Sunday, April 24, 2011
Event Study Analysis: Correctly Measuring the Dollar Impact of an Event, by Atanu Saha, Compass Lexecon, and Allen Ferrell, Harvard Law School; European Corporate Governance Institute (ECGI), was recently posted on SSRN. Here is the abstract:
Using the traditional event study approach in the context of securities litigation, the determination of the "materiality" of a firm disclosure hinges on the statistical significance of the abnormal share price change (i.e., return) on the disclosure day. To estimate per share damages, the abnormal return is then transformed to an abnormal dollar impact. It is often assumed that if the abnormal return on a disclosure day is statistically significant, so is the abnormal dollar effect. We demonstrate - first analytically and then through an empirical example - that need not be the case. We derive the proper t-statistic if one wishes to determine the statistical significance of an abnormal dollar effect. This has obvious implications for liability and damages in securities litigation matters.