The Research On Multi-Factor Stock Selection Strategy Based On Chinese Market Data

The concept of quantitative stock selection rose up early in abroad,although it started late in China,but in recent years it has developed and enriched in the Chinese market.Quantitative investment is a kind of investment method,in which we can search for efficient investment opportunities in the financial market and use computers or other devices to realize these approaches according to various trading strategies.It can capture valuable information in the market more quickly,and can overcome people's irrational and subjective emotions to a great extent,so as to improve the effectiveness and accuracy of the strategy.As one of the most stable and effective models of quantitative investment strategy,multi-factor stock selection model is frequently used in domestic and global investment community.However,due to the wide variety of factors in the stock market,investors often can not choose appropriate effective factors,and they have many problems in dealing with the combination of factors.At the same time,the existing domestic multi-factor quantitative research on ashare market usually focuses on the validation of Fama-French factor model in Chinese capital market,but fails to provide investors with a simple and effective way of stock selection.Therefore,this paper expects to find out the key factors that affect the return rate of the stock market in China,then construct a multi-factor stock selection model based on value investment strategy and verify its performance in the market,so we can find a rational stock selection method which is suitable for the investors.Based on the above realistic background and the research results of domestic and foreign scholars,this paper expounds the classical model and theoretical basis related to the multifactor stock selection strategy.According to the basic procedures of quantitative stock selection and the characteristics of Chinese market,the methods of selecting,testing,eliminating candidate factors and constructing investment portfolio are analyzed and adjusted in the paper,and the performance evaluation criteria of the model are determined.In the empirical test,we choose 15 factors which are widely used among the four kinds of factors,and use the single factor regression and Fama-macbeth method to test their validity,then five effective factors,including PS,PCF,PB,TAR and HSL,are selected to construct models of different strategies.These models include basic multi-factor models,multi-factor models adjusted according to monetary cycle and industry rotation,and multiple linear regression models based on fixed effects.In the follow-up empirical test and analysis,this paper compares the return rate and risk indexes of investment portfolios constructed according to various models.The empirical results show that the investment effect of the multi-factor stock selection model is better than the average level of the market,which validates the strategy is effective in Chinese a-share market,and it also proves that the domestic capital market has significant industry cycle and rotation effect in recent years.The validity of the multi-factor model can be improved by adjusting the positions of stocks in different sectors according to the industry cycle.In general,multi-factor stock selection strategy can not only achieve significant excess return,but also reduce investment risk,which provides effective support for investment decision in the capital market,and it also provides the research data and the stock selection logic for investors at different levels.