The opening of developing economies to international finance in the last three decades has led in a number of cases to severe balance-of-payments crises with large real costs. The Southern Cone crises at the beginning of the nineteen eighties, the Mexican crisis of 1994, the Asian crises of 1997, and the Argentine crisis of 2001, to mention but a few of them, all took place after the capital account had been liberalized. The literature dedicated to the empirical analysis of these events (Kaminsky & Reinhart 1999, Tornell & Westermann 2002, Calvo, Izquierdo & Mej?a 2004, among others) has identified a consistent set of stylized facts: the balance-of-payments crises go together with a real depreciation, a sharp drop of investment and a current-account reversal. Financial factors play a crucial role, and a lot of these currency crises were coupled with banking crises.

Some authors have also pointed to the role played by sectoral factors in these crisis episodes. Tornell & Westermann (2002) show that the relative size of the non-tradable sector usually increases before twin crises in middle-income countries. Calvo et al. (2004) find that the probability of a sudden stop is higher in economies where the absorption of tradable goods is small compared to the pre-crisis current-account deficit, a proxy for the size of a possible sudden stop. The rationale behind these findings is that any shock resulting in a lower demand for non-tradable goods has to be accommodated by a real depreciation in the short run. When the demand for non-tradable goods stemming from the tradable sector is large compared to the size of the non-tradable sector, it acts as a stabilizing buffer, so that the real exchange rate needed to close the gap is not very depreciated.