Dependence Patterns across Financial Markets: Methods and Evidence∗

Using the concept of a copula, this paper shows how to estimate association across financial markets, with a focus on the structure of dependence rather than the degree of dependence. A mixed copula model is constructed so that it can capture various patterns of dependence structures. An inferential apparatus for this approach is developed and the methodology is applied to estimate the dependence in several major financial markets. The empirical findings are shown to have some implications that seem important for a wide range of studies including risk management, portfolio choice, asset pricing, and tests for contagion in financial markets.