The Empirical Study of Dynamic VaR on the Index of Shanghai Security Market Based GARCH-EVT

This paper propose a method for estimating VaR and related risk measures describing the tail of the conditional distribution of a heteroscedastic financial return series. Our approach combines GARCH models to estimate the current volatility and extreme value theory for estimating the tail of the innovation distribution of the GARCH model. We use our method to estimate VaR and conditional expected shortfalls. Using backtesting of Shanghai stock market we find that there is serious ARCH effect on return rate of the stock market and GARCH-EVT model gives better estimates than GARCH-Normal and GARCH-t model.