News or Noise? The Stock Market Reaction to Different Types of Company-Specific News Events

Abstract

It is generally accepted that news, often defined as news stories in a professionally edited newspaper, moves stock prices. By exploring data from an online stock forum our study presents a novel approach to identify stock-related news events from an investor perspective as an alternative to traditional media sources. We investigate the market impact of different types of company-specific news events (e.g., news related to corporate governance, operations, or legal issues) on S&P 500 stock prices in order to discern genuine news that moves the market from insignificant noise without market reaction. We leverage computational linguistics methods to distinguish between good and bad news by controlling for the sentiment (i.e. the positive vs. negative tone) of different news stories. Our results show that the absolute value of cumulative returns prior to a news event are more pronounced for positive news than they are for negative news, suggesting more widespread information leakage before good news. We find that the market reaction differs substantially across various types of news events. In addition, a cross-industry comparison indicates that industry classification may partially explain the market reaction to the same event type.

Keywords: Twitter, Microblogging, Text Classification, Stock Market, News, Event Study, News Sentiment

JEL Classification: G14