An Empirical Research on the Chinese stock market based on VaR

This paper has introduced parameter VaR and non-parameter VaR methods, utilized the parameter VaR method to carry on a positive research to the security market of our country, and calculated VaR value of the composite index of Shenzhen Stock Exchange, so as to come to examine the valid degree of this method in risk management of market of listed companies in our country. The sample spacing involves the whole 242 bargain days in the year of 2005 and the confidence level is 95%. Through calculating, skewness coefficient of earning ratio S is 0.731 (skewness coefficient of normal distribution is 0) and kurtosis coefficient K is 5.96 (kurtosis coefficient of normal distribution is 3). The result shows: from the contrast of actual value and lower limit of predicted VaR value, actual index value for 10 days breaks below the prediction lower limit. Because the confidence level that we choose is 95%, namely real Shenzhen Stock Exchange index which break below the prediction is less than 12.1 days, and have only 10 days actually. It is obvious that the result of fitting of VaR method to risk of Shenzhen Stock Exchange composite index market is better and has an available value virtually.