Many authors have used some variables to examine the predictability of returns. These variables were price/yield ratio, book value/market value ratio, defult spread, term spread, interest rate, historical yield, and dividend yield (Aydoğan and Güney, 1997; Kendall, 1953; Fama ve Schwert, 1977; Fama and French, 1988; Fama and French, 1992; Kothari ve Shanken, 1997; Hodrick, 1992; Rapac and Wohar, 2005; Daniel and Titman, 2006). Utility maximization of the investors who prefer to invest their savings in stock exchange markets has been analyzed by many studies. Utility maximization can be regarded as same as maximization of returns.
But discussion on these variables which used to forecasting stock return is continued. In the literature the dividend yields of stocks correlate with their total returns. Theoretically stock return related to equity earnings and yearly dividend income. Dividend yield or dividend/price ratio simply shows how much dividends a stock is paying off for the price of the stock (Korukoglu and Korukoglu, 2005: 48). In literature dividend yield or dividend/price ratio is calculated by dividing dividends per share to stock price in definite period (Nielsen ve Olesen, 2000).
The academic literature on dividend yield or dividend/price ratio is very wide. Lewellen (2004) used financial ratios like dividend yield for predicting aggregate stock returns in the preiod 1946-2000. In the result of his study book-to-market and the earnings-price ratio predict ratios predict returns during the period 1963-2000. The evidence remains strong despite the unusual price run-up in recent years. Lettau and Ludvigson (2002) used data on aggregate consumption and measures of the dividend payments from aggregate wealth. This study’s findings imply that both the market risk-premium and expected dividend growth vary considerably more than what can be revealed using the log dividend-price ratio alone as a predictive variable.
Many investors try to create portfolios that give the maximum return using various criteria. Among these criteria, one of the most important is the dividend yield which is the main concern of this study. Instead of examining the dividend yields of each individual securities, we have examined the relationship between index yield (ISO 100 and S&P 500) and dividend yield of index to prove this relationship. Because of this, we analyzed the relation of dividend yield and index yield ratios in 1 month, 3 months, 6 months, 1 year, 2 years, 3 year, 4 year and 5 year periods after dividend yield calculated. In this study we used geometric yield of ISO 100 and S&P indexes. Analyze period is between January 1986 and March 2008. January 1986 is base year and one hundred after these index yields revised by inflation.
We made three groups of dividend yields of ISE 100. These are low (DY<39), medium (99). We used regression analysis for the relation of dividend yield and index yield. In these regression equation dividend yields are independent variable and revised index yields are dependent variable in 1 month, 3 months, 6 months, 1 year, 2 years, 3 year, 4 year and 5 year periods. Same method used for the S&P 500. But dividend yields are not grouped, because dividend yields fluctuated in 1.07 and 4.01, these yields are scatter in 1 and 2.
In these study we used estimation (regression) model of stock yields which used by Stambaugh (1986, 1999), Mankiw and Shapiro (1986), Nelson and Kim (1993), Campbell and Yogo (2006), Lewellen (2004).
Our results can be summarized as follows. Dividend yield of ISE 100 is distributed between 0,62 (February 2000) and 20,89 (February 1989), S&P 500 is distributed between 1,07 (September1989) and 4,01 (October 2006). The results show that the dividend yield which was previously assumed to affect the returns of stocks has been found to be an important decision criterion for stock investments.
