle, result from a perceived increase in permanent income in the wake of successful reforms, or from an anticipation of lower taxes following cuts in government spending. The evidence suggests, however, that the most successful reformers, such as the East Asian economies and Chile, have seen increased private saving rates even when these were very high at the outset (Corbo and Hernandez, 1994).

A number of analysts (for example, Kiguel and Liviatan, 1992; Calvo and Végh, 1994) have argued that a strong initial boom in consumption is a typical result of nominal-exchange-rate-based stabilization programs, reflecting a fall in interest rates that is often an initial effect of such programs, as well as anticipation on the part of the public that the program may not be sustainable. If the consumption boom persists—implying rapidly declining private savings—it can have important consequences for the sustainability of the balance of payments position and the credibility of the exchange rate anchor itself. For example, the ratio of private consumption to GDP in Mexico increased by 5 percentage points in 1990-93, compared with 1985-89 (see table). In four other countries with high real appreciation (Argentina, Hungary, Peru, and Poland), the increase in the consumption share between these two periods averaged 9 percentage points. In contrast, it fell by about 1 percent in the four countries in the table with the least real appreciation. Given that fiscal deficits improved by 2-3 percent of GDP, on average, in both high and low real appreciation countries, private savings rates must have fallen significantly in high appreciation countries.

Exchange rate appreciation/depreciation among major private capital recipients1/

(percent change, 1990-94)

Source: IMF Research Department.

1 Real effective exchange rate.

Inadequate investment in the tradables sector. Insofar as capital flows are attracted by increased efficiency, one would expect to see significant foreign and domestic investment directed to the traded goods sector, reflecting improved international competitiveness. If, however, capital flows finance a consumption boom and/or are driven by “herd instincts,” one might expect the inflows to find their way mostly to the nontradables sector, where relative prices and profitability have been bid up by high domestic demand (some of it induced by the capital flows themselves). Although comparable data on tradables and nontradables investment are not available, it is notable that the services sector, which is most closely associated with nontradables, expanded relative to GDP in the high appreciation countries but not in the low appreciation ones. Real growth of the service sector in 1990-93 in the high real appreciation half of countries shown in the chart exceeded overall GDP growth by 1.4 percent a year but only by 0.3 percent a year in the low appreciation half. High real appreciation countries also experienced poorer overall investment growth than low appreciation ones (see table).

Low export growth. This is a critical signal that a country’s ability to service mounting foreign obligations could be called into question by investors. In theory, perfect capital markets could supply virtually unlimited credit to a country expecting a future boom in exports but, in practice, creditworthiness indicators are based on the evolving track record of export growth and the country’s share in export markets. If the country is rated a high risk, investors will expect commensurately high returns, and rapid export growth becomes even more important to sustained capital inflows. As would be expected, export growth in the early 1990s showed a fairly close inverse relationship with real exchange rate appreciation, averaging 7 percent (in nominal dollar terms) in the high appreciation half of countries but 12 percent in the low appreciation half.

Unsustainable current account deficits. There is no hard and fast rule as to what constitutes a sustainable deficit, though a rule of thumb can be derived from setting a prudent target on foreign borrowing. If, for example, the target is the commonly cited prudent upper bound on the foreign liabilities/export ratio of 2, then the sustainable current account deficit in the long term, expressed as a proportion of exports, is equal to twice the export growth rate. Computations of sustainable current account deficits using this rough rule of thumb show that they vary greatly by country. The median level for large recipients of private capital in Europe and Central Asia is 2.5 percent of GDP, in Latin America 2.2 percent, and in Asia 8.9 percent. Among high appreciation countries, actual current account deficits in 1994 averaged about 2 percent of GDP greater than sustainable levels. But 1994 deficits in low appreciation countries were more than 6 percent of GDP below sustainable levels—a further degree of prudence no doubt reflecting the possibility that external conditions could lead to lower export growth than the rapid pace achieved by these countries in the early 1990s.

The rule of thumb applied here gives only a first indication. A fuller computation of sustainable deficits would account for other factors such as the level of foreign aid, which would increase the sustainable deficit; the maturity and initial level of foreign liabilities; the volatility of export earnings; and the import content of exports. Increases in all of these variables would tend to reduce the prudent liability/export ratio and the sustainable deficit.

Growing reliance on short-term flows. Short-term borrowing is appropriate for short-term needs such as trade, inventory, and working capital finance, but it is inappropriate for long-term projects, where it places borrowers in the precarious position of having to continuously refinance the investment. Growing reliance on short-term capital inflows (especially to finance current account deficits) may be a sign of strain. Although the borrower prefers long-term financing, lenders or direct investors are reluctant to comply because of perceived risk.

Key indicators for major recipients of private capital(percent)
Average real exchange rate change1Long-term private capital flows2Inflation 3Change in private consumption- GDP ratio 4Change in investment- GDP ratio 5Average annual export growth 6Current account to GDP ratio 7Deviation from sustainable current account ratio
Sources: International Economics Department, World Bank; and IMF for real effective exchange rates. Consumption and investment ratios and goods and nonfactor service export growth are measured in nominal dollar terms.


As percent of GDP, 1990-93.

Consumer price index, 1980-89.

1990-93 versus 1985-89.

1990-93 versus 1985-89.



Sources: International Economics Department, World Bank; and IMF for real effective exchange rates. Consumption and investment ratios and goods and nonfactor service export growth are measured in nominal dollar terms.


As percent of GDP, 1990-93.

Consumer price index, 1980-89.

1990-93 versus 1985-89.

1990-93 versus 1985-89.



Large currency exposures. Stabilization programs that use the exchange rate as a nominal anchor create a strong incentive for domestic firms to borrow at low interest rates in foreign currency, and for foreign investors and domestic financial intermediaries to lend at high interest rates in domestic currency, while carrying the currency risk. At the same time, the maturity structure of loans tends to become shorter as both lenders and borrowers try to minimize currency risk. The absence of forward markets aggravates the problem, since investors find it difficult to cover currency exposures. This behavior creates powerful impediments to devaluation, even when it is clearly needed. Large currency exposure is not so much an early signal of crisis as it is an indication of the disruption that might ensue.

Monetary tightening in industrial countries. Capital flows to developing countries may be adversely affected by monetary tightening in industrial countries. A consequent sharp rise in interest rates would reduce the relative attractiveness of developing country assets and would also be accompanied eventually by falling aggregate demand in industrial countries and slower export growth in developing countries. Heavily indebted countries would see a large deterioration in creditworthiness indicators. This scenario helped precipitate the debt crisis of the 1980s. However, the rise in interest rates in 1994 does not wholly conform to this pattern. Rate increases were moderate in comparison with those of the early 1980s, and their effect on creditworthiness was offset by global recovery and a surge in world trade.

The overall picture. This list of early warning signals is not exhaustive. Attention also needs to be paid to controlling quasi-fiscal as well as fiscal deficits, the appropriateness of monetary policy, and political stability. Other warning signals, such as high real interest rates to defend the currency, or slow economic growth resulting from currency overvaluation and high interest rates, may be late rather than early indicators of fundamental imbalance. Nevertheless, events since the start of the Mexican crisis have tended to confirm the value of the indicators discussed here. After a sharp pullback from most emerging markets in the immediate wake of the crisis, private capital flows to developing countries have recovered more quickly and substantially than seemed likely at the start of 1995. But the recovery has been most pronounced for countries with the best early warning indicators, such as the East Asian countries, and Chile and Colombia.

Suggestions for further reading:

    GuillermoCalvo and CarlosVeghInflation Stabilization and Nominal Anchors,Contemporary Economic Policy Vol 12 (April1994) pp. 35-45.

    V.Corbo and L.HernandezMacroeconomic Adjustment to Capital Flows: Latin American Style versus East Asian Style,Policy Research Working Paper No. 1377 (WashingtonWorld Bank1994).

    UriDadushAshokDhareshwar and RonJohannesAre Private Capital Flows to Developing Countries Sustainable?Policy Research Working Paper No. 1397 (WashingtonWorld BankDecember1994).

    RudigerDornbusch and AlejandroWernerMexico: Stabilization, Reform, and No Growth,Brookings Papers on Economic Activity: 1 (Brookings InstitutionWashington1994) pp. 253-315.

    MiguelKiguel and NissanLiviatanThe Business Cycle Associated with Exchange Rate Based Stabilizations,World Bank Economic Review Vol. 6 (May1992) pp. 279-305.

Other Resources Citing This Publication