The market model and the event study method: A synthesis of the econometric criticisms

Abstract

In recent years the applied finance literature has seen a proliferation of papers concerned with, or using the Event Study Method. An essential tool of this method is the use of the single index market model as the equilibrium returns generating process. Underpinning this model are a set of statistical assumptions, which though frequently acknowledged in the literature, are usually merely asserted, rather than empirically tested. In this paper a discussion of the market model and its crucial role in event studies is offered. The econometric problems which occur in both the estimation of the single index market model and in using this model to generate abnormal returns and cumulative abnormal returns and thus measure security price reaction to economic or accounting events is then outlined. The results of recent empirical investigations into the market model and the event study method are employed to discuss the applicability and validity of this method. These conclusions have important implications for market based empiricists, as doubt is cast on the validity of the statistical assumptions underlying the market model and the event study method and hence the use of these studies as a tool of applied financial research. “… the results from the market model should be discounted for policy purposes, on the grounds that there are unresolved methodological problems” (Conn, 1985, p. 54.) “Security return event studies are one of the most objective ways in which researchers can determine whether accounting or other events have an impact on investors.” (Thompson, 1988, p. 77).