Sentiment Factor And Stock Cross-section Return

Over the past decade,with the continuous development of computer technology,quantitative investment in overseas capital markets has become a mainstream form of investment.The increasing transaction volume and market heat of the domestic capital market also provides a good development environment for the development of quantitative investment.The traditional subjective judgment-based investment fund also faces the transformation of quantitative form.Based on the background of quantitative investment and development,We attempts to build a quantitative stock selection strategy model with the domestic A-share market.We builds the stock selection strategy based on a large number of backtesting studies,doing a lot of simulated trading tests,and comprehensively analyzing the effectiveness of the stock selection factor.With the help of Qian(2004)'s analysis framework of predictive factor in 2004,Our work was mainly to analysis the transaction data,and construct the simulation backtesting system and then construct the portfolio,realizes the whole factor research process.We studies the explanatory power of stock cross-sectional returns based on the Sperman Correlation and information coefficient,and further analyzes the risk characteristics of the factors,and digs out the components with high information value in the factors.Based on the microscopic transaction data of the capital market,we constructs the stock cross-section emotion factor,takes the investor transaction behavior as the research perspective,goes deep into the market transaction level,obtains the valuable investment information from the high frequency transaction data.The factor can be used to explain the excess return of stock partly through empirical test.Based on this factor,the performance of the stock simulation trading significantly outperformed the benchmark market index,which gained 50.7%of the annualized yield from January 1,2014 to August 25,2016,and markedly outperformed the benchmark index which was only gained 14.2%And the maxdrawdown on our investment portfolio is controlled.We try to compare the factors based on different parameters,study the differences between them,and make the best choice of parameters.In addition,the factor-building thinking is extended to the stock index futures market for timing strategy research,back to the September 11,2013 to August 25,2016,a total of 722 trading days,our straategy gained the cumulative return of 102.25%,the annualized return of 34.12%.From the market high-frequency trading data,we do find some of the elements that can explain the stock returns the next few day,the specificity of the factor indicates its existence a certain value,but the stock market changes rapidly,we can not based on a single indicator to its fullly explanation,the article aims to expand the factor-building thinking,investment decision-making research to broaden the road.