Multi-factor Quantitative Investment Based On Genetic Algorithm Strategy Practice And Application

In China's stock market,under the conditions of China's booming economy the number of stocks is also increasing.According to the data from the WIND database,there were 1086 listed companies in 2000,and 2,063 in 2010.On April 18,2017,there were 3,208 listed companies in Shanghai and Shenzhen.In July 2019,the shares of Science and Technology Board were added,which further increased the number of stocks,and this number will continue to increase in the future.How to get from these The selection of stocks with high investment value in stocks is an important research branch in finance.The stock market is fundamentally a capital market.In the capital market,how to use a suitable analytical method to obtain a portfolio of investment stocks with investment value is a problem faced by investors.This paper conducts multi-factor quantitative stock selection analysis of A shares of more than 3,000 listed company stocks,through the historical data of these stocks to carry out a large number of data processing and statistical tests to ensure the validity of stock selection,using selected factors as multi-factor stock picking From which stocks with investment value are selected as investment portfolios to obtain stable returns.This paper studies the factor selection technique of genetic algorithm based on the basic quality factor,value class factor and growth factor,selects the factor indicator of the multi-factor model of this paper,and filters each factor through historical data.And testing to build a more efficient multi-factor stock selection model,making the model more robust and practical.It can be seen from the backtest results that the multi-factor stock selection strategy based on genetic algorithm is effective in the A-share market and can obtain significant excess returns.However,this paper also has many shortcomings,such as the selection of factors is not large enough,and there is no in-depth analysis of the factors,but only a part of the analysis is selected within a limited range of factors.