Analysis Of The Volatility Of Financial Time Series

In this paper, we do some research on the volatility of the stock market in China, the logarithmic return series of the Shanghai Composite Index and Shenzhen Composite Index were selected for the sample sequence. Combined with the method of hypothesis testing and statistical inference analysis, autocorrelation, stationary and other statistical characteristics of the sample data is verified in the text, and found that the data are not normally distributed, but showed a peak thick tail nature. Specifically, in chapter5, we chose two groups of data, using seven different models for modeling, mining volatility con-ditional heteroskedasticity, and asymmetry and so on. The results showed that with the improvement of the model, the fitting results improved. For variance model, the GARCH model fitting effect is better than the ARCH model; TGARCH model and EGARCH model fitting effect is closer to, and are superior to the GARCH model; furthermore, the concept of fractional difference in the mean model, using the ARFIMA the model, based on re-establishment of variance model, the results for the ARMA model is superior to the mean model. Finally found the ARFIMA-TGARCH model is the best one, this is because the model fully extracts the long memory of the sequence and the leverage effect.