The role of information intermediaries in the capital market: Evidence from China

This thesis consists of three essays. In Essay 1, by developing a unique textual dictionary and implementing content analysis of firm-specific Chinese news articles, I examine the relation between news media and analyst forecast revisions in China. I document a significant positive relation between the tone (volume) of news articles and the direction (magnitude) of analyst forecast revisions. I also find stronger market reactions to, and higher accuracy of, revised forecasts for firms with more extensive news coverage. Further, the positive relation between news and analyst forecast revisions is more pronounced when firm reporting quality is low. Additional analysis shows that quantitative earnings-related news articles are more useful and informative to analysts. While the news tone for government-controlled firms is more optimistic, it seems analysts are able to unravel the possible positive bias and make forecast revisions of similar accuracy for both government-controlled firms and privately controlled firms. I further document a significantly positive relation between news articles by state-controlled newspapers and analyst forecast revisions, suggesting that state-controlled newspapers can provide information that has incremental value to Chinese securities analysts. However, the informativeness of information provided by state-controlled newspapers is reduced when they face greater political pressures. In Essay 2, using brokerage mergers and closures in the Chinese capital market as two natural experiments that lead to an exogenous shock to analyst coverage, I examine how exogenous reductions in analyst coverage affect firm innovation. Using a difference-in-differences design, I find that, on average, firms affected by broker mergers or closures experience a reduction in their innovation inputs and outputs. My results are robust to various measures of innovation, including the quantity and quality of innovation outputs, innovation efficiency, and R&D investment intensity. The effect of an exogenous drop in analyst coverage on firm innovation is more pronounced for firms with worse information environments and more severe agency conflicts, such as those covered by less skillful analysts (as proxied by analysts' general experience, house experience, brokerage house size, and forecast accuracy), have a small number of analysts following, and controlled by government shareholders. In Essay 3, I examine the effect of news coverage on corporate innovation in China, and how the characteristics of firms, news outlets and news articles affect this relation. I find that news coverage has a positive impact on corporate innovation measured by patent and citation counts, R&D investment intensity, and innovation efficiency. Despite enhancing in-house innovation, news spotlight reduces managers' needs in acquiring innovation from external channels. I also find that a firm's level of financial constraints and R&D intensiveness strengthen the positive effect of news coverage on innovation, whereas board independence, institutional ownership, corporate transparency, and state ownership attenuate this effect. My finding of a positive effect of news media on corporate innovation remains unchanged when using the 2SLS regression analysis to account for possible endogeneity issues. I further show that the informational and monitoring functions of news media are enhanced if the content of news articles is directly related to innovation and corporate governance, while the media's functions are impeded if newspapers are state-controlled, domiciled in the political center (i.e., Beijing) and socially connected with the firms under coverage