How to Find Stocks With Good, Predictable Capital Allocation
Someone emailed me this question:
“How do you evaluate the capital allocation skill by the management? I do so by looking at the FCF yields for the acquisitions and share repurchases or ROIC for internal investments.”
I want to focus on what will have the most influence on my investment in the company going forward. For this reason, I’m less interested in knowing quantitatively what the past return on investment of management’s actions were – and more in simply how management will allocate capital going forward.
Let’s start with who the manager is. If the manager is the founder, that’s the easiest situation. We can assume the founder will stay with the company for a long time. The average tenure of a professional manager – nonfounder – CEO at a Standard & Poor’s 500 type company is short. It’s maybe five years. If you think about that, it means odds are that the CEO you now see at the company you are thinking of investing in will be gone within two to three years (because chances are he’s already been running the company for two to three years by the time you buy the stock). It’s just not worth thinking about such a manager. In this case you’d want to focus on the board of directors or the chairman. Ideally, you want to find situations where the founder is still the CEO, the chairman or has some position in the company. This will make figuring out future capital allocation plans easier.
In situations where you don’t have a founder present, ongoing participation by a family is useful. In situations where you don’t have the presence of either a founder or his family at the company, you may still have first-generation managers who worked with the founders before they became CEOs.
Those three situations will make future capital allocation easier to predict. One, the founder influences capital allocation. Two, the controlling family influences capital allocation. Three, a manager who worked directly for the founder early in his career now influences capital allocation. If you don’t have any of those three scenarios, there are still two others that can lead to some predictability. You can have a long-tenured CEO. For example, the big ad agency holding companies are usually run by a CEO who has been at the company forever. I know Omnicom (OMC, Financial) best. The CEO there – John Wren – has been at the top position for 20 years. For most of those 20 years, the chairman of the company and the chief financial officer (CFO) were also the same.
In terms of capital allocation, if you have a lot of consistency in the offices of chairman, CEO and CFO, you’re able to more easily count on future capital allocation looking like past capital allocation. The last of the five scenarios that can lead to predictable capital allocation is the presence of a “refounder.” This is someone who comes in and reshapes an existing business through some sort of crisis. Again, this is an extraordinarily powerful figure. Not just a professional manager who will probably be replaced within the next two to three years.
Before worrying about whether capital allocation has been good or bad, worry about whether the future will be like the past. You can only count on future capital allocation looking like past capital allocation in five scenarios: 1) The founder influences capital allocation, 2) a controlling family influences capital allocation, 3) a former top lieutenant of the founder influences capital allocation, 4) a long-tenured CEO influences capital allocation or 5) a “refounder” influences capital allocation. You can look for some other signs beyond just these five scenarios, but first let’s take a moment to identify actual real-life examples of what I’m talking about.
I’ll go through some of the stocks I’ve talked about in the past. I just mentioned Omnicom. Omnicom meets a couple of these criteria. For a long time, the CFO, the CEO and the chairman positions didn’t turn over. The CFO eventually changed, but the CEO and chairman are long tenured. Omnicom as it exists today is pretty closely tied to the personal histories of the current CEO and the current chairman. You can argue about whether they are really founders – they didn’t found the agencies that form the organization – but they really determined the way capital was allocated from the very beginning.
Omnicom, as a holding company that allocates the free cash flow produced by the agencies that were merged to form it, was pretty much founded by people who still have roles at the company. This isn’t unusual among ad agencies. Both WPP (LSE:WPP, Financial) and Publicis (XPAR:PUB, Financial) also have people involved in capital allocation who have been at the companies forever. You can predict that future capital allocation at Omnicom, Publicis and WPP will be similar to past capital allocation.
In one of my most recent articles – on index funds – I mentioned three companies whose stocks I owned in the late 1990s: Village Super Market (VLGEA, Financial), J&J Snack Foods (JJSF, Financial) and Activision (ATVI, Financial). Village is controlled by the founding Sumas family. The family runs it day to day. There are something like five family members with top positions in the company. Historically, only some financial functions (CFO) and legal functions (general counsel) – and independent directors – have ever been held by nonfamily members, but that’s about as family controlled and family run as a public company can get.
J&J Snack Foods is still run by the founder. He’s been running the company for about 46 years now – most of his life and all of the company’s life. Capital allocation there should be the same in the future as it was in the past. Activision is an interesting case. Basically, the founder is still running the company. That company has a pair of people – President and CEO Bobby Kotick and Chairman Brian Kelly – who have been involved for a long time.
I’m giving you these examples because my advice to look for a founder, a refounder, a long-tenured CEO, etc., might seem overly restrictive. It’s not. I know a lot of S&P 500 type companies are now run by brief-tenured professional managers, some of whom were even hired from outside the company. In general, these aren’t the best companies to buy. It’s not hard to find companies that have a cleaner line of descent from founder to current CEO.
For example, I’ve written about three watchmakers: Swatch (XSWX:UHRN, Financial), Fossil (FOSL) and Movado (MOV). All three of these companies meet one of the criteria I mentioned. They also have different capital allocation from each other. Movado is conservative. Fossil is aggressive. It’s true, though, that over time most companies drift away from founder control, family control, etc. Of companies I’ve written about, the oldest family controlled example would be John Wiley & Sons (JW.A). That company is now probably something like 210 years old. Members of the Wiley family still have board seats and own a special class of voting stock. The family exerted some operational control for probably the first 190 years of that company’s history. That’s the most extreme example I can think of.
What if the stock you are looking at doesn’t fit into any of these five situations I’ve described? You are looking at a board, a CEO, a CFO, etc., who aren’t especially long tenured. They have no connection to the firm’s founder. They have no connection to the founder’s family. They may have been hired from outside the company. They haven’t spent their whole careers at the company.
Honestly, this usually means you can’t count on predicting how they will allocate capital, but there could be exceptions. I bought into Fair Isaac (FICO) just after the financial crisis. A big reason why I did so was the capital allocation I expected. I honestly believed that FICO was going to buy back a lot of its own stock. If I didn’t believe that, I might have been less inclined to buy the stock. The company was trading at a low price-to-free-cash-flow. This meant I could “double-dip” if the company – instead of paying free cash flow out in a dividend – used it to buy back its very cheap stock. The CEO was an outsider who had worked at IBM (IBM) before taking the position at Fair Isaac.
There were some clues that Fair Isaac would be buying back its own stock. Shares had peaked at 78 million in 2004. They then fell to 74 million (2005), 65 million (2006), 58 million (2007) and then 49 million (2008). It’s not common for a company to reduce share count by about 10% per year. The company also didn’t really talk about dividends as something it planned to increase despite having a ton of free cash flow to spend on it. As it turns out, Fair Isaac went on to reduce share count by about 6% per year from the time I bought it to today. That’s a rapid rate of share reduction, especially considering the stock became more expensive over time. It’s much easier for a cheap stock to reduce its share count than an expensive stock. Here, the unorthodox and repetitive nature of what Fair Isaac was doing is important.
Companies do buy back stock from time to time. What very few companies do is buy back stock each and every year while paying out close to nothing in dividends. Fair Isaac was doing that. There were years when it probably bought back $200 million in stock while paying less than $50 million in dividends. In other words, it was using more than 80% of the cash it “returned” to shareholders for stock buybacks. This is unorthodox because it’s not balanced the way most companies like to use both buybacks and dividends. I was willing to trust Fair Isaac’s capital allocation more than I would trust the capital allocation at other companies.
Finally, you can look at incentives instead of past actions. I focus on three predictors of future capital allocation: 1) The personalities involved, 2) the past actions of the corporation as a whole, and 3) the incentives of the decision makers. Copart (CPRT) is a good example of this. The key decision makers here were a founder and a long-tenured manager (and total corporate “insider”). Copart hadn’t bought back much stock when I first looked at the business. This was a little over five years ago. It had only carried out one buyback, but I knew two things from reading about the company. One, the people who ran Copart only cared about the narrow business Copart was in. They were very knowledgeable about the business, but they didn’t seem to have any interest in diversification.
Copart wasn’t going to be able to use all its free cash flow to grow the business. If it wasn’t going to make acquisitions, it would have to decide on piling up cash, paying out dividends or buying back stock. My bet was that the company would aggressively buy back stock from now on. Why? One, it had just done some buybacks. That was a big tipoff. Two, the top people in the company had signed an unusual compensation agreement that suggested they’d focus on buybacks. This is why you always read the proxy filings for a company in which you’re interested. The proxy document has info on bonus plans, etc. If you’re using EDGAR (the SEC website) the document you are looking for will be called the “14D.” Here is a description of the compensation agreement I was talking about (Willis Johnson is the founder of Copart):
From April 2009 to April 2014, Willis J. Johnson, our chairman, received no cash compensation in consideration of his services to Copart (other than a $1.00 annual payment). Instead, in April 2009, we granted Mr. Johnson (our CEO at the time) an option to acquire shares of our common stock, vesting over five years. This option became fully vested in April 2014.
Basically, both the current CEO and the former CEO (and now chairman) were given long-term options to acquire stock instead of being given any cash compensation. The company didn’t have a history of paying a dividend. It had started buying back some stock recently, and it granted the two people who were most important to the company – the CEO and the chairman/founder – five-year stock options instead of any other kind of compensation. Based on that, I was confident that the company was going to buy back a lot of stock. I was also confident that – whether they were right or not – the two people who knew the most about Copart were convinced its share price would be a lot higher in five years than it was at the time. Otherwise, they would have taken at least some cash rather than all stock.
This was an unorthodox compensation arrangement. It’s unlikely the board thought of it independent of the two executives. It’s also completely unthinkable that the two executives would feel obligated to accept zero cash compensation unless they were confident in the stock’s prospects over the next five years. As it turned out, Copart bought back about 5% of its stock per year over each of the next five years. The stock returned about 20% per year. You couldn’t have predicted these exact figures, but reading the proxy statement would be helpful.
I had also read some past interviews with the current CEO and the founder/chairman. Reading between the lines, I felt strongly that they were more interested in buying back stock than paying dividends. They just seemed parochial about their business compared to the way a professional manager at a big, bureaucratic S&P 500 type company thinks.
You can find other incentive systems in proxy filings and annual reports. For example, I have researched some companies that rely heavily on either total stock return (TSR) or economic value added (EVA) as the way they determine almost their entire bonus pool for employees from the very top to the very bottom of the organization. Companies that pay bonuses based on TSR or EVA aren’t going to issue stock that adds to equity if it depresses ROE. Companies that pay bonuses out of either total stock return or economic value added are basically targeting either return on equity (leveraged) or return on capital (unleveraged). They can be counted on to minimize the growth of assets generally and owner’s equity specifically.
So far, I’ve lectured you on how to figure out what I think really matters: How will the company allocate capital in the future? You asked how to judge past capital allocation decisions.
I judge each acquisition independently, but I’m most interested in the strategy of why they did what they did. I don’t necessarily care whether they got lucky. For example, I think Frost (CFR) has only so-so capital allocation. Frost’s strategy is good when it comes to the cultural fit of the acquisitions it makes. Frost mainly buys good, Texan banks that fit easily into their organization. They stay within their circle of competence. Frost is unlikely to make major capital allocation mistakes, but I don’t think Frost’s acquisitions will add value. Why not? The company sometime uses shares to make acquisitions. Other banks just aren’t as high quality as Frost. I’d rather Frost never issued shares.
I extend the same rule to other businesses I consider “great.” I own shares in BWX Technologies (BWXT). I would never want BWX to issue shares to make an acquisition. This is even more true at Omnicom. It’s extraordinarily true at Fair Isaac. Frankly, Fair Isaac’s business is so good that in the long run it would be hard for the company to ever come out ahead when giving up shares in itself to acquire anything else. If the stock is overvalued, it might be possible, but I doubt it. Capital allocation decisions that are simple, predictable and yet good are the best ones.
In the newsletter issue I wrote about ad agency holding companies, I compared capital allocation at Omnicom, WPP and Publicis. Publicis has a history of making poor capital allocation decisions. It overpays for things. WPP has a history of making good capital allocation decisions. It pays fair value or less for what it buys. Omnicom has a history of basically just buying back its own stock.
The truth is that as good as WPP’s capital allocation has been – and it’s been good – it’s hard to negotiate purchases of entire advertising-related companies at prices that are lower than these companies would trade for in noncontrol situations as public companies. In other words, control buyers are at a disadvantage to noncontrol buyers when it comes to ad agencies. You, the individual investor, can get a better price on a share of an ad agency you buy than WPP can get on a merger it has to negotiate with the seller’s board of directors. As a result, Omnicom has been able to match or top WPP’s return on its allocated capital while really doing nothing but dollar cost-averaging into its own stock. This is typical of the kind of capital allocation program that actually works.
You can even see this at Berkshire Hathaway (BRK.B). Warren Buffett (Trades, Portfolio) has sometimes used stock to make acquisitions. Berkshire’s shareholders would be as well off or better – and would sleep much more soundly – if Berkshire’s board of directors passed a rule banning Buffett from ever increasing Berkshire’s share count. Buffett is a great capital allocator. He’s the best of all time, and yet his record in using stock to acquire Dexter Shoe, General Re and Burlington Northern using Berkshire shares is mixed.
Burlington was a good acquisition, but considering where long-term interest rates were, how much debt Burlington itself can safely carry and Berkshire’s own financial strength, I’m not sure using stock worked that well. I’m sure he used stock because some of Burlington’s sellers were only willing to do the deal if it wasn’t all cash, but Berkshire could have borrowed and bought back its own stock to eliminate any dilution. In fact, I’m not convinced that Berkshire has ever added value by allowing its share count to rise, and I can point to cases where Berkshire destroyed value by increasing its share count.
When evaluating capital allocation, I mainly use long-term rates of return on things like retained earnings. I also try – whenever possible – to compare capital allocation at peers. For example, in the ad agency business, you’d think all companies were decent capital allocators if you just looked at their returns on capital or even return on retained earnings. The ad agency business is very forgiving. Even if you overpay to buy an agency, you’ll end up with a pretty value-neutral to value-enhancing purchase if you hold the acquired company forever. The economics of the ad business are good especially if you can borrow money to buy an ad agency you’re going to get a decent long-term return on your money, but the test is how good your acquisition is versus just buying back your own company’s stock. A return of 8% per year might look OK, but if you could have bought back your own stock at a return of 12% per year on the same day, it’s a bad capital allocation decision.
The real test of any capital allocation decision is the opportunity cost. We can always measure capital allocation decisions versus two very easy to measure alternatives: 1) Paying a cash dividend and 2) buying back stock. If you’re a good stock picker, you should prefer dividends. For example, if I really believe I can always make 10% per year or more on any money a company pays out to me, then the opportunity cost of them doing anything but paying out a dividend is never going to be less than 10% times one minus the current dividend tax rate.
Let’s say my dividend tax rate is 20%. In that situation, the opportunity cost for any capital allocation decision by a company I own can never be less than 8% per year. Remember, what matters for me is the cost of the capital allocation decision to me, not the cost to the company. From the company’s perspective, an 8% after-tax return on investment may seem like a perfectly good decision. To me, an 8% after-tax return on investment is never better than a cash dividend – and it is often worse. That’s because, over a long enough time period, I can almost always be assured a 10% annual return or better from my own stock picking.
The opportunity cost hurdle of not paying a dividend is actually pretty high. Right now, it’s not that easy to make more than 8% per year after-tax out of any merger unless you are funding that merger entirely with borrowed money. Of course, for great businesses (and especially great businesses with stock prices that are temporarily low), the real opportunity cost of any capital allocation decision is not buying back their stock.
I mentioned Omnicom. Let’s say Omnicom stock is trading at something like a 7% leveraged free cash flow yield. And let’s say the company can grow revenues – and free cash flow – by about 3% per year without additional investment. I’m not sure if those numbers are exactly right. They may be 1% to 2% too aggressive combined. I would guess Omnicom stock is now priced to return no less than 8% per year and no more than 10% per year. In this case, it’s possible the opportunity cost of Omnicom doing anything is the up to 10% return it could get from buying back its own stock. If I was advising the company on how to allocate capital, I’d say that any use of cash has to be expected to return 11% per year or more. Otherwise, the company should just buy back its stock.
There is no evidence that dividends can provide more value to shareholders than buybacks at Omnicom. The company should use all its free cash flow to buy back stock and not pay any dividend at all. I know that it will pay a dividend. Unfortunately, it will probably pay something like one-third of free cash flow out in dividends and two-thirds out in share buybacks. That’s just a guess. Theoretically, I’d prefer all free cash flow at Omnicom be used to buy back stock and none used to pay dividends. That’s because it’s hard to prove that the average individual investor would be better off getting a dividend from Omnicom than simply having their ownership stake in the company raised over time. And then, like I said, the hurdle you would need to hit for the use of cash inside the company would be 11% a year. That’s after-tax. And that’s pretty hard to hit. It’s hard to negotiate a control purchase of a private company or a public one at an 11% annual return, even in cases where there’s a lot of synergy.
The more the company spends on buying back stock the happier I’d be. Of course, there are cases where one company acquiring another can really add value. The two examples that come to mind are 3G’s acquisitions at companies like Kraft Heinz (KHC) and the acquisitions that Prosperity (PB) has done in Texas, but mergers at these companies haven’t been effective because the prices paid were low – they’ve been effective because you have a cost-conscious culture acquiring a business that hasn’t been run efficiently.
I’m sure you know 3G’s approach already. Prosperity’s approach is to remove a lot of office expenses at the branches it takes over. It also gets rid of a lot of the loan book. Basically it is acquiring deposits from the acquired company and then changing its cost and lending culture to reflect Prosperity’s.
Since Prosperity has lower office expenses and lower loan losses than the banks it acquires, it achieves cost “synergies” that aren’t really synergies at all. They’re just management-led improvements. I’m always in favor of using cash on these kinds of mergers. The catch here is that the management of the acquirer has to be better than the management of the acquired company. To benefit from this, you need to be invested in a company with an above-average management team that you know is going to stay in place for the long term.
Generally, Wall Street tends to notice above-average management teams and eventually bids up the prices of their stocks to above market price-earnings (P/E) multiples. It’s tough to profit from this approach unless you go looking for less well-known managers who are as good as the already famous ones.
Disclosure:Â Long Frost, BWX Technologies.