JOURNAL OF FINANCIAL AND QUANTITATIVE ANALYSIS November 1976 ABSTRACT: INFORMATION EFFECTS AND STOCK MARKET RESPONSE TO SIGNS OF FIRM DETERIORATION

INFORMATION EFFECTS AND STOCK MARKET RESPONSE TO SIGNS OF FIRM DETERIORATION Edward I. Altman and Menachem Brenner* The focus of our study concerns the information effect of newly reported data. Our time frame of reference is the period after the data become available. The information we analyze is generated by a multivariate model which utilizes publicly available data. We concentrate on an extremely poor performing group of companies where the new information indicates a change in status from a going concern to a potential bankruptcy; i.e., the firms possess characteristics similar, to other firms which were bankrupt in the past. Our sample, however, is comprised only of firms which, in fact, did not fail. The selected companies are then subjected to a series of tests that employed the residual methodology in different variants. Based on two-factor market model results it appears that the market's realization of the firms' deterioration, as reflected by subsequent excess negative returns, begins to manifest several months after the newly reported information indicated a problem situation. Our conclusions do not change when the same methodology was applied to a control sample of firms at the same time period with similar systematic risk. Tests of the differences in results from these two samples support the above findings. When we employed the single-factor model, the results were more consistent with other studies on market efficiency, i.e., the information is not new or when it is new the market reacts instantaneously. This result, however, was only for the market model residuals tested directly for the bankrupt-potential sample. When the difference in mean residual test was applied, the market model results were essentially similar to the other two models. It should be remembered that tests of market efficiency are weak tests in the sense that we are simultaneously testing several hypotheses; that the signals provided by the multivariate model are new information, that the market reacts instantaneously and the market model used to test market efficiency is correctly specified. New York University and Hebrew university, respectively.