Research On The Liquidity Risk Premium Of The Stock Market

China's stock market is developing well.Since the first public issuance of sound stock in November 8,1984,after many years of development,China's stock market has made a great chievements.Liquidity can have a huge impact on stock returns,the study of liquidity risk premium is of great significance.In this paper,it is studied that the liquidity risk premium in stock market,by using five model averaging methods.They are the weight selection method based on AIC criterion,the weight selection method based on BIC criterion,the weight selection method based on Mallows criterion,the weight selection method based on Jackknife criterion and OPT weight selection method.This paper is studied from two aspects.The first aspect,first,qualified samples and samples are selected.The samples are representative and there is no correlation between variables.Then,from the point of view of optimal rate,the forecasting effect of various models on the excess return of stocks is analyzed,from the point of view of the significance of the model average estimator,whether there is liquidity risk premium in stock market is tested.Next,in order to judge the advantages and disadvantages of the model average method,the model average method and the regression method are compared.Finally,These are compared that the several model averaging methods to find the optimal model averaging method and it is studyied the stability of each model average method.The second aspects,appropriate sample and variable are selected.Dividing stocks into three groups according to their liquidity,These are studied that whether there are significant differences in liquidity risk premium between different groups of stocks and the prediction accuracy of each group,by using the model average method.These can be proved: from the perspective of optimal rate,when liquidity is added to various models,the forecasting results of excess returns of stocks are more accurate.From the perspective of the significance of the model average estimator,the liquidity risk premium exists in the stock market.From the perspective of the optimal rate,the model average method is better than the regression method.The Friedman test results show that the five model average methods have significant differences.From the perspective of the mean absolute error or the optimal rate,the OPT method is the best model averaging methods.From the perspective of absolute error variance,the stability of the each model average method is very good,when the variance is almost large,the prediction of OPT is the most accurate.In addition,there are significant differences in the liquidity risk premium of each group stocks grouped by liquidity size,the low liquidity group has a greater impact on the liquidity risk premium.The accuracy of the prediction between the high liquidity group,the medium liquidity group and the low liquidity group is almost large and high.Friedman test results also show that there is no significant difference in the prediction accuracy of each group stocks.