Combining SentiWordNet lexical resources with Harvard IV-4 psychological dictionary, this paper computes negative sentiment score from the financial news articles one-month prior to the release of the quarterly earnings announcements. Within the event study framework, this paper uses negative sentiment score to (1) predict company quarterly earnings, and (2) to predict stock return including following daily return, weekly return and monthly return after the quarterly earnings announcements were released. Main finding of this paper is that negative sentiment score obtained from the financial news article forecasts lower company quartely earning. The relationship between negative sentiment score is strongest when predicting weekly return, but the relationship is not significant. It seems to suggest that market prices underreact to financial news. Lastly, negative sentiment score extracted from the Wall Street Journal performs the best when predicting stock returns.

, , , ,
V. Spinu
hdl.handle.net/2105/39432
Business Economics
Erasmus School of Economics

J. Li. (2017, July 24). Can Negative Sentiment Score of Company-specific Financial News Predict Quaterly Earning and Stock Return?. Business Economics. Retrieved from http://hdl.handle.net/2105/39432