A Predictability Research Of China's Stock Market Returns

Whether stock market yield can be effectively predicted is one of the hot issues in current financial research.Although economic financiers have made good progress in explaining economic phenomena,predicting the future is still a big challenge for scholars.At present,economists generally believe that once the market is fully effective,the stock market's rate of return is subject to a completely random walk,which cannot be predicted by any factor.Following this line of thinking,the predictability of stock market yields has returned to the study of the nature of stock price random walks.However,current research shows that some macro and industry policy factors affect the stock market price all the time.At the same time,investors' expectations and the company's ability to manage and manage are also reflected in the stock price.Therefore,some components of stock price fluctuations,such as cycles,trends,etc.,are predicted to exist,while others are random walks and cannot be predicted.This provides us with a new economic forecasting model that provides the possibility to predict stock returns within a certain margin of error.This paper proposes a new predictive test method,the IVX test for structural mutation correction,to test the predictability of China's stock market returns.The simulation results confirm that the IVX test of structural mutation modification can effectively correct the existence of IVX test in the case of structural mutation in the dependent variable,whether it is single breakpoint or multiple breakpoints,whether it is univariate regression or multivariate regression.The IVX test for structural mutation modification not only inherits the IVX test without considering the persistent prior information of the predictor,and is robust to any predictor variable belonging to the unit root process,the near unit root process,the near stationary process,or the stationary process.Moreover,when the dependent variable has a structural mutation process,the deficiencies of the traditional IVX method can be effectively corrected,and the result is more reliable.In addition,the Wald statistic of the IVX test due to structural mutation correction obeys a standard chi-square distribution,which is simple to calculate and can be extended to the multivariate model,which greatly improves the efficiency of the test.This paper selects 10 economic variables such as dividend payout ratio,dividend price ratio,dividend yield,earnings price ratio,book market value ratio,stock variance,net equity increase,turnover rate,inflation and currency issue growth rate as predictors.It is predicted that the A-shares of China's Shanghai and Shenzhen stocks will be weighted average by the total market capitalization.Moreover,considering the split share structure reform as a major event in the history of China's stocks,we will use the post-equity reform period as a sub-sample for comparative study.In order to make the conclusion of this paper more accurate,this paper divides all Chinese A shares into 13 industries according to the CSRC industry classification to test the industry's rate of return predictability,further improving our conclusions.In addition,the results of this study are compared with the OLS test results,the traditional IVX test results and some related scholars' research conclusions.This paper also analyzes the predictability of stock market returns with different maturity levels by constructing different test methods,which enriches the long-term and short-term research conclusions.The main conclusions of this paper are: stock variance and inflation have strong predictive ability in the short term;can significantly predict the predictive variables of different industries,but in general,the test results of all A-share portfolio yields are not much different.There are only differences in individual variables in individual industries;as the maturity level increases,the predictive power of predictors is significantly improved as a whole,and most predictive variables show an increase in the maturity level at all maturity levels.The characteristics of significant predictive ability are elevated or at any level;the predictive variables in the multivariate test are significantly less pronounced than the univariate test;the long-term predictive ability of some predictors is significantly weakened during the share-trading reform period,overall In general,the overall predictability of stock market returns after the split share reform is weaker than the full sample period.