Experts Differ on the Merits of Dividend-Paying Stocks
By: Ned Piplovic,
LAS VEGAS — A good argument can be made that dividends should be required to invest in stocks to help investors maximize returns but non-dividend-paying stocks also have merit.
Dividends advocate Rob Arnott, founder, Chairman and CEO of Research Affiliates, a research-intensive asset management firm that focuses on smart beta and asset allocation, debated Professor Joel Stern about the role of dividend-paying stocks in investment decisions. Arnott argued that dividend-paying stocks outperform non-dividend-paying stocks in the long term, but acknowledged that the tendency does not apply in all instances, whereas Stern said dividends are irrelevant to the overall performance of a stock.
Stern agreed with Arnott’s premise that dividend-paying stocks are good for investors but he contended that the sole determinant of stock price appreciation is the risk class to which a particular stock belongs. Risk on investment falls into three main categories: below average, average and above average.
Electric utilities carry low investment risk that is slightly above risk-free government bonds. In contrast, food companies carry a slightly higher level of risk than utilities but still offer investors less than average risk, Stern said.
Banks typically have average risk that exposes them to a broad cross-section of the entire economy. Cyclical industries, which include steel, cement and aluminum production, as well as the textile industry, auto industry, manufacturing, industrial machinery and industrial tools, carry risk that is higher than average, Stern added.
Each level of risk is associated with an expected rate of return on investment, Stern said. Category one investments, where returns on investments are higher than expected, are the key to increasing share prices. Category two investments carry average risk and do not create any special value. Category three projects yield lower-than-expected returns on investment and offer reduced potential value, he added.
Therefore, instead of investing earnings in category three projects, companies should use that money for other objectives, including paying out dividends. All companies in the same risk class will yield an identical rate of return on average, assuming these companies are part of a well-diversified portfolio across a particular risk class, Stern concluded.
Arnott agreed whether companies are dividend-paying stocks or not is irrelevant, in theory. He said that the Modigliani-Miller Theory on Dividend Policy proved mathematically that dividends have no impact on stock-price appreciation. Instead, Arnott said the valuation of a company can be determined with information on future growth and the discount rate, which is the interest rate used in discounted cash flow analysis to determine the present value of future cash flows.
However, Arnott also pointed out that Modigliani and Miller assumed perfect capital markets, no taxes, no transaction costs, no risk of uncertainty, a fixed investment policy and that company managers act solely in the best interest of the shareholders. With all those assumptions, the Modigliani-Miller model works only in theory and is not applicable in the real world where those assumptions never apply.
In the real world, return on any equity investment depends on dividend yield, real growth in earnings per share, inflation rate and shifts in price to earnings ratios over any given period. Given these four pieces of information, it is easy to calculate the market turn or rate of return for any stock, Arnott added.
Arnott presented a chart indicating that over the past 140 years, dividends contributed two-thirds of equity total returns annually. Earnings growth contributed only one-quarter of the average annual return during the same period. The chart showed an 8.7 percent average annual equity total return, with dividends contributing more than half — 4.5 percent in real terms. However, the real equity total return is only 6.7 percent when adjusted for inflation, indicating that dividends account for two-thirds of equity total returns.
Stern contended that compensation based on value creation, rather than earnings growth, dramatically improves employees’ performance and increases a company’s value. Therefore, investments in category one projects are better uses for company’s earnings than stock options incentives or dividends payouts.
Dividend payout is not a driver of stock price, but is merely a management tool for dividend-paying stocks to signal a company’s positive financial performance, Stern said. However, Stern estimated that management decisions contribute only about 25 percent to the price of the stock. Changes within a given company’s industry contribute another 25 percent and the majority contribution to the stock price change — 50 percent — is driven by the changes in the stock market overall.
Stern disputed Arnott’s criticism of the Modigliani-Miller study that the study’s conclusion is true only in theory for efficient financial markets. After their initial conclusions assuming efficient markets, Modigliani and Miller relaxed their assumptions to reflect more real-world conditions and came up with the same conclusion that dividends are irrelevant to company performance on average and over time.
Stern, who said he formerly served as a research assistant to Merton Miller, agreed that paying dividends is good management only if category one investments are not available.
Arnott contended that in the real world dividends are necessary tools to protect shareholders against mismanagement of a company’s earnings into investments that carry higher-than-average risk. Startup companies usually carry a high risk and a high potential reward on investment.
Therefore, startup companies should reinvest all earnings, Arnott said. However, mature companies must use dividend payouts to discourage management from investing in category two and category three investments that have lower return on risk than category one investments.
Arnott stated that he does not advocate paying out all earnings as dividends. However, companies that pay dividends outperform, on average, companies that do not pay dividends, Arnott concluded.
A survey of attendees at the FreedomFest debate in July 2016 indicated that most of the audience sided with Arnott.
Regardless of the debate results, each investor should make a decision on this subject based on individual investment goals, risk tolerance and overall financial strategy.