# Return on capital: the first number I check _Profit tells you a business made money. Return on capital tells you whether it was any good at it, and whether growth will make its owners richer._ Neil Woodford · 21 August 2026 · 4 min read ![Return on Capital Employed (ROCE)](https://cdn.sanity.io/images/v3acfbvo/production/14a2c3b62532dccf4d2cf1818dc23859373a8555-2240x1260.png?w=1600&fit=max&auto=format) --- No single number can tell you whether a business is worth owning. Not even three can, if I am honest. But if you put a gun to my head and forced me to, the three I’d pick are the price-to-earnings ratio, debt-to-EBITDA, and return on capital. These three don’t give you an answer in themselves, but they do generate a starting point in a decision-making process. _(Neil in the margin: EBITDA is earnings before interest, tax, depreciation and amortisation — a rough proxy for cash profit. Dividing debt by it tells you how many years of profit it would take to clear the borrowings; under 3 is generally comfortable, above 4 starts to worry lenders.)_ Each metric gives a static, partial picture of how a business is performing today, but the harder, more important work is judging the quality characteristics of a business and its growth potential. None of this is a precise science; valuing a business is an educated guess, which is why Graham, Munger and Buffett spent more time understanding businesses than on the arithmetic. _(Neil in the margin: The lineage of value investing: Benjamin Graham wrote the textbook, and Charlie Munger and Warren Buffett built Berkshire Hathaway on his ideas, gradually shifting from cheap statistics towards paying up for genuinely good businesses.)_ But if I were forced (again, against my will!) to choose just one of those three, it would be return on capital. In case it wasn’t clear enough already, this is a forced and artificial choice. Real judgement needs the full picture. But the reason return on capital earns that forced ranking is that it comes closest to telling you whether a business is actually any good at the thing it exists to do. ## What is return on capital employed? Return on capital employed, or ROCE, is the operating profit a business earns divided by the capital tied up in earning it: the money invested in its factories, stock, premises, brands, and technology, and the working capital that keeps it running. A business that makes £20 of operating profit for every £100 employed earns a 20% return. One that makes £4 earns 4%. Both might report rising profits, but only one, arguably, is earning an adequate return. _(Neil in the margin: The day-to-day money tied up in running the business — stock and money owed by customers, less money owed to suppliers. It is capital you can't deploy elsewhere, so it rightly counts against the return.)_ ## Why this one Profit on its own tells you a business made money last year. It says nothing about what that money cost to produce. A company can grow its profits every year but still be a poor home for your savings. To get a more rounded picture of how good a company is at doing what it does is to look at ROCE, but this only gives a static picture. What's also important to understand is the sustainability of high ROCE and whether a business can compound high returns by investing additional capital at that rate. ## A worked illustration Picture two manufacturers, each making £10m of operating profit. The first has £40m of capital employed, the second £150m. The first earns 25% on its capital, the second under 7%. Now let each reinvest £10m to grow. The first business adds capacity that, at its 25% return, will earn roughly £2.5m more a year. The second spends the same £10m to earn itself under £700,000. Both have "invested for growth", and both will announce higher profits next year. But the first owner's money is working close to four times as hard. This is one of the reasons why two businesses reporting identical profits can often be worth very different sums, and also why the return-on-capital figure is the one I want to help me make better-informed judgements about the right valuation for a business. ![Two ways to make £10M](https://r4at4qm6kmohrtvq.public.blob.vercel-storage.com/charts/chartDoc.roce-worked-example-65543be53fd5-light.png) ## The number has to be durable In a competitive world, high returns will in time attract competition. If a business earns a 25% return on capital but then attracts competition content to earn 15%, that is probably something to worry about. So an important question to address is not just whether a business earns a high return today, but whether it can continue to earn it. That turns the conversation to the moat: whatever it is that stops competition from dragging the return back down to the ordinary. _(Neil in the margin: Buffett's term for a durable competitive advantage — brands, switching costs, scale, network effects — that keeps rivals from competing your returns away. The metaphor is the water around a castle.)_ ## Where the number can mislead Return on capital is the first figure I check, and it's also one I read carefully, because it can be flattered in several ways. Debt is the commonest. If a company borrows heavily and shrinks the proportion of equity in the business, this can lift the apparent returns (if the returns are above the cost of debt) whilst, of course, quietly raising the risk. A high return bought with a fragile balance sheet is a different animal from one earned with an appropriate mix of debt and equity. Acquisitions are next. When a company buys another, the premium it pays lands on the balance sheet as goodwill. Strip that out, and the returns can look spectacular; leave it in, and they can look ordinary. Investors need to establish which version they are being shown and why. _(Neil in the margin: The accounting plug for paying more than a target's net assets are worth. Including it in capital employed makes the buyer's returns honest.)_ Then there is the single good year. A cyclical business at the top of its cycle can post a great return that it may not see again for a long while. One year tells you little; the return averaged through a full cycle tells you a great deal more. _(Neil in the margin: One whose profits swing with the economic cycle — miners, housebuilders, chemicals. Peak-year returns can look wonderful precisely because they are about to mean-revert, which is why averaging through the cycle matters.)_ And there are asset-light businesses, which can show large returns on a small capital base. This is often a real strength, but it can also mean that the business model can be easily replicated, and so in this context, the right answer needs further analysis before admiration. None of this makes the number useless. It just means it's the beginning of the work, not the end of it. When a return looks unusually high, the discipline is to find out whether it is real before you let yourself be impressed. ## What it changes Return on capital is not the only number that matters, and a high one is no reason to own a business at any price. What you pay decides whether a good business becomes a good investment. A useful sense check is to always benchmark the return you could earn for no risk at all, in cash, and compare that with what you are getting in what will inevitably be riskier.