Research On Smart Stock Selection Strategies That Introduce Dividend Factors And Corporate Quality Factors

China's investment market is growing,with the development of the economy and the improvement of openness,investors are also gradually turning to the stock market.The change of China's stock market is inextricably related to the dynamics of the whole country's market economy,which plays an important role in promoting the continuous development of our national economy.Compared with the more perfect capital market in the west,the development of the intelligent stock selection investment strategy field using machine learning and other methods in the stock market of our country is relatively late,and there are still some shortcomings.The traditional investment strategy relies more on personal experience,lacks objectivity,and is vulnerable to investors' hearts and minds.The influence of irrational behavior,such as theory,will also form huge time cost and manpower cost in the processing of a large number of data,which is not conducive to the rapid development of the investment market.In this context,the study of intelligent stock selection strategy began to rise.The intelligent stock selection strategy can avoid the disturbance of investor sentiment and affect the final result of the strategy,and the calculation of the model can reduce the time and manpower cost.Intelligent stock selection will be applied to a variety of mathematical econometric models,and the setting of parameters and the selection of indicators in the model will directly affect the results of the strategy.With the deepening of the research,the lack of quantitative models such as stochastic forest and neural network is lacking.Points are also exposed,because the stock information on the Chinese market is intermingled,the company has the existence of fraud,so that in the real A-share market,such as random forest,neural network and other intelligent algorithms will be effective,unable to reflect the real effect.It may lead to the failure of the model.Therefore,in order to solve the difficulties faced by intelligent stock selection and avoid the mistakes in decision-making of individual investors and major institutional investors,through the selection of indicators and the classification of enterprise quality in the market,Toimprove the accuracy of the stock selection strategy,so that the relevant models of intelligent stock selection can play a good role in the Chinese market,to help the vast number of investors to join together.Rational investment guidance has a very important practical significance.In order to improve the accuracy of the stock investment strategy and better forecast the portfolio,this paper uses the random forest model based on the first quarter data of 2167 stocks in the 2018 A-share market.BP neural network and GRNN neural network model train the model and understand the failure of the model,and then introduce dividend factor and enterprise quality factor.The dividend factor is added to the original index and the top 100 stock of listed companies in the "Blue Book of listed companies-quality Evaluation report of Chinese listed companies" is used as the stock pool of the model to solve the problem of stock selection in the A-share market.The problem of model failure improves the accuracy of stock selection strategy.Finally,three models,random forest,BP neural network and GRNN neural network,are used to introduce the stock selection strategy of dividend factor and enterprise quality factor.Each model selects five stocks to hold quarterly,and compares the returns of different models.The conclusion shows that the introduction of dividend factor and enterprise quality factor solves the problem of model failure of intelligent stock selection model in A-share market,and effectively improves the accuracy of stock selection strategy.At the same time,in the horizontal comparison of the three models,it is found that the stock selection income of the stochastic forest model with dividend factor and enterprise quality factor is better than that of BP neural network and GRNN neural network.Therefore,the stochastic forest model,which introduces dividend factor and enterprise quality factor,is chosen as the optimal stock selection strategy.In the future,we can use the optimized stochastic forest to set up the stock intelligent stock selection model for practical application.The stock selection strategy of this paper is more suitable for asset allocation and long-term value investment of buyer's institution because of its long-term holding period and no frequent trading,which is not conducive to the seller's organizationwhich gets profit from the commission and is more suitable for asset allocation and long-term value investment.Using the method of model quantification to make stock selection decision can avoid irrational investment behavior because individual investors who lack professional knowledge are affected by emotion.The emergence of intelligent stock selection makes up for the shortcomings of traditional stock selection methods.Can actively promote China's quantitative investment in the field of deepening development.The dividend factor added in the index selection and the classification of the stock pool by the enterprise quality factor can further solve the problem of the failure of the stock selection model in the A-share market,and eliminate the false and complicated information in the market.Improve the identification ability of stock selection model,make the stock selection strategy more effective,and have certain guiding significance to investors.