Discussion of “Do Investors Over-Rely on Old Elements of the Earnings Time Series"
2003, Contemporary Accounting Research
A study by Bloomfield, Libby, and Nelson 2003, Do Investors Overrely on Old Elements of the Earnings Time Series?, proposes a single explanation for the simultaneous existence of 2 seemingly contradictory financial market anomalies: underreaction to recent earnings changes and overreaction to sustained levels of performance. The study provides a clever and intuitively appealing bridge between the 2 anomalies. However, there are concerns about the contribution of the experiment. In particular, there is concern that the experiment is more a complex demonstration that humans are limited in their knowledge of business trivia and are horrible intuitive statisticians than a test of the theory that underweighting of old earnings information can lead to both the over- and underreaction anomalies documented in extant research.
Hopkins, P. E. (2003), "Discussion of 'Do Investors Over-Rely on Old Elements of the Earnings Time Series,'" Contemporary Accounting Research, Spring, pp. 33-46.