In the past, the dividend yield was considered a strong predictor for equity returns. For example, The Gordon Growth Model used to determine the “true” value of a stock based on a future series of dividends that grow at a constant rate discounted to the present time. There is a large amount of literature exploiting the properties of dividends and dividend yields to better understand the fundamentals of asset pricing both in the time series and cross-section.

However, academic research started to show deteriorated results in recent years. This also confirms authors Gray and Vogel in the “Dissecting Shareholder Yield“. The authors state that: “We confirm on a newer dataset what other research has found; dividend yield no longer works, but more complete measures of shareholder yield hold promise.” The problem of dividends is that they are not the only way that companies are able to distribute their excess cash back to the shareholders. The other possibility of how to distribute the cash is a stock repurchase. The stock repurchase is a simple operation, where the company buys back its own shares from the marketplace. This results in a reduced number of existing shares, making each worth a greater percentage of the corporation. Recently, this has become a favorite way for firms. Adding the fact that there is growing evidence that stock repurchases have substituted dividends in the last 15-20 years, it is clear that this topic deserves attention from the academic world.

This paper suggests a new combined ratio, the “net payout yield”, combined from both dividends and stock repurchases. This results in a very strong predictor of future equity performance, which can be even used in a profitable trading strategy that goes long a portfolio of high yield stocks and short a portfolio of low yield stocks while rebalancing these holdings on an annual basis.

* For those interested in systematic quantitative value factor ETF implementation, here is a link to the Alpha Architect Quantitative Value ETF (strategy background), our partner. *

Although for a long time, dividends were considered to be a strong predictor for a stock’s health, the recent research shows that this does not hold anymore, and potential investor has to use another predictor. In the past, the high dividend yield showed that a stock is cheap compared to the stream of cash going back to shareholders. Naturally, it would be logical to use the same logic but adjust the predictor to the actual situation. Net payout yield enhances this ratio because it also considers alternative forms of cash redistribution.
Research proved that the cross-sectional relation between total payout yield and returns is more distinct than the relation between dividend yield and returns. This conclusion is reinforced by Fama-MacBeth regressions of returns on beta, size, book-to-market, and yield variables in the paper. In these cross-sectional time-series regressions, dividend yields show an insignificant association with returns, whereas the payout measures exhibit highly significant associations with returns. Moreover, payout yields show no significant change in their dynamic properties, and consequently, their predictive ability remains intact across various time periods. It was also proved that payout yields exhibit significant out-of-sample predictability, whereas dividend yields do not. Last but not least, the trading strategy based on this idea is profitable and results in portfolios with negative market betas and negative loading on the size factor, suggesting that these returns are not likely to be explained by standard risk measures.