Analysts' Reactions to Warnings of Negative Earning Surprises

Abstract

We investigate analysts? reactions to qualitative warnings of adverse earnings, and attempt to reconcile analysts? more negative forecast revisions, as documented in previous research, and the apparently conflicting anecdotal evidence that suggests more positive responses to firms that warn. We conjecture that the inconsistency arises because warnings cause analysts to revise their earnings forecasts sequentially in response to two related signals (the warning and the earnings announcement). However, their responses to retrospective questions about the effects of warnings are the result of a simultaneous response to both signals. In our experiment, 28 financial analysts predicted future years? earnings in one of three conditions. Consistent with the anecdotal evidence suggesting that analysts evaluate warnings positively, analysts who simultaneously evaluated a warning and a later earnings announcement (the Simultaneous warning condition) forecasted higher future earnings than analysts who received no warning (the No Warning condition). However, analysts who first revised their forecasts based on the warning and then made a second revision based on the later earnings announcement (our Sequential warning condition) forecasted much lower future earnings than those in the Simultaneous warning condition, and slightly lower future earnings than in the No warning condition. This suggests that the act of sequential processing contributes to analysts more negative forecasts documented in previous archival research. Taken together, these results suggest that the positive impact of analysts' stated beliefs about firms that warn (as reflected in their responses to reporters? retrospective questions) are more than offset by the effects of sequentially processing a warning followed by an earnings announcement.

JEL Classification: M41, G29, C91