Research On Multifactor Stock Selection Strategy Based On Support Vector Machine

With the rapid development of the financial computer field,quantitative investment through quantitative and computer programming is playing an increasingly important role,and has been recognized by more and more domestic investors.It is no longer a strange word in China.Through the study of this paper,not only can test the validity of multi-factor stock selection model in the Chinese market,but also able to find out the effective factors of the Chinese stock market combination,facilitate market participants and regulators track market dynamic change,for investors how to choose from many listed companies worth investment stock portfolio provide reference,has practical significance.This paper takes CSI 300 as the research object,tries to find out the effective factor in the A-share market from many influence factors,uses the effective factor to build the multi-factor stock selection model of support vector machine.This paper considers the factors of growth,technology,value,mood,and momentum on stock prices,Using the factor IC,Pearson correlation coefficient,and factor exposure analysis to determine the factors that can significantly affect the stock yield,Finally,equity,turnover rate,ROE,PE,EPS were selected as the effective factor combination,Building a Gaussian kernel with optimal parameters,Select from January 1,2020 to January 1,2021 as the time interval for training back-testing,The performance was analyzed and evaluated according to different indicators such as sharpe ratio,maximum withdrawal,annualized return and Beta.Through the back test of the above process,the following conclusions are drawn: the effective factors of CSI 300 are equity,turnover rate,ROE,PE,EPS;the portfolio strategy constructed by the support vector machine model stock selection strategy can maintain the diversity of the portfolio,reduce risk,and ensure the Sharp ratio at a high level,high risk return ability and low retreat,relatively stable income,can perform the market,proving the effectiveness of the model.