Loading market quotes…

Subscribe free
Market Fieldnotes13 minFreeGBC Research

Economic moat: how to spot a competitive advantage

In short

An economic moat is a competitive advantage that protects a company’s profits from competitors for many years. The image comes from Warren Buffett’s letters to Berkshire Hathaway shareholders. In the financial statements it shows up as four signs: return on invested capital (ROIC) above the cost of capital almost every year for a decade, gross margins that move only a few points, prices that rise faster than inflation without losing customers, and customers who stay and spend more.

In 1972 Warren Buffett bought See’s Candies, a maker of boxed chocolates, through Blue Chip Stamps, for $25 million. The company sold 16 million pounds of candy a year, had sales of $30 million and earned less than $5 million before tax, on $8 million of capital.

By 2007 See’s sold 31 million pounds, less than double. Sales had grown to $383 million and pre-tax profit to $82 million, on $40 million of capital. In his letter to shareholders for that year, Buffett ran the totals: since 1972 See’s had produced $1.35 billion in pre-tax earnings while requiring only $32 million of additional investment.

Volume fell a little short of doubling while earnings rose more than sixteenfold. Those numbers hold proof of something competitors failed to copy for 35 years, and one division will find it.

When you read that a company “has a huge moat,” ask for three numbers: how much the capital earns, how stable the margins are, and who sets the price. If nobody can give them to you, the advantage is an opinion, and opinions are a poor reason to risk your own money.

What is an economic moat, and where does the term come from?

A moat is the ditch around a castle. Buffett has used the image since at least the 1990s: in his letter for 1993 he wrote that the might of Coca-Cola’s and Gillette’s brand names, the attributes of their products and the strength of their distribution systems were “setting up a protective moat around their economic castles.” In the letter for 2007 he gave the tighter version: “A truly great business must have an enduring ‘moat’ that protects excellent returns on invested capital.”

A business that earns high returns draws competitors, who cut prices until those returns are normal again; Buffett writes that competitors will repeatedly assault any castle earning high returns. The moat can be the lowest cost in the industry (he cites GEICO and Costco) or a powerful worldwide brand, such as Coca-Cola, Gillette and American Express.

Morningstar turned the idea into a rating with five sources of advantage: switching costs, cost advantage, intangible assets (brands, patents, licenses), network effect and efficient scale, meaning a market that only supports a few players. It calls a moat wide when its analysts think it can last at least 20 years, and narrow when it should hold for at least ten.

Which numbers reveal a competitive advantage?

Return on invested capital (ROIC)

ROIC (return on invested capital) measures how much after-tax operating profit a company produces for every dollar tied up in the business: operating income × (1 − tax rate) ÷ invested capital, where invested capital is shareholders’ equity plus financial debt minus cash.

Take a hypothetical company with $800 million in revenue and a 17% operating margin: operating income is $136 million. At a hypothetical 21% tax rate, $107.4 million is left. With $420 million of equity, $180 million of debt and $100 million of cash, invested capital is $500 million and ROIC is 21.5%. If you have never read these statements, start with our guide on how to read the income statement, balance sheet and cash flow statement.

ROIC has to be compared with the cost of capital, the return shareholders and lenders demand for financing the company (a hypothetical 9% in our examples). A moat shows when ROIC stays above the cost of capital for many years in a row, recessions included.

Stable gross margins

Gross margin is what is left of revenue after the direct cost of what the company sells. A business with pricing power passes higher raw material costs on to customers and its margin barely moves; one without it watches the margin shrink with every wave of costs. Measure the gap between the highest and lowest gross margin of the past ten years: a few points signal control, fifteen or twenty are the signature of a cyclical industry. Operating margin and revenue growth get a guide of their own.

The capital it takes to grow

See’s produced $42 of pre-tax earnings for every dollar of additional capital. For a listed company, the closest signal is cash conversion: free cash flow divided by net income. If it sits around 80% or higher for years, the company grows without reinvesting almost everything it earns. Below 50%, much of the accounting profit ends up in plants, inventory and receivables, and the advantage is expensive to keep.

Moat or cycle? How to tell an advantage from a golden year

Before you look at the chart, one question. In year 5, hypothetical Company B earns 33% on its capital and Company A earns 20.4%. Looking at that year alone, which of the two has a moat? Write your answer down, then look at all ten years.

A moat looks like a high, flat lineA moat looks like a high, flat line. ROIC of two hypothetical companies over ten years, against a 9% cost of capitalFIG. 2 · MOAT OR CYCLE?GB.CAPITALA moat looks like a high, flat lineROIC of two hypothetical companies over ten years, against a 9% cost of capitalCompany ACompany BCost of capitalROIC0%10%20%30%40%50%Year 1Year 4Year 7Year 10B: from 33% in year 5 to4% in year 9A: never below 17.8%Hypothetical example for illustration. ROIC = after-tax operating profit ÷ invested capital.
Figure 2 One golden year looks like a competitive advantage. Ten years of data separate a moat from a cycle.

In year 9, B earns 4%, less than half its cost of capital. That is the profile of a cyclical business, which earns a fortune when demand and prices are high and almost nothing when they turn. A never dropped below 17.8%.

Ten years of two hypothetical companies, hypothetical 9% cost of capital
MeasureCompany ACompany B
Average ROIC19.6%14.6%
Highest ROIC21.5%33.0%
Lowest ROIC17.8%4.0%
Years above cost of capital10 of 105 of 10
Gross margin range2.6 points18 points

B’s average ROIC, 14.6%, sits above 9%, and it hides five years out of ten below the line. Anyone who bought in year 5 would have watched earnings melt away within three years. That is why the scorecard at the end counts years above the line and ignores the average. If you picked B, you reasoned like someone buying a cyclical company in its year of record profits.

How do you measure pricing power?

Pricing power is the ability to raise prices without losing enough customers to cancel out the gain. Go back to See’s and do the division we left hanging. In 1972 average revenue per pound was about $1.88 ($30 million divided by 16 million pounds); in 2007 it was $12.35 ($383 million divided by 31 million), 6.6 times as much. Over the same period the US consumer price index, on an annual average basis, went from 41.8 to 207.3: about five times.

Price outran inflationPrice outran inflation. How many times each line grew in 35 yearsFIG. 1 · SEE’S CANDIES, 1972 → 2007GB.CAPITALPrice outran inflationHow many times each line grew in 35 yearsPounds of candy sold×1.9US consumer prices×5.0Capital required×5.0Revenue per pound×6.6Total revenue×12.8Pre-tax earningsover ×16.4Source: Warren Buffett’s 2007 letter to Berkshire Hathaway shareholders; CPI-U annual average, FRED/BLS.Pre-tax earnings were under $5 million in 1972, so that multiple is a floor.
Figure 1 Volume barely doubled, price per pound beat inflation and capital grew roughly in line with prices: an economic moat seen through the numbers.

See’s raised prices faster than inflation for 35 years and nearly doubled the pounds it sold. Capital went from $8 million to $40 million, five times, in line with consumer prices: in real terms it barely moved. In 2007 pre-tax earnings were 205% of the capital employed. The combination to look for is price per unit beating inflation, volumes holding up and capital that doesn’t balloon.

For a listed company you can run the same test when it reports volumes (units, subscribers, passengers): divide revenue by volume, year by year, and compare growth in revenue per unit with inflation. Check product mix too: selling more expensive products can lift the average without any price increase on the same product. In management’s discussion of results in the annual report, look for price, mix and volume, the terms many consumer goods companies use to break down their growth. A company that grows only through volume and cuts prices to defend it is buying market share and paying for it in margin. Our guide to researching a company through its SEC filings shows where to find them.

How do you read customer retention and switching costs?

Subscription businesses often report net revenue retention (NRR): what last year’s customers pay today compared with what they paid then, after churn, downgrades and upsells. In a hypothetical example those customers paid $10 million a year; then churn takes away $0.6 million, downgrades $0.2 million and upsells add $1.8 million. The same group now pays $11 million: NRR of 110%, gross retention (losses only) of 92%. Above 100%, a company grows without signing a single new customer. Every company defines it its own way, so read the footnote.

Behind high retention there is often a switching cost, what it costs a customer to change suppliers. Picture a company paying $100,000 a year for software while a rival offers the same service for $80,000. Data migration costs $50,000, and training and downtime another $30,000 by estimate. The $20,000 annual saving pays back the migration in two and a half years and the full bill in four.

20% cheaper, more expensive over three years20% cheaper, more expensive over three years. A software customer weighing a rival, over a three-year horizonFIG. 3 · THE COST OF SWITCHINGGB.CAPITAL20% cheaper, more expensive over three yearsA software customer weighing a rival, over a three-year horizonRival: 3 years of fees$240,000Data migration+$50,000Training and downtime (est.)+$30,000Total if it switches$320,000Total if it stays: 3 × $100,000$300,000Hypothetical example. Current vendor $100,000 a year, rival $80,000. Training and downtime are estimates;no discounting of cash flows.
Figure 3 Switching costs in a hypothetical example: the rival is 20% cheaper per year and more expensive over three years.

Over three years, staying costs $300,000 and switching $320,000. At an annual fee of about $106,700, 6.7% more, staying would cost roughly the same as switching over three years. That is how switching costs turn into pricing power. Over five years the math flips ($480,000 against $500,000), and the rival may offer better features. Switching costs buy time, and the proof a company uses that time well is retention that stays high after price increases.

Which numbers fake a moat?

Besides Company B’s cyclical peak, there are three false positives.

The customer worth a quarter of revenue

US accounting rules (ASC 280) require public companies to disclose any single customer worth 10% or more of revenue, with the amount, though not always the name. A company growing 40% because one customer multiplied its orders depends on a negotiation you can’t see.

Growth paid for with margin

If revenue grows 30% and gross margin loses five points, check discounts, input costs and sales mix. Those two figures alone do not show what caused the margin decline or whether the competitive advantage has weakened.

The famous brand without pricing power

A brand’s fame shows up in surveys; pricing power only shows up after a price increase. If volumes hold and margin rises, the brand is working for the company; if every increase needs promotions to win customers back, fame never became an advantage.

Is a moat enough to buy a stock?

No. Business quality and share price are two separate questions. At 40 times earnings, each $100 of share value corresponds to $2.50 of annual accounting profit. At 20 times, it corresponds to $5. Neither figure is a cash payout or a promised investment return. Even with the moat intact, at 40 times earnings have to double just to reach the starting point of someone who paid 20. The multiple is explained in our guide to the price-to-earnings ratio. Obvious moats tend to be priced richly, because the numbers are public; the better hunt is for an advantage still forming, with ROIC rising year after year and margins settling.

The scorecard organizes evidence, rather than certifying a moat. Distinguish an undisclosed metric from one inapplicable to the industry: retention does not measure software and steelmakers in the same way. Compare scores only across the same applicable rows. For ROIC, state whether invested capital is averaged or measured at period end, which adjustments are used and how much cash is considered excess; an assumed WACC remains a test assumption.

Frequently asked questions

What is an economic moat?

It is a durable competitive advantage that stops competitors from eroding a company’s profits. The term comes from the image Warren Buffett used in his letters to Berkshire Hathaway shareholders: a castle that earns high returns attracts attacks, and the moat protects it. In the numbers it shows up as a high, stable ROIC over many years.

What are the sources of a competitive advantage?

Morningstar identifies five: switching costs, cost advantage, intangible assets (brands, patents, licenses), network effect and efficient scale. A company can have more than one. For each, name the metric that supports it, compared with its history and competitors, such as retention for switching costs or revenue per unit against inflation for a brand.

How do you calculate ROIC?

Multiply operating income by one minus the tax rate and divide by invested capital, which is equity plus financial debt minus cash. With $136 million of operating income, a 21% tax rate and $500 million of invested capital, ROIC is 21.5%. Compare it with the cost of capital and track it over many years.

What is the difference between a wide moat and a narrow moat?

In Morningstar’s classification, a wide moat is an advantage its analysts expect to last at least 20 years, and a narrow moat at least ten. These are estimates about the future, reasoned opinions that make a useful starting point and still need checking against the financial statements.

Can a moat disappear?

Yes. New technology, rule changes or a management team that squeezes customers can shrink it within a few years. The first signs are ROIC falling for several years in a row, gross margin losing points and retention slipping, which is why an evidence scorecard needs rereading with every annual report.

Is a company with a moat always a good investment?

No. The moat is about the quality of the business, while the return on your investment also depends on the price you pay. At 40 times earnings, each $100 of share value corresponds to $2.50 of annual accounting profit. At 20 times, it corresponds to $5. These are not cash payouts, and even a solid advantage may not justify a price that is too high.

Sources and further reading

Related research
Operating margin: formula, examples and operating leverageMarket Fieldnotes · 05.10.2026P/E ratio: price-to-earnings formula, limits and trapsMarket Fieldnotes · 05.10.2026Portfolio diversification: correlation and hidden exposureMarket Fieldnotes · 05.10.2026How to compare ETFs: holdings, overlap, costs and the KIDMarket Fieldnotes · 05.10.2026Famous investors: 10 ideas to apply to your own accountMarket Fieldnotes · 05.10.2026 Full archive →
FREE · EVERY SUNDAY

Get the next Radar
before Monday’s open.

The first name, the reasoning, and our public research.

FREE · EVERY SUNDAYGet the watchlist