The Study Based On The GARCH Model Of The Csi 300 Index Volatility

China launched the csi 300 index On April 8, 2005, it can more fully reflect our country securities market one of the main features of the stock price changes, so it is the main reference object and evaluation standard for our country stock market investors and is also China’s financial institutions reference for policy makers to make decisions. As the subject matter of stock index futures, the csi 300 index volatility research is very necessary.First of all, this thesis introduces the relevant theories of the csi 300 index, shows the GARCH models such as the ARCH model, GARCH model and asymmetric effect model- TGARCH model and EGARCH model, gives the expression and parameters of these models, the characteristics of each model, and uses the method of maximum likelihood estimation to estimate the GARCH(p, q) model parametersSecondly, this thesis collects the day’s closing price of Shanghai and Shenzhen 300 index from January 4, 2011 to February 20, 2014, and the original data are logarithmic yield differential treatment, and then compared the logarithm yield sequence correlation, stationarity and the ARCH effect of the inspection. Test results find that: the logarithm yield sequence has rush fat tails and volatility clustering and asymmetry.Finally, this thesis respectively with residual obeyed normal distribution, t distribution and GED distribution of GARCH(1, 1) model, the EGARCH(1, 1) model and TGARCH(1, 1) model of the csi 300 index logarithm yield an empirical analysis is simulated, the results found that compared with the normal distribution, the distribution of the residual subjected to the t distribution and GED GARCH class model fitting effect is better; On the basis of the same kind of distribution, EGARCH(1, 1) model and TGARCH(1, 1) model’s fitting effect is better than the GARCH(1, 1) model.