What a moat is, in numbers
Capitalism works by competing excess profits away. If a packaging plant earns 25% on its capital, someone builds another plant, prices fall, and returns drift toward the cost of capital, which is the return investors require and the point at which growth stops adding value. A company that keeps earning well above that level for a long time is being protected by something, and investors call that something a moat. The measure is the spread between return on invested capital and the weighted average cost of capital, which the what is ROIC article computes; the evidence is that spread persisting through a full business cycle and through at least one serious attempt by a competitor to take it.
The five sources
| Source | What it is | How it shows up | How it fails |
|---|---|---|---|
| Intangible assets | Brands, patents, licences, regulatory approvals | Price premium over identical products; gross margin above peers | Patent expiry; a brand that stops meaning anything to a new generation |
| Switching costs | The pain of moving to a competitor: retraining, data, integration, risk | Very high customer retention; price rises that stick | A new standard or platform that makes migration easy |
| Network effects | The product gets better as more people use it | Winner-take-most share; rising margins with scale | A rival network reaching critical mass; regulation of the dominant one |
| Cost advantages | Lower unit costs from scale, process, location or access to a resource | Lowest prices and still the best margins | A new technology that resets the cost curve; scale that stops mattering |
| Efficient scale | A market only big enough for one or two profitable players | Stable duopoly or monopoly returns; no entrants despite good margins | Market growth that invites a third player; a regulator that forces one in |
Naming the source matters because it predicts the failure mode, and a thesis is a statement about what has to stay true. A moat built on a patent has a known expiry; one built on switching costs is threatened by whoever makes switching cheap; one built on scale is threatened by a technology that makes scale irrelevant.
The test: three questions, ten years of data
- Has ROIC stayed above WACC for ten years? Not on average; in each year, including the bad ones. One strong decade with a collapse in the middle is a cyclical business, not a moated one.
- Can the company raise prices? Read the revenue breakdown in the filings and the earnings calls for price and volume. A business that raises prices 4% and keeps its volume has pricing power; one that raises prices and loses customers is being tolerated, not preferred.
- Is share stable or rising while margins hold? Rising share with falling margins means the company is buying its position; stable share with stable margins under a known competitive attack is the moat doing its job.
PackCo, the fictional packaging maker used across these guides, earns a 14.5% return on invested capital against a 9% cost of capital, a spread of five and a half points. If that spread has held for a decade through two downturns, PackCo has a moat, probably a cost advantage from scale and plant location, since packaging has no brand and little switching cost. It is a narrow one: a competitor with a newer plant nearer the customers could take it. Zenith Software, the other fictional company, earns far higher returns on little capital, and its moat is switching costs; its failure mode is a rival that makes migration painless.
Moats and valuation: usually already in the price
A moat changes a valuation in one place: the number of years over which above-average growth and returns can be assumed before fading to the economy's. In the DCF model example, the explicit forecast gives Harbor Foods five years of 8% growth before fading; a company with a proven moat might deserve ten or fifteen, and the difference is worth a great deal. But the market knows this too. Moated companies trade at high multiples precisely because their moats are visible, and the implied growth rate method will usually show that the price already assumes the moat holds for a long time. The investment question is rarely whether a company has a moat; it is whether the moat is wider or longer-lasting than the price assumes, which the priced for perfection article treats from the other side.
How moats erode, and the early signs
- Gross margin slips first. Pricing power leaves before volume does. Three years of gross margin drifting down while revenue grows is a moat being paid for.
- Returns on new capital fall below the average. The business keeps its historic ROIC on its old assets while each new dollar earns less; the ROIC calculator article explains the incremental figure.
- Customer concentration rises. A moat that depends on one or two customers is their moat, not the company's.
- Management starts talking about the moat. Companies that have one rarely need to say so; the word appears in calls when the evidence is leaving.
- A cheap substitute appears and is dismissed. The dismissal is the tell. Moats are rarely breached by a better product at the same price; they are breached by a good-enough product at a much lower one.
Writing the moat into a thesis
A moat belongs in a thesis as a condition with a number attached, not as an adjective. "PackCo has a scale moat" cannot be checked; "PackCo's ROIC stays above 12% and its gross margin above 30% through the next downturn" can, every quarter, from the filings. The investment thesis template has a place for exactly that condition, the when to sell article treats its failure as a sell signal rather than a reason to hope, and the investment thesis guide puts the whole method together. The value trap article describes what the same company looks like after the moat has gone and before the price admits it. VALYSTR's Synthesis workspace checks thesis conditions like these against each company's filed numbers as they arrive.
Check the moat against the filings
Synthesis tests the conditions in your thesis, such as a floor on return on capital or gross margin, against each company's filed numbers as they arrive.
Frequently asked questions
What is an economic moat?
A durable advantage that lets a company keep earning returns above its cost of capital despite competition. The evidence is a return on invested capital that stays above the weighted average cost of capital for a decade or more, with pricing power and stable market share. The five usual sources are intangible assets, switching costs, network effects, cost advantages and efficient scale.
How do you tell if a company has a moat?
Compute return on invested capital against the cost of capital for each of the last ten years and look for a spread that never closes, including in downturns. Then check for pricing power in the filings and calls: price increases that hold without losing volume. Finally confirm that market share and margins have held through a known competitive attack. A story without those numbers is not a moat.
Does a moat make a stock a good investment?
Not by itself. Moats are visible, so moated companies usually trade at prices that already assume the advantage lasts for many years. The investment question is whether the moat is wider or more durable than the price implies, which a reverse DCF can test, and whether the moat is eroding in ways the market has not yet noticed.
What is the strongest type of moat?
Network effects and high switching costs tend to be the most durable, because they strengthen as the business grows and are expensive for customers to leave; cost advantages from scale are strong but can be reset by new technology; brands and patents are powerful but have known expiry risks. The source matters less than whether the ten-year numbers confirm it.
How do economic moats erode?
Usually quietly: gross margin drifts down as pricing power weakens, returns on newly invested capital fall below the historic average, a cheap substitute appears and is dismissed, and customer concentration rises. By the time revenue or share falls visibly, the moat has often been gone for years, which is why the conditions in a thesis should be the early measures rather than the late ones.