Multi-factor Model Stock Selection Plan Planning Based On Industry Rotation Strategy

In recent years,quantitative investment has attracted more and more attention from institutional investors and hedge funds because of its characteristics of discipline,systematic,timeliness and decentralization.From the effectiveness of China's securities market and the development experience of foreign securities markets,there is no doubt that the development prospects of quantitative investment are worth looking forward to.However,there are still some shortcomings in domestic quantitative investment products,such as small overall scale,single quantitative strategy and differentiation of strategic performance.It is of great significance for the development of quantified investment.Studying new ways of quantified investment and excavateding new ideas of modeling are of great significance.Among many quantitative strategies,multi-factor stock selection strategy has attracted many researchers' attention because of its stability and comprehensive advantages,and has performed well in the market.Based on the characteristics of the model,this paper optimizes them in three aspects,and in the stock pool,selecting high-quality components of CSI 300 Index to study.Firstly,in the selection of factors,in addition to the financial and technical factors that researchers used extensively,the price index,real estate sales situation,fiscal revenue and expenditure,total import and export value are added to enrich the information reflected by factor model;secondly,in view of the industry rotation effect of A-share market,industry rotation strategy is used to assist in screening stock pools before establishing multi-factor model.Thirdly,in the calculation of multi-factor model,the stochastic forest algorithm is used to simplify the steps of traditional multi-factor model,change the previous factor screening methods and modeling process,and use the way of training while screening,the screening method is more scientific and reasonable.In the selection of machine learning algorithm,the advantages and disadvantages of support vector machine,random forest and logistic regression are compared,and the results show that random forest is most suitable for performance and stability.Based on the above planning ideas,a multi-factor quantitative stock selection scheme is successfully designed by using machine learning method,which surpasses the excess return rate of the Shanghai-Shenzhen 300 Index.After 10 holdings,the total return of the selected stock portfolio is 13.6%,the annual composite yield is as high as 27.58%,Sharp ratio is 1.29,information ratio is 4.32,far exceeding the benchmark Shanghai-Shenzhen 300 index yield.