Analysis of China Stock Market: Volatility and Influencing Factors

The volatility of stock market return is the main technique measurement in the risk management. This paper chooses GARCH class models to estimate in-sample period and to forecast out of sample period. The return of daily data was collected from Shanghai Stock Exchange, during the time period from January 1st 1997 to April 30th 2007. Three error measurement methods: ME, MAE and RMSE were used to evaluate the forecasting ability of GARCH class models. The empirical result indicates that EGARCH-M model is the best one for estimation of in-sample period. Asymmetric model or simple GARCH model is explained better for out of sample forecasting.