1. Return on capital, and why it reverts to the mean

Level 5: Competition and duration

You know how to read a price as a forecast and break it down into sales, margins and investment. What is missing is the question that decides whether that forecast is sustainable or a fairy tale. Where does the money the market is taking for granted over the next fifteen or twenty years come from? From a business that earns more than it costs to fund itself. And that is precisely what the rest of the world spends its days destroying. This module is about the number that measures that advantage, and about the most stubborn regularity in fundamental analysis: that the number goes down.

What return on capital really measures

Return on invested capital (ROIC) compares after-tax operating profit with the money tied up inside the business to produce it: premises, machinery, inventory, what customers owe and have not yet paid, the acquisitions already digested. It answers an owner's question, not an accountant's: for every hundred euros locked up in here, how many come out in a year?

Three common confusions worth killing off at the root:

• It is not the margin. A business can have a splendid margin and a mediocre return if getting there meant burying a fortune in assets. And the other way round: a distributor on a 2% margin can have a very high return if it turns its capital over twenty times a year.
• It is not return on equity. That one inflates with debt without the business improving one jot. Return on capital looks at the business; return on equity mixes the business up with the financing structure.
• It is not earnings per share. It says nothing about the capital it took to generate it, and it is the easiest metric of all to dress up.

There is value only above the cost of capital

Here is the arithmetic to commit to memory. Take a business with a thousand euros of invested capital and a cost of capital — what those who put up that money demand, shareholders and lenders combined — of 8%.

• If it earns 80 euros a year, its return is 8%. It hands back exactly what is asked of it. It has created zero value: it is worth its thousand euros of capital and not a cent more.
• If it earns 150, its return is 15%. Every year it leaves 70 euros above what was demanded, and that excess, only that excess, is what can justify it trading well above the capital it holds inside.
• If it earns 40, its return is 4%. It destroys value at a rate of 40 euros a year, even though its income statement says profit and even though the headlines call it a profitable company.

Stop at the third case, because it is the one most people misread. Making money and creating value are different things. Accounting profit does not deduct what shareholders' capital costs, and that cost exists even though it never arrives as an invoice. A business with a positive profit and a return below its cost of capital is worth more shut down than running.

The most stubborn regularity in fundamental analysis

Now the uncomfortable part. If a business earns 25% on capital in a world where money costs 8%, somebody else notices. A competitor comes in. A supplier raises its price. A large customer demands a discount. An executive walks out and sets up the same thing. That excess return is, to everybody else, a neon sign saying there is easy money here.

The aggregate result, measured over and over across decades of listed companies — Michael Mauboussin's work on the reversion of return on capital is the usual reference — always draws the same picture: rank companies by return, follow each group for ten years, and the top group comes down, the bottom group comes up, and both drift toward the average. They do not converge entirely, that nuance matters and it is the next module, but they close far more of the gap than almost anyone expects.

Marathon Asset Management gave the mechanism its name in their writing on the capital cycle: high returns attract capital, new capital adds capacity, capacity sinks prices, returns fall, capital flees, capacity is withdrawn and returns climb again. Nobody has to be stupid: it is enough for each participant to behave reasonably while looking only at their own project.

How fast it happens

What changes a valuation is not that reversion exists, but how long it takes. The usual measurements on the US market point to much of the advantage of the most profitable group eroding over something like a decade, and to staying at the top for twenty or thirty years running being exceedingly rare: a handful of companies per generation, not a whole category.

Translate that into reverse-engineering the expectations. When you take a price apart and find it only adds up with a return on capital of 30% sustained for twenty years, you are not looking at an optimistic sales forecast. You are looking at the claim that this particular company belongs to that historical handful. Perhaps it does. But the burden of proof has moved: it is no longer enough to say the sector is growing, you have to explain what stops anyone else from coming in for two decades.

A real case: when moving a container cost ten times more

Container shipping between 2020 and 2023 is a capital cycle compressed into three years. With supply chains jammed, moving a container on the main routes came to cost something in the order of seven to ten times the normal rate, and the shipping lines went from mediocre years to unprecedented profits. Maersk reported a net profit of the order of $29 billion in 2022, more than it had earned in several previous years put together.

And what did the industry do with that money? Order ships, in record numbers. When those ships began to be delivered and demand normalized, freight rates went back to ordinary levels and profits collapsed with them.

Take those figures as an order of magnitude, not as audited data. What is not approximate is the pattern: anyone extrapolating that return in 2022 was not being optimistic, they were ignoring that the return itself was the cause of its own disappearance. This is education about how businesses behave, not an opinion on any company.