Why returns matter
Since it’s something I get asked a lot – understandably, since I never used to use return metrics at all in my investing immaturity – I thought I’d do a quick post on why returns metrics matter so much, how I calculate returns, and what they mean. I use return on capital on more or less every company I look at, with a few exceptions, as I think it’s a figure which tells us a lot about how the company is positioned and what we can expect.
Returns on capital
Return on capital matters because it’s the metric which makes the most economic sense. Bear with me as I whizz through the economic intuition. Every (good!) investment – we’ll take a widget factory – earns a return. To calculate a return on capital, we need to know how much capital is employed in the business first, and secondly how much operating profit the business makes. If the business has a factory worth £5m and £5m worth of equipment and working capital (almost every company has to keep some stock around, for instance), it has £10m of operating capital. If it makes £1m a year in operating profit, return on capital is 10%. Simple enough.
Why is this figure interesting? Well, more so than any other metric – a lot of investors like to use margins, for instance – return on capital tells us about the competitive position the business we’re looking at is in. A high return on capital denotes, if it’s sustained over a long period of time (and therefore not noise – data always has lots of noise!), a competitive advantage.
It’s not strictly the return on capital per se which denotes a competitive advantage, though; it’s the difference between the return on capital and the cost of capital. The cost of capital is something of a more intangible metric, but it’s easiest to think of it as compensation for an investor’s risk. Utility companies tend to have lower cost of capitals because their returns aren’t particularly volatile – come rain or shine, they have pricing power and a customer base which needs electricity and water. Debt holders (say, banks) and equity holders (us – shareholders) demand less return in exchange for less volatility. This partly explains why small-caps are usually cheaper (cheaper = higher returns, ceteris paribus) – they’re more volatile.
Back to our example of the widget factory – if they can borrow money at 6% and invest in the business to earn 10%, we’ve got an attractive proposition on our hands as shareholders. The difficulty, of course, is in sustaining a premium over your cost of capital. If you earn a premium on your cost of capital, competition is incentivised. Capital always flows to where the returns are – that’s how capitalism works – and, thus, rival widget factories may open up. Competition reduces margins or splits revenues, and returns on capital tend toward cost of capital.
So it is in most cases – theory and empirical research shows that high returns on capital trend towards the mean over time. An investor’s dilemma basically becomes working out whether a high cost of capital is sustainable, or to what degree it is, or if a low cost of capital will permeate. The expectation, I think, should be for both to move toward the mean, barring any strong evidence either way – say a dominate management stake who show no interest in improving returns for shareholders on the downside, or a business model which effectively entrenches the company with their customers on the upside.
It sounds relatively simple, then; look for companies with high and sustainable returns on capital and growth potential, since growth potential means they can take extra money (either by debt or equity, or simply by reinvesting earnings) and compound at a rate higher than something like a simple P/E metric might suggest. The first barrier to that is the most obvious – companies like this aren’t usually cheap. These are high quality companies, and usually command prices as such. The closest I ever felt I came to having this ideal – a really high quality company at a cheap price – was Howden Joinery. A combination of legacy issues and the clouding over of its true potential by a number of transient factors saw their price nosedive in the recession, and it’s more or less tripled since.
The more practical issue is around the calculation of return on capital. There are three main points of contention I always wrestle with when I’m looking at a company, and I don’t really have any good answers to any of them. They’re all simply guesses at the underlying economics.
Firstly, on property/plant/equipment; how close is the book value to fair value for their assets? Companies that have held factories on balance sheet for long periods of time, for instance, have often depreciated them down to a very low value – when, in actuality, the value of the property has appreciated. The difference is manifold, and understating the value of their assets serves to make the company look better than it should. After all, if our widget company could sell its factory for £20m, the rational thing to do would be to simply sell it and return its money to investors – it isn’t earning enough of a return to justify that investment. Using gross PPE before depreciation helps, but it also helps to know what the assets are. Some assets genuinely do have a short lifespan, which depreciation nicely approximates.
Secondly, and in some cases a complete non-issue, cash on balance sheet. Some cash is nearly always required for businesses to operate – they need to meet short term liabilities, lubricate the wheels between payments and incoming cash and so on. Seasonality issues mean that some companies will have more than others – remember that the balance sheet is a snapshot at one point in time, and it’s often timed quite carefully from the company’s point of view. 3 months earlier, their cash pile might have been sorely stretched. On the other hand, the cash on the balance sheet might all be excess – completely unecessary for the functioning of the business, and just there because management want to make acquisitions or grow organically, or are just hesistant to pay it out as dividends. This is also a possibility. The designation of cash as excess of otherwise changes your operating capital metric, too. Most companies give an idea as to what their cash is needed for, and I find the best way of deciding is usually to simply look at the past. Cash balances as a percentage of either revenues or working capital sans-cash, for instance.
Finally, the ever present operating leases. As I’ve mentioned before, I capitalise these in a way similar to that which Moody’s, the ratings agency, uses. I say that because there’s a nice pdf summing up their methodology here. Loosely, though, it’s attractive because – unlike some other ways of capitalising operating leases – it tries to capture the economic value of the asset as if it had been purchased by the company. This differs from other methods, which attempt to simply sum up the obligation of lease payments and discount them to present value. This will serve to understate how much economic assets the company is employing.
Tediously, capitalising operating leases like this means you have to adjust the income statement somewhat, too, including your measure of operating profit – since we’re sort of pretending the company purchased the asset, some of what the company used to call ‘operating rent expense’ is now what we’d call depreciation – an operating expense – and some of it is the effective interest in that rent.
Doing so, though, means you have a far better – and far more reasonable return on capital metric – one that better matches economic reality.
That’s why RoC is a key part of my investing toolbox. Metrics like P/E and P/TBV – particularly together, since they hint at the RoC (without lease adjustments) – give an indication, but not a particularly close picture. If investing is staying one step ahead of the game and realising where value is less obvious, I think refining the way you understand returns is a particularly good way of doing that.