Luke Shulver
←All writing

Who Owns the Payment Rails · Part 1

7 min read

I Thought Stripe Would Own Payments. I Was Half-Right.

PaymentsStablecoinsStripeStrategyFintech
An open padlock with payment rails flowing out to a dollar coin, Visa and Mastercard — the ownership Stripe couldn't lock

Earlier this month I wrote that the settlement layer had picked a side, and I allowed myself a small victory lap: I'd argued stablecoins would quietly become the rail, and Open USD landing with 140-plus backers proved the point. I stand by that call.

But a follow-up is only worth writing if it also owns the part I got wrong — and there's a big one. I was right that stablecoins would win. I was wrong about who would win with them. I thought the answer was Stripe. Not Stripe-as-a-participant. Stripe-as-the-owner: one company, every layer, a walled garden with a Stripe-shaped moat.

That specific bet didn't survive contact with the announcement. Here's the case I'd built, why it was wrong, and the thing I under-weighted that explains it.

The empire read

For about eighteen months I read Stripe's moves as an empire play, and the acquisition trail made it easy.

  • Recko (2021) — reconciliation. The back office.
  • Lemon Squeezy (2024) — merchant-of-record, which is to say tax, compliance and liability. The paperwork layer.
  • Bridge (announced October 2024, closed February 2025) — stablecoin issuance and orchestration, at ~$1.1 billion, the largest acquisition in Stripe's history.
  • Privy (2025) — embedded crypto wallets. The on-chain user layer.

Terms on three of those four were never disclosed, so I'll resist the temptation to quote a tidy "total spend" figure — only the Bridge number is public. But you don't need the sum to read the direction. Stack them and you get a company deliberately assembling every layer from the buy button to on-chain settlement: checkout, reconciliation, issuance, wallets.

Then came the tell that sealed it for me. Stripe didn't just want to move money over other people's chains — it went and helped build one. Tempo, a payments-first Layer-1 co-incubated with Paradigm, announced September 2025 and live on mainnet by March 2026: stablecoin-native, gas paid in dollars, tuned for payments throughput rather than general-purpose smart contracts. When a company owns the issuance layer and the wallet layer and then puts a chain underneath them, the story writes itself. Stripe was going to be the rail. One solution, every role, a moat you paid a Stripe toll to cross.

I was confident. I was also wrong about the shape.

What Open USD actually said

On 30 June, Open USD launched — and it was close to the opposite of a walled garden.

It's a shared stablecoin operated by an independent entity, Open Standard, governed by a board of its partner companies rather than a single issuer. The reserve yield — the money a stablecoin actually makes — flows back to the partners, not the issuer. It launched multi-chain, first on Solana with Base, Stellar, Polygon and Stripe's own Tempo among the chains to follow. And the roster is 140-plus names that include Stripe's most obvious rivals sitting at the same table: Visa, Mastercard, Coinbase, BlackRock, BNY.

Stripe's role in all this? It made Open USD the default stablecoin for businesses on its platform. Not its own proprietary coin on its own exclusive chain — a shared standard it doesn't control, running across chains it didn't build, alongside the competition.

That is not the move of a company trying to own the vertical. That's the move of a company that concluded it couldn't — and decided being the biggest tenant in an open building beat owning a smaller one alone.

Why the walls came down: Visa and Mastercard

Here's the force I under-weighted. While I was busy watching Stripe stack acquisitions, the card networks were making the same bet with far more leverage — and, it turns out, a bigger chequebook.

Visa spent 2024–25 wiring stablecoins into its own rails rather than fighting them: its Tokenized Asset Platform launched in late 2024, and by December 2025 it had live USDC settlement in the US running at roughly $3.5 billion annualized. Mastercard built out its Multi-Token Network and plugged in a whole shelf of stablecoins. And then it did the thing that should have recalibrated me on the spot: in March 2026 it agreed to buy the stablecoin-infrastructure firm BVNK for up to $1.8 billion — a larger stablecoin acquisition than Stripe's Bridge deal.

The networks weren't ceding the rail to Stripe. They were outspending it.

And they hold the one asset Stripe's stack can't clone quickly: distribution. Minting a stablecoin is easy. A stablecoin that a merchant will actually accept and a bank will actually distribute is a network-effects problem that took Visa and Mastercard the better part of sixty years to solve. A proprietary Stripe chain, however elegant, starts that clock at zero.

So the rational endgame was never "Stripe replaces the networks." It was a table with Stripe, Visa and Mastercard all seated and a shared coin none of them fully owns — which is more or less exactly what reporting foreshadowed. As early as 3 June, CoinDesk had the three of them converging on a joint stablecoin platform. Open USD is what a truce looks like when no single player can win outright.

The tell I missed

The honest part: the signal was sitting in plain sight, and I filed it under the wrong heading.

Tempo was never actually structured as Stripe's private chain. It's an independent company — it raised $500 million at around a $5 billion valuation from outside investors, and Fortune reported that Stripe and Paradigm didn't even put capital into that round. Its launch partners included Visa. The "open, multi-party" instinct was baked into the architecture from day one; I just kept reading it as a moat because the acquisition trail told a tidier story and I let the tidier story win.

That's the mistake worth naming, because it's a repeatable one: I weighted the aggressive, legible signals — billion-dollar buys, a company building its own blockchain — over the quieter structural ones sitting right next to them. The loud facts pointed at a walled garden. The quiet facts pointed at a shared utility. The quiet facts were right.

Half-right, and why it still matters

I called this "half-right" deliberately, because the correction isn't a total loss for my read — or for Stripe.

Being the default is its own kind of power. Stripe making Open USD the out-of-the-box settlement token for the millions of businesses on its platform is an enormous distribution win, ownership or not. It traded control of the coin for ubiquity of the coin, and in a market where the whole game is which token gets picked by default, that may well be the better trade. Stripe didn't lose the rail. It just isn't the toll booth I assumed it was building.

And the thing I actually got right still stands, sharper for the correction: stablecoins did become the settlement layer. I simply had the ownership model inverted. The rail isn't a fortress with a single gatekeeper. It's a shared standard the incumbents agreed to co-run precisely because none of them could own it alone.

The lesson I'm keeping: in payments, distribution beats ownership, and "the default" beats "the owner." The winning move was never the walled garden. It was to be the most-used tenant in an open one — and to be honest enough about it afterwards to say the garden was never yours to wall off in the first place.

I got the winner half-right and the shape wrong. If you're building in this space and reading the same tea leaves, I'd genuinely like to compare notes on what the next wrong-but-instructive call turns out to be.

Next in this series → Part 2 — SWIFT 'Went Crypto', But Not How You Thought — the banks' answer to the same question.

Sources

A follow-up to Open USD: The Settlement Layer Just Picked a Side — and Part 1 of a short series on who ends up owning the payment rails. I work on the delivery side of frontier tech; these are my own views, not my employer's.

Written by Luke Shulver — Operations Manager at Labrys.

Get in touch→