Someone who reads my blog emailed me this question:
“What is your take on a Dividend Growth strategy? Buying large cap businesses with a decent initial yield and a track record of dividend growth.
It's seems very attractive to me to own a portfolio of large, safe US based companies that have solid dividend growth prospects.
If an investor is patient and chooses his portfolio carefully, they may be able to get 10% "yield on cost" after only 10 years. That seems like a powerful investing position to be in. And the "yield on cost" number gets insane after 20 or 30 years.”
A dividend growth strategy makes sense. So, does a high-quality strategy. A predictable businesses strategy. A consistent EPS growth strategy. And even a consistent share buyback strategy. People are attracted to the dividend approach more than some other strategies. So, I’d just caution that there’s nothing special about dividends. A company that increases its dividends per share by say 7% a year every year while increasing its EPS by about 7% a year on average over every rolling five-year period is a good stock if you buy it at a good price. But, is it a good stock because it pays a high dividend? Is it a good stock because it grows its dividend? Or is it a good stock because it grows its free cash flow?
Consistent dividend growth tells you about two things: One, it tells you about historical free cash flow growth. In the long-run, dividends tend to track free cash flow. Two, it tells you something about capital allocation. And then a dividend yield tells you something – not a lot, but something – about valuation. So, if you buy a company with a decent enough dividend yield in year one and the promise that it will keep growing that dividend – you should know you are buying a stock that is: cheap enough (as shown by yield), has high enough free cash flow growth (otherwise it wouldn’t be raising the dividend), and has good enough capital allocation (since it is choosing to pay more and more in dividends each year).
To be honest, I don’t think the dividend part of this is important at all. In other words, if you could find a stock with those three traits: 1) Good enough price 2) Good enough annual growth and 3) Good enough capital allocation – you should buy it regardless of whether it pays a dividend or not. Berkshire Hathaway is a good example. Berkshire doesn’t pay a dividend. It hasn’t paid a dividend in 50 years. So, it’s not any sort of dividend stock. But, Berkshire is reasonably priced. It has good growth in intrinsic value for a stock its size. And it has great capital allocation. So, Berkshire checks all the same boxes that I think drive the success of a dividend growth strategy. And yet Berkshire doesn’t pay a dividend.
The same is true of companies that consistently buy back their stock instead of paying much in the way of dividends. Omnicom (OMC, Financial) is a good example. Omnicom is an ad agency stock. Historically, those stocks have been underpriced relative to the market. They’ve traded at a premium P/E. But that P/E hasn’t been quite “premium” enough to stop the group from outperforming the overall market over decades. This is because they have a high combination of free cash flow yield and free cash flow growth. We can approximate this by just looking at free cash flow yield and sales growth. Margins are stable at most ad agencies. So, in the long-run, free cash flow tracks sales. Let’s say you find an ad agency that trades at 20 times free cash flow and grows sales by 4% a year. A price-to-free-cash-flow of 20 isn’t low. And a growth rate of 4% isn’t high. But, ad agencies don’t have to retain any earnings to pay dividends, buy back stock, etc. So, the stock will return about 9% a year in this scenario. It will have 5% of its stock price to pay out in dividends, use on share buybacks, etc. this year. And the sales (and thus free cash flow) at the company will grow organically at 4% a year. So, that’s 5% plus 4% equals 9%. The S&P 500 isn’t going to do better than 9% a year. So, an ad agency priced at 20 times free cash flow and growing as slow as 4% a year can still outperform the S&P 500. If the stock is priced to outperform the S&P 500, it’s better for continuing shareholders (those that don’t sell the stock) for the company to devote all its free cash flow to buying back stock instead of paying any dividend at all.
When I was writing a newsletter, we picked Omnicom. And we talked a little about capital allocation. Why did Omnicom tend to outperform WPP and why did WPP tend to outperform Publicis? If you looked at the long-term return of these stocks, the order was Omnicom, then WPP, then Publicis. Why? It had nothing to do with the underlying businesses and everything to do with capital allocation. Omnicom spent more on buybacks than other companies. WPP spent more on acquisitions done at low prices. And Publicis spent more on acquisitions done at high prices. That explained basically all the difference in performance between the stocks. If WPP had just bought back its own stock instead of making any acquisitions – it would have done a little better in terms of the total return of the stock. It made smart acquisitions at smart prices. But, the return on those acquisitions simply couldn’t match the return on an actual stock buyback of WPP’s own shares. Likewise, Publicis could have improved its total return for shareholders if it only did lower priced acquisitions or if it only bought back its own stock. We are talking about a difference of a couple percentage points a year over a decade or two. The difference between having say a 9% return on everything you don’t pay out in dividends and an 11% return on everything you don’t pay out and a 13% return on everything you don’t pay out is huge. And that’s the kind of differences we are talking about. Similar businesses with similarly priced stocks can return 9% a year or 13% a year depending on whether they are totally disciplined or totally undisciplined about capital allocation.
What does this have to do with dividends? Let’s use the 9%, 11%, and 13% example I just gave. It’s imperfect. Because all of these companies did pay some dividends. So, the return on earnings not paid out in dividends is not exactly the return on the stock. But, in the very long-run your destiny in a stock you hold forever is going to be the return the company gets on its own money. Dividends are a way to mitigate this. Dividends are a safety valve.
Companies have free cash flow. They must allocate that free cash flow. If they have debt, they can pay that down. And they can pile up cash. But, these are only short-term swings. A company can’t pay down debt forever or add to cash forever. That’s not realistic. So, free cash flow can only be used for 4 things. One, it can be used to expand the core business. Two, it can be used to acquire other businesses. Three, it can be used to buy back the company’s own stock. Four, it can be used to pay out a dividend. That’s it. How a company uses its free cash flow is critical.
Dividends are rarely the best use of free cash flow. But, they’re also rarely – in fact, almost never – the worst use of free cash flow. AT&T (T, Financial) is trying to buy Time Warner (TWX). The deal includes a cash component. Now, the share part of the deal is complicated to value. If AT&T is overvalued relative to Time Warner, this can be a good deal for its shareholders as far as the stock part of the deal is concerned. But AT&T is also using cash to make this offer. How likely do you think it is that AT&T using that cash to buy Time Warner will create more value for shareholders than simply paying that much cash out in a special dividend? I think it’s very, very unlikely that the cash part of the Time Warner offer can be more valuable to an AT&T shareholder than a special dividend would be. That means it would be better for shareholders if AT&T paid out more in dividends and spent less cash on making acquisitions.
That’s true for a lot of companies. There are mergers that create value. There are plenty of such mergers. They tend to be horizontal mergers that are done all in cash. So, if Berkshire Hathaway wanted to buy Progressive using cash – that would be an example of a horizontal merger. There are synergies between GEICO and Progressive. They are the two biggest companies in the direct auto insurance business. So, GEICO (really Berkshire) buying Progressive using cash could easily be a good deal. AT&T buying Time Warner is much harder to do in a way that can possibly be worth as much to shareholders as a simple special dividend for the same amount of cash would be worth. So, dividends are a safety valve. They are a guard against bad capital allocation. Dividends are also a hindrance to great capital allocation though.
Let’s use my example of 3 ad agencies over 10-20 years. Company A buys back its own stock. It gets a return of 13% a year from doing this. Company B makes inexpensive acquisitions. It gets a return of 11% a year on this activity. And Company C makes expensive acquisitions. It gets a return of 9% a year from this activity. The companies it buys aren’t bad businesses. They are good businesses. And there may even be synergies in buying them. But, because Company C is always paying a premium of 30% to 50% over the P/E ratio its own stock trades for – it can never do as well buying other companies as it could do buying its own stock.
What does changing each of these company’s capital allocation policies towards just dividends do? The more Company A spend on dividends, the more its total return as a stock will drift toward 10%. The same is true for Company B and Company C. Investors can’t make 13% a year – especially after paying a tax on dividends – by taking a dividend from Company A and buying something else in their brokerage account. So, a company like Omnicom makes its shareholders poorer to the extent it pays out more in dividends. A company like Publicis might make its shareholders richer the more it pays out in dividends. I think the more AT&T were to pay out in dividends, the richer it would make its shareholders. But this is only because I think it’s hard for AT&T to come up with anything to do with its billions in free cash flow that can return more than the company’s shareholders could earn for themselves buying other stocks with a dividend.
Healthy dividend growth and a high dividend yield are fine things to look for. But, high free cash flow growth and good capital allocation are what underlies this. Berkshire isn’t wrong not to pay a dividend. I wouldn’t necessarily want a company like ATN International (ATNI, Financial) to pay out more in dividends – although it has raised dividends at a good clip – because I think the management at that company tends to make decent capital allocation decisions. The less I trust management to make good capital allocation decisions, the more I want to see paid out in dividends. But there is no reason to prefer dividend growth over share buyback growth. In fact, if the stock you are interested in is a good business in a good industry – it should spend more on buybacks and less on dividends. Bank of Hawaii (BOH, Financial) is a bank. Most banks focus on paying out dividends and buy almost no stock back. I’d rather BOH focused entirely on buying back its own stock and didn’t worry about dividends at all. The company targets a tangible equity to total assets ratio. It keeps about 7 cents of tangible shareholder’s equity for every $1 of total assets it has. Everything else is surplus. As long as I knew BOH wasn’t going to buy any other banks and as long as I knew it wasn’t going to allow tangible equity to rise much above 7 cents a share for every one dollar of assets – I’d be fine with either dividends or stock buybacks. The bank would have to grow whatever it chose to do (dividends or buybacks) at about the rate of deposit growth. If it failed to do this, it would start building up too much tangible equity.
So, again it’s a growth and capital allocation issue. Consistent dividend growth is a sign. Growing dividends are a symptom both of decent growth and decent capital allocation. If a company can’t grow at all – it can’t increase its dividends. And if a company has truly atrocious capital allocation – it’s not going to allocate anything to dividends. But I don’t think dividends per se are important.
It’s a good idea to investigate any company that has consistently grown its dividend. But, you shouldn’t focus too much on the dividend. Instead, you should focus on what the price to free cash flow is.
What should you ask instead? Ask: What’s the free cash flow yield rather than what’s the dividend yield. Ask what’s the free cash flow growth rate rather than what’s the dividend growth rate. And ask how good is capital allocation overall rather than just how high is the dividend payout ratio.
Talk to Geoff about Investing in Stocks that Grow Their Dividends
Disclosure: None
