Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)

First Advisor

John Howe


The purpose of this dissertation is to reexamine the dividend information hypothesis and to provide additional evidence on the nature of the information revealed by dividend initiations. Evidence on (1) announcing firms' stock price reactions, (2) analyst earnings forecast revisions for announcing firms, (3) rivals' stock price reactions, and (4) analyst earnings forecast revisions for rivals will be presented. On average, the stock market reacts positively to the announcement of dividend initiations and the degree of this reaction can be significantly explained by three information proxies: dividend yield, firm size, and trading volume. Moreover, I find that dividend initiation announcement month earnings forecasts by analysts are systematically revised upward for a sample of dividend initiating firms on the I/B/E/S tape. These results are consistent with the hypothesis that a dividend initiation announcement conveys favorable information about the announcing firm's earnings prospects. However, this information may not be entirely firm-specific. It may have an industry component and therefore have a contagion (positive) effect on other firms in the industry. On the other hand, the information may imply that rivals face an unexpectedly competitive and prosperous firm. In this scenario, the dividend initiation will affect rivals negatively. The second part of this dissertation examines the external nature of the information revealed by dividend initiations. I analyze earnings forecast revisions and stock price reactions for industry rivals upon the announcement of dividend initiation. The evidence shows no stock price reactions for rivals. I also observe no significant analyst forecast revisions for rivals during the announcement month. This finding supports the hypothesis that dividend initiations reveal firm-specific information only. Cross-sectional analyses of rivals' stock price reactions and rivals' earnings forecast revisions indicate that they are not sensitive to bull/bear market conditions or growing/declining business cycles at the announcement.