Information dissemination, market efficiency and the joint test issue

Despite the popularity of event studies in capital market research, the inherent joint test issue has surprisingly received very limited attention in the existing literature. Specifically, the state-of-the-art consists in merely asserting the descriptive validity of a specification of the two-parameter market model or even cruder models such as the constant mean model and the model of market adjusted returns-hardly an epitome of sound research practice.The most salient contribution of this work resides in the suggested two-step approach to event studies. Indeed, the proposed approach requires that the conduct of the formal event study be systematically preceded by the search for the best fitting pricing model. By proceeding this way, the beauty defect inherent in the de facto Brenner (1977) approach, consisting in using a battery of return-generating processes for robustness purposes, comes to be mitigated. This is because the suggested two-step procedure ably disentangles the deficiency of the underlying return-generating process and the intrinsic inefficiency of the market to hand. The empirical results attained by Mr. Basse Mama suggest that, for the retained sample, the price formation mechanics on the German stock market are inconsistent with the semi-strong form market efficiency. This study adds to the literature in that it uses an updated sample of ad hoc announcements pursuant to § 15 WpHG. What is more, the joint test issue even in short-window event studies appears to be more crucial than many researchers have apparently assumed.