Contrarian On China Stock Market

This paper use accountant and revenue data in China stock market to exam the contrarian strategy.We choose BM《book value of equity to market value of equity》, E/P(earning to market value of equity),C/P(cash flow to market value of equity), GS(growth rate of sales),ROE(return on equity),(BM,GS),(E/P,GS),(C/P,GS),(C/P,ROE) and(BM,ROE) to form the value stocks and the glamour stocks.We find in most case the value stocks perform better than the glamour stocks,but this is not true for GS and(E/P,GS) indexes.For them,the glamour stocks have more return than the value stocks.In the long run,higher BM and C/P stocks are statistically distinct from the lower,then other indexes are not.Then we compare numerical value of the indexes,find the high indexes may go down and the low indexes may go up. Compared the return of GS portfolios with their characteristics,we believe this illustrate the expected error of investors that they believe the past high growth rate would continue.We also form the portfolios on two-dimension classification.In the long run,the return of value stocks is statistically different from the glamour,but their performances are not better than one-dimension classification.As a whole,BM is the best index to classification.Based on the extra return,we use CAPM,Fama-French three factor model and other models including turnover rate to explain the reason.We find R~2 is very low in CAPM,saying that this model have no explanation.Three factor model perform better, but R~2 is low too.βin all of the portfolios have no statistically significant predictive power,and in most portfoliosβare negative,showing that glamour stocks are risker than value stocks.Size factor and book-to-market(BM) factor are both significant,the coefficients of size are either positive or negative,and the coefficients of BM are positive.Considered the high turnover rate in China stock market,we add turnover factor to CAPM and three factor model,and they become two factor model and four factor model based on turnover rate.The two factor model have more predictive power than CAPM.Turnover factors have statistically significant predictive power,the coefficients of turnover factor,λ,in most portfolios,are positive, showing that high turnover rate lead to high return.Added size factor and BM factor, turnover factors have statistically significant predictive power either,butλs are negative,showing high turnover rate lead to lower return.Besides,the predictive power of two factor model and four factor model is better after adding turnover factor, but not too much.We can conclude that turnover factor is not the most important factor to explain the extra return.