Recent events in Asia must have convinced anyone who still needed convincing that financial crises in emerging markets must be a major priority for research in economics. Economies that until recently had been hailed as paragons of good macroeconomic management suddenly find themselves mired in a massive collapse of asset prices and economic activity. This situation is unprecedented in scope, but not kind. Other financial meltdowns, particularly in Chile in 1982 and in Mexico in 1994, were just as surprising to observers and just as difficult to rationalize using standard models.
A model that provides a useful formalization of this phenomenon must have the following features:
1. It must not rely on government misbehavior to generate the crisis. A striking fact of recent crises is that government budgets were either in balance or showed surpluses. This has been stressed by Velasco (1987) for the case of Chile, by Sachs, Tornell and Velasco (1996a) for Mexico and by Radelet and Sachs (1998) for Asia. This means that “first generation” models of currency crises, (pioneered by Krugman, 1979) which rely on large money-financed fiscal deficits to generate reserve erosion and an eventual currency crash, are not well suited to explain these recent crashes.
2. It must be general enough to accommodate a wide variety of macroeconomic circumstances. As Frankel and Rose (1996) and Sachs, Tornell and Velasco (1996b) have shown, there is no unique pattern of behavior for basic macroeconomic variables in the buildup to a crisis. Sometimes the current account is in deficit, but not always. The same is true for private consumption and investment. In the Frankel and Rose (1996) study of macroeconomic behavior over a large set of currency crises, there is often a contraction in output the year of the crisis but, as the authors themselves point out, causality could run in either direction. This suggests that “second generation” models (pioneered by Obstfeld 1994), in which the government devalues in reaction to mounting unemployment and/or a growing external imbalance, are not too useful either.
3. It must be specific enough to explain why in some of these macroeconomic scenarios a crisis occurs, and in some it does not. It must answer the question of why now and not before. Take the case of the current account, whose behavior is often blamed for currency crashes. East Asian countries often had large current account deficits in the 1980s and early 1990s, but the crash did not happen until now. Why? It must also answer the question of why here and not elsewhere: in 1996 Malaysia, Korea, the Philippines and Thailand had large current account deficits (above 4 percent of GDP), but so did Brazil, Chile Colombia, and Peru. Yet the crisis happened in Asia and not in South America.