Expectations and Interpretations : Testing Confirmatory Bias Against the Representativeness Heuristic in Financial Market

This study evaluates the importance of confirmatory bias and the representativeness in interpreting corporate earnings news. We find that: (i) confirmatory bias and the representativeness heuristic affect how analysts interpret earnings announcements; and (ii) confirmatory bias has greater impacts on one-quarter-ahead forecasts than the representativeness heuristic does, but reverse is true for forecasts of longer horizons. We find firms with more exposure to confirmatory bias experience a stronger under-reaction in quarterly earnings forecast revisions, a weaker over-reaction in long-term growth forecast revisions, and a stronger postearnings-announcement-drift. We find the opposite for firms with more exposure to the representativeness heuristic