Memoirs of a Money Manager: The Case for Dividends

Payday never gets old. When Joe Plumber gets his paycheck, he feels that a happy day has come. Investors, similarly, count their happy days when dividends arrive in their mailboxes—or more likely their brokerage accounts. How do investors evaluate companies to maximize those happy days?

THE ROLE OF DIVIDENDS IN TOTAL EQUITY RETURN

Dividends are the harvest investors collect from their investment, like bees make honey from nectar and farmers gather crops on autumn days. Dividends contribute a significant portion to total stock return. Dividend return is the humble giant in the stock return arena, as Dr. Robert D. Arnott, the former editor of the Financial Analyst Journal, pointed out. Years ago, he documented the investment return statistics of the U.S. stock market from 1802 to 2002 (see chart below), which delivered an annual return of nearly 8%. The breakdown of the return showed that dividends returned 5% per year during the 200-year period, dwarfing the combined returns of inflation, price multiple expansions (or change in valuation) and dividend growth. Another study, based on Professor Robert Shiller’s data from 1871 to 2010, also concluded that dividends accounted for approximately 50% of total stock return.

Institutional and novice investors alike have been stunned when presented with the data of compounded dividends. This dividend effect is often forgotten during the euphoria of later-stage bull markets.

Dividends don’t normally show their significance in short-term trading accounts, but, when total return data are dissected by decade (see S&P 500 Price Return and Dividend Return by Decade, 1871-2010), its importance becomes obvious.

Price movements have had their heydays, but only in a few brief periods. They occurred during the robust bull market cycles of the 1920s, 1950s, 1980s, and 1990s. Otherwise, dividend returns, like a stabilizer, were the pillar of total returns regardless of the market situation. Dividends were even more valuable during market slumps, such as the 1930s and 2000s. During these periods they were the investor’s airbag, acting to cushion severe price losses.

The dividend’s significance isn’t unique to the U.S. equity market. The authors of the monumental investment book Triumph of the Optimists calculated and documented world investment return history over 101 years. They arrived at similar conclusions that researchers of the U.S. market did: no matter where you invest in equities, dividends matter.

DIVIDENDS VERSUS EARNINGS

Corporate accounting and reporting practices have long been criticized for distorting economic reality and giving management the power to massage financial information and boost short-term earnings. Potential distortions include unearned revenue, under-reported expenses, exaggerated non-recurring items and undisclosed liabilities. In addition, accounting rules allow industries to calculate “profit” differently and tax rules can vary depending on the industry. Finally, these rules have changed over time and vary by country. Thus valuation models built on earnings hinder reliable comparisons of companies over time, across industries, and between countries.

Since dividends are cash that has been paid to the investor, they can be compared across companies, industries and countries. We believe that dividend-based company valuation comparisons provide a superior analytical basis with fewer differences of opinion than the use of earnings.

The use of dividends rather than earnings in equity valuation is a disadvantage when evaluating companies that do not pay cash dividends. Old fashioned security analysts would say not to bother with such companies because they are speculative and not investment grade. While there are exceptions, there is more than a grain of truth in this observation.

DIVIDENDS VERSUS BUYBACKS

It has been argued that the recent tendency for many companies to not pay cash dividends but rather to buy back their common equity in the market is a superior method of returning money to shareholders. This view holds that buying shares increases the earnings-per-share (by decreasing the shares outstanding) and, assuming the price-earnings ratio is unchanged or higher, the stock price goes higher. These wonderful gains are supposed to benefit the shareholder through unrealized capital appreciation with no tax liability, whereas the receipt of dividends incurs a tax. An additional benefit to the shareholder of stock buybacks is that the shareholder can determine the timing of tax obligations by choosing whether to sell or hold, whereas shareholders cannot control the timing of dividend payouts.

The arguments for the superiority of stock buybacks from the shareholder point of view are compelling, but noticeable drawbacks emerge when we examine these ideas from the enterprise point of view. First, stock repurchases are not necessarily a good signaling device to shareholders about the future prospects of the business compared with year-after-year small but consistent increases in cash dividends. The decision of company management to initiate or increase cash dividends is an implicit forecast of corporate confidence about its future.

Stock cash dividends are much more predictable than buybacks. Stock buybacks may be announced but never completed. The cash used to purchase shares in the marketplace is residual from operations after management has budgeted for capital expenditures, potential acquisitions and cash dividends, if any.

This leads to some questions: What about the timing of stock buybacks? What reason do we have to believe that corporate management is any better than the average market timer and is just as likely to buy high and avoid buying low? The evidence is absolutely none. Indeed, stock buybacks tend to lag the major equity market indexes; corporate management behaves like novice investors, buying in rising markets and avoiding purchasing in bear markets. A study by Legg Mason found that:

“...the year-to-year changes in capital expenditures and dividends are much more modest than those for M&A and share repurchases. Indeed, M&A and buybacks follow the economic cycle: Activity increases when the stock market is up and decreases when the market is down. This is the exact opposite pattern you’d expect if management’s primary goal is to build value.” [Mauboussin, June 11, 2012 Legg Mason]

Our view is that for the purposes of valuing public companies, the amount, growth and history of cash dividends is a more useful tool than earnings when we wish to compare companies over time, across industries and in different countries. Earnings are useful when dividends are not being paid, but care should be taken to ensure that the management is reinvesting the cash flow above the cost of capital. When companies are engaged in non-strategic mergers and acquisitions and stock repurchase programs, they are signaling that the existing business has little growth opportunity. In such cases we would prefer management return the money to shareholders through cash dividends rather than try to manage an acquired business or market time their stock buybacks.

DIVIDENDS AND EQUITY VALUATION

We have looked at dividends from the perspective of equity investor returns, but dividends are the backbone of stock valuation. This concept was pioneered by John Burr Williams in his 1938 text The Theory of Investment Value. Dr. Williams articulated that a stock price, unlike a painting or an ounce of gold, should be evaluated in terms of its future cash flows. In 1962, Professor Myron J. Gordon expanded the idea to create the well-known dividend discounted model (DDM), or the Gordon Growth Model, which judges the value of a stock on a future stream of discounted dividend distributions.

A special case: When the model’s dividends grow at a constant rate (g) and investors expect to receive a fixed rate of return (r), then the model is collapsed to an elegant equation:

If we shuffle the components around both sides of the equation, the return on stock (r) is restated as dividend yield (D/P) plus dividend growth (g).

This simple equation offers a convenient way to estimate a share’s total return, which is the sum of dividend yield and dividend growth. Dividends, we know, are the residuals of corporate earnings. Their nature is slow, stable, and more predictable, and often predetermined by company management. They are the antithesis of frenzied, volatile, mercurial Wall Street.

Stock market prediction has been a popular pastime since trading was created. Now, with this equation, forecasters could play with it and look for earnings, instead of dividends, to get a more predictable result. What could be done is to replace dividends (D) with earnings (E) and a payout ratio (b). Such changes create a whole paradigm shift to focusing on the earning power that drives stock values.

Around the time when the Gordon model was born, Professors Modigliani and Miller engineered a theorem, dividend irrelevance, which further muted the importance of dividends. The theorem shaped corporate behavior, encouraging firms to fully utilize profits to enhance shareholder value, such as conducting mergers and acquisitions and share buybacks to increase the returns on equity. Since then, dividends have fallen out of investors’ favor.

But there are downsides to using earnings in valuing stocks. First, the stock valuation model, as we discussed earlier, is constructed from known dividends, rather than unknown earnings. Despite the earning power to predict future returns, earnings are often subject to wide manipulations. In the Financial Times September 18, 2014 blog, posted by Andrew Smithers, he plotted the profit gap between published profits and profits that were reported in the national accounts. The discrepancy was enormous: American companies’ published profits were three times larger than the profits reported through national account systems. The two figures were similar until 2000, when the published profits began to rise much faster.