Discussion of: “Testing behavioral finance theories using trends and sequences in financial performance,” (by Wesley Chan, Richard Frankel, and S.P. Kothari)☆

Abstract

Chan, Frankel, and Kothari (CFK) present new empirical evidence which they argue rules out a broad class of behavioral models as potentials explanations of long-horizon return predictability. I provide a different interpretation of their findings. The CFK evidence and other extant empirical evidence is inconsistent with investor misinterpretation of earnings information as a source a long-horizon predictability, as CFK claim. However, this evidence points to misinterpretation of non-accounting based information as a source of the predictability. This type of predictability is consistent with the predictions of other behavioral models, such as that of Daniel, Hirshleifer and Subrahmanyam (1998).