Leading Indicators of Currency Crises in Emerging Economies

This study identifies common features of currency crises in 15 emerging countries over the period 1980-1998. By analyzing such features, we build an early-warning system aimed at predicting looming crises in probabilistic terms. This work departs from the existing literature in several ways. First, we use quarterly data, in contrast to other studies, which are based on monthly or annual data. This allows us to characterize crises more accurately and also to analyze the behavior of leading indicators as actual crises approach. Second, the overvaluation of currencies is assessed by using real effective exchange rates, instead of the usual bilateral rates. In addition, capital control dummies are included in the set of explanatory variables and contagion indicators are constructed. Finally, we use the Fisher linear discriminant analysis technique. The model yields a relatively good - and unbiased - ratio of correct predictions: four out of five crises are predicted correctly and only one out of five non-crises is predicted as a crisis. These results compare favorably to those of other models. For early warning systems, there is a fundamental trade-off based on the Bayes' formula in a context of rare events: to a certain extent, one has to choose between a high ratio of good classifications of crises and a low ratio of false alarms. Furthermore, using Bayes' formula allows us to calculate the posterior probability that a given emerging economy will be in a period of currency crisis within a one-year horizon