I recently wrote a post called Why Dividends Matter in which I attempted to articulate why dividends are significant, as opposed to simply looking at total return (dividends plus price appreciation). I was inspired by commentary by Larry Swedroe, a well-known advisor and columnist who believes that dividend-focused investors are misguided in their specific preference for cash distributions.
Dividend investors often say that one of the appeals of this approach is that you are not depleting your capital stock when you receive income. During a down market, for example, investors who sell shares to provide income can end up selling an ever-larger fraction of their portfolios to maintain their income streams. The total-return advocates talk about selling shares as a form of "self-created" dividend. The total-return folks think the dividend enthusiasts are creating an artificial distinction. Larry makes the case as follows:
"This whole idea about 'not having to sell assets' is why dividends are good makes no sense whatsoever. That's just basic finance or math. A dividend is the equivalent of disinvestment in a company's stock (the value falls) which you can replicate by selling the number of shares needed to get the same cash as the dividend - or you can reinvest the dividend and that puts you in exactly the same place as if the company never paid them out."
I see the basic argument here, but I also see a fundamental problem. It is absolutely true that when a company pays a dividend, the stock price drops by the amount of the dividend. What is missing in this logic however is clear if you think in terms of a stock representing a percentage of ownership in a company. If you own a certain number of shares, you own a specific percentage of the company's assets and have a proportionate claim on future earnings. Before and after the company pays a dividend, your percentage ownership stays the same. If you sell shares to generate income, your percentage ownership in the company falls over time. This is patently true, but seems to get missed. In other words, dividends are not equivalent to a "disinvestment" in a company's stock in terms of your percentage ownership and claim to future earnings. On a mark-to-market basis, they are equivalent. Thinking in terms of exactly what an investor owns, they are not the same.
A company's value is the sum of all of the assets that it owns or has some form of claim on. This includes physical assets, know how, relationships, goodwill, human capital, intellectual property and cash. As a shareholder, you own the fraction of the firm that your shares represent and you have a claim on the same fraction of all future earnings. When a company pays a dividend, the dividend is paid out of cash. The market value of a share drops by an amount equal to the dividend because the company's cash holdings are smaller by that amount. An investor's stake in all of the other assets of the firm remains the same. You own the same fraction of the assets that make a firm a firm. If you sell shares, you are selling off your share of all assets of the firm proportionately. You own less of the factories, the intellectual property, the supply chains, etc. These really are different things.
If a company pays a dividend and investors reinvest that dividend immediately, the value of the position before and after the dividend payment and reinvestment are the same, but what you actually own is different. After the transactions (dividend+reinvestment), cash on hand is a smaller part of the total value of the company and you, the investor, own a larger stake in the other classes of assets.
Quality of Earnings
The "quality" of earnings refers to how consistent and reliable the earnings from a company are. A company with low-quality earnings is much more likely to see its earnings in subsequent quarters be volatile. A company with high-quality earnings is more likely to provide a consistent future stream of earnings. A 2009 paper titled What Do Dividends Tell Us About Earnings Quality by Douglas Skinner of the University of Chicago and Eugene Soltes of Harvard Business School concludes that companies that pay dividends have a higher quality of earnings than those that do not. This paper, published in the Review of Accounting Studies in 2011, is summarized thus:
We find that the reported earnings of dividend-paying firms are more persistent than those of other firms and that this relation is remarkably stable over time. We also find that dividend payers are less likely to report losses and those losses that they do report tend to be transitory losses driven by special items. These results do not hold as strongly for stock repurchases, consistent with them representing less of a commitment than dividends.
I find this conclusion to be very intuitive. Companies that pay dividends know that their investors care about consistent performance. For this reason, they are more likely to make decisions that will result in reliable earnings through time. Managers at non-dividend payers know that their investors are concerned only with price appreciation and price gains will occur only when there is a belief that the earnings per share (EPS) are rising. The only way for the EPS to rise is for the company to invest in new opportunities. New investments may or may not be as attractive as the company's current business, but management really has no choice.
Dividends also tend to reduce agency problems between shareholders and corporate managers. Managers are supposed to act in the best interests of shareholders, but changes in executive compensation over the last several decades have created some conflicts of interest. These issues, studied over decades by Harvard's Michael Jensen and USC's Kevin Murphy, are significant and widespread. When executives are largely compensated with stock options - and we have seen a major increase in this component of compensation over the past twenty years - their interests are less aligned with those of shareholders. Specifically, managers with large option grants have an incentive to drive the price of shares upwards as quickly as possible and this gives them a disincentive to pay dividends. Dividends result in lower share prices while buybacks result in higher share prices. The value of stock options increases with a buyback and decreases with a dividend. Which one will managers choose?
Studies of the efficacy of share buybacks as a way to increase shareholder value tend are not inspiring. Companies often engage in share buybacks as part of "earnings management," a process by which a company makes its earnings reports appear more consistent and robust.
In Part 1 of this post, I focused on the fact that the ability to predict the component of returns associated with dividends is much higher than that from price gains. In Part 2, I have briefly summarized some of the other research on what dividends mean for investors. I do not believe that investors should avoid companies that don't pay dividends. There are perfectly good reasons why a company chooses not to do so. Young companies that are still developing their business models need their earnings to grow. There are industries in which it is required that a firm invest heavily in research and development and thus may need to reinvest earnings for this purpose. The point, however, is that dividends are an important and meaningful piece of information for investors. In addition, I have attempted to articulate why there is an obvious difference between selling shares to provide income and using dividends for income.
In Part 3 of this post, I will discuss the issue of the role of dividends in predicting future returns in more detail and with additional research.