The information content of interim and final earnings announcements

Abstract

This study examined the effect of unexpected earnings, dividend change, and surrogates for the quality of earnings on abnormal returns. Since both earnings and dividends are announced contemporaneously in New Zealand the market reaction to unexpected earnings per share and dividend per share categories were the main effects examined. The study examined whether the market reaction to unexpected earnings and dividend change was moderated by company size. It also examined two surrogates for quality, (1) reliability (audited versus unaudited earnings announcements), and (2) the timeliness of the earnings release.

A sample of 147 companies listed on the New Zealand Stock Exchange was selected, with up to 798 announcements from 1982 to 1987 available. Sub group analysis was also undertaken. This study used daily data therefore a more precise examination of the market's reaction was possible compared to previous New Zealand research literature on this topic.

Other novelties for New Zealand based event study research included the examination of the effect of both interim and final earnings announcements and the employment of a more refined dividend change variable to better capture any possible signal from this variable.

A surrogate of quality used in this study was the audit status of the earnings announcements. All interim announcements were unaudited in this study whereas final preliminary earnings announcements were announced audited or unaudited and the market reaction to the impacts of non-qualified audited announcements was considered. The market to the timeliness of earnings announcements was also examined. Both factors have not been previously researched in a New Zealand setting.

Consistent with previous research, the main driver of the explained portion of abnormal returns around the announcement date was found to be unexpected earnings. The results showed a positive abnormal returns reaction to unexpected earnings for both interim and final announcements.

The impact of the change in dividend variable was at best marginal around the time of announcement. There was some limited evidence that dividend change moderated the unexpected earnings/abnormal return relationship over a longer event window. In contrast to previous research, company size did not appear to moderate the Unexpected EPS results at the time of the announcement however a size effect was noted over a longer event window and smaller firms' post earnings drift was typically higher than larger firms. The presence of an audited announcement did not appear to have any impact on the market's reaction to unexpected earnings. There was some evidence that timeliness moderated the unexpected earnings relationship/abnormal returns relationship. Note that the ANOVA results showed the inclusion of other variables beyond Unexpected EPS provided little additional explanation as measured by R2.

Finally, the results also offer some further evidence of a post earnings drift effect in New Zealand.