The Application Of VaR In Chinese Stock Market With Historical Simulation Method And An Improved Approach

VaR indicator is one of the most popular risk indicators in financial risk management field now.As an important computing method of VaR indicator,Historical Simulation Method can not respond to the change of volatility of stock market.This is why it fails in measuring stock market risk accurately.It is of significance and value to research and find ways to improve the defect of Historical Simulation Method.In the theoretical part,this paper firstly reviews the related literatures of Parameter Method,Historical Simulation Method and Monte Carlo Simulation Method for calculating VaR indicator.Secondly,the definition,calculation principle,calculation method and test method of VaR indicator are introduced.Finally,on the basis of previous studies,an improved Historical Simulation Method based on volatility adjustment coefficient is proposed.It combines the GARCH model with the Historical Simulation Method,so that the VaR indicator calculated by the it can be adjusted with the fluctuation of the market.In the empirical part,this paper chooses the closing price data of Shanghai Composite Index and Shenzhen Composite Index from December 31,2005 to January 17,2018,calculates 90%,95% and 99% VaR indicators with Historical Simulation Method,and tests the accuracy of this method in 255 trading days and 500 trading days respectively.The test results show that 99% VaR indicator calculated by it is accurate,while 90% and 95% VaR indicators overestimate stock market risk.Then,after checking the stationarity,autocorrelation and heteroscedasticity of the two indices return series,the GARCH model needed for the improved method is established,and the accuracy of 90%,95% and 99% VaR indicators calculated by the improved method is empirically analyzed.The results show that the 90% VaR indicator of Shanghai Composite Index are inaccurate in 500 trading days,and the remaining VaR indicators are accurate.The above empirical results show that the improved method improves the inherent defects of the Historical Simulation Method to a certain extent,and is more accurate in measuring the risk of China’s stock market.