By LARRY SWEDROE
Nobel Prize-winning economists Franco Modigliani and Merton Miller are perhaps most famous for their proposition that, in the absence of taxes (and a few other assumptions, such as efficient markets), it does not matter what capital structure a company employs to finance its operations. They theorised that the market value of a firm is determined by its earning power and by the risk of its underlying assets, and that its value is independent of the way it chooses to finance its investments or distribute dividends. Of course, in the real world, companies and investors face taxes, which gives corporations an incentive to use leverage (due to the tax deductibility of interest), at least to an optimal level. The Modigliani/Miller theorem is often called the “capital structure irrelevance principle”. In other words, without considering taxes, whether a company uses cash to pay a dividend, repurchases shares or makes an investment, its cost of capital (and thus the return to shareholders, the flip side of the cost of capital) is unchanged. This has been the operating model in finance for more than 60 years.
Supporting evidence for the capital structure irrelevance principle was provided by Philip Straehl and Roger Ibbotson, who examined this issue in their study The Long-Run Drivers of Stock Returns: Total Payouts and the Real Economy, which appeared in the third quarter 2017 issue of CFA Institute’s Financial Analysts Journal. They observed: “Buybacks have a fundamentally different on the return generation process than dividends do. Although payouts via dividends increase the income return, buybacks increase the price return (per share) because a buy-and-hold investor’s share in a company is increased.”
Straehl and Ibbotson explained that traditional return models, like the dividend discount model (DDM), are often wrongly applied in practice because investors tend to combine current dividend yields with historical per-share growth rates when forecasting expected returns. This can lead to underestimating forward-looking per-share growth because buybacks increase a buy-and-hold investor’s proportion in the company over time. In other words, buybacks increase per-share growth. Examining data from 1871 through 2014, their findings led Straehl and Ibbotson to conclude: “The total payout model represents a viable alternative to such traditional supply models of stock returns as the DDM, providing a framework for deriving macro-consistent forecasts of long-run stock returns as well as superior forecasts of short-term expected returns.”
It’s always good to see that the empirical evidence supports the economic theory. With that said, Miller and Modigliani (1961) suggested that dividend changes may reveal managers’ information about the firm's future earnings prospects to investors—corporate managers view dividends as a persistent commitment to pay out cash, and thus dividends are used to signal information as to future earnings, and changes to dividends are made only when anticipating a permanent change in earnings.
Charles Ham and Zachary Kaplan contribute to the literature on dividends with their May 2022 study, Dividend Levels (Not Changes) Signal Future Earnings, in which they argued that “managers adjust dividends to a level they will be able to sustain moving forward, so the dividend level conveys information about permanent earnings. While earnings vary from year-to-year, the dividend level provides investors with managers’ expectation of cash inflows available to fund the payment.” They tested this hypothesis by regressing future earnings on the dividend level as well as extensive controls for earnings expectations based solely on information in financial statements. They also examined the implications of the dividend level for expected returns “because if the market assigns a high discount rate to a firms’ earnings, we would expect that firm to have a high dividend yield and also generate higher returns in the future.”
Their data sample covered the period 1972-2020 and included stocks on the NYSE, AMEX and Nasdaq (excluding financials and utilities). They controlled for variables including firm age, asset size, asset growth, accruals, leverage and retained earnings. Following is a summary of their findings:
The dividend level has incremental information about future earnings for at least five years—dividends communicate information about profitability.
The earnings information is incremental to that in future earnings released after the dividend declaration but before the ultimate future earnings date — dividend levels from year have information about earnings in year +5 incremental to the information in earnings itself over the years +1 through +4. This provides strong evidence that dividends convey long-horizon earnings information to investors.
Including dividend changes in the regression did not affect their findings—the dividend change’s information content is largely subsumed by that of the dividend level.
Consistent with the permanent earnings information embedded in dividends rather than the availability of cash that can be paid to investors, dividends produced stronger effects than repurchases.
Dividend levels have stronger associations with future earnings following dividend increases — when a firm increases the dividend, it renews its commitment to the new dividend level, increasing the informativeness of the dividend.
The dividend level is not associated with future earnings following a dividend cut, when the firm breaks its commitment to the dividend level.
Dividend levels convey more information about future earnings when recent returns have been negative, consistent with dividends providing information about sustainable earnings as opposed to earnings growth.
Higher future earnings for higher-yielding firms translated into higher returns—high-yielding firms outperformed those with low yields, with the top-quintile of dividend-yielding firms outperforming those in the bottom quintile by a significant (at the 1 percent confidence level) 0.44 percentage point per month. The outperformance was exclusively during periods of negative market returns—their outperformance helped mitigate market risk and did arise from assuming it. Yield had a negative association with future returns when market returns were positive.
The dividend yield had stronger (weaker) implications for future returns when there had been a recent dividend increase (decrease).
The outperformance of high-yielding firms was concentrated in months in which returns over the most recent six months were negative — high-yielding firms with negative recent returns experienced a bounceback in price as those sustainable earnings were reported.
Consistent with more transitory information in returns, momentum strategies earned lower returns in dividend payers.
Dividend payers tend to have high operating profit, low asset growth and high book-to-market (three of the five factors in the Fama-French five-factor model), so we should expect them to have higher returns than non-payers because of their firm characteristics. While high-dividend payers significantly outperformed, low-yielding firms (quintiles 1-2) performed similarly to non-payers.
Their findings led Ham and Kaplan to conclude: “Dividends have significant information about long-horizon earnings, incremental to that in a host of financial statement items.” They added: “Dividend levels convey sustainable earnings to investors and dividend changes convey a shift in that level.”
Investor takeaways
There is nothing magical about dividends because they are the financial equivalent of buybacks. However, Ham and Kaplan showed that while dividends and repurchases are similar from a capital structure perspective (i.e., both remove cash from the control of managers), from an information perspective they differ because the commitment to the dividend provides information about expected earnings, while repurchases tend to be more transitory in nature. They also showed that high dividend payers (which tend to be value stocks) also have defensive characteristics — they tend to outperform when market returns are negative but underperform when market returns are positive.
I would add that while the structural shift in payout policies that occurred in the 1980s, favouring more tax-efficient share buybacks over dividends, clearly has impacted how investors receive returns, it has not impacted the total return investors receive. If it had, the lower dividend payouts we have been experiencing would have led to lower returns—which has not been the case. In other words, the shift in payout policies (the form that investors receive returns) has changed, but the total return investors receive has not been impacted by that shift.
It is also important to note that dividend-paying strategies negatively impact not only tax efficiency but diversification as well because today only about 40 percent of companies pay dividends. That is why my firm, Buckingham Strategic Wealth, does not consider dividends in investment strategies. Instead, for investors seeking higher returns, it focuses on adding exposure to factors that have shown premiums that have been persistent, pervasive, robust to various definitions, survive transactions costs and have risk or behavioral explanations for why we should expect the premiums to persist in the future. Those factors are size, value, momentum and profitability/quality.
© The Evidence-Based Investor MMXXIV. All rights reserved. Unauthorised use and/ or duplication of this material without express and written permission is strictly prohibited.