Modelling Currency Crises in Emerging Markets: A Dynamic Probit Model with Unobserved Heterogeneity and Autocorrelated Errors*
用动态Probit模型分析新兴市场货币危机的原因，发现实际、货币、债务和全球变量以及过去银行危机和贬值经历都会影响危机概率，且不同汇率制度下影响因素不同。
Abstract The paper investigates the causes of currency crises in emerging markets. We estimate the probability of a currency crisis by applying maximum smoothly simulated likelihood to a dynamic LDV model. This approach allows us to take explicit account of the existence of intertemporal links between crises. The results show that currency crises are influenced by real, monetary, debt and global variables. Past banking crises are significant determinants of the probability of currency crises. Moreover, countries that sharply devalued in the past are less prone to experience another currency crisis. We find evidence of unobserved heterogeneity, which may reflect differences in the countries’ institutional/historical background. Finally, the determinants of currency crises differ by type of exchange rate regime.