EDITOR’S NOTE: Goizueta Newsroom regularly shares news on the school, its faculty, staff, students and alumni. But we also take pride in sharing vital pieces of research from the world-class business school faculty. Look for regular updates on published pieces of research like this one…
Title: Payout Yield, Risk, and Mispricing: A Bayesian Analysis
Authors: Jay Shanken, the Goizueta Chair in Finance, and Ane Tamayo (London School of Economics) Jay Shanken was recently named one of the World’s Best Business School Professors by Poets and Quants (2012).
Journal: Journal of Financial Economics (2011)
Of the various financial ratios that have been studied for potential insights into the predictability of aggregate stock returns, the dividend yield has received the most attention. A simple ratio in which annual dividends per share are divided by current stock price, the dividend yield tells investors how much cash flow they are receiving per dollar invested in a stock. Research has generally shown a positive relationship between dividend yield and subsequent returns; one key study in 1996, moreover, even suggested that predictive regression results that usually would be dismissed as statistically insignificant can, nonetheless, have important implications for asset allocation. In a recent paper, Jay Shanken, the Goizueta Chair in Finance, and a colleague from the London School of Economics sought to expand analysis on the subject by considering return predictability when market risk is not assumed to be constant over time. The assumption of constant risk, the researchers noted, has been a key factor in most other academic studies on the subject.
“A Bayesian analysis of return predictability with changing risk seems long overdue,” the researchers wrote, “particularly given the fact that persistent time variation in risk is one feature of the data about which there is no doubt.”
To conduct the empirical analysis, the researchers opted instead of the dividend yield ratio to use payout yield, which captures the recent propensity of firms to substitute share repurchases for cash dividends. They developed a model that can simultaneously evaluate hypotheses about predictability, mispricing and the risk-return tradeoff while also allowing for changes over time in both risk and expected return. Using this model to analyze stock returns from 1940 to 2004, the researchers set out to answer a series of questions such as: Would the beliefs and investment decisions of classical or behavioral investors change after observing the same evidence on stock return predictability? They posed the question in terms of economists Eugene Fama and Richard Thaler, asking what would happen to their expected utilities if Fama were forced to hold Thaler’s portfolio and vice versa. Fama helped formulate the efficient-market hypothesis, which posits that financial markets are rational and feature equity prices that largely reflect public information about a stock. Thaler, meanwhile, was instrumental in establishing the field of behavioral finance, which considers the role that psychological and emotional factors play in economic decisions.
In the end, the researchers’ data fit with previous studies in suggesting substantial predictability. A change in the risk premium of about 3.25 percent correlated with a one percentage point change in payout yield, the researchers found. They noted little association between yield and the conditional market variance, which led to the conclusion that yield-related predictability was dominated by a mispricing effect. It’s important for investors to consider persistence in risk in asset allocation, the researchers wrote, but changes in risk played a minimal role in the relationship between yield and asset allocation. Prior beliefs about both mispricing and the nature of the risk-return tradeoff mattered for asset allocation, according to the study results. Portfolio choices based on informative prior beliefs displayed superior performance than allocations based on diffuse priors, the researchers found, although informative prior beliefs that allowed for some mispricing did better than those that dogmatically ruled out the possibility of mispricing.
“We find that prior beliefs about mispricing matter ex ante, in the sense that an investor with a prior that accommodates mispricing experiences a substantial drop in utility if forced to hold the portfolio of an investor who presumes there is no mispricing (and vice versa),” the researchers concluded. “Beliefs about the nature of the market risk-return tradeoff similarly impact ex ante utility, as do beliefs about the persistence of ex ante risk. Different beliefs about the manner in which mispricing influences the relation between yield and expected return matter far less, presumably because the gap between those initial beliefs is reduced through the process of learning from the data.”