Journal Issue

Choosing an exchange rate regime to deal with capital account shocks: pick your poison

International Monetary Fund. External Relations Dept.
Published Date:
March 2004
  • ShareShare
Show Summary Details

What is the optimal exchange rate regime—rigidly fixed, freely fluctuating, or a managed float? Historically, most emerging market economies have opted to peg their exchange rates, but such regimes are more susceptible to the speculative attacks associated with substantial capital outflows. Some have suggested that emerging market economies should permit more exchange rate flexibility, but greater flexibility can have other, negative consequences for economic performance. In a recent IMF Working Paper, Shigeru Iwata (Professor, Department of Economics, University of Kansas) and Evan Tanner (Senior Economist, IMF Institute) assess the trade-offs.

When confronted by adverse capital account shocks, central banks must choose how such shocks are transmitted to the domestic economy: they must, in effect, “pick their poison.” As one option, central banks can defend the exchange rate by raising interest rates, but doing so is likely to adversely affect economic activity and financial sector balance sheets. Alternatively, central banks can let the currency depreciate, but this can spur inflation and it will increase the domestic currency value of the country’s foreign currency liabilities. If a central bank wanted to limit both interest rate and exchange rate changes, it could instead let international reserve flows absorb the shock. That is, the central bank would sell foreign exchange reserves to maintain the currency’s value—but this policy may be risky and will, after a period, become unsustainable.

Using a sample of three emerging market economies that have faced external shocks over the past decade—Brazil, Mexico, and Turkey—Iwata and Tanner analyze how the shocks were channeled into international reserve movements, interest rate movements, and exchange rate depreciation. They then assess the impact of such shocks on the domestic economy. For each country, Iwata and Tanner analyze selected subperiods that correspond to distinct, well-known policy regimes. For Mexico, they contrast the low-inflation, managed exchange rate period before the Tequila crisis (January 1988 to October 1994) with the postcrisis period of more flexible exchange rates (January 1995 to December 2000). Similarly, in Turkey, they look at two periods separated by an exchange rate crisis—namely, January 1987 to December 1993 and May 1994 to January 2001. For Brazil, data are available only for the exchange rate band period (August 1994 to December 1998).

Characterizing regimes

Characterizing a country’s exchange rate regime can be difficult. Indeed, a country’s stated exchange rate policy, as presented in the IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions, can differ substantially from the policy it actually pursues. The same report now also shows how IMF staff classifies countries’ actual exchange rate policies. Exactly how to classify an exchange rate regime—that is, its degree of flexibility—is a difficult question. In the approach adopted by Iwata and Tanner, they first isolate econometrically shocks to the capital account, where a positive (negative) shock is associated with a capital inflow (outflow). They then estimate parameters that correspond to the fractions of a given shock that are channeled into reserve movements, interest rate changes, and currency depreciation. Their estimated parameters can be combined to form a single indicator of exchange rate flexibility: the exchange rate is considered to be freely floating if the entirety of an external capital account shock is channeled to exchange rate movements. Moreover, the estimated parameters also reveal what combination of reserve and interest rate movements a central bank uses when it defends the exchange rate.

Their estimates suggest that prior to Mexico’s 1994 exchange rate crisis, the entirety of a given external shock was absorbed by sales or purchases of international reserves. But after the crisis, Mexico’s policy changed. During 1995-2000, the fraction of an external shock that was absorbed by reserve movements fell to about 80 percent; to compensate, interest rates and exchange rate depreciation each absorbed about 10 percent. Thus, while exchange rate flexibility increased, the authorities nonetheless continued to actively manage the exchange rate.

During Brazil’s managed float period (1994-99), the authorities permitted about 80 percent of an external shock to be absorbed by reserve movements; of the remainder, about 18 percent was absorbed by interest rate hikes, so that the exchange rate absorbed only about 2 percent, on average, of a given shock.

In Turkey, as in Mexico, there is a striking contrast in central bank policy before and after 1994. In the earlier period, the authorities channeled over 90 percent of a shock into reserve movements, but only about 54 percent afterward. During the later period, the Turkish authorities relied to a greater degree on interest rates to defend the domestic currency. The degree of overall exchange rate flexibility stayed about the same: exchange depreciation reflected about 6 percent of a given shock in both periods.

Assessing the domestic impact

What effects, if any, did the capital account shocks have on these three economies given their various exchange rate regimes? Looking at two key variables—industrial output growth and inflation—Iwata and Tanner find strong evidence for Mexico and Turkey that the effects of an external shock depended on the exchange rate regime chosen by the authorities. While in Mexico external shocks had no statistically significant effect on output growth prior to the 1994 crisis, significant effects were evident (with a three-month lag) after the crisis. This finding supports the idea that the exchange rate regime affects the transmission of capital account shocks to the domestic economy. The central bank did not permit interest rates to respond to capital account shocks in the precrisis period, but did so in the post-crisis period. Thus, stabilizing the exchange rate through interest rate movements (as happened in the postcrisis period) might have made output growth more volatile than it would have been otherwise.

The effects of external shocks on Mexican inflation also differed substantially between the exchange rate regimes. Prior to the 1994 crisis, when exchange rates were tightly pegged by reserve movements, positive (negative) external shocks to the capital account had positive (negative) effects on inflation. Iwata and Tanner say that this is a logical outcome because a positive shock to the capital account will cause the real exchange rate to appreciate. If the nominal exchange rate is pegged—as it was before the crisis—this appreciation takes place through domestic (nontradable) inflation. However, in the postcrisis period, when exchange rates were more flexible, the real appreciation associated with an increase in capital inflows instead took place through a nominal appreciation. This reduced inflation through the tradable goods component.

In Turkey, differences were also evident from one regime to another. During the earlier period, capital account shocks were primarily channeled by the central bank to reserve flows and, with a lag, to the exchange rate. Here, such shocks, when positive, significantly raised output growth, with a lag of four months. However, in the later period, shocks were absorbed to a much greater extent by interest rates, thereby affecting output about two months sooner and by a somewhat greater magnitude. The behavior of inflation also mirrors the exchange rate regime. In the earlier period, a positive capital account shock would significantly increase inflation, with a one-month lag. By contrast, in the later period, external shocks had no significant impact on inflation. This finding reflects the more active interest rate policy during the later period.

Iwata and Tanner also address a counterfactual question: how would inflation and output have behaved in these countries if the central bank had chosen a different exchange rate regime? Their simulation has important implications for inflation: if the exchange rate regime were to become more flexible—that is, if external shocks were to be absorbed to a greater degree by exchange rate movements rather than interest rate movements—the inflation rate would become even more sensitive to capital account shocks because of the exchange rate pass-through mechanism. By contrast, if such shocks are transmitted to interest rates, with less exchange rate flexibility, such passthrough would be less, making inflation less sensitive to capital account shocks.

Iwata and Tanner’s results suggest that capital account shocks can have substantial effects on inflation and growth. The nature and magnitude of the effects depend on the exchange rate regime. When capital account shocks are channeled into the exchange rate, they are passed through to inflation. If such shocks are channeled into the interest rate, exchange rate pass-through effects are substantially reduced. The interest rate movements may also be transmitted to output: an increase in interest rates, for example, is likely to reduce output growth. Inescapably, Iwata and Tanner argue, authorities must pick their poison: in the event of an adverse shock, they must choose between higher inflation and lower growth.

Copies of IMF Working Paper No. 03/92, “Pick Your Poison: The Exchange Rate Regime and Capital Account Volatility in Emerging Markets,” by Shigeru Iwata and Evan Tanner, are available for $15.00 each from IMF Publication Services. Please see page 52 for ordering details. The full text is also available on the IMF’s website (

Other Resources Citing This Publication