In this paper, the author outlines a dummy variable technique that is a convenient procedure for obtaining cumulative prediction errors and related test statistics. By appending a vector of (0,1) dummy variables to the right‐hand side of the market model, results usually obtained in two steps can be obtained in a single multiple regression. The primary advantage of this technique is that both prediction errors and correct test statistics may be obtained from any standard regression package.

No abstract

No abstract

In this paper, we use stock market data to examine the intra‐industry effects of the July 5, 1982, closure of the Penn Square bank. A sample of 54 bank stocks is divided into four portfolios: industry, money center, Texas, and upstream. The latter group consists of banks that had purchased loans directly from Penn Square. Our objective is to determine whether FDIC Chairman Isaac's decision to close, rather than merge, Penn Square had an industry‐wide contagion effect or a firm‐specific information effect. We conclude that the stock market reaction to the Penn Square closure represents a rational investor response to new bank‐specific information.

Abstract.We examine several event-study test statistics that can be used to detect abnormal performance during a multiperiod event window. We demonstrate that one of the most commonly used test statistics does not, under the assumptions made, have the distribution claimed (standard normal), and thus tests using it will be biased. The magnitude of that bias is shown to increase with the length of the event window and can generally be expected to lead to excessive rejection of the null hypothesis. We also compare the relative power of alternative test statistics that are normally distributed and are straightforward to apply.