The point of the article was that on a total return basis, it appears that certain once-consistent dividend growth investing strategies have underperformed as of the last few years. Despite numerous articles touting dividends being at all-time highs (see this WSJ article as one example), those once consistent strategies were trailing the S&P or even down on an absolute basis. Dividends themselves reflect the willingness and paying power of the underlying corporations. We have emerged from a period of necessary cash hording by companies, and the dividends remain one of the best ways to return that cash to shareholders. However, supply and demand of the stocks themselves will dictate total return. Stocks or strategies that are out-of-favor will see greater supply and less demand – which will affect the total return.
More than one commentor suggested total return is not a necessary measure for many dividend investors. Rather, current income and growth in dividends might be the utmost factor, if not the only factor. I could not disagree more.
Total return must be the overriding factor for any investor to determine wealth and income creation. Dividend yield as a stand-alone measure tells you nothing about wealth creation. A salary is wealth creation – labor for dollars. If I gave the local manufacturer $50,000 in exchange for working a year on his assembly line, and he paid me $50,000 in salary, I would have an income. However, if after that year, he shut the business down, I have created nothing for myself. I would have been better off doing something else with my dollars and my time.
The same goes for stocks. A 10% income stream from a dividend paying stock that loses 10% in price isn’t really income, it’s a return of your orginal capital that is taxed like income. Consequently, it’s actually a loss. It may not be a “realized loss,” but a loss nonetheless. Therefore, total return has to be the ultimate measure.
Otherwise, the article generated some interesting comments and questions. The one strategy that we found that maintained its robustness and consistency of late was an odd metric. We looked at companies that had consequtive years of dividend growth, and found that companies that had the least absolute dollar differential between the most recent annual payout and that of five years ago resulted in an attractive return profile. For example, the highest possible rank would go to a company that had only a five cent differential in payouts (increase by 1c/yr for 5 years). The results suggested to me, the markets favored established dividend payers, but were not paying for dividend growth. Commentor Jeff Paul, made an interesting observation/question.
I like the analysis of different metrics vs performance. I’ve read other research more recently that found dividend growth rate to be the main driver of total return, which seems to conflict with these results, since it ranked the dollar change from low to high (not high to low), though that doesn’t necessarily reflect on the percentage change. Can you rerun the strategy on a percentage basis and see if it makes any difference? … [S]ince a 10-cent per share increase might rank low on your list, but it could be a high div growth rate depending on the current dividend. You can either calculate the total %chg (as in your example) or the compound annual growth rate in the dividend. The results should be the same in terms of any patterns. The research I’ve seen suggests that the stocks with the higher DGR should have higher returns. Your example showed lower dividend $ growth amounts had higher performance.
Five-year dividend compound growth rate as one of our original strategies that we evaluated (it’s the second strategy in the original article), and it’s results while good over an extended period of time, were poor over the last three years. For that reason, I didn’t expect that percentage change in dividends would change the results dramatically, but here they are:
On a ranking of high to low, looking at companies with the highest percent change improves performance over that of a dollar change. However, in the reverse, ranking low to high (thus valuing stability over growth), the dollar measure remains the best (with percent difference only trailing slightly). Particularly note the Sharpe ratios over time. Either way, ranking the measure towards low growth beats high growth, particularly driven by the most recent years.
% Change Dividend/Shr Growth, Hi to Lo
One Final Thought
Some readers supposed I was bashing dividend strategies. In fact, I am not.
Whether a particular investing style is in- or out-of-favor in not a comment on whether it is a good or bad strategy, but rather an observation as to how the market is viewing those strategies. But nothing says the market is right! If anything, finding strategies that are out-of-favor as a contra-indicator may be a good strategy in of itself!