In my first SeekingAlpha article a few weeks ago, I argued for casting a wider net in search of good dividend growth stocks, suggesting that it would be reasonable to look at stocks that either didn't have a quite unbroken record of yearly dividend increases, or which yielded less than the typical minimum income requirement, but which were growing their dividends more rapidly than the average dividend growth stock. In this article, I explore the second thought in more detail.
To create a universe of stocks suitable for this discussion, I started with the June 2013 edition of David Fish's "CCC" list, found here, and set out to find stocks, which roughly fit the description of yield beneath the typical minimum of dividend-growth investors (2.5%) but high dividend growth (a 10-year compound dividend growth rate of 10% or more.) This produced an initial list of 97 companies.
A few other limitations were then imposed to refine the list:
1. A yield of 1% or more, to eliminate stocks which would have difficulty growing a very small payout to the levels that dividend investors typically require, in a decade. (=84 stocks, 13 eliminated.)
2. Dividend growth of 10% or more in the last year. This was to eliminate stocks whose dividend growth rate had obviously decreased from previous levels (=60 stocks, 24 eliminated.)
3. Stocks with a 5/10 year Acceleration/Deceleration rate (CCC column AK) of 0.9 or more. This was a further measure to eliminate stocks whose dividend growth appeared to be materially weakening. (=40 stocks, 20 eliminated.)
Here is the resulting list of High Dividend Growth, Low Current Yield stocks:
(10-year CAGR from Low-Risk-Investing.com. All other figures from the June 2013 issue of David Fish's CCC list.)
To provide a benchmark for the HDGLCY (let's just say HG) list, I decided to compare its performance with David Van Knapp's recent list of the stocks that are the most popular with Dividend Growth Investors. Using the Van Knapp list has the advantage of reflecting the "conventional wisdom," while not incorporating the peculiarities of any individual list. However, the Van Knapp list brings with it some of its own problems, since it includes the remnants of recent spin-offs [e.g. Phillips 66 (PSX)], some stocks without a substantial dividend history [e.g., Apple (AAPL)], one speculative high-yield play which has recently cut the dividend [American Capital Agency (AGNC)], some 'til-death-do-us-part low-growth old favorites [Paychex (PAYX) and Pfizer (PFE)], one Canadian stock (BCE), and some victims of the 2008 financial crisis. The fairest solution to using the Van Knapp list but taking out the "eccentrics" seemed to me to be including only those stocks which also appear on the CCC list. So my ultimate list of Dividend Growth Favorites (DGFs) is as follows:
(10-year CAGR from Low-Risk-Investing.com. All other figures from the June 2013 issue of David Fish's CCC list.)
That's a lot of data. For ease of comparison, let me restate the averages before I begin my discussion:
Let's begin with the most obvious generalization: the HG stocks yield about 1.7% less than the DGFs, but their total return is almost 4% greater. In other words, greater long-term total return makes up more than double for the lesser dividend yield. Investors in the distribution phase may object, saying that they need income now. But, as I will demonstrate below, there's a reasonable way to get stable income at the level of the DGFs from these stocks. Further, investors in the accumulation phase, i.e. those with a longer time horizon, or those who can afford to assume a little more risk, can point to offsetting advantages. Let's start by considering some of these.
First, consider what I think is one of the most underappreciated columns in the CCC list, "5/10 A/D." This means the 5-year (compounded) average increase in the dividend, divided by the 10-year average increase. Thus, if the average rate of increase in the dividend is accelerating over the past five years, compared to the past 10, the number is greater than 1; if it's decelerating, it's less than 1. So as the summary table shows, the HG stocks are on average actually accelerating the growth of their dividend over the most recent five-year period, while the DGF stocks are showing smaller and smaller increases. Given the standard generalization that the larger companies get, the slower their percentage growth, this result strikes me as rather remarkable.
And, it seems to me, decelerating dividend growth should worry the holders of the DGF stocks, particularly in cases when the dividend growth in the last year or two has slowed to low single digits [as it has for DGFs AT & T (T) and Verizon (VZ).] One common argument for dividend-growth investing is that dividend growth will generally exceed inflation, keeping the investor's head above water despite increases in the cost of living. But there is certainly an argument to be made that inflation is higher than the officially-reported figures, and higher than the rate of dividend growth of T and VZ. If future inflation turns out to be higher than the optimistic government forecasts (and when aren't government forecasts optimistic?), then any dividend growth rate that is only a couple percent more than the reported rate of inflation is ominous news indeed for an income investor.
But on the other hand, what's not to like about an accelerating trend of dividend growth? Unless we get hyperinflation, it's almost certain to stay ahead of the cost of living. And while an accelerating trend obviously can't continue forever, high growth can often continue much longer than many people would imagine. One of the more thoughtful comments on my first article said "if you're basing an investment decision on a 5- or 10-year CAGR streak [of dividend increases], you've missed 5-10 years of the streak before you begin to enjoy the expected advantages," and intuitively that seems like a plausible objection. But the fact remains that, on average, the 40 stocks in the HG list are showing accelerating dividend growth, after a 10-year period of 10% CAGR of the dividend - and, their estimated annual earnings growth rates for the next five years average more than 10%. Possibly, you can have your cake and eat it too. As an ignorant California boy, I remember seeing my first Wal-Mart more than 30 years ago in Elkins, West Virginia, and thinking "I wonder if they're ever going to expand to the west coast?" And last year, Wal-Mart's dividend was still increasing 11.1% on earnings growth of 11%.
So here's the basic concept for an advocate of the HG stocks. Dividend growth investors want a company that rewards shareholders by paying continuously increasing dividends (an indicator of stable growth), but the amount of the dividend will necessarily vary with the growth stage of the company. Very generally, in early stages, companies will be plowing most of their earnings back into expansion. At this stage, dividends will be minimal, and the reward for ownership of shares of a successful enterprise will be mainly an increase in the price of the stock. At the next stage, as companies reach "critical mass," they are able to start returning more of their profits to stockholders in the form of dividends, and the increase of the dividend from a very low base can be quite dramatic; but price appreciation is still by far the largest part of total return. Finally, for the "mature" companies, dividends come to represent a significant part of the total return (even though according to most studies I am aware of, price appreciation still represents a majority - about 60% - of total return.)
So to the extent this generalization describes the growth of a typical company, the HG investor is simply looking to climb on the train a little earlier than the investor who is interested only in enterprises which have already reached Grand Central Station. Some companies always seem to have a low dividend, not because the dividend isn't growing substantially, but because the earnings and share price keep growing just as rapidly. So for the possibility of sharing in this kind of robust and extended youthful growth (while of course not disdaining full maturity if and when it arrives) the HG investor is willing to take more of his return in the form of price appreciation, and less in dividends. And his reward for doing so, according to the comparison I have provided above, seems to be about an extra 4% of total return.
But, the DGF investor will reiterate, What about current income? And what about the risk of price fluctuation? After all, they will argue, dividend payouts are much more stable than stock prices; if income has to be derived from selling shares when the market is down, it might require selling off an inordinate amount of principal to retain the same (inflation-adjusted) level of income. This would in turn lead to a reduced level of future income, which would be unacceptable.
But before we accept this conventional wisdom, allow me to present an example for consideration. For a benchmark, let's consider the standard dividend-growth approach: imagine buying 1000 shares of a model DGF stock at $10. As we can see from the table above, this average DGF stock will yield 3.41%, and will offer a compound annual total return of 12.6%. (That is, 9.19% of total return will come from price appreciation, and the other 3.41% from dividend payout.) The average dividend increase last year for the DGF stocks was 11.1; the 3- and 5- year increases are quite similar; but the 18.1% 10-year average looks to be distorted by some outliers. So we use the average of the 1-, 3- and 5-year increases and project the average annual dividend increase of the DGF at 11.2%. We then mechanically project these increases 10 years into the future; e.g. one year's appreciation of the initial $10,000 investment at a rate of 9.09% is $10,919, while one year's compounding of the dividend rate brings it to 3.82%. Of course, such mechanical projections are artificial, but rough approximations will suffice for the point I want to make. Here is a 10-year table of returns for the model DGF stock:
Turning now to the case of a model HG stock, we can see from the previously presented summary table that this stock will yield 1.69%, and will offer a compounded total return of 16.4%. (That is, 14.71% of total return will come from price appreciation, and the other 1.69% from dividend payout.) And the dividend will grow at a compounded annual rate of 22.5%, based on the average of the 1-, 3-, and 5-year returns. (This may sound unbelievably large, but consider that 10 stocks on the HG list have actually exceeded this rate over the past decade, including CSX (CSX), a stock that also appears on the DGF list. Walgreen (WAG), another DGF, barely misses at 21.2. And it is much easier to grow the dividend rapidly when starting from a low base.)
So the problem for someone advocating investing in HG stocks is to provide a model, which can provide a plausible equivalent for the income profile of the model DGF stock, without taking on too much additional risk. Well, actually, we can provide an exact equivalent. Suppose we set aside a small portion -say 10% of the initial $10,000 investment - as a "cushion" from which the difference in dividend payout would be made up, refreshing the cushion whenever it is about to run out. A table of the results would then look something like this:
To put it into words: Instead of buying 1000 shares at $10, the HG investor would purchase only 900, leaving a "cushion" of $1000. In the first year, the yield of 1.69% on an investment of $9000 would be $152, compared to the $341 the DGF investment would have earned. So an additional $189 would be drawn from the cushion to make total income equal to that of the DGF investor. Repeating this process, the cushion would be virtually depleted by the end of the 5th year, so to replenish it, 50 more shares of the HG stock would be sold at the current price of $19.85, bringing the cushion back to $1049. After the 6th-year $184 "make-up" payment to equal the income of the DGF investment, the cushion would be $865, and the process would continue until the 10th year. At that point, the cushion would again be considerably depleted, but the need for a makeup payment would have vanished, because the dividend payout of the HG stock would have caught up to that of the DGF. To summarize, the 10-year result would be:
Now, one can argue about whether the dividend of the HG stock can really continue to grow at the assumed 22.5% rate over a decade (even though there are examples which support the possibility.) But this really does not affect the point I am trying to make. The other component of total return, price appreciation, produces such a large difference over 10 years that there is plenty of slack to imagine replenishing the cash cushion more frequently and still producing a favorable result. One could probably also replace the model DGF stock with a somewhat higher-yielding one and successfully play catch-up, though catch-up might take a little longer than a decade.
So, the artificial character of this example notwithstanding, I think there are several important takeaways. Beyond yet another demonstration of the power of compounding, the most important point to me is that even a relatively modest cash cushion can give the income investor a very stable bridge to buying lower-yield but higher-growth stocks, with equivalent cash flow. If someone really needs every last dollar of the $364 that a $10,000 investment in the average DGF will provide in its first year, then setting aside $1,000 (i.e., foregoing $36.40 in dividends) to give a competing HG investment a few years to "get its footing" will not be an option; there is no getting around the fact that the HG investment will not provide as much initial income. But realistically, should anyone be investing $10,000 if they can't afford to keep a $1,000 backstop? It's a pretty severe restriction on designing an income-investing strategy if we have to tailor it to someone who needs an extra $36.40 of income.
A cash-cushion approach also does a very good job of protecting against unfavorable market conditions. For mathematical simplicity, my comparison of model HG and DGF investments has assumed straight-line growth, based on long-term compounded average growth rates, which fictionally replace yearly ups and downs, bull and bear markets, etc., with a number as if long-term growth was produced by a perfectly smooth rate of growth each year. But assume a bear market: then the cushion I am hypothesizing would be able to sustain income at the DGF level for five years - much longer than the 18-month duration of the average bear market. Of course, if the cushion ran out at just the wrong time, there would be the problem of replacing the income for say a year - but that could still be done with just 2 percent or 3 percent of original principal.
(Of course, we shouldn't contemplate the worst without also considering the best - say, that the stock appreciated an unexpected amount, and perhaps even became overvalued. In this case, it would be easy to declare oneself a "special dividend," and set aside a cushion that would guarantee the projected income, and perhaps considerably more, for several additional years.)
Another important takeaway from my example of deriving a desired level of income from a HG stock is the approach, seen at the end of year five, of selling shares of appreciated stock to maintain a cushion. This highlights another issue, which I think has led to excessive focus on the DGFs. With the higher level of income which these stocks provide, it is said, you can put yourself in a position of deriving your entire income from dividends, never needing to sell off principal - as if never selling stock to provide income was an inviolable rule. In fact, I think selling off principal makes excellent sense in the right circumstances, and that the conventional wisdom should be directly challenged.
The resistance to selling seems to arise from failure to distinguish between appreciated and "core" principal. Of course, invading principal that is necessary to generate the minimum level of income one desires is short-sighted and dangerous; but that is not what I am arguing for here. In my example above of drawing some income from price appreciation, principal increases from a starting amount of $10,000 to an end value of $27,612 in ten years. It is just that a little bit of principal is held back at the time of the initial investment to guarantee cash flow, and then appreciated principal is shaved off the top as the total value of the investment increases. It just seems incredibly one-sided to insist that all income must come from dividends, when the fact is that a majority of total return comes from price appreciation, even with mature companies. Some people want to pass their entire principal on to beneficiaries, which is wonderful - but this goal hardly needs to be regarded as a foundational principle of dividend growth investing. Someone who believes in the basic idea of investing in stocks which provide a reliable and growing stream of income, but maintains a cushion and/or derives a percentage of their income from cashing out appreciated principal, can surely claim to be a legitimate dividend growth investor as well. A fortress mentality is not required.
In summary, then, I believe that by beginning with a small cash cushion - a prudent general principle of personal finance - one can plan a HG investment which provides cash flow equal to that of a DGF investment, while producing superior long-term total returns. Is it more work? Well, once or twice in a decade, the HG investor who needs to keep cash flow steady would have to rouse himself to sell small blocks of stock, and pay a small commission to his discount broker. But realistically, within the context of an entire portfolio, even this would probably not even be necessary. For instance, in his recent report on his model dividend growth portfolio - a portfolio which includes many members of the DGF list - David Van Knapp reports making a number of changes over the years, selling stocks which came into conflict with his various "Constitutional" rules. In many cases these were not losses; they were simply situations where he saw a change of business model, a reduction in the dividend growth rate, overvaluation etc. These kinds of routine adjustments should provide adequate opportunity for taking profits to supplement income. While the idea of not having to give a moment's thought to the flow of income is attractive to the couch potato in all of us, putting a portfolio completely on autopilot is probably not a good idea anyway, no matter what type of stocks one chooses to own.
Ultimately, however, the major argument in favor of making some investments from the HG list (or something like it) has to be that it opens up a broader horizon where many great investments can be found. For instance, companies such as IBM, Deere, and Target are not exactly fly-by-night operations. For those who require at least a decade of yearly dividend growth, 25 members of the HG list are Contenders, and six more are 25-or-more-year Champions. Thirteen have compounded total returns of 20% or more over the last decade.
Based on the method I have presented for stabilizing income even as stocks with lower current yield are purchased, I believe that such stocks deserve serious consideration by any dividend growth investor who does not need to be deploying their last dollar to earn immediate dividend income.
This article has presented the case for including HG stocks in a dividend-growth portfolio in the most abstract and general terms. In future articles, I intend to present the results of some backtesting, as well as case studies of individual companies.
Additional disclosure: Disclosure: I, or family trusts I control, are long APD, CVX, INTC, JNJ, KO, LMT, MCD, MSFT, O, and UTX from the DGF list, and CCE, CVS, IBM, JW.A, TGT, and UNP from the HG list.