What do twins share? A joint probit estimation of banking and currency crises

We study the determinants of twin crises and investigate the direction of causality between banking and currency crises in emerging markets. We model banking and currency crises as dynamic events, correlated over time, and jointly estimate their probability using panel data simulation techniques. We show that banking and currency crises are closely intertwined and are driven by common fundamentals. Banking crises exhibit strong state dependence, suggesting that countries that have experienced a banking crisis in the past are more prone to experience another crisis. Finally, we find evidence of unobserved heterogeneity and autocorrelation in the error term structure.