Short-Term Reactions to News Announcements: What Do Investors Learn from Them?

Abstract

This paper extends models of learning about profitability to the case of voluntary disclosure. Using short-window regressions, I first find that the whole company-specific news flow, not just earnings announcements, has the property of providing useful information about other stocks. Consequently, betas of announcing stocks increase significantly in months with some news, especially if the volume of news is unusually high. However, I also find that voluntary disclosure differs from earnings announcements in one important respect. While for the latter the sign of the news is not important, it is only negative non-earnings news that has a significant impact on beta. This is consistent with theoretical predictions concerning incentives of managers to withhold negative news, which subsequently leads to disclosure "bunching".

Keywords: public news, learning, tone, market beta, CAPM

JEL Classification: C33, G12