| With the development of science and technology and the availability of network big data,media news,as an important information transmission tool,plays an extremely important role in people’s life and investors’ investment behavior.In this context,the media has contributed to the development of the discipline of behavioral finance by delivering information to investors because the information it contains can increase investor awareness and attract investors’ attention.At present,researches on media information and financial market emerge in an endless stream,mainly from three levels of media coverage,media coverage sentiment,and media opinion differences.However,constrained by limited attention,the limited attention of individual investors,investors are more likely to use each category of investment to simplify their asset allocation than to invest it in a single stock.The industry factor is also an important factor affecting stock returns.Buying stocks in a certain industry is also a very important category.Some scholars even believe that the industry is the most important category of investment,attracting a large number of asset managers and investors.Not only that,stocks in the same industry have similar fundamentals and are subject to common macro policies and shocks,which represent industry-specific risks.In addition,the media has a spillover effect on the stocks of the same industry,and the media is more likely to report some industry information than company information.However,previous studies have studied the impact of media information on stock prices from the level of individual stocks,markets and countries.As an important perspective in asset pricing,industry is still lacking in the industry level to explore the relationship between media information and stock returns.research on the relationship between them.Therefore,this paper studies the impact of media coverage,media coverage sentiment,and media opinion divergence on cross-sectional stock returns from the industry level.The data source for this article comes from China News Network.The industry classification in this paper uses Shenwan’s secondary industry classification,and uses crawler technology and machine learning methods to construct industry media reports,industry media sentiment,and industry media opinion divergence indices,and based on this,study their impact on cross-sectional stock returns.impact and draw the following conclusions:First,this paper examines the effect of industry media coverage on cross-sectional stock returns.First,the number of reports of basic industry news,this paper constructs the amount of abnormal reports of industry news(industry media reports).Second,based on portfolio analysis and Fama-Mac Beth(1973)regression,we find that grouping by industry media coverage can achieve significant positive cross-sectional returns,and stocks with low industry media coverage can achieve significant gains in the following month.Positive returns,this conclusion is in line with the investor perception hypothesis put forward by Merton(1987),that is,investors have lower perceptions of stocks with low media coverage,so relative to stocks with high media coverage,stocks with low media coverage exist risk premium.Again,this paper finds that this result holds after controlling for market cap,book-to-market value,momentum,industry value added,and individual stock media coverage.In addition,we found that after controlling for common risk factors,portfolio analysis based on industry media reports can still achieve excess returns of more than 0.50% per month.Finally,this paper examines the reversal and persistence of stock returns driven by industry media coverage and finds that the negative relationship can persist for two months,with a reversal in the third month.In addition to this,in robustness checks,this result persists after we replace metrics reported by industry media,calculate portfolio returns using a market capitalization-weighted approach,and use industry-level Fama-Mac Beth(1973)regressions.Second,this paper examines the effect of industry media sentiment on crosssectional stock returns.First,based on the Naive Bayes method,we distinguish news sentiment into positive and negative sentiment,and based on this,construct a measure of industry media sentiment,which is the number of positive news per stock per month minus the The number of negative news,divided by the sum of the number of positive and negative news.Second,based on portfolio analysis and Fama-Mac Beth(1973)regression,we find that grouping by industry media sentiment can achieve significant positive cross-section returns,and stocks with high industry media sentiment can achieve significant positive returns in the following month.This result is more pronounced in stocks with smaller caps and lower institutional investor holdings,meaning it is largely driven by investor sentiment.Again,this paper finds that this result holds after controlling for market capitalization,book-to-market value,momentum,industry value added,and individual stock media sentiment.In addition,after controlling for common risk factors,the portfolio analysis based on industry media sentiment can still obtain an excess return of about 0.50% per month.Finally,this paper studies the reversal and continuation of stock returns driven by industry media sentiment,and finds that the positive relationship can last for six months without a reversal phenomenon,which means that media news contains Information about company fundamentals.Beyond that,this result persists after replacing machine learning methods,replacing metrics reported by industry media,and using industrylevel Fama-Mac Beth(1973)regression.Third,this paper examines the effect of industry media disagreement on crosssectional stock returns.First,based on the Naive Bayes method,we distinguish news sentiment into positive and negative sentiment,and based on this,we construct a measure of industry media disagreement,which is the second-order moment of industry media sentiment.Second,based on portfolio analysis and Fama-Mac Beth(1973)regression,we find that grouping by industry media disagreement can achieve significant positive cross-sectional returns,and stocks with low industry media disagreement can achieve significant gains in the following month positive returns.This result is in line with Miller’s(1997)view that when there are short selling restrictions,differences of opinion in the market lead to lower expected returns(higher prices).Again,this paper finds that this result holds after controlling for market capitalization,book-to-market value,momentum,industry value added,and individual stock media disagreement.In addition,after controlling for common risk factors,the portfolio analysis based on the differences of opinion in the industry media can still obtain an excess return of about 0.60% per month.Finally,this paper examines the reversal and persistence of stock returns driven by industry media disagreement and finds that the negative relationship can persist for two months,with a reversal in the third month,implying that the result is dominated by driven by investor sentiment.In addition to this,this result persists after using the market capitalization weighted calculation method and using the industry-level Fama-Mac Beth(1973)regression.The theoretical significance of this paper is as follows:(1)It expands the research on the measurement dimension of media information.Previous studies on media information have focused on one point in the number of media or text features to study its impact on stock returns.This paper supplements the research on media information by measuring the impact of media information on stock returns from three perspectives:industry media coverage,industry media sentiment,and industry media opinion differences.(2)Expanded research on industry phenomena in asset pricing.In the past theoretical and empirical literature,it is found that the industry plays an important role in asset pricing,such as industry momentum,industry book-to-market value ratio,industry concentration and industry expected return,etc.have an important impact on stock prices.Research on the impact of media coverage,industry media sentiment,and industry media disagreement on stock returns extends the research on industry phenomena in asset pricing from a media perspective.(3)Expanded the research related to media information in asset pricing.Previous studies have found that media coverage,media sentiment,and media disagreement can have an important impact on stock returns,and this finding exists at the individual stock level,market level,and country level.The study extends the study of the impact of media coverage on asset pricing.The practical significance of this paper is:(1)It has a guiding effect on the investment behavior of investors,especially individual investors.The investor structure in the Chinese market is dominated by individual investors.Due to the lack of professional ability,individual investors are greatly influenced by emotions,resulting in poor investment performance.This research studies stock returns from three perspectives: industry media reports,industry media sentiment,and industry media opinion differences,which can help investors better use media reports to improve investment decision-making behavior,which has implications for individual investors’ investment behavior.guiding role;(2)It has an enlightening effect on the supervisory behavior of supervisors.Financial market supervision is of great significance for maintaining financial stability and preventing the occurrence of systemic financial risks.This study finds that the media plays an important role in the operation and profitability of the stock market,so it has important implications for regulators.Regulators should strengthen the supervision of domestic and foreign media to stabilize the domestic financial market and prevent illegal acts of manipulating the media to earn excess profits;(3)To supervise the development of enterprises.In the past,companies were evaluated from financial statements and other indicators.This paper finds that the media plays an extremely important role in the company’s stock price changes.This will urge companies to pay attention to their own image,and has certain inspiration and guiding significance for listed companies on how to rationally use social media news to improve their financing capabilities. |
PDF Full Text Request