Stock Selection Base On Multiple-factor Quantitative Models

After the financial crisis, the international and domestic capital markets haveexperienced a catastrophe, almost all funds have suffered different degrees of loss,Ordinary investors near us also suffered significant losses. Value of the investment orqualitative investment appears to be so weak a financial catastrophe like this, but asopposed to quantitative investment has sprung to maintain a good rate of return in thecapital market. With the theoretical basis of mathematical finance, behavioral finance,such as the continuous development of computer science and artificial intelligencecontinue to have provided the necessary and sufficient conditions for the developmentof quantitative investment. Quantitative investment has become an inevitable trend inthe growing number of investment targets investment community.In this paper, a method based on multi-factor stock selection models, throughquantitative methods detailed insights into the part of the financial indicators andtransaction data, combined with statistical tests, in order to help investors find themost investment value of the stock portfolio. First introduced some relevantknowledge, domestic and international developments, such as quantitative investmentliterature review. Second introduce some basic concepts to quantify multi-factor stockselection models involved in the classification of multi-factor stock selection models,as well as some of the policy model concepts. Followed by the establishment of amulti-factor stock selection models to quantify, to get a comprehensive evaluationmodel, and verify its effectiveness.According to economic logic and market experience, and preliminary screeningdata availability according to the Chinese market for25candidate factors, usingJanuary2000to December2009A-share market is part of its financial data andtransaction data effectiveness of inspection; examination evaluation system includes acompound average annual income of each factor, the excess revenue, earnings andscores correlation analysis and probability of outperforming the market, get aneffective factor7the test results.7Effective factors by calculating the correlationmatrix to score ROE changes in addition to the redundancy factor. Finally get earnings yield, market capitalization than the carrying amount of a cash yield, ROAchanges, PEG, turnover does not change in six effective redundancy factors.Such as the establishment of a comprehensive weight Z score model to determinethe optimal number of stocks to50portfolios, portfolio adjustment period of onemonth. Using January2010to December2013A-share market data on the model wastested by comparing the income and portfolio income in the same period of the CSI300, as well as risk measurement model to overcome the reference frequency andsuspicious, can prove that the use of multi-factor model portfolio is effective, so as toachieve the purpose of providing investment information and advice to stockinvestors.