In the last lesson we said high profits “get competed away” as if it were obvious. It isn’t — it’s a precise economic process with a precise destination, and unless you can state the mechanism exactly, you can’t tell which businesses will escape it. So this lesson does two things: it builds the engine of competition piece by piece until you can see why the gravity pulls so hard, and then it hands you the single test that defines a moat. Everything else in the course — the five moats, the mirages, the erosion — hangs off this one engine.
Return on capital vs. the cost of capital
Start with the two numbers the whole model turns on. A business takes in capital — money from owners and lenders — and uses it to earn a return. The return on invested capital (ROIC) is just that return as a percentage: earn 100 of capital and your ROIC is 15%. Against it sits the cost of capital: what that money costs you, which is really what your investors could earn somewhere else at similar risk. If investors could get 8% elsewhere, your cost of capital is about 8% — that’s the bar you must clear just to justify existing.
The gap between them is the entire story:
- ROIC > cost of capital. You’re earning more than your capital costs. This surplus is the excess return — economic value created, the treasure in the castle. It is also a flashing sign that says attack me.
- ROIC = cost of capital. You’re earning exactly the going rate. The business is fine, the lights stay on, but there’s no surplus — nothing for a rival to gain by copying you, and nothing for a moat to protect. Economists call this “normal returns,” and it is the default destination of competition.
- ROIC < cost of capital. You’re destroying value — earning less than your capital could make elsewhere. Do this long enough and the capital leaves.
Why the destination is 'normal returns'
Here’s the logic in one breath. Excess returns attract entrants. Entrants add supply and compete on price, which pushes returns down. New entrants keep coming as long as there’s any excess left to grab. They stop only when the excess hits zero — when ROIC has fallen to the cost of capital and there’s no longer a reason to enter. So the stable end-state of unprotected competition is precisely the point where excess profit has vanished. That’s not cynicism; it’s just where the entry pressure runs out.
Competition is gravity
Picture excess return as a ball you’ve carried to the top of a hill. The market’s natural tendency — entrants, imitators, price competition — is gravity, forever pulling that ball back down to the valley floor of zero excess. Left alone, every ball rolls down. The only way a business stays up the hill is if something physically holds it there. That something is the moat. Without it, the question isn’t whether your excess return decays to zero, only how fast.
This reframes the whole job. People instinctively ask “is this a good business?” The sharper question is: “what’s holding this business up the hill, and how strong is it?” A wonderful product with nothing holding it up is a ball already rolling. A mediocre product wedged behind an immovable rock can sit up the hill for decades. Let’s make the two forces — the gravity of competition and the grip of a moat — something you can actually watch.
Excess returns over time
Excess returns: compound up the hill, or melt to the valley
A firm starts earning fat returns above its cost of capital. Set how strong its moat is and how hard rivals push. When pressure beats the moat, the excess melts to commodity zero; when the moat wins, it compounds.
Moat 3/10 against pressure 6/10: the firm starts 20 pts above its cost of capital and ends at 1 pts after 24 years — a mirage — fully competed away to commodity returns.
Notice what the island makes concrete. With a weak moat against strong pressure, the curve dives to the commodity baseline — that’s the default fate. Crank the moat above the pressure and the curve bends upward: the advantage doesn’t merely survive, it compounds, because some moats (a growing network, accumulating scale) get stronger the longer they run. That asymmetry is the heart of the course: most advantages erode, a rare few widen, and the entire value of a business over its life is decided by which curve it’s on.
On the island, you set a high moat strength against low competitive pressure and the excess-return curve bends upward over time instead of staying flat. What real-world phenomenon is this picturing?
The one test that defines a moat
Now the payoff. With the engine in view, a moat has a clean, operational definition — the test you’ll apply to every business for the rest of this course and the rest of your life:
The moat test
A moat is an advantage that survives perfect imitation. Ask: if a well-funded, competent rival copied this company feature-for-feature tomorrow — same product, same prices, same people — would the advantage still be there? If yes, the advantage lives in something the rival can’t copy by copying the product (a network, switching costs, scale, a unique asset) → moat. If no — if perfect imitation erases it — then whatever you had was copyable, and copyable advantages get competed away → not a moat.
This test is sharper than it looks because it forces you past the product. A glorious product fails the test: copy it exactly and the glory is gone. A network passes it: a rival can clone your app perfectly and still face an empty network while yours is full, because the value isn’t in the code, it’s in the crowd already there. Run the test and you stop being dazzled by quality and start seeing structure.
Apply the moat test. A ride-hailing app has a beautifully designed interface that users love. A rival hires a top studio and clones the interface pixel-for-pixel, matching it exactly. Which has a moat — and why?
Because the temporary excess return is still worth a great deal — and because the race to earn it is exactly what drives progress. An innovator earns fat margins for the months or years before rivals catch up; that prize is the incentive that funds the innovation in the first place (straight from your Incentives course). Society gets the better product and the eventual low price once competition arrives. The moat lens doesn’t say innovation is pointless — it says innovation alone rarely produces a durable advantage. To keep the prize past the catch-up point, you need the innovation to build something un-copyable: a network, a dataset, a brand, a scale position. The clever firms don’t just innovate; they use the temporary lead to dig a moat before the lead runs out.
Why “we’ll just run faster” isn’t a moat
The most tempting non-moat is effort. “Our advantage is that we execute better / ship faster / work harder than anyone.” Feel the problem against the engine: effort is a relative race on a copyable dimension. Your rivals can also work hard, hire well, and ship fast — and the moment they do, your edge reverts. You met this exact creature in the Red Queen Effect: running flat out to hold a position that others are sprinting to match. It can keep you alive (fall behind on execution and you die), but it doesn’t hold the ball up the hill, because there’s no rock — just your own legs, which tire, and rivals whose legs are just as good.
A moat is the opposite of a treadmill. The treadmill demands continuous effort to stay in place; the moat holds you in place structurally, whether or not you sprint. That’s why Buffett prizes businesses a moat protects even from mediocre management — “so good an idiot could run it, because sooner or later one will.” A castle that needs a hero on the walls every single night isn’t well-defended; a castle with a wide moat is safe even when the guard falls asleep. Structure beats heroics.
A logistics company says its moat is that 'our team simply out-executes everyone — we're faster and more disciplined than any competitor.' Why is this not a moat in the strict sense?
Recap
- Every business lives between two numbers: return on capital (what it earns) and cost of capital (what its money could earn elsewhere). The gap is the excess return — the treasure, and the target.
- Competition is gravity. Excess returns attract entrants who compete them down; the stable destination of unprotected competition is normal returns (excess = zero). The question is never whether an unprotected advantage decays, only how fast.
- A moat is whatever holds the ball up the hill against that gravity. A rare few moats don’t just hold — they compound, widening the lead over time (networks, scale).
- The moat test: an advantage is a moat only if it survives perfect imitation — if a rival copying the company feature-for-feature still couldn’t erase it.
- Effort is not a moat. Out-executing rivals is a copyable, relative race (the Red Queen treadmill). A moat is structural — it protects you even when you ease off.
You now have the engine and the test. Next we open the toolbox: the five real moats — the specific structures that pass the test — each with how it works, a worked example, and exactly how a clever attacker can still breach it.