Study On The Style Factors Of Barra CNE6 Model

As of December 31,2021,with the introduction of the top-level design related to the trial of registration-based system and the expansion of direct financing ratio,China’s A-share listed companies have exceeded 4,600,and the total market value of the stock market has exceeded CNY 90 trillion,becoming the second largest stock market in the world.Different from previous trend of simultaneous rise and fall in the past,the structural characteristics of the current market are prominent,and subjective investment is increasingly difficult,so quantitative investment has become a new choice for investors.In order to study the significant style factors in A stock market in recent years,this thesis takes China’s A share market as research object.The research direction is the style factors of Barra CNE6 risk model(2018),which mainly includes volatility factor,momentum factor,size factor,quality factor,liquidity factor,valuation factor,analyst factor,growth factor,dividend factor,etc.The research framework is an analysis of market data from January 1,2010 to December 31,2021 through the MSCI Barra China Equity Model(Barra CNE6).First,based on the pre-processing of style factors data,this thesis conducts single-factor and multi-factor Fama-Macbeth regression significance analysis on style factors of the whole market to find out the effective factors with significant explanatory for the attribution of Chinese stock market performance.Meanwhile,this thesis compares the significant differences of style factors under different market values by grouping market values.Second,the regression analysis of factors significance is carried out for gem and SCI-tech Innovation Board respectively,so as to compare the change of factors significance before and after registration system.At last,factors are re-screened through yield test,IC test,stratification test and Fama-Macbeth factor goodness of fit analysis.Five factors,including volatility factor,scale factor,quality factor,liquidity factor and valuation factor,are eventually selected.According to the historical backtest results of the five significant style factors screened out and all nine style factors constructed by the return moving mean model,it is found that there is no significant information loss.On this basis,this paper builds a stock selection model through equal-weight scoring method,and finally shows that the stock selection strategy achieves a cumulative return rate of 379.21%,an annualized return rate of 13.94%,and an excess return rate of 333.47% during the period from January 1,2010 to December 31,2021.