SUMMARY: This event study is the first to examine the price effect of dividend announcements on the Copenhagen Stock Exchange (CSE). Almost all Danish dividend announcements (98%) coincide with the release of the yearly accounting report, and we find that an unexpected dividend increase leads to a significant announcement effect of 2.6%. Controlling for simultaneous earnings announcements we find a significant announcement effect of the same order. This announcement effect is at least in line with results reported for UK and US markets, but differs from results reported for Continental Europe where the announcement effect typically is insignificant or negative.We find that an unexpected dividend decrease leads to an insignificant announcement effect of -0.9%. Controlling for simultaneous earnings announcements we find an insignificant announcement effect of the same order. This announcement effect differs from results reported from UK/US as well as results reported from Continental Europe. In UK/US the announcement effect is numerically larger and significantly negative. In Continental Europe the announcement effect is typically positive although insignificant.We also find that unexpected dividends and unexpected earnings both contain significant information in explaining the announcement effect of the yearly accounting report. Furthermore we find significant interaction effects of the simultaneous announcement of dividends and earnings.There have been published several studies regarding the announcement effect of unexpected earnings on CSE. This study complements these studies since we are able to control for unexpected dividends. Controlling for unexpected dividends this study lends support to a more optimistic view regarding market efficiency than the most recent study does.Finally, we find indications that the owner structure may explain the asymmetric announcement effect of unexpected increases/decreases in dividends. Overall, the findings are in support of a signaling hypothesis. However, there are indications that this explanation does not hold for firms dominated by large shareholders where the results are more in line with results reported from Continental Europe.
Nationaloekonomisk Tidsskrift, 2006, Vol 144, Issue 2, p. 137-168