Putting the dividend-price ratio under the microscope

We analyze the time-series properties of the dividend-price ratios (DPRs) of 11 developed countries since the early 70s. Despite its frequent use in research as a valuation method for stock prices and a determinant of stock returns, previous studies suggest that there is mixed evidence of the time-series properties of DPRs predicted by economic theory. We argue that this mixed result is attributable to the sample size used in previous studies. Here, we have opted to implement the panel data approach (i.e., N>2) to increase the total size of observations rather than relying on the traditional method (i.e., increasing the size of T). In this way, we can increase the total number of observations without increasing the likelihood of structural breaks. For this purpose, we implement the panel unit root test taking account of cross sectional dependence, and obtain clear evidence of stationary DPRs. Thus, we conclude that a significant one-to-one long-run relationship exists between stock prices and dividends for the countries in our samples.