Linguistic Tone in Earnings Announcements: News or Noise?

Abstract:     
This paper examines whether, and under what conditions, the linguistic tone (“soft information”) of management’s quarterly earnings press releases predicts firm’s valuation fundamentals and affects stock prices and stock price volatility. We use textual-analysis programs to extract two dimensions of managerial soft information - net optimism and certainty - from more than 20,000 corporate earnings announcements from the period 1998 to 2006. We show that soft information predicts valuation fundamentals (the levels of, and uncertainty associated with, future earnings of the firm), and that it affects asset prices both within the announcement event window and during the 60-trading day post-announcement period. Consistent with a standard Bayesian learning model, two key aspects of the information environment influence the linguistic tone’s impact on share prices: (i) the noisiness of the prior and contemporaneously available hard information; and (ii) the likely credibility of the net optimism itself. We also find that linguistic certainty attenuates the post-announcement period’s response to unexpected net optimism, is associated with contemporaneous announcement period idiosyncratic volatility, and predicts future idiosyncratic volatility after controlling for the simultaneously issued “hard” earnings information and other variables.

Number of Pages in PDF File: 65

Keywords: soft information, earnings announcements, post-earnings drift, cheap talk, earnings quality, information uncertainty, momentum, voluntary disclosure

JEL Classification: G14, D82, M41

working papers series