Analysis of dividend announcement effects on stock returns: An application on Borsa İstanbul companies

As an attempt to explain dividend policies of firms, many theories and empirical models have been emerged in the finance literature. Although there is an intensive amount of studies on this subject in the literature, still no consensus has been reached among academic circles on this subject. In other words, "Dividend Puzzle" has not been resolved yet. There are mainly two approaches when looking at studies which are examining effect of dividend distribution policies on company stocks. First approach argues that there is no relationship between dividend policies and the market value of a firm while the other approach defends that a relationship can be built for dividend policies and the market value of a firm. The aim of this thesis is to examine the effect dividend announcements on stock returns by using companies data registered in BIST. For this purposes, a total of 548 dividend announcements from 136 companies have been used. The effect of those announcements on stock prices has been tested by regression analysis using event date windows created around announcement date. Analyses have been carried out for each cash dividend, stock dividend and hybrid dividend distribution forms. According to results of cash dividend analyses, there is a significant and positive relationship between dividend yield and abnormal returns for the days after and including announcement date. This relationship starts at the day of announcement, reaches its peak during the first two days, and continues at least 15 days. Dividend yield per se is a better indicator than nominal value of dividend for investors. Investors prefer stocks with high dividend yields. Thus, those stocks have more returns compared to market index of BIST 100 returns. However, significant relationships could not be reached by stock dividend and hybrid dividend analyses. Investors are indifferent to stock dividend and hybrid dividend announcements around announcement date. In all analyses, any significant relationships could not been set up in the days before the announcement date. Hence, it has been concluded that dividend information which is an important information that can effect stock price has not been leaked from inside of the company to market prior to announcement date and hence there is no insider trading. Key Words: Dividend, Dividend Policies, Dividend Yield, Announcement Effect, BIST