Research And Application On The Method Of The Correlation Measurement Between Stocks

The stock market had steped into a new bull market since 2014.The Chinese indulge in stock investment and their demand for technology analysis of stock is increasing sharply.As the tool to research the relationship between asset price or yields or industry classification and pricing,the correlation between stocks was regarded by investment institutions and individual investors as an important reference basis to measure the stock market risk and the effectiveness of a portfolio.Generally,domestic or foreign scholars use the correlation coefficient to measure the correlation between stocks.The greater the correlation coefficient,the stronger the correlation between stocks.However,the significant bull market or bear market in China will lead to the over similar rise or fall pheno menon among the stocks in different sectors that means unilateral rise or fall phenomenon in stock market causing the excessive linkage effect between stocks which will lead to the estimates of the stock correlation coefficient deviates from the real value and affecting the industry classification,investors’ decision-making investment and the effectiveness of the portfolio.Therefore,it is important to research the way to avoid the excess linkage effect between stocks and find out the the real correlation coefficient between the stock returns.In order to solve the problem,this paper studied the stock correlation analysis method of OLS,CAPM model and the correlation between stock returns under different stock market characteristics firstly.we take the l ogarithmic yield of the A-share as the empirical object to verify the universality of the excess linkage effect in the stock market by the hadoop algorithm.Secondly,the paper transforms the traditional CAPM formula to eliminate the market systemic risk c aused by the excess linkage effect,and constructs the improved stock correlation measurement method.For the excess linkage effect in financial time series,the paper constructs a new method of the stock correlation measurement which effectively avoids th e effect of over-linkage effect under OLS and market characteristics by means of manual interception,automatic interception by volume,and interception of correlation coefficient curve of rolling time window.Finally,the paper empirically studies the A-share stock,and compares the correlation coefficient between the new method of the stock-to-stock correlation measurement and the traditional method to verify the rationality and validity of the new method.At the same time,the article will apply the new method of the correlation coefficient measurement between stocks which avoids the excess linkage effect as much as possible to the complex network on the correlation study between stocks.The new measurement method is more conducive to improving investor investment decisions,the effectiveness of portfolio selection and the accuracy of industry classification.