[Access article in PDF]
On the Consequences of Sudden Stops
Money . . . has oppressed nearly all people in one of two ways: either it has been abundant and very unreliable, or reliable and very scarce.
Sudden stops in capital flows—and the ensuing current account reversals they induce—have been at the center of economic policy discussions since the outbreak in the mid-1990s of the series of financial crises that plagued emerging market economies. Sudden stops spared no region and have been particularly prevalent in both Asia and Latin America.
As a result of these crises, policymakers and researchers alike directed their attention toward identifying the causes and designing policies to prevent crises.1 In fact, policy circles placed substantial effort in the development of a system of early warning signals, under the presumption that some key country fundamentals would be sufficient for identifying future crises.2
The reality of capital market behavior soon showed, however, that a number of institutional and regulatory factors might easily spur contagion across seemingly unconnected economies, often with little relation to the [End Page 171] quality of domestic policies.3 Chile, a model of macroeconomic prudence, suffered a massive sudden stop in the aftermath of the Russian crisis, and even high income economies such as Portugal and Spain were unable to avoid a sudden stop in the wake of the collapse of the European Monetary System earlier in the decade.
If sudden stops are to become a permanent feature of the landscape facing emerging market economies, equal emphasis should be placed on understanding how a sudden stop affects an economy and how the costs of such an event can be minimized. Casual evidence suggests that countries may experience large discrepancies in the aftermath of a sudden stop. For example, a comparison of Asian and Latin American countries, such as we make in this paper, shows that the former tend to adjust to a sudden stop via fast export growth; as a result, Asian recessions have been short-lived and recoveries swift. In contrast, adjustment in Latin American economies has occurred via import and demand contraction, while exports have remained stagnant even in the wake of significant exchange rate devaluations. Latin American recessions have thus been protracted relative to the Asian experience.
A number of relevant questions emerge from these observations. How frequently are economies likely to be exposed to sudden stops in capital flows? Why does an economy adjust to a sudden stop in capital flows via export growth or import contraction? How does this translate into gross domestic product (GDP) growth or investment decisions? What causes this different behavior? What are the implications and policy lessons? Such questions constitute the focus of this paper.
A large body of literature focuses on understanding the reasons for a sudden stop or designing appropriate prevention measures to avoid such events.4 Much less work aims at identifying the characteristics that may determine a less painful aftermath to the crisis, yet understanding whether the adjustment comes through output and export growth or through domestic absorption contraction has important policy implications.
An analogy may help illustrate the objective of our exercise. Airbags are not useful for preventing accidents, but they help reduce the costs associated with them. In this paper we are not concerned with the question of [End Page 172] how an economy comes to suffer from a sudden stop (the accident), but with assessing what sort of airbags might minimize the damage. In other words, we look at what characteristics of an economy might make a sudden stop less painful.
Recent empirical literature on crises in emerging markets addresses a number of related phenomena, and many papers provide precise definitions that allow analysts to separate these phenomena into different events with different causes and consequences. Currency crises refer to the demise of an unsustainable peg. Large depreciations refer to significant realignments in nominal and real exchange rates. Speculative attacks are measured as combinations of exchange rate, interest rate...