← Cross-cutsCross-cut · Payments

The take-rate spine

Seven Teardown studies cover the businesses that move money — Visa, Mastercard, PayPal, Stripe, Coinbase, Nubank and Ramp. Read together, they line up along a single number: the take rate, the slice each one keeps of the volume it moves. That number sorts the cluster into fat-margin toll roads and thin-margin disruptors — and reveals the one structural threat all seven share.

The whole cluster sorts by one number

Payments looks like many businesses; it is really one variable measured many ways. At one end, the networks: Visa keeps ~50% net margins and Mastercard ~58% on a take rate of a fraction of a percent — but across colossal volume ($10.6T of gross dollar volume at Mastercard). At the other end, the infrastructure disruptors: Stripe’s ~2.9% sticker price nets only ~0.40% after interchange pass-through, and Ramp keeps just ~0.8% of the $200B it processes. The networks own a thin slice of an enormous river; the disruptors own a slightly different thin slice and dress it in a software multiple. Everything else in the cluster is a variation on where you sit relative to that toll.

Three ways to monetize the same flow

The seven split into three models. The networks (Visa, Mastercard) are asset-light toll roads with no credit risk — the highest-quality, most defensible seat, which is why their studies call the moat “essentially unassailable.” The software-led players (Stripe, Ramp, and PayPal’s checkout) ride interchange but bolt on higher-margin software to lift a thin blended take. And the balance-sheet monetizers (Nubank, Coinbase) barely use the take rate at all: Nubank earns ~83% of revenue from credit, not payments, and Coinbase from trading fees, custody and stablecoin interest. The cross-cut’s first insight is that “a payments company” can mean a toll collector, a software vendor or a lender wearing a card — and they carry completely different risks.

The threat they all share: account-to-account rails

Read the seven studies’ risk sections together and the same name keeps appearing: real-time, account-to-account (A2A) rails that route around the card entirely. The proof point recurs in study after study — Brazil’s Pix already processes more transactions than Visa and Mastercard combined there, reaching ~70% of Brazilians. Pix, India’s UPI, and the US FedNow are the structural threat every one of these businesses names, because they attack the toll at its source. The cluster splits on how fast it bites: the networks’ studies lean “slow tide” (A2A wins thin domestic debit-like flows, not the rich cross-border ones), while the disruptors are more exposed because their thin take has less cushion. It is the rare risk that hits the fat-margin incumbent and the thin-margin upstart at the same point.

PayPal is the warning; the networks are the fortress

The cleanest contrast in the cluster is PayPal versus the networks. All three are incumbents, but PayPal sits in the one spot that gets disintermediated: the checkout button, ~30% of its volume but 65%+ of its gross profit, now decelerating to ~1% growth as Apple Pay (the iOS default) and Shop Pay (embedded in Shopify) capture it. The market re-rated PayPal from a growth stock to a ~8.7× value stock — roughly 80% below its 2021 peak — while Visa and Mastercard hold ~28× multiples. Same word, “incumbent,” opposite fates: the difference is whether your position is a network others must plug into, or a button others can replace by default. That distinction is the cluster’s core lesson about where payments value is safe.

Where they agree — and where they split

All seven accept that digital payments keep growing and that the take rate is under long-run pressure — from regulation (the DOJ’s Visa debit suit, the ~$200B swipe-fee settlement), from A2A rails, and from stablecoins. They split on two things. First, valuation: the networks trade at quality premiums the studies call earned; the private disruptors (Stripe at ~$159B / ~23× revenue, Ramp at ~$44B / ~29×) carry marks their own teardowns flag as running ahead of verifiable economics. Second, what happens in a downturn: the asset-light networks have no credit risk, while Nubank’s 33% ROE rests on a young, unsecured credit book facing its first full Brazilian downturn, and Coinbase’s earnings still track the crypto cycle. Everyone is moving money; the question each study answers differently is how much of it they get to keep, and for how long.

The cluster at a glance

CompanyNetwork / scaleModel & take rateGrowthBiggest risk
VisaNYSE:V257.5B txns · $40B net revToll on volume · ~50% margin+11% net revDOJ debit suit · A2A rails
MastercardNYSE:MA$10.6T GDV · 175.5B txns~0.31% net take · ~58% margin+16% net revSwipe-fee settlement · Pix
PayPalNASDAQ:PYPL~$1.8T volume · 438M accountsTake rate · checkout = 65%+ of GP+4% (checkout +1%)Apple Pay / Shop Pay disintermediation
StripePrivate$1.9T TPV (+34%)~0.40% net take + software~+36% net rev (est.)$159B mark vs thin core
Coinbase GlobalNASDAQ:COIN$5.2T volume · ~84% of US BTC-ETFFees + USDC interest + custody+9% FY · −30% Q1'26Crypto cycle · USDC concentration
NubankNYSE:NU131M customers · $41.9B deposits~83% credit/interest · 33% ROE+45% revFirst full Brazil credit downturn
RampPrivate~$200B volume · ~1.5% of US biz~70% interchange · ~0.8% net+~170% volume$44B mark · rate compression

Figures as of each study’s stated date (2026-06); Stripe and Ramp are private (estimated). Take rates and multiples are on mixed bases and not strictly comparable; see each teardown for sourcing and the full weighing.

The seven studies — and the question each turns on

Visa Inc.NYSE:VIs Visa's dominance still a wide-moat compounder, or is that very scale now its biggest liability as antitrust, routing legislation and account-to-account rails converge on the toll?Read the full weighing →MastercardNYSE:MACan interchange-funded card economics keep compounding while regulators cap fees, merchants push the CCCA, and governments stand up instant-payment rails that bypass the cards?Read the full weighing →PayPal Holdings, Inc.NASDAQ:PYPLIs PayPal a cash-gushing franchise being slowly disintermediated at checkout, or a deep-value turnaround the market gave up on too early?Read the full weighing →StripePrivateCan a low-margin payments business sustain a software-multiple $159B valuation while moving upmarket into pricing pressure and betting on stablecoins and AI commerce?Read the full weighing →Coinbase Global, Inc.NASDAQ:COINHas Coinbase escaped crypto's boom-bust gravity, or just built a more diversified, better-regulated version of the same cyclical bet?Read the full weighing →Nubank (Nu Holdings Ltd.)NYSE:NUCan a young, fast-growing, unsecured-heavy credit book hold up through a full Brazilian credit downturn while also funding Mexico and a new US bank?Read the full weighing →RampPrivateCan a business still ~70% reliant on thin (~0.8% net) card interchange sustain a $44B software-multiple valuation as rates stay pressured and a bank-owned Brex sharpens?Read the full weighing →

This is the kind of reading the Desk does for you

A cross-cut takes seven teardowns and one toll and asks what they say together. The Desk does the same for the companies you actually own — your thesis, the rivals and rails that move it, and the tripwires that would change your mind.

See the Desk →