Forecast And Application Of Shanghai And Shenzhen 300 Index Returns

This paper makes an empirical study on the returns of the CSI 300 index from the perspectives of international capital market volatility,domestic macro-economy,corporate financial situation and investor sentiment.The results can help policy makers accurately grasp the impact of foreign macro and investor sentiment volatility on the stock market.In this paper,the CSI 300 index is selected as the research sample and 24 forecast variables are selected at the same time.After the data is transformed into stable,the stepwise regression method is used to screen variables,and three significant factors,namely,the volatility index(EMV),the inflation rate(CPI)and the investor sentiment composite index change(sent),are obtained.At the same time,the monthly data of the stock time window(2006-2018)is divided into three parts In the training period and the verification period,the rolling prediction is carried out outside the sample,and the direction accuracy,root mean square error(RMSE),mean absolute error(MAE)are compared with the historical moving average prediction method;secondly,the logistic model,the mixing model and the neural network model are established by three factors to compare;finally,the Shanghai 50 index,the Shanghai Composite Index,the Shenzhen Composite Index and the China composite index are respectively compared For the stock index,add Shanghai and Shenzhen 300 volatility and change data samples to conduct robustness test in June 2019.Through above empirical research,we drawn the following conclusions: the current U.S.stock market volatility index based on News newspaper mining has a negative correlation with the next CSI 300 index yield;the current inflation rate has a negative correlation with the next CSI 300 index yield,and the rising inflation in the current period will cause investors to worry about raising interest rates or tightening policies,which will lead to the decline of the next yield The change of investor sentiment has a positive predictive effect on the return of CSI 300 index,that is,the higher the investor sentiment,the higher the next period's return;secondly,the accuracy of CSI 300's return is predicted to be higher than 50% by using the three significant factors of EMV,CPI and sent,The mean square error and the mean absolute error are smaller than the historical average method.Thirdly,in practical application,according to the simulation results of transaction strategy,the Sharpe ratio of the model is better than the historical moving average method.Fourth,the direction accuracy of the prediction using logistic is decreased,while the mixing model does not significantly improve the goodness of fit and prediction effect,and the nonlinear BP neural network does not significantly improve the prediction effect.Fifthly,Shanghai 50,Shanghai Composite Index,Shenzhen Composite Index and China stock index have passed the robustness test;the verification window has been extended to June 2019 to carry out the prediction,which has also passed the robustness test;adding Shanghai and Shenzhen 300 volatility to carry out the rolling prediction,it is found that the goodness of fit inside and outside the sample has improved,but the accuracy of the direction has declined,but the prediction effect is better than the historical average method;trading strategy model The results show that the three factors are better in practical application,and also pass the robustness test.