Research On Fund Performance Based On Model Averaging Method

Since entering the 21 st century,the developing status of Chinese Securities investment funds is favorable,The open-end fund Frequently appear in the public view and develops into a mainstay of the fund market.Fund performance evaluation system plays an important role in the development of the industry.Now the method of fund evaluation is diverse,when using the method of model selection,the basis of choosing is not clear enough.Now considering the method of model averaging applied to fund performance is expected to improve prediction effect.The fund performance evaluation and research based on the method of model averaging mainly from three aspects: firstly,collecting the factors that have a significant impact on fund performance,gathering as far as possible little correlation between factors,and easy to construct linear model;Secondly,introducing S-AIC,S-BIC,JMA and OPT four method of model averaging.We regard the monthly net growth rate of one fund as dependent variable,and other selecting five factors as the independent variable,including RMRF?SMB?HML?CR and Risk,then to build a number of alternative linear model,for each model in accordance with the standards of the four different weights to the empirical analysis.We select the open-ended funds which build before June,2010 as the research sample,and the evaluation time interval is from January 1,2011 to March 31,2016,we forecast the issue of the nine fund net value of the monthly growth rate which is from April 1,2016 to December 31,2016.The absolute error and the optimal rate can be as prediction accuracy to measure the effect of prediction.Results show that whether from the mean absolute error or the optimal rate,OPT,the best method of model averaging has the best prediction precision on fund performance,and it has the best prediction effect and the best stability.At last,we conclude that the five factors have some effect on the fund performance based on OPT.