The timing and value of forecast and recommendation revisions

Abstract

This paper evaluates the information content of analysts’ one-quarter-ahead earnings forecast revisions and recommendation revisions at various points in event time relative to earnings announcement dates. Across three sets of tests, we find that the revisions are least informative in the week after earnings announcements and that the information content of revisions generally increases over event time. We find a sharp increase in the information content of upward forecast revisions and recommendation upgrades in the week before earnings announcements, but we do not find a similar increase for downward revisions or for recommendation downgrades.