This paper defends stock dividends as significant and important investment portfolio components. Dividends and related financial dynamics are covered in several terms: "bird in the hand" psychology, dividend signaling and average returns.
Given that even professional investors are subject to emotional investing decisions as well as confusing skill with luck, dividends provide a steadying effect on portfolio returns. Dividend payments eliminate some of the buy/sell decisions that can haunt investors and drive them to emotional and self-sabotaging investment moves, especially in a volatile market.
Generally speaking, the more an investor is subject to emotional extremes and money worries, the more he should rely on dividends. As an investor gains experience, perspective and in-depth understanding of the market, he can smoothly transition to higher-risk small caps without excessive worry -- and without excessively bad trading decisions -- when it comes to overall returns.
Bird in the Hand
One prominent aspect of investor psychology is the allure of "a bird in the hand is worth two in the bush." I believe the sentiment encompasses two related mindsets:
- Uncertainty of capital gains that have to be realized vs. after-the-fact certainty of dividend income. ("After-the-fact" because there is no guarantee a dividend will continue.)
- An irrational but pernicious notion that dividends are "given" while capital gains are "cashed in" for a decreased stake in the company.
The first point is well-known. Capital gains need to be realized, meaning that the underlying security needs to be sold. If a liquidity issue or a shortage of "greater fools" handicaps that sale, the gains will dwindle until a buyer is found. That may not be until the investor is in the red. Such a sell-squeeze is very unlikely to be an issue with an index fund or highly liquid and well-known stocks, but it is something that investors keep in mind when dealing with smaller, less liquid stocks. By contrast, dividends can, in effect, salvage capital losses. This is a substantial allure of dividends that enhances their appeal relative to a similar-priced stock operating entirely by price changes. It is entirely possible for an investor to profit from dividends even if the underlying stock never gets above buying price. Naked capital gains do not have this prospect of a hedge against losses.
The second point is of course false, financially speaking. Dividend money has to come from somewhere. However, the psychological reality of dividends is a deposit into the investor's account without any action on the investor's part. An identical capital gain or "homemade dividend" would entail closing a trade, thereby losing ownership of some of the underlying shares.
A homemade dividend can be constructed by selling some stock to give the same effective realized gains and purchasing power that a dividend would deliver. For instance, an investor may want to simulate a 5% annual dividend from a $100,000 portfolio. To do so, he would sell 5% of the stock, generating $5,000 in capital to spend or reinvest as he wishes. Of course, upon doing so, the rest of his portfolio decreases to $95,000. The thing is that a "real" dividend encompasses exactly the same financial dynamics. The only difference is that the 5% (or whatever percentage) payout and corresponding portfolio decrease is completed automatically instead of at the investor's whim.
Nevertheless, dividends are perceived to be icing on the cake on top of whatever unrealized capital gains exist for the underlying stock. The implicit loss of potential income that, instead of given through dividends, might have been more efficiently used if reinvested into company operations or a higher stock price is technically accurate but psychologically distant.
One aspect of dividend investing that I find interesting is the notion of investing by dividend signaling. An increased dividend allegedly signals optimism and corresponding rising stock prices, with the opposite for a smaller or discontinued dividend. The following charts follow dividend and stock price changes for a few well-known industry leaders.
(click to enlarge)
All data for the above charts was gathered from Yahoo! Finance for each respective company.
The stock prices used are those recorded on the last trading day of a given year, usually December 28-31. These charts seek to capture a long-term trend at the expense of significant intra-year fluctuations, impacts of quarterly results and other short-term events. I feel this approach is justified since the goal is a long-term perspective. We need to see whether or not a trend plays out over a time frame that encompasses several boom/bust cycles. The 2000-2013 timeframe accomplishes this. The percentages for changes in stock price and dividends were kept independent to illustrate the magnitude (or lack of) of dividend changes with respect to prices for each corporation.
For example, assume a stock was trading at $20/share at the close of one year and $21/share at the close of a second year. In the same timeframe, the annual dividend for that stock rose from $0.50 to $1.00/share. The graphs would record a 100% gain in dividends relative to the starting dividend value of $0.50 and a 5% increase in stock price relative to the stock's starting price of $20/share.
The combined stock-dividend gain (in below table) would be calculated from a starting point of $20/share to $22.5/share (+$1/share capital gains + $0.50 + $1.00 dividend/share). A movement from $20 to $22.5 is 12.5 percent total gains. This way, readers see an allegedly powerful dividend "signal" of +100% resulted in a respectable, though relatively weak boost of 12.5% to overall returns.
Of course this approach is necessarily approximate. Results would vary tremendously if dividends were reinvested into the original stock or another security. Tax treatment, inflation and timing within a year would all significantly affect returns. Nevertheless, as a first approximation, the above-listed companies give a hint of dividend effectiveness.
For the purposes of the below table, the final 2013 price and dividend summation from 2000 to 2013 - assuming no reinvestment, tax or inflationary effects - is added, with the sum subtracted from the original 2000 stock price. This result is then divided by the original price and converted to a percentage for overall returns if an investor had bought at the close of 2000, sold at the close of 2013 and did not reinvest dividends.
P13: Stock price at close of 2013
Po: Stock price at close of 2000
Dsum: Sum of dividend payments for a given stock between 2000 and 2013
2000-2013 Total Gain, per above equation
2000-2013 Dividend Increase
Exxon Mobil (XOM)
Precision Castparts Corp (PCP)
Dominion Resources (D)
Goldman Sachs Group (GS)
McDonald's Corp. (MCD)
The above equation was personally derived, though I am sure it is present in many financial publications and studies. The table was constructed from applying the equation to Yahoo! Finance data for 2000-2013 stock price and dividend behavior for the mentioned companies.
As the above charts and data indicate, Exxon Mobil has a consistent signaling history. Taking into account the inherent errors and approximations in constructing the above information, Dominion Resources and Goldman Sachs have tentative correlations between dividend and stock price increases. However, Precision Castparts and McDonald's are difficult stocks to fit into the signaling theory. PCP recorded a very modest dividend boost between 2000 and 2013. Precision Castparts dividends doubled in value, while the stock price increased by nearly 1200 percent. McDonald's behaved in an opposite manner, with MCD dividends hinting a magnificent stock price boost in the range of 1300 percent gains. Instead, MCD stock rose a relatively small 242 percent.
Of course one could find many securities that fit the dividend-signaling pattern as well as many that don't. It must be emphasized, though, that given the "deviants" from signaling, it seems that dividend signaling is not reliable.
To summarize, the notion of accurate dividend signaling has little basis in reality. Interested readers may consult this recent study by the University of Chicago on the reliability of dividend signaling.
Relying on "average returns" in the context of large price swings common to the stock market can be an expensive mistake. This can get especially painful since market dips and other large-scale economic troubles have a pretty tight correlation. Just when an investor may need money to get him/her through tough times such as unemployment, the market dips. Now any stock sales will generate far less gains, if any, than if the investor sold during an earlier bull market or if he waits out the dip to sell again on the next stock uptick.
Sometimes, expenses and pressing financial obligations may not allow the kind of nonchalant patience needed to justify an "average return" perspective.
Dividends can help since, as mentioned earlier, they soften capital or income losses and provide relatively stable income. Though it is true that a company can cut dividends during a market downturn, a diversified dividend stock portfolio is very unlikely to see all its dividends cut. After all, companies have an interest in retaining shareholders, even (especially?) during bear markets. Though I can't claim knowledge about the inner workings of upper-management decision-making tendencies, I feel that executives are not likely to act too brashly in cutting off what they know is an enticing aspect of their stock.
For example, of the above-mentioned companies, note that Exxon Mobil maintained a steady and increasing dividend payout through the 2001-2002 and 2008-2009 recessions. McDonald's did likewise, as did several other companies not charted here, such as Boeing (BA), Caterpillar (CAT) and Loews (L).
Several conclusions come to mind:
1. Given the inevitably irrational behavior of investors, especially as regards panic-selling during recessions and bear markets, dividend income should be a part of an investor's portfolio. Realistically, constructing homemade dividends assumes a degree of rationality that real-life investors are unlikely to possess. During bull markets, dividends can be reinvested in the original or more speculative securities for greater overall gains. During recessions, dividends can provide financial cushioning for other necessities of life besides investor returns.
2. "Average return," especially in the context of a homemade dividend, is an unrealistic fallacy since there is lots of stock price variation at multiple time scales. Over/undervalued markets often present a situation in which "average" expectations will widely deviate from the financial reality an investor may face.
I won't pretend to know dividend-related economics at a professional level. Nevertheless, some research and facts of life about irrational behavior leads me to believe that, though arguably less "efficient" than capital gains, dividend stocks have a place in an investor's portfolio.