The Adjustment Of Fama-French Three Factor Model And The Study Of Its Adaptability On SSE Stocks

In the recent three decades,China's stock market has made considerable progress in terms of regulations as the economic power of China strengthened steadily,becoming one of the most important component of the financial system on earth.However,since the stock market in China started to develop about one hundred years later than those in the developed countries,there are still many problems in China's stock market,such as the lack of proper proportion of institutional investors,serious speculation,and immature value-investment idea.Besides,local investors often fail to consider the market emotions,transaction behaviors and Chinese characteristics when analyzing the investment of stocks.Therefore,whether the widely used Fama-French model could be appropriate for China's stock market and whether new factors should be added to this model have become one important topic for many researchers.Fama-French model is a classic model built by Fama and French in 1992 to study the difference between the excess return of difference stocks based on transaction data and phenomena in America,while when Chinese researchers try to apply it Into China's stock market,the study outcome does not turn out to be consistent.In recent days,the idea of "Internet +" has attracted an increasing amount of attention.The huge amount of data provided by the Internet has contributed much to studies in all fields and helped to create more value for the society.As for the China's stock market,individual investors take on a large proportion,and the Internet can provide the mass data for analyzing the effect of investors' transaction behaviors and expectations on the return of stocks.As a result,the Big Data 100 come in to being.As the first index based on big data,it provides better portfolios that can yield more,therefore more investors are starting to choose stocks based on the Big Data 100,which attaches much significance to the study of the relationship between Big Data 100 and the excess rate of return.To begin with,this dissertation tests whether there are BM effect and scale effect in China's stock market,which turns out to be true that there are similar effects in China's stock market to those in America's.Then we add the Big Data 100 to improve the Fama-French model.Through the Pearson test,it turns out that there is no linear relationship between the influencing factors,and then we use the data between March 2010-February 2016 as the sample and the listed companies of SSE A shares are divided into 16 different segments according to their main business.Next,the Fama-Macbeth regression method was used to test the influence of each single factor.It was found that the three factors--scale(SMB),book market value ratio(HML)and big data 100(BD)--can be closely associated with the return.Therefore,a new four-factor model is built based on these three factors and the market factor.By using the Fama-Macbeth method again,we found that the new model has a strong ability to explain most of the stock segments.Among the 16 segments,the intercept of three segments are not 0,the market factor coefficients of three segments are not significant,the scale factor coefficients of four segments are not significant,the book market value ratio coefficients of seven segments are not significant,the Big Data 100 factor coefficients of five segments are not significant.It indicates that there are differences among the risk factors of different segments in China's stock market.For example,financial industry is affected by external economic factors such as international economic trends;Water,environment,public facilities management,health,and social work industries are more likely to be affected by gambling-style speculation;and information transmission,software and IT service industries would be often influenced by national policies.But overall,the four factors turn out to be good for most of the portfolios,and the Big Data 100 factor can explain the excess returns in both the single factor test and the multi-factor test.It shows that the Big Data 100 can affect the return of stocks,which is of great significance for investors to construct appropriate portfolios,to predict the trends of stock prices,and to cultivate good trading habits and investment ideas.