Earnings announcement timeliness and investor wealth.

摘要

The research tests for the existence of a systematic relationship between earnings announcement timeliness and the wealth characteristic of the announcement user set. Prior research by Cready (1988) indicates that the user set for earnings announcements may be wealthier, per capita, than the general population of investors. Hakansson (1977) suggests that higher-wealth investors may be more likely than lower-wealth investors to utilize substitutive predisclosure earnings information. Two testable implications are derived from the integration of Cready (1988) and Hakansson (1977). One, the user set for a more timely announcement should be wealthier, per capita, than the user set for a less timely announcement. Two, the user set for a predisclosure earnings signal should be wealthier, per capita, than the user set for the related later announcement signal.; In testing the hypotheses, signal utilization is proxied by the observed choice to engage in event period trading. User wealth level is indicated by transaction size (i.e., number of shares and dollar value). User set wealth composition is indicated by the sign of event period abnormal mean transaction size. A significant positive (negative) value is interpreted as indicating that the user set is wealthier (less wealthy), per capita, than the population of potential users. An announcement signal is considered more (less) timely if it is made earlier (later) than expected. The predisclosure signal considered is the passing of an expected announcement date. For each hypothesis test, the event period abnormal mean transaction size is predicted to be greater (i.e., more positive) for the more timely signal.; The findings support the first hypothesis, but not the second. Abnormal mean transaction size is found to be significantly greater for early announcements than for late announcements. The observed timeliness effect appears to be particularly strong in the case of annual announcements. The effect does not appear to be driven by other factors on which the groups of early and late announcing firms are found to differ. The lack of support for the second hypothesis may be attributable to the selection of a relatively weak signal with a difficult-to-specify event timing.