The impact of press releases on stock prices

Abstract

Press releases are found to be important events that represent a potential explanation for up to 24% of the major stock price movements. The classi ers are able to predict price movements larger than 3% with up to 60% precision. The event study con rms that press releases have a statistically signi cant e ect on stock prices on the rst day. Trading strategies were de ned and shown to be viable for the period tested, but the post-event patterns that motivate them are statistically insigni cant. To better understand stock price movements, and the ability to forecast them, press releases from companies on the New York Stock Exchange are used as data. Classi ers are applied to predict whether the stock price will react strongly to a press release or not. An event study is done to investigate patterns prior and subsequent to a press release being published. Algorithmic trading strategies are de ned and tested based on the results.