Dependence structure analysis between stock index futures and spot markets in the case of the “Golden week” effect

The dependence structure analysis of a financial time series of returns is significant when applied to contemporary financial risk management. Copula function is a flexible and effective tool to be used on modeling the financial model and risk management. This paper aims to set up the dependence structure between CSI300 index and futures by Copula-ARMA-GARCH models and find out which copula can provide a better fit to the empirical data in case of the “Golden Week” effect. Moreover, we analyze the degree of linear dependence, rank correlation and tail dependence between CSI300 index and futures. The empirical results indicate that there is high degree of dependence between CSI300 index and futures. The asymmetric tail dependence description is better, and tail dependence is significantly high. It also demonstrates that the “Golden week” effect could decline the rank correlation and slightly weaken the dependence between CSI300 index and futures.