A little over 10 months ago I discussed the advantages and potential disadvantages of buying ConocoPhillips (COP) before the Phillips 66 (PSX) split. The article can be found here but to-date it has been my most popular article, so don't fret over whether or not you should back-track to peruse it. Within this article one advantage that I detailed was that the original COP Company would continue to pay the same $2.64 yearly dividend - thus any PSX dividends would be a pure bonus. While doing this analysis it occurred to me that this movement might unjustly affect Conoco's record of dividend increases. While it would be true that the amount of dividends per COP share would stay the same at $0.66 a quarter, long-term shareholders would also receive a "bonus" dividend from PSX. In addition, new shareholders would have a significant income yield increase due to the lower share price implied by the split.
Like many dividend growth questions, the solution came from the Champion, Contender and Challenger work of Seeking Alpha's David Fish, which can be seen here. Through a direct response David replied that COP would keep its record of increasing dividends, as the dividend paid per COP share actually increases once the split-adjustment is made - case closed. However after remembering this "special"treatment of dividend scenarios I was recently directed to another thought experiment. What happens if the dividend per share increases but the total dividends paid actually stayed the same? It sounds like an obscure case, but it's theoretically possible through the work of share repurchases. Well just for you, Seeking Alpha community, I went out on a "dividend per share increases, total dividends paid doesn't increase" scavenger hunt. Here's what I found:
The Sherwin-Williams Company (SHW)
Boom. Now if you're not impressed quite yet, you might be shortly. Without the knowledge that SHW was a perfect example of what I was looking for, I surveyed perhaps 100 securities for the matching characteristics. Now I want to be overwhelmingly clear that this is a very special case, but there are two points that we can take from it. First, finding such an example suggests that my ideology wasn't just theoretical, but it also has real world application. Second while it is indeed a rare case, that doesn't preclude it from having a more wide spread implication. So what's so special about the Sherwin-Williams Company? That's a good question and I'm eager to answer. SHW has not only paid a dividend but has also increased it for the last 34 years. Over the last decade these increases have come in at an average rate of just under 10% a year. And if you've ever painted or seen something that has been painted before, you probably know why: Sherwin-Williams is a well trusted and dependable paint source. Let's take a look at the dividend data for Sherwin-Williams over the 2008 to 2011 period:
|Dividends per Share||$1.40||$1.42||$1.44||$1.46|
|Total Dividends Paid||$165,111,000||$162,561,000||$156,424,000||$153,512,000|
Now to be honest I'm not sure if you're going to be as impressed as I was, but in reality this example is better than the one that I had initially imagined. Let's take a look at what's going on here. In each year from 2008 to 2011, the dividend paid per share increases. Yet even with increasing the dividend each year Sherwin-Williams is able to pay less in total dividends. The underlying reasoning behind this idea is the fact SHW is repurchasing shares in conjunction with the dividend payout policy. The math allows that enough shares were repurchased to increase the dividend per share yet decrease the total payout. It should be noted that the shares outstanding are average numbers thus one cannot simply multiply the total shares by the dividend paid per share in order to find the total dividends paid, although it correlates nicely.
An obvious criticism to this scenario is that sure the dividend is increasing, but only minutely. In fact the 1.4% average growth rate over this period is almost as dismal as Consolidated Edison's (ED) recent increases, of which I have expressed repeated caution. But look at what happens without share repurchases:
|Total Dividend Paid||$165,111,000||$162,561,000||$156,424,000||$153,512,000|
|Dividend per Share||$1.41||$1.39||$1.34||$1.31|
Instead of a 1.4% dividend increase, the dividend would actually decrease by about 2.4% a year; the same annualized amount that the total dividends paid decreases by. Within this example we can see the benefit that share repurchases has on dividend growth. That is, while it is true that any funds used for share repurchases directly takes away from the available amount of dividends to be paid out, this doesn't mean that they do not add value to dividend growth. More specifically, while you can debate about the suitability of having dividends or share purchases in reality they are complimentary.
Moving away from this specific illustration we can find a more realistic or "real world" application. For example, if you move to the dividend history of SHW in 2007 we find a dividend of $1.26 per share, an average of 127,222,000 shares outstanding and $162,301,000 in total dividends paid. Now it is true that the total dividends paid increased by 1.7% in 2008, but the dividend per share increased by 11.1%. This was a direct result of the average shares outstanding moving from about 127.2 million to 116.8 million. In fact it would take an additional $17.8 million to make the same dividend increase without using share repurchases. (Granted the amount spent in share repurchases exceeds the additional hypothetical dollars needed to increase dividends, but it is paramount to consider that there are reasons beyond supplementing dividend growth to use share repurchases)
If we move to 2012 we see a quarterly payout of $0.39, or $1.56 a year. This represents a 6.8% yearly increase over 2011. However, as SHW is still repurchasing shares, the total dividends paid will likely not have to increase by this amount. If we look at the most recent 10-Q we see that there were 101,446,643 average shares outstanding and SHW had paid $80,438,000 in total dividends through June 30th. Keeping the shares constant, this would indicate about $79.1 million more in dividends that need to paid, for a total of about $159.6 million in 2012. While this is above the $153.5 million number seen in 2011, this is only a 4% increase in total dividends paid against the 6.8% per share dividend increase. The spread widens if SHW repurchases more shares in 2012. It's easy to see the friendly connection between share repurchases and dividend growth.
In addition to finding the SHW case, I also happened upon the Joy Global (JOY) case. Joy Global has kept the same $0.175 quarterly dividend since September of 2008. Over this period we see that total dividends paid were $71.6 million in 2009, $72.1 million in 2010 and $73.3 million in 2011. As a consequence we see that the average share count went from 102.4 million in 2009 to 104.9 million in 2011. That is, the exact opposite scenario as Sherwin-Williams occurred. While the debate can be made between whether one prefers dividends or share repurchases; in my opinion I would much rather have an increasing dividend and a decreasing amount of outstanding shares in lieu of a stagnant dividend and an increasing share count. In fact extrapolating from the Joy Global case, it's not unimaginable that a company might pay more total dividends yet the dividend per share would decrease.
Certainly the Sherwin-Williams case was unique, but that doesn't mean that we can't learn something from it. Instead of viewing dividend growth and share repurchases separately, both of which have their independent merits, perhaps we consider them in tandem. For example, I have previously detailed how Target's (TGT) generous share buyback program will make its goal of $3 in dividends and $8 in earnings by 2017 easier to formulate; less shares, equal or growing earnings, and consequently a lower bar to jump over for long-term shareholders. There is a reason that almost all annual letters contain a semblance of the phrase: "We returned XX amount of dollars to shareholders through the use of dividends and share repurchases." Whether you prefer share buybacks or dividend growth, there's no reason that you can't be friends with both.