Uniformly least powerful tests of market efficiency

Defenders of market efficiency argue that anomalies involving long-term abnormal returns are not robust to alternative methodologies. We argue that because various methodologies use different weighting schemes, the magnitude of abnormal returns should differ, and in a predictable manner. Three problems are identified that cause low power in value-weighted three-factor time series regressions when abnormal returns following managerial actions are being estimated. We illustrate the sensitivities in the context of the new issues puzzle as well as with simulations. More generally, multifactor models as currently used do not, and cannot, test market efficiency. (C) 2000 Elsevier Science S.A. All rights reserved. JEL classification: G12; G14.