The Estimation Of Chinese Stock Market's VaR Using Switching Regime Models

In 2007,Chinese stock market has experienced a great development. On October 16, Shanghai composite Index reaches to the highest 6124.04 points. The annual increase is 98.5 percent. But in November,the market fell 1082.99 points,or as much as 18.19 percent,and becomes the largest one-month decline since 1995.The surge and crash phenomenon means that our stock market exists enormous risk,effectively preventing and controlling the risk is imminent. This paper uses quantified models to measure the stock market. It could provide the basis to the regulators and investors for analyzing and defusing the stock market risk.In our real life, financial time series often show that the conditional mean and conditional variance have different state. So in this paper, we consider switching regime models based on the discrete Markov process to measure the risk of our stock market. On the basis of the previous studies,we give a new model-Switching Regime ARCH-M model and the its parameters estimation method. We also do empirical analysis with return data of Shanghai Composite Index and Shenzhen Composite Index, and demonstrate the corresponding model's practical application in VaR estimation. We compute the VaR of three models in different confidence level and do return tests with VaR results. It proves that switching regime ARCH-M model has feasibility and validity in estimating Chinese stock market risk.