Content area
Full Text
There is evidence that dividend omissions provide information about a firm's cash flows and earnings. Several studies document a negative abnormal return following dividend omission announcements (Ghosh and Woolridge [1988, 1991] and Healy and Palepu [1988]). Earnings declines typically precede dividend omissions. The traditional dividend signaling hypothesis postulates that omission announcements signal unanticipated declines in future dividends.
This hypothesis, however, overstates the observed market reaction to omission announcements. William Christie [1994] reports that the negative information effect of a dividend omission is not as severe as one would predict, given the pattern for reductions in dividends of less than 100%. A comparison of dividend omissions to dividend reductions reveals that wealth losses attributed to dividend omissions amount to only 70% of wealth losses expected from a sizable dividend reduction. Although there is little doubt that a dividend omission is bad news, the dampening valuation effect associated with omission announcements is puzzling.
We seek to resolve this puzzle by focusing on the link between earnings and dividend omissions. Consistent with the signaling hypothesis, omissions could convey two types of earnings information.
First, they could provide information about declines in future earnings. The "future earnings hypothesis" postulates that managers use dividend omissions to convey to the market unfavorable information about future cash flows. This is consistent with the idea that omissions are costly and credible signals of the firm's value.
Kalay [1980] finds that firms omit dividends when future cash flows are not expected to be sufficient to sustain the particular level of payout. Work by Lintner [1956] suggests that dividend changes are a function of long-term sustainable earnings. When managers are no longer confident that future earnings will support these payments, they omit dividends. Missing future dividend payments imply greater earnings variability after the omission, an observation consistent with the finding that omissions are followed by increases in return variance as well as greater dispersion of analysts' earnings forecasts (Sant and Cowan [1994]).
The second type of earnings information that dividend omissions convey concerns past earnings or pre-omission earnings. Healy and Palepu [1988] report that dividend-omitting firms experience systematic earnings patterns. Omissions are preceded by substantial declines in earnings, declines that do not persist beyond the year of the announcement. If earnings patterns are fully...