One of the more annoying consequences of the recent downturn has been the growing din of voices that are preaching the cult of dividends. A certain talk show host named Jim, for instance, is constantly citing the dividend yield as a reason to own or not own a particular stock. Here on Seeking Alpha, where there are some highly-focused authors who have been sharing various nuances of dividend investing for years, we have seen many articles lately that rank stocks on dividend yields without much else. Fortunately, most of the dividend-related articles do include "quality" considerations as well, but I worry that by focusing on just one parameter investors may end up making poor choices.
Let me go on the record and say I am not anti-dividend. I do believe, though, that the dividend yield is a secondary valuation tool and that it must be considered in the context of the payout ratio and the overall capital structure of the company. I also like the idea of using dividend-related information to judge the quality of a company. Here, I am talking about the number of years of consecutive or increasing dividends or perhaps the growth rate in dividends.
So, what do I have against using dividend yield as a primary criterion? Quite simply, the raw number doesn't mean a lot by itself. First, a great company that pays no dividend, like Apple (AAPL) or Intuitive Surgical (ISRG) gets excluded from consideration. Second, a high dividend yield may be a sign of risk - remember what great yields the banks had in 2007? The payout ratio, the capital structure and the cyclicality of the company's earnings must be considered.
In order to illustrate how the dividend yield is not the primary influence of a stock's price, I have plotted the price returns since 6/30 of all 60 Industrial stocks in the S&P 500 against the current dividend yield. Notice how random it is. Click to enlarge:
The average dividend yield is 2.4% - this includes 3 that pay no dividend. The average price decline has been 18.7% (through 9/6). While it's fairly obvious to me that some big dividend-payers didn't protect and that some low dividend-payers fared better than average, I believe that there is no clear relationship.
I decided to stratify the group by dividend yield in order to hone in on other characteristics. I broke the group into 3 - those yielding less than 2% (21), those yielding 2-3% (21) and those yielding above 3% (18). The largest members of the first group are Danaher (DHR), Fedex (FDX) and Precision Castparts (PCP). the largest members of the second group are United Technologies (UTX), 3M (MMM) and Caterpillar (CAT). The largest members of the last group are GE (GE), United Parcel (UPS) and Honeywell (HON).
While it might have produced slightly different and more precise results to use initial yields, there isn't much benefit to doing so since the stocks moved roughly in the same magnitude. Click to enlarge:
Here, the story is a little more clear. The higher the dividend yield, the worse the return. I have included market cap, net debt to capital and PE as well. The three groups are very different. The lowest-yielding set consists of smaller companies with better balance sheets. On a PE basis, they are valued a bit more richly.
I chose Industrials for this exercise because so many of the members of this group (95%) do pay a dividend. For those who might choose to make an investment primarily on the basis of dividend yield, recent history provides evidence that the results of such a strategy might not be so great. Sometimes the lower-yielding stock is a better one, and sometimes the higher-yielding one is not. Bottom-line: Dividend yield should be a secondary consideration when buying stocks.