Endogenous contagion – a panel data analysis

Abstract

This paper proposes a panel data model to analyze contagion in a multivariate framework. The model distinguishes between vulnerability and contagion, and provides a time series of contagion. The most important feature of the model is the endogenous determination of contagion without an a priori and potentially arbitrary specification of the crisis period. In addition, the model can distinguish between positive and negative contagion, and no assumption needs to be made about the source of the crisis. Eleven stock markets from the Asian region are analyzed during the Asian financial crisis, and contagion is found to be significant in four broad periods. These episodes are split equally between positive and negative movements. Anecdotal evidence is matched to the significant incidences of contagion, and it is found that events surrounding Hong Kong equity markets are key drivers of contagion.