|dc.description.abstract||Dividends are the portion of the company’s earnings paid to the shareholders. Many researchers have used different estimation techniques to estimate the smoothness of dividends, determinants of dividends and the impact of dividend announcements on share prices. They used these estimation techniques in order to understand better the reason that companies smooth their dividends, be able to know what determines the dividend decisions and determine if the market reaction to the dividend announcements can be predicted. However, these investigations have been unsuccessful in finding an agreement on the above three issues. Furthermore, the concentration of such empirical investigations has been on the developed markets; rather few studies have been conducted on the developing markets such as Jordan.
This study has tried to fill three main gaps in the previous literature by empirically investigating the smoothness of dividends, the determinants of the two dividend decisions and the impact of dividend announcements on share prices. In addition, different empirical methods have been used in the three empirical chapters: firstly, fixed and random effects techniques are used to check the smoothness of dividends; secondly, the results of three estimation techniques (OLS, Tobit and Heckman’s simultaneous technique) are compared to estimate the determinants of the two dividend decisions; finally, the standard OLS event study methodology is compared with a hybrid method (OLS/EGARCH) to check the impact of the dividend announcements on share prices.
The results of the first empirical chapter confirmed that the non-financial Jordanian companies smooth their dividends in a moderate rate. In addition, our results contradict the signalling theory; we find that large companies smooth their dividend faster than small ones. Furthermore, in line with the agency cost theory; low leveraged firms smooth their dividends faster than high leveraged firms. Also, our results confirmed that highly profitable companies smooth their dividend more and this comes in line with the signalling theory. The second empirical chapter concluded that Hechman’s model is the better technique to estimate the determinants of the two dividend decisions. The main determinants are profitability, growth, leverage and size. The last empirical chapter showed that the Hybrid method provided a higher level of significance and concluded that dividend has no impact on share prices on the dividend announcements day. However, the results suggested that there was a significant information leakage prior to the announcements. In addition, the market reacts significantly one day after the dividend announcements and the third day.||en_US