Empirical Research Of Building Robust Growth Stocks Pool Based On The Statistical Method

Under the background of the prosperity and development of the world, our stockmarket gradually grow up with the establishment of our market economic system. In thestock market, the investor is indispensable role, at the same time, the long-termreasonable return on investment is the key to guarantee the investor to invest the stockmarket. Throughout the entire investment process, whether gain or not is also be themost important thing. At present, there is no lack of theoretical and empirical analysisof the equities research at home and abroad, however, foreign stock investment methodis not suitable for the weak validity China’s securities market, and the domesticresearch is focus on the analysis of the affecting factors for stock prices, so the methodof choosing stock is very lack. Above all, this article will use statistical method to putforward a set of effective way to choose stocks, then to help investors to select thesuitable stocks for themselves from more than2,000stocks in the A-share marketquickly and effectively.In order to select the valuable stocks to invest, the good operation of listedcompanies is a guarantee of the stock to get a good income, so that it is necessary toanalysis the company’s fundamentals. Based on the theory of value investment, thispaper uses fundamental analysis to introduce the factors that can affect the stockfluctuations, and selects27quantifiable indicators from the factors to establishevaluation system of the listed companies by the evaluation indexes of company andindustry, then takes advantage of this evaluation system to study the means of selectingthe stocks with stable income and set up stock pool. At first, this paper uses R softwareto carry on the principal component analysis to the evaluation system. In accordancewith the requirements that the cumulative contribution rate is more than80%, the13factors are be screened which can describe the evaluation system. In the next place, thisarticle uses logistic regression twice to analysis these13factors and listed company’schange within60days. Based on the analysis result, the five principal componentfactors which are significantly associated with stock volatility can be confirmed. In thethird place, the article uses the five principal components to establish the comprehensivefactor score model for stock selection.This article takes use of the stock selection model for the following four industries:medical, information services, agroforestry and fishing, textile and clothing, then chooses stocks from the357listed companies, and uses two ways to analyse empiricalresults: one is to select each industry’s top10stocks; the other one is to select twocompanies from each industry to establish stock pool. From the empirical results, wecan know that the rising number of companies accounts for60%to70%of the totalnumber. In accordance with the price, some reasonable profit will be geted. Therefore,the empirical results verify the validity of the stock selection model that can be used tochoose the stocks with long-term stable profits, which is of great significance forpractical application.