Learning dynamics and endogenous currency crises

This paper introduces adaptive learning into the third-generation currency crisis model of Aghion, Bacchetta, and Banerjee (2001, Currency crises and monetary policy in an economy with credit constraints, European Economic Review 45, 11211150). Adaptive learning might reflect, for example, uncertainty about the economy's exposure to adverse balance sheet effects. Even when equilibrium is unique, we show that the learning algorithm's escape dynamics can produce the same kind of Markov-switching exchange rate behavior that is typically attributed to sunspots or herds. An advantage of our learning model is that currency crises become endogenous, in the sense that their stochastic properties can be related to assumptions about learning and other structural features of the economy.