Chinese Media, News Characteristics and Information Efficiency of Capital Market
Yang Yulong,Wu Wen,Gao Yongjing,Zhang Qiannan
DOI: https://doi.org/10.16538/j.cnki.jfe.2018.06.008
2018-01-01
Abstract:Since reform and opening-up, Chinese media industry has undergone profound development. The forms of media have become more diversified. However, the increase in the amount of media is accom-panied with some problems, such as the erosion of independence due to business benefits, and the relatively low quality of employees in media industry. Therefore, readers are often confused about how to select reliable media. To address these issues, we need scientific research to provide us with guidance, and help us under-stand the information characteristics of media.As a combination of policy-oriented media and market-oriented media, Chinese media industry is unique for its dual-track feature. The policy-oriented media is supported by government finance in some way, and mainly focuses on government policy. In contrast, the market-oriented media is self-financing, and pays more attention to the information demands of capital market. Prior literature finds that media coverage can help im-prove the quality of corporate governance, and enlarge the information efficiency of capital market. But for in-vestors, they care more about how to select media or news articles to facilitate their decision making. In other words, they want to identify the very media or news article that has a substantial impact on the information in-tegration of capital market. Therefore, further research should take account of media types and characteristics of news articles, in order to figure out the functional distinction between different types of media.Based on the dual-track feature of Chinese media industry, this paper discusses the differences and reas-ons of performance in facilitating information integration of capital market between policy-oriented media and market-oriented media. By comparing the use of positive and negative words in news articles, we find that positive reports dominate in all media. But the proportion of positive reports is larger for policy-oriented me-dia than for market-oriented media, while the proportion of negative reports goes opposite. We also find that the proportion of positive reports gradually declines from 2000 to 2013, while that of negative reports gradu-ally increases, and the above trends are more significant in the market-oriented media. Given news articles are carriers of information, so the differences of news articles' features between two types of media must lead to different performances in disseminating information.Using Python programming language and Genius Finance News Database, we sort and analyze all the news articles related to the listed firms. Consequently, we construct a database of positive and negative media reports, which covers the period from 2003 to 2013, and consists of 137 unique media. With a sample from China A-share listed non-financial firms from 2003 to 2013, we find that the media coverage can improve the information efficiency of capital market through the channel of negative news rather than that of positive news. Moreover, the above effects are mainly driven by the policy-oriented media, while the market-oriented media barely shows no influence. To explore the mechanism behind those relationships, we quantify the depth of information discovery and the accuracy of information dissemination for each news article, using the meth-od of textual analysis. Further empirical evidences indicate that the policy-oriented media discovers informa-tion deeper and disseminates information more accurate than market-oriented media, and this is the reason why policy-oriented media performs better than market-oriented media in facilitating the information integration of capital market.Our results imply that negative news is more valuable for investors' decision-making. Especially, we find that the impact of media on information efficiency of capital market is mainly driven by the policy-oriented media, which is helpful for investors to narrow their scope of media selection. Meanwhile, our results can help regulatory authorities evaluate and select media platform, and assist them to provide media with guidance in respect of news content.Technically, using the method of textual analysis, we get a database of positive and negative vocabulary which is compatible with Chinese news reports. Based on the vocabulary database, we further analyze the news articles in Genius Finance News Database by Python programming language. Finally, we construct a database of positive and negative news reports, which covers the period from 2000 to 2013. Moreover, we em-ploy the model of Latent Dirichlet Allocation to measure the depth of information discovery and the accuracy of information dissemination, which helps us explore the reason why the policy-oriented media performs bet-ter than the market-oriented media in terms of improving information efficiency of capital market.