An empirical investigation of UK stock market reaction to dividend announcements in a complex signalling setting

This dissertation examines the impact of the dividend announcement on the share prices of a large sample of UK firms within a multiple signal setting. The dividend announcement is viewed as simply one among many signals which are emitted by firms to convey new information about company performance to outside investors. The analysis takes place (i) at the time of the dividend announcement and (ii) over a longer time span, including six months before and nine months after the dividend announcement date. The investigation uses a battery of tests which included (a) an event study test based on abnormal returns, (b) a regression-based test which directly examines the data for evidence of interaction between the dividends and earnings news, (c) an analysis of variance (ANOVA) test on the mean values of the accounting ratios of the different groups of companies and (d) a simultaneous-equation approach which attempts to model interdependency between a firm's dividend, financing and investment policies. The main findings indicate that (i) investors appear to react as though dividends convey useful information to the market, (ii) the market reaction is strongest when the dividends and earnings signals corroborate one another, (iii) the different dividend-earnings groups in the sample exhibit distinctive patterns in their financial accounting profiles in the years before the dividend announcement and (iv) the firms in the sample do not seem to choose their dividend policy independently of all other strategic decisions but instead appear to take account of their investment needs and fund-raising potential when deciding what dividend to pay.