INTRA-INDUSTRY INFORMATION TRANSFERS ASSOCIATED WITH MANAGEMENT EARNINGS FORECASTS.

摘要

This research provides a theoretical basis for expecting a stock price impact of a firm's disclosure of changes in expected earnings on other non-disclosing firms in its industry (an information transfer) and proposes tests of the proposition that management forecasts of earnings generate such transfers. Cluster analysis procedures are used to construct homogeneous industries for which the transfer effect is hypothesized. Parametric tests are proposed that mitigate the effects of cross-sectional correlation of market model residuals within the industry. In addition, a model is formulated that seeks to explain cross-sectional differences in the magnitude of observed information transfers by economic relationships between the disclosing firm and non-disclosing industry co-members. The magnitude of the information transfer is expressed as a function of the magnitude of changes in earnings expectations of the disclosing firm, the market share variability of the reporter, and a measure of the relative diversification (homogeneity) of industry co-members. The results of empirical tests of these models fail to reject the null hypotheses for information transfer existence and relationship to the sign and magnitude of changes in earnings expectations conveyed by management forecasts (for homogenous industries as defined by cluster analysis). At the 4-digit SIC code level, the results support the hypothesis that the magnitude of information transfers is positively related to the sign and magnitude of changes in earnings expectations conveyed by management forecasts. In addition, the results support the hypothesis that the magnitude of information transfer is inversely related to the relative diversification of industry co-members. The results do not support market share variability as a determinant of the magnitude of information transfers.