Word Power: Content Analysis in the Presence of Competing Information
34 Pages Posted: 29 Jan 2019
Date Written: January 15, 2019
Abstract
Earnings announcements are the first notice of a firm’s financial performance and preempt the informational value of subsequent 10-K filings. This study evaluates tone measures that use earnings announcement returns or 10-K filing returns to interpret the relative strength of positive and negative words used in annual reports. I find that market reaction to earnings announcements are more informative than 10-K filing returns, and as these filing returns misclassify the relative strengths of words and consequentially tone measures. Market reaction to earnings announcements allows quantifying tone and elaborating the interpretation of financial language use better. In fact, my approach yields stronger correlation of qualitative information to first and second moments of stock returns, that is, future returns and volatility, and future profitability.
Keywords: textual analysis; tone; earnings announcements
JEL Classification: G14, G19, G30, M40, M41
Suggested Citation: Suggested Citation