Research On Factor Model Of Chinese Stock Market Based On Fama-French Model

The improvement of the system,the enrichment of the system and the opening of the market make the Chinese stock market enter the stage of rapid development.Due to the important role of the stock market,the research on the influence factors of stock pricing and risk premium has been a hot topic in the financial field.In an article published in 1996,Fama and French proposed that the changes in the returns of American stocks could be well explained by the three-factor model composed of market factors,size factors and book-to-market ratio factors.In 2015,based on the analysis of stock data in the financial market and the research results of others,the company’s fundamentals were added to the three-factor model,and the fama-french five-factor model was constructed by considering the factors of profitability and investment style.Fama and French used the five-factor model to carry out empirical tests in the stock markets of North America,Europe and asia-pacific region,and the model had good explanatory ability and applicability.However,there are disputes over the applicability of the three factors,namely market factor,size factor and book-to-market ratio factor,to the Chinese stock market,and whether the new profitability factor and investment style factor of the five-factor model are applicable to the Chinese stock market compared with the three-factor model.The momentum factor is the most widely concerned.With the development of big data on the Internet and the gradual popularization of the application of behavioral finance,the method of factor construction can also be extended to introduce emotional indicators for empirical research on Chinese stock market.In this thesis 240 months ’data of the comprehensive a-share market from May1999 to April 2019 were selected as research samples.Firstly,it empirically analyzed whether the five factors were effective.It divided the five factors into groups by means of size and book-to-market ratio,profitability and investment crossover,constructed the factors by means of 32(9),and tested the investment portfolio by means of 55(9),(9)(9)222 and 22(9).And then check whether China a-share data has significant effect of momentum(reversal),respectively,in 3,4,5,6,9,12,18 months as the observation period,such as weight,value of weighted two winners(losers)that constitute the weight ratio of site selection and 30%,10%,two points method to build momentum factor,with 55(9)methods,a total of 75 investment portfolio is weighted yields as explained variable regression,The momentum factor constructed by selecting significant portfolio proportion up to78.67% in the 4-month observation period-market capitalization weighted,30%quantile was added into the factor model,and it was concluded that the momentum factor improved the explanatory ability of the model through F statistics and goodness of fit test.The stock recommendation report opinions used in the construction of sentiment factors were graded and ranked on a monthly basis to construct the sentiment factors.Finally,the momentum factor and emotion factor were analyzed to improve the explanatory ability of the original five-factor model.The proportion of the non-significant constant term increased from 17.33% of the original five-factor model to 93.33%,indicating that the risk exposure was covered.On the whole,adding new momentum and emotion factors can greatly improve the explanatory ability of excess returns.The modified goodness of fit increases from0.35 to 0.5 to 0.45 to 0.6,with obvious improvement effect.