Copula-GARCH Analysis of Chinese Stock Market Dependence Structure

In recent five years, Chinese stock market experienced unprecedented prosperity and slump, which provides valuable data for research on market action in extreme situations. This paper analyzes the correlations between returns of live indices (industrial index, financials index, metals index, property index, Shenzhen Composite Index) and SSE Composite Index from 2006 to 2011. We adopt Copula method combined with GARCH-t process to construct a Copula-GARCH model and use this model to analyze static and time varying correlations. The static analysis shows that t-Copula functions tit the significant tail dependence best. The dynamic analysis shows that correlation parameters of each index have similar trends, but the levels of variations are different which indicates that the macro-environment exerts severer influence on financial and property sectors in last five years.