Research On Multi-factor Stock Selection Strategy Based On CSI 300 Stocks

The stock market is a "high-risk,high-yield" field.In addition,the complex and volatile nature of China’s stock market makes it difficult to obtain stable returns only by using traditional investments.How to effectively avoid risks and select stocks with excess returns from a wide range of stocks has always been the concern of investors.In this context,designing trading strategies based on the statistical analysis results of mathematical models can effectively reduce investment risks caused by human common deviation and irrational subjective factors in the investment decision-making behavior.This paper uses the ranking and scoring method to screen effective factors,establishes an analytic hierarchy process weighting model and designs a multi-factor quantitative strategy based on the CSI 300 constituent stocks.This paper uses the historical data of the Shanghai and Shenzhen 300 constituent stocks from January 1,2010 to December 31,2016,and uses the IC mean,IR value,monotonicity of permutation,and excess returns as the measure of the effectiveness of the factors.A total of 24 candidate factors in 7 categories,such as valuation,growth,profitability,asset structure,technology,scale,and consensus expectations,were tested for validity.Finally,9 valid factors were screened out,which are P / E ratio,turnover ratio,net profit growth rate,P / B ratio,operating cash flow growth rate,trading volume of the day,consistent expected return on net assets,operating income growth rate and consistent expected net profit.On this basis,this paper constructs a multi factor weighting model based on AHP,and selects the historical data of Shanghai and Shenzhen 300 shares from January 1,2017 to December 31,2019 to compare the operation effect of equal weighting and AHP.The results show that AHP is better than equal weight method.Finally,this paper uses the AHP weighting model to design a multi-factor quantification strategy based on the CSI 300 constituent stocks.During the strategy design process,based on the principle of risk dispersion,the functional relationship between individual stock positions and portfolio risk is calculated,and the relevant parameters of the strategy are optimized,including stock position adjustment ranking,stock industry distribution restriction,stop loss and stop gain.In addition,considering the price impact of strategic position adjustment,based on the statistics of strategic historical transaction data,this paper studies the corresponding relationship between strategic capital scale and individual stock position.Under the condition that the trading amount of stock position adjustment does not exceed 1.5% of the trading amount of the day,the functional relationship between capital scale and individual stock position is established.The back test results show that the factor stock selection model and quantitative strategy in this paper can achieve stable alpha,which has certain reference value for the majority of investors and is conducive to cultivate the value investment concept and improve the level of investment management.