On the Dynamic Dependence between US and other Developed Stock Markets: An Extreme-value Time-varying Copula Approach

This paper examines the dynamic dependence structure between US and four developed stock markets, namely, Japan, United Kingdom, Germany and France during a recent period including the global financial crisis 2007-2009. The econometric approach is based on the extreme-value time-varying copula functions. Specifically, the marginal distributions are reproduced by an extreme-value based model, while the joint distribution is explored using the time-varying Normal and Symmarized Joe Clayton copulas. The empirical results show that the dynamic dependence between US and Japanese stock markets is symmetric, while that between US and European stock markets is asymmetric. In particular, this dynamic dependence increases during a crisis and seems to be related to the geographic position.
JEL Codes: G15; C32; C10.
Keywords: Dynamic Dependence; Stock Markets; Extreme Value Theory, Time-varying Copulas.