Mastercard

Six weeks ago we analyzed Visa and called it a wonderful business at a full price — a wide-moat toll booth clipping a few cents on every swipe, with no credit risk and no margin of safety. Mastercard is the other toll booth on the same road, and this report exists to answer one question we deliberately left hanging: Mastercard has always cost more than Visa. Is the premium worth it?

The short answer is that the question has quietly expired. For most of the last decade Mastercard traded at a visible premium to Visa, and it largely earned that premium by growing faster. But by the middle of 2026 the two have converged: both now trade at roughly 30 times trailing earnings, and on next year's estimates Mastercard is actually the marginally cheaper of the two. So the real question is no longer "is the premium justified" — it is the more interesting one of which twin you would rather own, at what is now essentially the same price.

The mechanics are identical to Visa's, and worth restating because almost everyone gets them wrong. Mastercard does not lend money and carries no credit risk — that sits with the bank that issues the card. It does not keep the interchange fee, the ~1–2% "swipe fee" that so enrages merchants — that flows to the issuing bank. What Mastercard actually earns is a thin network toll: a few basis points on every dollar and a few cents on every transaction it switches across its rails. On $10.6 trillion of annual volume, those few basis points compound into $32.8 billion of revenue at a 59% operating margin. It is one of the great business models ever devised, and there are exactly two companies on Earth that run it at global scale.

"The £10 billion case that became £200 million." — the UK Merricks class action, once touted at up to £17bn, settled in 2025 for a fraction — one of two litigation clouds that cleared over Mastercard in 2026

There is one genuine way in which Mastercard is a better buy than Visa was when we wrote it up, and it is not about the business — it is about the weather. When we analyzed Visa, two enormous legal clouds hung over the entire card industry: a UK consumer class action valued in the tens of billions, and a US merchant-interchange settlement that a judge had thrown out. Both cleared in 2026, and both far below the feared numbers. The UK case settled for £200 million against a £10-17 billion headline; the US settlement was approved in June 2026. That does not make Mastercard cheap — but it does mean you are buying it with less overhang than Visa carried, at a price that is no longer a premium. That is the whole of the opportunity, and it is a modest one.

Mastercard's history rhymes with Visa's, because they were born as rivals doing the same thing — and the key event in both stories is the same: the moment each stopped being owned by banks.

The demutualization is the whole game, and it is worth dwelling on because it explains why these two companies are such extraordinary businesses. As a bank-owned association, Mastercard existed to serve its member banks — it was a utility, run roughly at cost. The moment it went public and cut the banks' control, it was free to price the network for profit, and the profit turned out to be enormous, because the banks still needed the rails and had nowhere else to go. Visa and Mastercard between them had spent decades building an acceptance network — tens of millions of merchants, billions of cards, a trusted global brand — that no new entrant could replicate. The 2006 IPO simply handed the keys to that fortress from the banks to public shareholders. Everything Mastercard has earned since is rent on infrastructure that was built, in effect, by the whole banking system and then privatised.

Mastercard sits with Visa, Coca-Cola and Costco at the very top of our understandability scale, and for the same reason: you can explain the entire business to a ten-year-old and be done in thirty seconds. Money moves; Mastercard takes a tiny cut; it never lends and never loses money if a cardholder defaults. The one honest complication relative to Visa is that Mastercard has chosen to build a much larger services business alongside the pure network — cybersecurity, fraud scoring, data analytics, consulting — which now supplies about 41% of revenue. That is a deliberate strategic bet (more on it in Part VI), and it makes Mastercard a fraction less pure and a fraction harder to model than Visa. But the core remains a toll booth, and toll booths are the easiest wonderful businesses in the world to understand. The difficulty here has never been the business. It has always been the price.

The services business is the whole strategic story, and it cuts both ways. Visa is a purer switching network; Mastercard has chosen to build a large, fast-growing services layer on top of its rails — selling banks and merchants the tools to fight fraud, understand their data, and run loyalty programmes. In FY2025 that business was $13.3 billion, about 41% of revenue, growing 23% a year. The bull case is that this is brilliant: it deepens Mastercard's relationships, diversifies it away from pure transaction fees (and therefore away from interchange regulation and disintermediation risk), and grows faster than the core. The bear case is quieter but real: services carry lower margins than clipping a basis point on a switched transaction — they involve people, and consulting, and acquisitions — which is the entire mechanical reason Mastercard's operating margin (59%) runs several points below Visa's (~65%). Mastercard is, quite deliberately, trading margin for growth and optionality. Whether that is the right trade is the crux of the twin comparison, and reasonable people land on opposite sides of it.

Mastercard's moat is, essentially, Visa's moat — because they are the two halves of the same duopoly, protected by the same forces. We will not repeat the Visa analysis at length; the point worth making here is that being the #2 network takes almost nothing off the moat, because of how payment-network economics work.

1 · Two-sided network effects. Consumers carry Mastercard because tens of millions of merchants accept it; merchants accept Mastercard because billions of consumers carry it. Each side reinforces the other, and the loop has been compounding for sixty years. A new entrant would need to sign up both sides simultaneously, at global scale, against an incumbent that already has both — which is why, in six decades, the number of global general-purpose networks has stayed at roughly two (plus China's UnionPay, walled off at home).

2 · Acceptance ubiquity and trust. The value of "it works everywhere" cannot be built quickly at any price. Neither can the brand trust that lets a traveller in a foreign city tap a card and know it will be honoured.

3 · No credit risk, capital-light economics. The moat is not just wide, it is cheap to defend: Mastercard earns a ~52% return on invested capital because it needs almost no capital. It runs software and a brand, not branches and loan books.

Does being #2 matter? Barely. In most network businesses the leader takes all; in this duopoly, both networks are universally accepted, so a merchant that takes Visa takes Mastercard too, and a consumer's wallet holds both. Mastercard is smaller than Visa by volume (about $10.6 trillion to Visa's ~$17 trillion), but it is not disadvantaged by it in the way a #2 usually is — it grows faster, not slower, and it is more international. The one genuine crack in the moat is the same one Visa faces, and it is the subject of Part VII: the rails can be bypassed at the domestic margin by account-to-account systems like Pix and UPI. I score the moat a 10 — one of exactly two moats of its kind on the planet, and being the smaller of the two takes essentially nothing off it.

This is the report's reason for existing. For years the choice between Visa and Mastercard came with a price tag attached — Mastercard cost more, so you had to decide whether its faster growth was worth the premium. In 2026 that framing is gone: the premium has compressed to almost nothing. So the decision is now purely about what kind of toll booth you want, at the same price. Here is the honest scorecard.

The pattern is clean once you see it. Visa is the bigger, purer, higher-margin network — if you want the most entrenched toll booth and the fattest margin, and you are content with low-double-digit growth, Visa is your business. Mastercard is the faster, more international, more services-diversified one — if you want a little more growth and a little more emerging-market runway, and you will accept a few points of margin to get it, Mastercard is your business. Neither is better in any absolute sense; they are two expressions of the same wonderful model, tuned differently. The trade Mastercard makes — margin for growth and diversification — is a perfectly defensible one, and its services business is a genuine hedge against the two things that could hurt a pure network: interchange regulation and disintermediation.

Our own read, holding both up against each other, is mildly in Mastercard's favour at this moment — not because it is a fundamentally superior business, but because of three timing facts. First, the premium you used to pay for Mastercard's faster growth is gone, so you are getting that growth for free relative to history. Second, the two great litigation clouds cleared in 2026 (Part XI), removing an overhang that still hung over the sector when we wrote up Visa. Third, on next year's estimates Mastercard is fractionally the cheaper of the two. None of that is enough to make it a table-pounding buy — both remain full-priced — but if a reader asked us "I want to own one toll booth and the price is the same, which one," we would lean, gently, to the faster twin. And we would note, as we must, that Berkshire looked at both and sold both in early 2026.

The one genuine threat to this business — the same one that hangs over Visa — is that the rails themselves get bypassed. When you pay by Brazil's Pix, India's UPI, or America's FedNow, money moves directly from your bank account to the merchant's, in real time, without touching a card network, and therefore without the interchange economics that the whole industry rests on. In Brazil, Pix went from launch in 2020 to ~140 million users within two years and now dominates domestic person-to-person and much of retail payment. Stablecoins threaten a similar bypass for cross-border. This is not a fringe worry; it is the reason a wonderful business trades at "only" 27 times earnings instead of 40.

Our judgement is the same as it was for Visa, and it is deliberately unexciting: the disintermediation threat is real at the domestic margin and manageable everywhere else. In the markets where governments have built free instant-payment rails, the card networks will grow more slowly — that is simply true, and it is why neither Visa nor Mastercard will compound at the rates they did in the 2010s. But the idea that the networks get replaced globally underestimates how much of their value is in the things A2A systems do badly: cross-border acceptance, fraud liability, chargebacks, and the sheer trust of a brand honoured in every country on Earth. And Mastercard's response — spending to become the settlement layer for stablecoins rather than pretending they will not happen — is exactly what you want to see. The services business is the deeper answer still: a fraud-scoring and data company does not much care whether the underlying payment travelled on a card or on Pix. Mastercard is, more than Visa, building the business that survives its own core being disrupted. That is worth something, and it is part of why the market no longer demands a discount to own it.

The capital allocation is close to textbook for a capital-light compounder, and there is little to criticise. Mastercard converts more than 100% of net income to free cash flow (FCF per share of $19.88 against EPS of $17.47), needs almost none of it to run the business, and returns essentially all of it — buying back nearly a fifth of its shares over the decade and raising a small dividend at mid-teens rates. The bolt-on acquisitions have been disciplined and concentrated in the services business (fraud, data, open banking) rather than empire-building. The one thing worth flagging for a value-minded owner is simply that a buyback is only as good as the price paid, and Mastercard has been repurchasing stock at 25–35 times earnings throughout — accretive to earnings per share, yes, but not the value-creation bargain that buybacks at 12× would be. That is a mild note, not a criticism: with a stock this expensive and a business this capital-light, there is genuinely nothing better to do with the cash. I score management a 9 — a superb operating and capital-allocation record, marked a notch below perfect only because so much cash is being returned at full prices, which is unavoidable rather than a mistake.

There is very little to say against these numbers, which is precisely the problem — everything good about them is known, and priced. Revenue has tripled in a decade while the operating margin sat near 55–59% the entire time; net margin is 46 cents on the dollar; return on invested capital is an almost absurd 52%, because the business needs virtually no capital to run — it is software and a brand. Free cash flow exceeds reported earnings ($19.88 per share against $17.47), the mark of a genuinely asset-light model with no capital-expenditure drag of the kind that distorts Walmart, Lilly or the AI hyperscalers. The balance sheet is a fortress (Altman-Z of 10.4, negligible net leverage). And the highest-margin line, cross-border, is growing fastest at +15%.

The only honest asterisks are two, and both are features rather than flaws. First, the operating margin of 59% trails Visa's ~65% — not because Mastercard is worse-run, but because it has chosen to carry a larger, lower-margin services business, a deliberate trade for faster growth. Second, the book value per share is tiny ($7.54), which makes the price-to-book ratio of 71× look alarming and is completely meaningless: years of buybacks have retired equity faster than it accumulates, so book value is not a sensible yardstick for a company like this — ignore it, exactly as you would for Visa. What the numbers describe is one of the highest-quality businesses in existence. What they cannot tell you is whether $540 is a sensible price to pay for it. That is Part X.

Mastercard produces an unusual result on our board, and it is worth stating up front because it is the single most important valuation fact here: the discounted-cash-flow model values the stock at $545.54, against a price of $539.66 — a difference of one percent. For once, the model and the market agree almost exactly. Across two dozen analyses we have grown used to the DCF screaming that quality compounders are 40–60% overvalued (it said so about Costco, Lilly, Amazon and Novo); here it does not. That does not make Mastercard cheap. It makes it fairly valued — which, for a business this good, is itself notable.

Put the pieces together and Mastercard is the most defensible full-price stock we have valued in a while — which is a genuinely different verdict from "overvalued," even if it lands in the same "Watch" bucket. The trailing multiple of 31× looks expensive and is; but the forward path is reasonable — 27× next year, 20× by 2028 — for a business growing revenue mid-teens at a 59% margin with a duopoly moat. The DCF, which has been merciless to every other compounder on this board, calls it fairly valued to within a percent. The analyst community sees ~21% upside and not one of the 64 covering it has a sell rating or a target below the current price. And the two legal clouds that darkened the sector when we wrote Visa have lifted.

So why not a buy? Because "fairly valued" is not "cheap," and our method insists on a margin of safety before it pounds the table. At $540 you are paying full price for a wonderful business and being handed no discount for the genuine risks — the domestic-A2A erosion, the cross-border cyclicality, the reintroduced routing legislation, the simple fact that a stock at 31× trailing earnings falls hard when growth disappoints. Wonderful businesses at fair prices are perfectly good things to own, and if you are a long-term compounder who buys quality and holds for a decade, we would not argue with accumulating here. But the disciplined entry — the price that turns "fair" into "attractive" — sits toward $470, about 24 times next year's earnings and right at the 52-week low, a level this stock traded below within the past twelve months. I score valuation a 5, the same as Visa, and note that it is the more comfortable 5 of the two, because for once the model is not flashing red. The verdict, like Visa's, is Watch — but watch it a little more warmly.