| This paper establishes a Fama-French three-factor model with time-varying parameters.The results show that in Fama-French three-factor model,this method can be used to estimate and test the time-varying of factor loads.The parameters of each component of Shanghai 50 Index and Chuang 300 Index in China are significant in the single test,which indicates that the parameters of individual stocks change with time.It shows that it is necessary to consider the time-varying of the relevant factor model of China's stock market and make contributions to the study of the factor model.Taking SSE 50 Index and GEM 300 Index as the research objects,this paper investigates whether the parameters of Chinese stock in this factor model change with time.The results show that the estimated parameters of each component stock of SSE 50 Index and GEM 300 Index change with time in individual stock test and portfolio joint test,respectively.On this basis,this paper further chooses stocks according to the non-time-varying parameter factor model and the time-varying parameter factor model,constructs portfolio one and portfolio two,and constructs portfolio three according to different forecasting models.Besides,TM-FF3 model and HM-FF3 model are used to evaluate the profitability and timing ability of the three options.It is found that the profit and timing ability of model 2 are far more than those of model 1 and portfolio 3.Therefore,based on the conclusion of this study,when adopting the multi-factor stock selection model,fund companies should consider the impact of time-varying parameters.Through case study,this paper finds that the stock selection model with time-varying parameters can not only obtain higher excess return,but also make up for the shortcomings of the non-time-varying parameter stock selection model which needs to consider the estimated period.It can simplify the operability of the stock selection model and put forward relevant improvement suggestions for quantitative stock selection. |
PDF Full Text Request