Economia 4.2 (2004) 204-211
[Access article in PDF]
[Return to Article]
José de Gregorio: This is a very interesting paper, an insightful contribution to the research on external adjustment. Most of the existing literature focuses on two issues, namely, current account reversals and currency crises. This paper looks at capital account reversals, or sudden stops according to Calvo's taxonomy, from a new angle More concretely, this paper analyzes the consequences of sudden stops on growth and examines policy and structural preconditions that may ameliorate the costs of sudden stops. In addition, the ensuing current account reversals that often, but not always, follow a sudden stop are evaluated in terms of export expansion versus import compression.
The paper is full of interesting and thought-provoking evidence. Instead of summarizing it, I discuss the issues and questions that are raised and left open to discussion.
Sudden Stop, Current Account Reversal, or Both?
In principle, the central issue is what triggers a reversal in the current account or in the capital account—in other words, whether a sudden stop of capital inflows leads to a current account adjustment or the other way round. The authors focus on sudden stops that may induce a current account reversal. The paper starts by defining a sudden stop as an episode in which there is a significant decline in the capital account, defined as a contraction that exceeds one standard deviation with respect to its mean and that amounts to at least 5 percent of GDP. A sample of about 3,600 yearly observations yields 313 episodes of sudden stops. In the full sample of sudden stops, only forty-eight (15 percent) did not require current account adjustment.
It is not clear to me what the authors mean by whether the sudden stop did or did not require current account adjustment. In my view, they are generally part of the same phenomenon, but the authors are implicitly [End Page 204] assuming that causality runs from sudden stop to current account contraction. I am not convinced that the definition used in the paper is sufficient to argue that the cause of the adjustment derives from the sudden stop. I would agree that the paper cautiously leaves open the conclusion that current account reversals could be caused by sudden stops, but the focus is ultimately on sudden stops as the key cause of adjustment.
In a flexible exchange rate regime, changes in the capital account must be offset by changes in the current account, and vice versa. Inferring causality sets up an identification problem. Although one can argue that there are additional mechanisms through which it is possible to observe differences in the timing of the evolution of the current account and the capital account, this is not enough. The first, and most obvious, way to unlink the timing of the current and the capital accounts is a change in reserves. This allows the adjustment to the sudden stop to smooth the adjustment of the current account. The second would be to consider only private inflows, with the current account smoothed through official flows such as support from the International Monetary Fund (IMF); this is not the definition considered in the paper, however. Finally, the unlinking could be done through transfers, which could be relevant in small, poor countries. I therefore do not think that a capital account reversal happening at the same time or before a current account adjustment is definite evidence that this is a sudden stop.
An interesting case is the 15 percent of the sample that has a sudden stop without a contemporaneous current account adjustment. I suspect that in many instances this may be a problem of timing. For example, there are two cases in Latin America in which a sharp reversal in the current account came a year later. Argentina experienced a sharp sudden stop in 2001, but the huge 11 percent change in the current account deficit came in 2002. The domestic adjustment is not properly observed when one looks at simultaneous yearly data. Something similar happened in Ecuador. A sudden stop occurred in 2000, but the change...