Misspecification of the market model: the implications for event studies

Abstract

Event studies are increasingly found in the applied finance literature. They are employed to analyse the market reaction to events and thus to investigate market efficiency. The paper is concerned with misspecification testing of the single-index market model which is conventionally employed to generate returns within the event-study method. Weekly data from the constituents of the FT-SE 100 are used in order to subject the single-index market model to rigorous misspecification tests. It is concluded that misspecification is endemic in the market model and that this is likely to lead to incorrect estimation of abnormal returns, and therefore erroneous conclusions concerning the market impact of the event in question.