The VAR Study Of CSI 300 Index With Two Types Heteroscedasticity Model

Since the reform and opening up,the securities market of China has rapidly developed.Daily turnover in Shanghai and Shenzhen from 2004 of more than 1000 million yuan to 6,000 billion yuan in 2014.In March 24,2015 Shanghai and Shenzhen daily turnover is over 1.4 trillion yuan. With the continuous improvement of the stock market,a variety of derivatives of financial securities have been launched.the CSI 300 stock index futures has been launched in April 2010 and bond futures in September 2013 officially listed on the China Financial Futures trading.In February 2015,the SSE 50 ETF options has been launched.In this paper,risk measurement of CSI 300 stock index futures was analyzed with empirical analysis method.The two different models-GARCH variance model and the SV model have been used,GARCH model parameters were estimated using Eviews software,and based on Bayesian estimation Markov Chain Monte Carlo method SV model parameter was estimated,combined with Openbugs software.The yields of CSI 300 Index was fitted,and mean square of yields was obtained,and then with parameter method the VAR parameter value of CSI 300 Index was got. GARCH-N, GARCH-T, GARCH-GED,SV-N,SV-T,SV-MT,SV-MN seven heteroscedasticity models have been tested with the fundamental frequency detection method.The results show that the method using the VAR measure the risk of the CSI 300 Index,SV-MT model is the most accurate.