The analysis based on BP and GARCH combination model in volatility of the securities market

In order to fully exploit the volatility inherent principle in the return of the securities market, describing asymmetry volatility and heteroscedasticity of the securities market by GARCH method, and combining nonlinear approximating and generalization capability of the neural network, respectively set up GARCH(1,1) and GARCH-BP models for empirical analysis return rate of Shanghai stock index. The results demonstrate that, two models have significant fitting capability, but the latter is more accurate.