| After more than.30 years of development,stock index futures become one of the most widely used financial derivatives with hedging risk,stabilizing market and other functions,but problems such as leverage margin system bring greater investment risk.As the main method of risk measurement,VaR is widely applied at home and abroad. And the research of volatility is the core of risk management.currently VaR method based on volatility model is a good measure of market risk.This paper focuses on CSI 300 stock index futures market and Hong Kong’s hang seng index futures market for a comparative study of VaR risk measurement. With disturbance assumption such as normal distribution,T distribution, volatilities of two markets were obtained via GARCH and SV models representing two types heteroskedasticity models,which were applied to estimate VaR.Then Kupiec failure frequency detection method was used to test validation of each model VaR sequence.The empirical results show that both market yields exist fat tail and leverage features.In the selecting futures market risk models,in addition to T distribution,VaR method based on GARCH models is consistent with two markets of risk characterization.But for SV model,Leverage SV model is the best model to character CIS 300 stock index futures market,and SV-MT model performs better for hang seng index futures market. |
