Font Size: a A A

An Empirical Research Of Alpha Strategy Based On Multi-factor Stock Selection Model

Posted on:2020-03-28Degree:MasterType:Thesis
Country:ChinaCandidate:S ShiFull Text:PDF
GTID:2480306302478684Subject:Business Administration
Abstract/Summary:PDF Full Text Request
With the development of domestic securities market,the main body of market investors is gradually institutionalized,the regulatory system is gradually standardized,and the trading system is gradually finalized,especially all kinds of financial derivatives tools have been launched successively,which makes the domestic quantitative investment enter the accelerated development period.At present,domestic quantitative investment adopts various strategies,including high-frequency trading strategy,quantitative arbitrage strategy and alpha strategy.With the gradual introduction of two financing businesses,futures products and options products,alpha factors in the market are more easily separated.In the future,it will be the development direction of quantitative investment to build alpha strategy by analyzing market factors.Based on this,this paper will combine the characteristics of China's stock market,through the analysis of market factors,build a combination of multi factor stock selection model,and then obtain stable excess return from the market.This paper uses Barra's method of obtaining pure factors for reference,considers the particularity of industry and market value in domestic stock trading,and adopts industry neutral and market value neutral methods to select factors in factor analysis,so as to obtain the actual market income brought by pure factor exposure.The main contributions of this paper are as follows: 1.Define and select 8 categories of 42 market segmentation factors,constitute factor base and analyze a single factor.Through the correlation analysis of each factor and its individual stock return,the IC value is obtained,and then the effectiveness of the factor is evaluated;through the T value of regression analysis and the significance evaluation of the regression return value,the stability evaluation of the factor is combined with the IR value;through the historical Cumulative Return trend of the factor,the monotony of the factor is analyzed.Combined with the significance,monotony and effectiveness of the factors,we can find the effective market return factors in the factor pool.2.Draw on the method of obtaining pure factor in Barra,take Shenwan industry as the classification standard,return the industry as dummy variable and market value as independent variable,and then obtain the neutral pure factor value.By adopting the method of industry and market value neutralization,we can analyze the market characteristics of factors from the perspective of pure factors.3.By ranking the factors according to IC value and IR value,we can get the effective factor combination;by scoring individual stocks according to factor value,we can build a stock pool through comprehensive score ranking.4.The alpha strategy is established by constructing pure multi empty portfolio and hedging portfolio,and the effectiveness of the strategy is verified by income analysis and data outside the sample.To sum up,based on the practical application of the domestic market,this paper introduces 8 categories and 42 segmentation factors through the establishment of a two-tier structure of factor library,verifies the effectiveness,significance and stability of each factor,and selects the relatively high-quality factors.By constructing factor combination and stock pool,and combining short strategy,alpha strategy with stable return is realized.Finally,the validity of the relevant alpha policy is verified by the policy out of the sample.The multi factor analysis and research framework constructed in this paper can realize the process of factor stock selection,and has application value for the research of factor quantitative investment in the industry.At the same time,according to the test results of alpha strategy,it can be concluded that the alpha strategy constructed by the multi factor stock selection model can achieve relatively stable returns.
Keywords/Search Tags:Multi-factor Stock Selection, Alpha Strategy, Single-factor Test, Scoring Method
PDF Full Text Request
Related items