Review of Quantitative Finance and Accounting, 17: 361–376, 2001
2001 Kluwer Academic Publishers. Manufactured in The Netherlands.
Prior Information and the Market Reaction
to Dividend Changes
ROGER J. BEST
Department of Economics and Finance, Harmon College of Business Administration, Central Missouri State
University, Warrensburg, MO 64093. Fax: (660) 543-8465
RONALD W. BEST
Department of Accounting and Finance, Richards College of Business, State University of West Georgia, Carroll-
ton, Georgia 30118. Fax: (770) 836-6774
Abstract. We investigate two hypotheses regarding the information content of dividend change announcements.
The ﬁrst is that the “importance” of information signaled by a dividend change depends on the reliability of earnings
forecasts existing before the dividend announcement. The second hypothesis is that the stock price reaction to
dividend change announcements is related to earnings forecast error as of the time of the dividend announcement.
Our results reveal that dividend increases convey more information for ﬁrms in which ﬁnancial analysts least
accurately predict earnings. The results also indicate that dividend increase and decrease announcements provide
market participants with information which, on average, allows them to differentiate between ﬁrms on the basis of
future earnings realizations. These differential information effects are shown to be robust to price, size, dividend
yield, and overinvestment effects.
Key words: dividends, analysts’ forecasts, signaling
JEL Classiﬁcation: G14
The information content of dividends has been the source of debate since Pettit (1972) and
Ball and Watts (1972). Many empirical studies document a signiﬁcant stock price reaction
to dividend changes. This power of dividend changes to move prices is often interpreted as
evidence that dividend announcements provide information about the ﬁrm’s expected cash
ﬂows (e.g., Pettit, 1972; Charest, 1978; Aharony and Swary, 1980; Asquith and Mullins,
1983; Eades, 1982; Brickley, 1983; Lipson et al., 1998). Alternatively, Bajaj and Vijh
(1990, 1995) show the existence of dividend yield, price, and size effects associated with
dividend-change stock price reactions. They interpret these effects to be evidence of dividend
clienteles. Further, Lang and Litzenberger (1989) ﬁnd a differential market reaction across
ﬁrms segmented by Tobin’s Q, which they use as an indicator of the expected proﬁtability of
future investment. They interpret the results as supportive of an overinvestment hypothesis
and inconsistent with cash ﬂow signaling arguments.