A Research On Stock Selection Strategy Based On Fundamental And Technical Factors

By the end of 2019,the total size of global quantitative investment funds had exceeded US$3 trillion,showing explosive growth both in terms of volume and market value.With the listing and trading of the CSI 300,CSI 50 and CSI 500 stock index futures,quantitative trading has attracted more and more attention and research from the fund team.However,at the same time as the global wave of quantitative trading,the overall returns of domestic quantitative funds have performed poorly,and there is a large performance gap between products,and there is a phenomenon of"two-eighth differentiation"^[1,2].One of the main reasons is that the multi-factor stock selection model currently used in China is mostly imported from overseas,which does not implement the localization of the strategy well and lacks innovative trading ideas.In order to explore a quantitative trading strategy suitable for the Chinese A-share market,this article builds a multi-factor quantitative stock selection model based on the scoring method,using the factor's IR value as a weight,and combines the famous Foster Friess positive growth strategy's ideas to conduct empirical research on quantitative trading.This article uses the JoinQuant quantitative investment platform as the main operating software,and takes the constituent stocks of the CSI All-Share Index as the research object.The full sample interval is set from November 3,2008 to February 6,2020,for a total of 11 years of sample data.First of all,after testing the effective factors,three time windows were selected to represent the bull,the shock and the bear market,respectively.The traditional multi-factor model was constructed to find the differences and similarities of the effective factors in different market environments.Then,the effective factors with significant performance are selected into the optimal factor set,and the optimal factor trading strategy is constructed by the scoring method.The operating effect of the strategy is comprehensively analyzed from the perspective of returns,risks,and positions,and the article tests the robustness of the strategy by moving the backtest interval.Finally,the weak part of the strategy is improved and debugged,and the expected return rate,asset-liability rate,and sales gross profit rate are innovatively used as screening factors to construct a growth optimal factor composite strategy.The empirical results show that the improved strategy is significantly better than the original strategy,whether it is an excess return,sharpe ratio,or maximum retracement index.The main conclusions of this paper are as follows:First,the correlations between various factors that affect stock returns are different in different market environments or different time windows.Second,the idea of Foster Friess positive growth strategy can be well applied to the Chinese stock market.Third,from the performance of certain time periods of the strategy,compared with the regression method,the scoring method can better adapt to the domestic A-share market by using the IR value of each factor as the weight of the factor.Fourth,from the perspective of the sensitivity of stock returns to factors,growth,cash flow,and technical factors perform generally,but the circulation market value,the 20-day average turnover rate,the ratio of 20-day average turnover rate to 120-day average turnover rate,and the fixed asset turnover factor can be used as a good stock selection factor for the Chinese stock market to construct a trading strategy.