Font Size: a A A

Research On The Efficiency Of Multi-Factor Model Based On The Sentiment Of Investor

Posted on:2021-10-14Degree:MasterType:Thesis
Country:ChinaCandidate:Q W WangFull Text:PDF
GTID:2480306230496314Subject:Finance
Abstract/Summary:PDF Full Text Request
In the stock market,the issue of stock pricing is an important issue that is a common concern of academics and the industry.The classic capital asset pricing model has been expanded into the famous Fama-French three-factor model and the Fama-French five-factor model in the research of Fama and French.It can effectively measure the relationship between stock returns and risks,and explain the risk premium of the stock portfolio.In 2018,Fama proposed a six-factor model formed by adding a momentum factor to the five-factor model,further improving the model's interpretation ability.Based on Fama's six-factor model framework,this paper uses natural language processing technology to extract sentiment indicators from the market,incorporates them into a six-factor model to form a new multi-factor model,and explores the effectiveness and differences of each factor in China's A-share market.Based on this multi-factor model,a quantitative stock selection strategy is constructed,which is not only of theoretical significance but also practical significance.This article selects the CSI 300 stock data from January 2011 to November 2019 and the stock bar data of corresponding stocks as the research object.By constructing sentiment indicators and using monthly data of listed companies,a Fama-French six-factor model is constructed,and emotional factors are incorporated to form an improved multi-factor model.Then conduct an empirical study on the multi-factor model on the 25-grid stock portfolio.The empirical results show that the multi-factor model can better explain the fluctuation of the yield of the A-share CSI 300,and the joint assumption of excess returns is not significant.Finally,on the basis of the multi-factor model,the stock selection model is implemented using regression andranking methods,and a quantitative strategy is constructed,and the timing index is realized using emotional indicators.The return from strategy backtesting is significantly higher than that of the CSI 300 index.In summary,the multi-factor model constructed in this paper not only verifies the validity of the model on A shares,but also completes the construction of investment strategies in practice,and exerts the theoretical practical value.
Keywords/Search Tags:Investor sentiment, Fama-French six-factor model, natural language processing, Quantitative investment
PDF Full Text Request
Related items