Visa’s settlement pilot is quietly becoming multichain plumbing

While consortium stablecoins take the headlines, the settlement layer at Visa has been compounding in the background, and the compounding is the story. The network's stablecoin settlement programme, which lets issuers and acquirers meet their obligations to Visa in tokens rather than wires, expanded to nine supported blockchains in April with the addition of five networks. US domestic stablecoin settlement, switched on in December 2025, has scaled through its first months at a pace that surprised even sympathetic observers, with industry tracker Transak putting the annualised run rate near 7 billion dollars by April.

Settlement never trends, which is precisely why professionals watch it: consumer features announce strategy, but settlement changes commit balance sheets, and balance sheets do not participate in hype cycles. It also decides everything, because it is where the money actually moves.

Why this beat every flashier roadmap

It is worth pausing on why settlement, of all the crypto ambitions the payments industry has entertained, is the one that quietly worked. Consumer crypto payments fought human habit and merchant indifference simultaneously; settlement asked nothing of consumers and offered treasurers a measurable number. Tokenised securities waited on market structure that regulators are still drafting; settlement slotted into obligations that already existed, between parties that already trusted each other, governed by rulebooks that already worked. The general lesson is the one infrastructure history keeps teaching: new rails win fastest where they replace a cost invisible to the public and enormous to the operator. Pre-funded settlement float was exactly that cost, and stablecoins were exactly shaped for it.

What card settlement actually is

To see why this matters, unwind what happens after a card tap. Authorisation is theatre for the databases; no money moves. The movement comes later, when the network nets the day's flows: each issuer owes the system for its cardholders' spending, each acquirer is owed for its merchants' sales, and the net positions settle through designated banks, currency by currency, on banking days, at appointed hours.

That machinery works, and it costs. Participants pre-fund settlement accounts so the money is guaranteed to be there at cut-off, idle balances whose only job is punctuality. Cross-currency positions add correspondent legs and their fees. And the calendar rules everything: value initiated on Friday evening completes when the settlement day next opens, while obligations accrue continuously. The system's reliability is bought with trapped liquidity, and every treasurer who runs a card programme can tell you the number to the cent.

What the pilot replaces

Visa's programme substitutes a token transfer for the wire leg: an issuer or acquirer meets its net obligation by sending stablecoins, or receives its position the same way, over a supported chain, around the clock. The mechanics that make it credible are the boring ones: the network still nets, the obligation is still to Visa, the participant still faces one counterparty. Only the settlement asset and its calendar change.

The nine-chain roster is the tell that this has left the lab. One chain is a demonstration; nine is an operations decision, made because participants hold liquidity in different places and the network stopped arguing with them about it. Rubail Birwadker, Visa's head of growth products and strategic partnerships, framed the April expansion in exactly those terms: more networks means partners "can choose the networks that best fit their needs." Settlement infrastructure earns adoption by being indifferent to its users' preferences, and indifference at the chain layer is what the expansion actually shipped.

The token layer's quiet concentration

One structural fact shapes the whole programme: the settlement flows run today overwhelmingly through USDC. That was the pragmatic choice, the licensed dollar token institutions could approve, and it created a dependency the networks have plainly decided to manage rather than deepen. Read the year's moves as one portfolio: expand chains so no network holds the programme hostage; pilot with multiple issuers, Circle and Paxos both appear across the two card networks' announcements; and back a consortium token, Open USD, that would make the settlement asset itself member-governed. The sequence is textbook infrastructure strategy, standardise on what exists, then commoditise it, and the issuer economics of settlement flow, the most durable stablecoin demand there is, sit at the end of it.

Five years from stunt to plumbing

The programme's history is short and unusually linear. In 2021 Visa announced it would accept settlement in USDC from a crypto-native issuer, a first that was widely read as marketing. By 2023 the pilot had crossed to the acquiring side, with merchant acquirers including Worldpay and Nuvei taking settlement payouts in stablecoins, and expanded beyond its first chain to faster ones. December 2025 brought the domestic US launch, moving the concept from cross-border novelty to home-market infrastructure. April 2026 added five chains in one release. Each step looked incremental; the compound effect is a settlement capability that now spans the network's core market, multiple public blockchains, and both sides of the interchange relationship.

The 7 billion dollar annualised figure should be handled with its provenance attached, it comes from an industry tracker rather than the network's own disclosure, but its magnitude is believable and its direction is not in dispute. Against the trillions Visa settles, it is a rounding error. Settlement migrations are always rounding errors until the treasury departments finish their spreadsheets; then they are policy.

The working-capital arithmetic

The spreadsheet in question is simple enough to sketch. A settlement participant sizes its pre-funding to peak obligation plus buffer, across every currency it settles. Move a share of that to tokens and the buffer shrinks toward actual flow, because the top-up is minutes away at any hour rather than a banking day away at the wrong hour. Weekend float compresses. Cross-border legs lose their correspondent tail. For a large issuer or acquirer, the released liquidity is measured in days of settlement volume, capital that stops being furniture and returns to work.

That, and not ideology, is why the phrase settlement optionality recurs in every network announcement: the networks are not selling participants a belief system, they are letting CFOs discover the arithmetic and volunteer.

Who moves first, and why

Migrations like this have a predictable adoption order, and it is visible already. Crypto-native issuers went first because their treasuries were already on-chain; settlement in tokens removed a conversion rather than adding one. Merchant acquirers with global payout businesses came next, using token settlement to shorten the leg between network and merchant in markets where banking hours are the bottleneck. The heavy middle, large bank issuers, moves last and matters most: their volumes dwarf everyone's, and their constraint was never interest but permission. Each regulatory document of the past year, the GENIUS Act, its rulemakings, MiCA's full application, converts another compliance objection into a configuration decision. Watch the first top-ten US issuer to disclose token settlement in a filing; that is the moment this stops being a pilot in anyone's vocabulary.

Mastercard, in parallel

The story is a duopoly's, not one company's. Mastercard has been building the same capability along the same timeline, piloting stablecoin settlement for payments and payouts with Circle and Paxos handling the token legs. Raj Dhamodharan, who runs blockchain and digital assets there, locates the frontier in the same place: the next phase of adoption is about utility, "especially in settlement, where timing and liquidity matter most."

The competitive meaning of the parallelism is that token settlement is not a differentiator either network expects to own; it is table stakes both are racing to make ordinary. The differentiation lands one layer up, in whose stablecoin carries the flows, which is precisely why both networks now sit inside the Open Standard consortium building a token of their own. Settle in tokens first; then make the token yours.

The nine-chain operations reality

A word on what supporting nine chains actually obliges, because the number is doing quiet work. Each network added is a distinct finality model, fee market, outage history and upgrade calendar that a settlement operation must monitor with the same seriousness it applies to a settlement bank. Treasury teams need signing infrastructure, address governance and incident playbooks per chain; risk teams need concentration limits across them; auditors need evidence trails from systems built long before auditors were imagined. The card networks are among the few institutions on earth with the operational budget to normalise all of that, which is itself part of the strategy: every chain the network industrialises becomes safe terrain its participants no longer need to industrialise themselves. Infrastructure leadership in payments has always meant absorbing complexity so members do not have to; nine chains is that tradition, restated in new syntax.

The regulatory unlock underneath

None of this scales without the year's legal groundwork. Banks and regulated issuers could not hold and transfer settlement stablecoins at size while the assets' American status was undefined; the GENIUS Act's licensing regime, and the Treasury rulemaking opened this same week, convert the category from a supervisory question mark into a defined instrument with named regulators. The same convergence, MiCA in Europe, Singapore's draft law, is what lets a global settlement network treat tokens as a currency-grade asset across jurisdictions. Half a decade of pilots and the pilot never needed scale; it was waiting for the statute that made scale legal.

The limits worth stating

Honesty about the residual frictions sharpens the thesis rather than weakening it. Chain operations are a new risk discipline for settlement teams: nine networks means nine failure domains, fee markets and finality models to monitor. Token concentration cuts the other way, today's flows run overwhelmingly through one issuer's asset, and the networks' own consortium token is, among other things, an answer to that dependency. And the accounting and capital treatment of token settlement balances at regulated participants is still being written into rulebooks, which is the kind of detail that gates the final order of magnitude.

None of these are objections. They are the snag list of a migration that has already been decided.

What issuers and acquirers should do with this

For settlement participants the near-term agenda is unglamorous and worth stating. Model the pre-funding release corridor by corridor, because the business case is a liquidity calculation, not a technology one. Stand up token treasury operations in sandbox now, custody, signing, monitoring, so the capability exists before the CFO asks for it. Engage the network's pilot teams early: supported-chain and supported-token rosters are being shaped by participant demand, and the participants at the table are choosing defaults for everyone who arrives later. And put the accounting questions to auditors this year, not at go-live; capital and balance-sheet treatment of settlement tokens is exactly the kind of issue that costs a quarter when discovered late. The organisations that treat 2026 as the preparation year will spend 2027 configuring. The rest will spend it explaining.

The number to actually watch

Run-rate headlines will keep arriving, but the metric that will mark the real inflection is duller: the share of net settlement, by value, moving in tokens on non-banking days. Weekend and holiday settlement is the one service the legacy stack structurally cannot render, so its token share starts near total and its growth is pure adoption signal, uncontaminated by substitution effects. When the networks, or their large participants, begin disclosing that figure, the migration will have entered the stage where it gets measured, and in payments, what gets measured gets budgeted. The pilots are over in every sense but the accounting.

What it means

The public story of stablecoins this week was consortiums and neobanks. The durable story is here, in the obligation layer of the card system: once issuers and acquirers can hold and settle in tokens at any hour, the case for pre-funded accounts parked across time zones dissolves on its own schedule, quarter by quarter, CFO by CFO. That is not a crypto product. It is the treasury architecture of the card industry being rebuilt underneath everyone standing on it, nine chains at a time.