In a competitive market, high profits are supposed to attract competition until those profits disappear. That is the mechanism working as designed. A moat is whatever stops it from happening: the structural reason a company can keep earning unusually high returns while rivals try and fail to take them away. It is the single most important question in long-term investing, because without a moat, even a great year is just a head start that competitors will eventually erase.
The five types of moat
Almost every durable advantage falls into one of five buckets. Recognising which one a company has matters, because they erode in different ways and on different timescales.
| Type | How it works | Typical example | How it usually dies |
|---|---|---|---|
| Cost advantage | The company can produce or distribute more cheaply than anyone else, so it can undercut rivals and still profit. | Large-scale retail and logistics networks. | A new technology or business model resets the cost curve for everyone. |
| Network effects | Each additional user makes the product more valuable to every other user, so the leader gets stronger as it grows. | Payment networks, marketplaces, social platforms. | The network fragments, or users adopt several networks at once. |
| Switching costs | Leaving is expensive in money, time, risk, or retraining, so customers stay even when a rival is cheaper. | Enterprise software embedded in daily operations. | A competitor absorbs the migration cost, or standards make data portable. |
| Intangible assets | Patents, licences, regulatory approvals, and brands legally or psychologically block substitution. | Pharmaceutical patents, premium consumer brands. | Patents expire on a known date. Brands decay slowly, then quickly. |
| Efficient scale | The market is only large enough to support one or two profitable operators, so entering is irrational. | Regional utilities, pipelines, some rail networks. | The market grows enough to support another entrant, or regulation changes. |
How to see a moat in the numbers
A moat is a qualitative idea, but it leaves a quantitative fingerprint. The clearest one is a high return on invested capital sustained over many years. If a business consistently earns far more on the capital it deploys than its rivals do, something structural is protecting it, because otherwise that gap would have attracted enough competition to close it.
- Return on invested capital that stays well above the company's cost of capital for a decade or more. One good year proves nothing; persistence is the evidence.
- Gross margins that hold steady or improve while competitors' margins compress. This is pricing power showing up in the accounts.
- Market share that holds or grows even while rivals spend heavily to attack it.
- The ability to raise prices roughly in line with or ahead of inflation without losing customers.
- Low customer churn, or revenue that recurs without having to be re-won each year.
The order matters. Start with returns on capital, because that is the outcome a moat produces. Then look for the mechanism that explains it. A company with high returns and no identifiable mechanism is usually enjoying a cycle rather than holding an advantage.
What is not a moat
A large part of moat analysis is discarding things that look like advantages and are not. Each of these is regularly cited as a moat and none of them survives contact with a determined competitor.
- Being big. Scale only helps if it lowers unit costs or strengthens a network. Size that does neither is just more of the same problem.
- A good product. Products are copied. The question is what stops a competitor selling the same thing more cheaply next year.
- Brand recognition. Everyone recognises plenty of brands they would abandon instantly for a lower price. What matters is whether the brand supports a price premium.
- First-mover advantage. Being early is only valuable if it lets you build one of the five real moats before rivals arrive. Otherwise it means paying to educate a market your competitors will then sell into.
- A talented management team. Genuinely valuable, and genuinely mobile. People leave; structural advantages do not.
- A temporary technology lead. In most industries a lead is measured in months. It becomes a moat only when patents, switching costs, or network effects convert it into something durable.
Moat width and moat direction
Two questions matter and they are separate. How strong is the advantage today, and is it getting stronger or weaker? The second usually matters more to an investor, because the current advantage is already reflected in the price while the direction of travel often is not.
Moat erosion: what to watch for
No moat is permanent. Technology dissolves cost advantages, patents expire on a fixed date, regulation changes, and tastes shift. The useful thing is that erosion shows up in the numbers well before it shows up in the narrative.
- Gross margin drifting down over several years, particularly if revenue is still growing. It means the company is buying growth with price.
- Rising customer acquisition cost, or marketing spend growing faster than revenue.
- Churn ticking up, or contract lengths shortening.
- Market share slipping to a specific named competitor rather than to the market in general.
- The company competing on price for the first time in a segment it used to dominate on quality or lock-in.
- Management explaining a weak quarter with factors it described as irrelevant a year earlier.
History is full of companies that dominated a category, held early patents in the technology that replaced it, and still lost. The moat did not fail because management was stupid; it failed because the thing it protected stopped being what customers wanted. Monitoring a moat is not a one-time check, it is a question to re-ask every year.
Why the moat outranks the quarter
A company can post a great quarter on a crumbling moat, or a weak quarter on an unbreakable one. For a long-term investor the moat wins the argument every time, because it determines whether the business can absorb a setback and keep compounding. A wide moat buys time: time to fix a mistake, reinvest, and come back. A company without one has to be right every year.
This is why serious analysis spends as much energy on competitive positioning as on the financial statements. The statements describe what already happened. The moat is the closest thing available to evidence about what happens next, which is the only part you are actually buying.