In 30 seconds
- An economic moat is a durable competitive advantage that keeps rivals from taking away a company's high returns.
- Its five sources are intangible assets, switching costs, network effects, cost advantage and efficient scale.
- You measure it by its footprints: a ROIC well above the cost of capital for years, and margins that hold up through recessions.
- Moody's kept an operating margin of 42.6% in 2008 and 43.4% in 2025, with a ROIC of 23.9% in 2025.
- Intel shows how a moat breaks: its gross margin went from 61.7% in 2018 to 32.7% in 2024, before the losses arrived.
Introduction
In 2008, the year its business landed at the center of the financial crisis, Moody's brought in $1,755 million of revenue, 22% less than in 2007, and still earned $748 million in operating income: an operating margin of 42.6%. In 2025, on $7,718 million of revenue, that margin was 43.4%. Seventeen years, a crisis, a legal settlement running into the hundreds of millions and competition from all comers, and the business still earns the same on every dollar it sells.
That's a moat. In his 2024 letter to shareholders, Warren Buffett listed Moody's alongside Apple, American Express and Coca-Cola among the household-name businesses Berkshire owns a piece of, and noted that many of them "earn very high returns on the net tangible equity required for their operations." This lesson explains where that kind of advantage comes from, how you measure it and how a company loses it.
Explanation
Buffett summed it up in his 2007 letter: "A truly great business must have an enduring 'moat' that protects excellent returns on invested capital." The idea is simple. A business that earns a lot attracts competitors, and competitors cut prices until returns become ordinary. The moat is what stops that from happening: an advantage a rival can't copy, or one that isn't worth the cost of copying. In the same letter he added that "a moat that must be continuously rebuilt will eventually be no moat at all."
Picture the only pharmacy in a small mountain town. It isn't any better than the ones in the city, but the town can't support two: whoever opens the second one will lose money, and the first one knows it. That moat doesn't come from the pharmacist's talent but from the size of the market. Moats come in several kinds, and it pays to know which kind each company has, because each one breaks in a different way.
The five sources of a moat
Morningstar, which rates the moats of thousands of companies, identifies five. Pat Dorsey, who ran its equity research, popularized them in "The Little Book That Builds Wealth" (2008).
1. Intangible assets: brands customers pay more for, patents, licenses. In the 2007 letter Buffett cites "a powerful world-wide brand" like those of Coca-Cola, Gillette or American Express.
2. Switching costs: changing suppliers is so expensive or so risky that customers don't do it. A bank won't swap out the software that runs its books just to save 10%.
3. Network effect: the service becomes more valuable the more people use it. More merchants accept Visa because more people carry it, and vice versa.
4. Cost advantage: producing more cheaply than anyone else. In 2007 Buffett's examples are GEICO and Costco, each the "low-cost producer."
5. Efficient scale: a market with room for only a few players, like the mountain-town pharmacy: pipelines, airports or credit ratings.
What is not a moat: a trendy product, good management, fast growth or a large market share without high returns. Any of those can last a year or ten, but none of them protects the business on its own.
How to measure it
A moat leaves footprints, and there are three to look for:
1. ROIC above the cost of capital, sustained. If a company earns 25% on its capital when that capital costs 9%, and keeps it up for ten years, something is stopping competitors from matching it. It's the metric from the lesson on ROIC.
2. Stable margins. A margin that holds up through recessions and new competitors signals pricing power. The sector guide on Kaplio's stock pages puts it this way: "5y consistent gross margin = moat signal."
3. Market share that doesn't erode. A high share that holds up without price cuts.
Formula
Return spread = ROIC − WACC
ROIC = NOPAT / (Shareholders' equity + Total debt)
NOPAT = Operating income × (1 − effective tax rate)
Margin stability = operating margin for each year over 10 years or more, including at least one recession
Value created per year = (ROIC − WACC) × Invested capital
Example
The table below compares the ROIC of four companies that make their living from financial information: Moody's, S&P Global, FactSet and Intercontinental Exchange. Read it as cents of after-tax operating income for every dollar of capital. And before you judge, ask yourself what's sitting in the denominator.
Before you look: Moody's and S&P Global run almost identical businesses: they're the two big credit rating agencies. Do you think their ROIC will be similar? The answer is in step 5.
Full analysis: Moody's (10-K annual reports, millions of dollars)
Step 1. What kind of moat? Moody's has two businesses: it rates the debt issued by companies and governments (Moody's Investors Service), and it sells data and analytics software (Moody's Analytics). The first has the clearer moat. An issuer pays to be rated because investors, and many fund management mandates, require a rating from a recognized agency. That's an intangible (reputation) plus efficient scale: according to the SEC staff report of January 2025, as of December 31, 2023, the three big agencies (S&P, Moody's and Fitch) accounted for 94.15% of all outstanding ratings issued by the agencies registered in the US.
Step 2. ROIC versus the cost of capital. In 2025 Moody's had operating income of $3,351 million and paid $668 million in taxes on 3,130 of pre-tax income (21.3%). NOPAT: 3,351 × (1 − 0.213) = 2,636. Invested capital at the end of 2025: 4,054 of shareholders' equity + 6,994 of debt = 11,048. ROIC: 23.9%. With a cost of capital of 8-10%, the range we saw in the ROIC lesson for a large listed company, the spread comes to roughly 14-16 points.
Step 3. Is it sustained? Calculated the same way, Moody's ROIC was 25.5% in 2019, 23.8% in 2020, 22.5% in 2021, 14.8% in 2022, 17.2% in 2023, 20.0% in 2024 and 23.9% in 2025. The 2022 dip lines up with a bad year for debt issuance: revenue fell from $6,218 million in 2021 to 5,468 in 2022. Even then, ROIC didn't drop below 14%.
The operating margin tells the long story: 50.1% in 2007, 42.6% in 2008, 38.3% in 2009, 42.3% in 2015, 45.7% in 2021, 34.4% in 2022 and 43.4% in 2025. In 2016 it sank to 18.1%, but because of one specific charge: $863.8 million of expenses for the settlement with the US Department of Justice and several states over its pre-crisis ratings. Without that charge, the 2016 margin would have been 42.0%. A settlement of nearly 900 million, and the business intact: that tells you a lot about the moat.
Step 4. How much capital does it need? In 2025 it generated $2,901 million of operating cash flow and spent 326 on property, equipment and software, or 11%. The rest goes back to shareholders: $1,607 million in buybacks in 2025. Average diluted shares fell from 236.6 million in 2010 to 179.9 million in 2025.
Step 5. The denominator trap. Those buybacks have an accounting side effect: Moody's shareholders' equity was negative in 2010, 2011 and from 2014 through 2017 (−$1,225 million at the end of 2016). With negative equity, ROE means nothing and ROIC has to be read with care. S&P Global has the opposite problem: after merging with IHS Markit in 2022, its goodwill went from $3,506 million at the end of 2021 to 34,545 at the end of 2022 (36,475 at the end of 2025, against equity of 31,127). If its ROIC in the table comes out well below Moody's, that doesn't mean its moat is worse: its capital includes what it paid for the acquisition.
Step 6. Your verdict. No answer key here: apply the scorecard at the end of the lesson to Moody's, then to a company you know well.
Real-data example
| Company | Ticker | ROIC |
|---|---|---|
| Moody's Corporation | MCO | 23.9% |
| S&P Global Inc. | SPGI | 10.2% |
| FactSet Research Systems Inc. | FDS | 16.0% |
| Intercontinental Exchange, Inc. | ICE | 7.0% |
How to read it
How to read it
Rough ranges for judging a moat with numbers (not a Kaplio rule):
| Signal | No moat | Narrow moat | Wide moat |
|---|---|---|---|
| ROIC − cost of capital | zero or negative | positive, for a few years | 5 points or more for 10 years |
| Operating margin in a recession | collapses or turns into losses | drops sharply, then recovers | barely moves |
| Capex as a share of operating cash flow | almost all of it | about half | a small fraction |
Adjust for the sector. In the sector guide on Kaplio's stock pages, a "healthy" ROIC starts at 15% in technology and industrials and at 12% in pharma and in general. In financial services, where Moody's sits, the sector guide on the stock page leaves ROIC out and looks at operating margin (healthy at 30% or more) and ROE (healthy at 15% or more). At banks, the moat shows up in ROE and in the cost of deposits, not in ROIC.
How Kaplio's methods see it. Kaplio's Buffett method has no "moat" checkbox: it looks for the footprints instead. Its 0-100 quality score, which ignores the price, asks for "sustained return on equity (ten-year median), stable margins, retained earnings that create value and contained debt," and it places banks, insurers and commodity businesses outside the circle of competence. A company with a wide moat usually scores high on quality, but the verdict also depends on the price: a great company that's expensive ends up at "would wait."
Pitfalls and limitations
1. Mistaking the past for a moat. Ten years of high ROIC tell you there was a moat, not that there will be one. Ask what sustains it and what could break it.
2. The shrunken or bloated denominator. Buybacks shrink equity and inflate ROE and ROIC (Moody's); expensive acquisitions bloat it with goodwill (S&P Global). Compare against capital excluding goodwill as well.
3. The one-person moat. In 2007 Buffett rules out "the business whose success depends on having a great manager." If the advantage walks out the door when the founder does, it wasn't a moat.
4. The technology moat. In industries that change quickly, the advantage has to be won again every few years, which is exactly what Buffett says a moat is not.
Case in point: Intel, a moat that broke
For decades Intel made the most advanced processors at the best cost in the industry: scale plus its own manufacturing technology. In 2018 its gross margin was 61.7% ($43,737 million on 70,848 of revenue). In July 2020 it announced a delay in its 7-nanometer process, while another manufacturer (TSMC) was already producing more advanced chips for its rivals. Gross margin fell to 55.4% in 2021, on revenue of $79,024 million, and to 32.7% in 2024, on 53,101. Operating income went from $19,456 million in 2021 to a loss of $11,678 million in 2024, and the year closed with a net loss of $18,756 million. The declared dividend per share fell from $1.46 in 2022 to $0.38 in 2024, the year the company suspended it. In 2025 revenue was $52,853 million and gross margin was 34.8%.
The signal showed up in gross margin before it showed up in earnings. It wasn't one piece of bad news: it was a cost advantage that depended on always staying ahead in technology, the kind of moat that has to be rebuilt every few years. Kodak, which dominated film photography and filed for Chapter 11 bankruptcy protection in January 2012, is the classic example of the same mistake with a different technology.
Self-assessment scorecard: ten questions for judging a moat
1. Can you say in one sentence what stops a rival from copying the business?
2. Which of the five sources does it come from, and is there more than one?
3. Does ROIC beat the cost of capital by at least 5 points over ten years?
4. Have you checked whether ROIC is inflated by buybacks or depressed by goodwill?
5. What did the operating margin do in the last recession?
6. Does market share hold up without price cuts?
7. How much of operating cash flow goes to capex just to defend the business?
8. Does the advantage depend on one person, or on a technology that changes fast?
9. What would have to happen for the moat to break, and would you see it first in gross margin?
10. Is the price already assuming the moat will last forever?
Score yourself: one point for every answer backed by a figure from the accounts or a concrete fact. Eight or more, and you have a thesis. For question 10, the lessons on terminal value in a DCF, the P/E ratio and enterprise value (EV) will help: a wide moat bought at too high a price can still be a bad investment.
Practice on Kaplio
Frequently asked questions
What is an economic moat?
It's a company's durable competitive advantage: whatever keeps rivals from copying the business and taking its profits. Buffett called it a "moat" in his letters, like the one around a castle. You can spot it in a ROIC well above the cost of capital for years: Moody's, for example, had a ROIC of 23.9% in 2025.
What is Warren Buffett's moat concept?
It's the advantage that protects high returns from competitors. In his 2007 letter, Buffett wrote that a truly great business must have an enduring moat that protects excellent returns on invested capital, and gave as examples the low costs of GEICO and Costco and the brands of Coca-Cola and American Express.
What are the types of economic moats?
Morningstar identifies five sources: intangible assets such as brands, patents and licenses; switching costs; network effects; cost advantage; and efficient scale, a market with room for only a few players. A company can have more than one. Each kind breaks in a different way, so it pays to know which one you're looking at.
How do you measure an economic moat?
Ask what stops a rival from copying the business, then check the footprints in the accounts: a ROIC above the cost of capital for at least ten years, margins that hold up through recessions and market share that doesn't erode. Then look for what could break it: new technology, a regulator or one key person leaving.