VaR is a new risk management method. It can describe the market risk concisely and clearly. Besides it can be applied to many fields. It can use the simple language to describe risks by pointing out the maximum loss under certain confidence level. AR model is commonly used in stationary time series, can be well fitted with self characteristics of time series regression. GARCH model can describe the heteroscedasticity of financial time series very well. It is widely used in the risk value of the stock market. We use logarithmic yield to describe the risk value. Then we selected the daily closing price of Shanghai Composite Index, Shenzhen Component Index, Growth Enterprise Index to describe the stock market situation. Finally we use AR-GARCH model to study the risk value of domestic stock market. |