Collapsing controlled exchange rates: how well can we predict and explain them?

Theoretical models that link the probability of a collapse of a fixed exchange rate to fundamental economic values such as domestic credit, the trade balance, international reserves and relative prices are now standard, but empirical testing of these models has nevertheless been very limited. This paper applies such a model to the Mexican experience between 1988-1995 and to several European currencies that participated in the European Exchange Rate Mechanism (ERM) in the ‘crisis' year of 1992.

The results for the Mexican case are encouraging. Mexico experienced several periods of speculative pressure on the peso between 1988 and 1995, and the model predicts collapse probabilities larger than zero for most of them. Just before the collapse of the exchange rate in December 1994, the model indeed indicates a probability of collapse equal to one. The results for the European currencies are mixed. The speculative pressure against the Spanish peseta is predicted well. The problems with the Italian lira and the French franc were predicted as well, although the timing of the events was not perfectly indicated. For the British pound, the Belgian franc, the Irish pound and the Danish krona the model results are unconvincing.