Preventing crises caused by a large depreciation of exchange rates is one of the top agenda items for policy makers. This dissertation investigates the occurrence and prevention of crises from several perspectives: Chapter 1 analyzes currency crises from the firms' perspective; Chapter 2 studies a twin banking and currency crisis from the banks' and the monetary authority's perspectives; and Chapter 3 empirically researches the effects of different shocks and the mechanism through which shocks can be propagated.
Chapter 1 analyzes the role of firms during currency crises. The literature assumes that currency depreciation has only negative effects on the economy. I also incorporate positive effects by introducing exports into the third generation model developed by Aghion, Bacchetta and Banerjee (2001) (the ABB model). I derive an intuitive formula that describes each country's structural vulnerability and graphically explain how this vulnerability plays an important role.
The ABB model has assumed that only firms take the exchange rate risk. However, a currency crisis tends to occur concurrent with a banking crisis. In Chapter 2, I develop the first theoretical model in which banks take the exchange rate risk in the framework of the ABB model, and I show how a twin banking and currency crisis occurs. I also analyze several monetary policy instruments to prevent a twin crisis and find that not only the interest rate defense, which has been argued as a primary policy response in the literature, but also the reserve requirement policy can be effective for avoiding the crisis.
Based on the theoretical model developed in this dissertation, Chapter 3 empirically analyzes the effects of interest rate policy, real and financial shocks and structural vulnerability, which propagates the effects of shocks on the economy, on exchange rates. Using panel data on 51 emerging countries from 1980 to 2011, I employ instrumental variable methods with the generalized method of moments and the two-stage least squares estimators to control the endogeneity of monetary policy response. I provide the first empirical evidence that a productivity shock is quantitatively important for currency crises and the effects of this shock depend on structural vulnerability.