Working Paper no. 171
This paper reviews the theory of balance-of-payments crises, and its implications for identifying potential leading indicators of crises. It discusses and evaluates three different empirical approaches to balance-of-payments crises: the signalling, discrete-choice, and structural approaches. Despite claims of success in predicting currency crises, we note some serious theoretical and empirical qualifications which throw these claims into question. Nevertheless, we conclude that a range of indicators supported by theory may still be useful for policy-makers interested in preventing financial instability.