Multi-Factor Alpha Strategy Design Based On Growth Enterprise Market

Over the past 30 years,China’s stock market from scratch,attracting countless people to join the waves of gold.With the development of China’s stock market,the increasing complexity and efficiency of the stock market has challenged the ability of traditional qualitative investors to go it alone.Quantitative investment is the combination of financial theory and computer program,and the integration of hardware and software.Compared with overseas mature markets,China’s stock market is less efficient,so China’s market is more suitable for the use of quantitative investment strategy.As an important financial engine of high-quality development in China,GEM market has a shorter listing time,smaller market size and higher stock price fluctuation than main board market.In particular,since the GEM registration system,the number and scale of listed companies continue to expand,“Excellent innovation,high growth”characteristics.It is worth exploring whether the multi-factor Alpha strategy is applicable to the GEM market.This thesis divides fundamental factors into 10 categories: profitability,growth ability,operational efficiency,earnings quality,safety,corporate governance,valuation,analysts,shareholders,and scale;Based on the construction method and core logic of price and quantity factors,they are divided into 8 categories,namely momentum and reversal,liquidity,volatility,volume price correlation,chip distribution,fund flow,northbound funds,and margin trading factors.After testing and verification,26 factors were ultimately selected as candidate factor pools.In addition to simple factors,linear purified profit factors,one month amplitude adjusted momentum factors were also attempted to be constructed Six relatively stronger composite factors,including the steady acceleration of operating profit.Firstly,this thesis conducts data pre-processing,mainly including outlier,standardization and neutralization,and tests the validity of single factor in the candidate factor pool,and determines the factor pool after preliminary screening;Secondly,the clustering algorithm of the machine learning library is used for classification and verified through the IC thermal map;Once again,multiple methods were gradually adopted to synthesize the highly correlated single factors into a single class factor,and a stock selection model was established;Finally,backtesting and the results showed that the multi factor Alpha strategy was effective in the GEM market.The constructed multi factor stock selection model could outperform the market and achieve stable excess returns.The research results of this thesis can provide some inspiration for investors to construct effective factors,and also contribute to the more rational and healthy development of the GEM market.