Bridge over ocean
1 November 2013 CFA Institute Journal Review

Demographics and the Long-Horizon Returns of Dividend-Yield Strategies (Digest Summary)

  1. Rajiv Kalra

Given the preference of older investors for higher-dividend-paying stocks, the excess returns for these stocks should be positively related to increases in the proportion of older investors. The author contends that in accordance with the behavioral life-cycle model, older investors in retirement spend their accumulated wealth but prefer to spend dividend income, not capital gains. The author’s findings confirm this behavioral life-cycle hypothesis, even when he controls for a number of other factors that are widely believed to influence returns.

What’s Inside?

The life-cycle model of income and consumption suggests that as investors move into retirement, they start to consume their accumulated savings. The behavioral life-cycle model postulates that people compartmentalize their income: Capital gains are considered something to save, whereas current income is meant for consumption. The author suggests that for this reason, older investors prefer dividend-paying stocks. He hypothesizes that as the proportion of older investors increases, so does the demand for higher-dividend-paying stocks. Using all US exchange-traded stocks over a 75-year period, he shows that long-run returns of dividend-yield strategies are positively related to changes in the relative proportion of older investors in the overall population.

How Is This Research Useful to Practitioners?

Previous researchers have demonstrated that investors consider dividends and capital gains to be nonfungible; their preference for one kind of income over another may depend on where they are in their life cycle.

The author examines whether the returns on higher-dividend-yielding stocks are influenced by the proportion of older people in the population. He shows that an increase in the proportion of older investors is associated with higher relative returns for high-dividend-yielding stocks. A 1% increase in the older population is associated with a 222%, 200%, and 172% increase in the 10-year excess returns of high-dividend-yielding stocks versus low-yielding stocks excluding zero-dividend stocks, low-yielding stocks including zero-dividend stocks, and the market, respectively. These findings hold even when controlling for such variables as Fama–French factors, the inflation rate, the consumption growth rate, and interest rates.

Because of the aging population in the United States and much of the Western world, the author’s findings should be of interest to practitioners responsible for choosing investment strategies. In addition, the author argues that older investors may switch to high-dividend-yielding stocks rather than liquidating their entire portfolios during retirement. Consequently, the forecast of an asset market meltdown arising from the liquidation of retiring Baby Boomers’ investment portfolios is not reasonable.

How Did the Author Conduct This Research?

The author assumes that the relative price returns of high-dividend-yielding stocks depend on, among other things, changes in the proportion of older investors. To minimize the effects of short-term and often random stock price fluctuations and to acknowledge that older investors hold stocks for longer periods of time, he uses a moving 10-year window to test his hypothesis.

The data set includes all the US exchange-traded stocks for the period of 1937–2011 and is categorized into quintiles based on a stock’s annual dividend yield. To examine the performance of high-dividend-yielding stocks, the author calculates 10-year excess returns on three dividend-yield investment strategies: (1) return on the first quintile minus the return on the fifth quintile, excluding zero-dividend stocks; (2) return on the first quintile minus the return on the fifth quintile, including zero-dividend stocks; and (3) return on the first quintile minus the return on the market portfolio.

US Census Bureau population data are used to compute the ratio of individuals 65-years-old or older to the total population, and the change in demographics is the change in that ratio over 10 years. The author then uses ordinary least-squares (OLS) regressions and finds statistically significant positive relationships.

He uses data from Robert Shiller’s website to compute the term structure of interest rates and short-term interest rate movements over the last 10 years. Data from the Citizens for Tax Justice website are used to determine the tax clienteles as the difference in the highest marginal tax rate on dividend income versus capital gains.

To test the robustness of his findings, the author separately controls for (1) Fama–French factors, including the excess return on the market portfolio (RmRf), the return on small stocks minus the return on large stocks, and the difference in the returns of growth stocks and value stocks; (2) the change in the inflation rate; (3) the change in real per capita consumption; (4) the change in interest rates; and (5) tax clienteles. He uses multivariate OLS regressions on overlapping data (with appropriate adjustment of the standard errors) to estimate the relationship between the excess returns of the three dividend-yield investment strategies and the change in demographics in the presence of the control variables.

In all of the regressions, the demographics variable remains a significantly positive determinant of the 10-year excess returns of the three high-dividend-yielding investment strategies.

Abstractor’s Viewpoint

The author’s research is perhaps the most comprehensive study of the effect of demographic changes on the returns of dividend-yielding stocks. To ensure that his findings are not influenced by other factors, he tests the relationship by controlling for a number of factors believed to influence stock returns. Although I very much enjoyed reading the article, it could have been improved by including fewer figures, tables, and details of results, as well as by a more careful editorial review. Nevertheless, the findings of the research will be of interest to academics, practitioners, and corporate managers.

We’re using cookies, but you can turn them off in Privacy Settings.  Otherwise, you are agreeing to our use of cookies.  Accepting cookies does not mean that we are collecting personal data. Learn more in our Privacy Policy.