Shanghai Stock Exchange and Shenzhen Stock Exchange, after twentyyears of development, although great progress has been made, but the twomarkets still have some problems, such as market structure is not balanced, theoperation is not standardized. Especially in1997and2007financial crisis,China’s stock market has shown dramatic fluctuations trend. We study thecorrelation analysis and prediction of the volatility on the Shanghai andShenzhen stock markets in this paper. Firstly we use EMD to divide theShanghai and Shenzhen stock price index into several components at differentfrequencies, then according to the different components of the cycle, mean,variance, etc. to build a combination of features, extract three time seriesincluding market volatility, major events and trends affecting. Secondly we useCo-integration test, Granger causality test and other methods to study thecorrelation of three time series of two markets in Shanghai and Shenzhenrespectively, research results of the model show that Shanghai and Shenzhenhigh linkage exists. Finally we construct three different SVM prediction modelsto get the predictive value of each sequence, and get the final prediction bycombining three predictive values. The results show that the accuracy of thismethod is higher than single SVM prediction model. |