Visa opens its settlement records so lenders can finance stablecoin cards

What happened

Visa announced on 8 September 2026 that it will pair its own settlement data with onchain lending infrastructure to finance the working capital of stablecoin-linked card programmes. The wire release, headed 'Visa Brings Onchain Lending into Everyday Payments' and datelined San Francisco on 8 September, describes 'a new approach to onchain credit designed to help stablecoin-linked card programs and fintechs access working capital using onchain lending infrastructure and Visa data'. The problem it addresses is a funding gap that exists in every card programme and is sharper in this one. A programme must settle with the network before it has collected from its cardholders, and the instruments that ordinarily bridge that gap, warehouse lines and securitisation, are documented facilities sized for large portfolios. Visa's own case study, published the same day and running to seven minutes, states the mismatch plainly: 'A program in its first year might need a few million dollars, drawn and repaid on a daily cycle, secured against receivables it has only recently begun to generate. The fixed costs of documenting a traditional facility are hard to justify at that size, and the performance history a lender would want to underwrite against does not yet exist in a form that is easy to assess.' Many of these programmes settle every day of the week including weekends and holidays, which no conventional facility is built for. The mechanism Visa is backing is a stablecoin-denominated revolving credit facility run by Credit Coop, secured on the settlement receivables themselves. Three parts make it work. Repayment is enforced in code: 'Settlement receivables flow through Credit Coop's Spigot smart contract before reaching the borrower's operating account, and the Spigot routes repayment automatically from incoming proceeds', which Visa compares to 'a lockbox operating under a deposit account control agreement (DACA), enforced programmatically rather than through manual sweeps'. Underwriting runs on Visa's own records: 'As a registered Third Party, and with each program's authorization, Credit Coop receives the program's daily Visa settlement files directly through a secure data pipeline', and facility sizing, disbursement and repayment verification are calibrated against those files alongside the onchain repayment history. And the price has moved: 'borrowing costs for participating programs have come down by as much as 30%.' The figures are the reason the announcement matters. Visa says more than 160 stablecoin-linked card programmes were live on its network in its fiscal second quarter, that payment volume on those programmes grew nearly 200 per cent year over year, and that 'stablecoin settlement volume recently surpassed a $20 billion annualized run rate, up more than 15x year over year'. On the credit side it says the model 'has supported more than $2.5 billion in cumulative financed settlement volume since 2023 with zero defaults across participating facilities' and has 'processed more than 3,000 borrow events and 9,000 repayment events programmatically onchain'. The market it places this in is measured from its own dashboard: 'since 2020, more than $694 billion in stablecoin-denominated loans have been sent through onchain lending protocols, creating a global credit market that operates 24/7', a figure the release attributes to the Visa Onchain Analytics Dashboard. The case study names the borrower that accounts for most of the record. Rain, which it describes as 'a Visa Principal Member that powers stablecoin card programs around the world', has funded its daily Visa settlement obligations through a Credit Coop revolving facility since August 2023; the table Visa prints for it gives approximately $2bn of cumulative settlement volume financed, zero defaults, more than 2,000 borrow and 7,000 repayment events, and interest paid to date of '$1.58 million and up'. Karta, a premium Visa credit card for travellers operating under Rain's BIN, is given as the proof of the graduation the whole scheme is aimed at: it 'launched and scaled on a Credit Coop facility at a point when its performance record was still being built', and in June 2026, on the strength of 10x growth in 2025, announced a $140m raise made up of a $15m Series A led by Galaxy Ventures and a $125m institutional credit facility from Community Investment Management. Moto and Xplace are named as two further programmes under Rain's BIN on the same infrastructure. What Visa says comes next is a move from a drawn facility to funding on demand: 'a program's daily settlement file can trigger a same-day disbursement for exactly the net amount owed, sent to the Visa settlement address', so that a programme pays for capital only while it is deployed and 'the window is hours rather than days'. Two people are quoted. Rubail Birwadker, Visa's global head of growth products and partnerships: 'Stablecoins are not only changing how money moves, they're creating opportunities to rethink the financial infrastructure that supports payments... We're seeing how trusted payment data and onchain technologies can work together to unlock new forms of liquidity, helping businesses access capital in ways that are more transparent, programmable and aligned to the speed of modern commerce.' Chris Walker, founder and chief executive of Credit Coop: 'Payment companies have always had good collateral in their settlement receivables, but no way to show lenders how it performs in real time. By combining Visa settlement data with onchain infrastructure, we can evaluate live performance, enforce repayment from the settlement flow and extend capital onchain from participating lenders as a program grows.' Visa's head of crypto, Cuy Sheffield, put it to CNBC in more provisional terms on the same day: 'We've been running a pilot with a company called Credit Coop that is enabling a credit facility for stablecoin-linked card providers, which we think is a positive step forward for how onchain credit can start to come into our network.' He also said that 'Stablecoin-linked cards are in hypergrowth mode', with new issuers joining and launching cards every week.

Why it matters

The interesting thing here is not that a payment network is using a blockchain. It is that Visa has made its private settlement records into the underwriting input for someone else's loan book, and has done it in a structure where the collateral cannot be diverted because the receivable never reaches the borrower's account unencumbered. That is a genuinely new asset, and the description Visa reaches for is the right one: the Spigot is a lockbox under a deposit account control agreement, written as a smart contract instead of as a document. The reason that matters commercially is stated in the case study and is about fixed costs rather than technology. Receivables finance works at scale because the legal apparatus of a warehouse line is expensive to build and cheap to run. A programme that needs three million dollars drawn and repaid daily cannot carry that apparatus, so it does not get financed, whatever the quality of its receivables. Automating the control and the servicing moves the minimum viable facility size down, which is why the cost of borrowing for participating programmes has fallen by up to 30 per cent. Second, the figures are the first credible size for a market this desk has been covering without one. A $20bn annualised stablecoin settlement run rate at Visa, up more than fifteenfold in a year, is the number to carry forward: it is not stablecoin supply, which is what the industry usually quotes, but stablecoins actually clearing through a card network's settlement leg. Against a total stablecoin float of roughly $305bn, it says that the share of stablecoins doing retail payment work is still small and growing very fast. The 160 programmes and the near-200 per cent growth in payment volume come from the same place, with the caveat below. Third, this is a bank disintermediation story told from inside the incumbent. The case study is careful to say that 'none of this displaces the institutional credit market. It adds to it', and offers Karta's path from a small revolving facility to a $125m institutional one as the model. But the sequence it describes is that onchain credit seasons the asset, the performance record is built in public, and the bank arrives afterwards to take the mature exposure. The banks' own answer to stablecoins, the twenty-one-institution consortium this desk covered on 1 September, is about issuing the money. This is about lending against the receivables the money generates, and the firm doing it is a protocol nobody outside the sector has heard of. Fourth, it connects two clusters in this corpus that have not been read together. Everything about this depends on the receivable being legally solid and the control being enforceable, which is the same question ESMA asked of tokenized equities on 10 September and OSFI answered for deposits in Canada the same day. The Spigot works because a claim on a settlement flow is an ordinary receivable, subjected to an ordinary security arrangement, that happens to be administered by code. Nothing in it requires the law to recognise a token as anything new. That is the pattern of the week across three continents: the things that work are the ones that change the plumbing and leave the legal characterisation alone. Finally, the footnotes deserve to be read as part of the document. They show that the headline growth figures date from a quarterly earnings call in April and that two of the claims were still being checked when the page went out. For a story whose whole point is that better data produces better credit, that is worth noticing.

What is not settled

The numbers are not as current as they read. Visa's own footnote says the programme count and the payment volume growth come 'from the Visa fiscal second quarter 2026 earnings call, April 29, 2026', which is four and a half months before publication, and then adds: 'Stablecoin settlement run rate per Visa; figure to be confirmed with Investor Relations ahead of publication.' The second footnote says the Credit Coop figures are 'provided by Credit Coop (CMBT Labs Inc.); onchain event counts as of August 19, 2026. Zero-default status to be reconfirmed by Credit Coop immediately prior to publication.' Both instructions are still on the published page, so a reader cannot tell whether either check was carried out. The zero-default record is also a claim about a portfolio nobody outside has seen: there is no facility list, no drawn balance, no arrears schedule and no independent audit, and the concentration is severe, with roughly $2bn of the $2.5bn attributable to a single borrower over five years. Two accounts of Credit Coop's scale are in circulation and they are not the same measure. Visa says $2.5bn of cumulative financed settlement volume since 2023; CNBC says the firm 'has processed $2.7 billion in total volume on its platform'. The second is larger and broader and neither party reconciles them. CNBC also moves Visa's growth figure onto the wrong quantity, writing that Visa operates more than 160 programmes, 'a nearly 200% increase year over year', where the release attaches the 200 per cent to payment volume and gives no growth rate for the count. The $694bn of stablecoin-denominated lending since 2020 is sourced to Visa's own dashboard, which renders client-side only and served this desk no figures at all; the lending page names Allium as its data provider and gives no definition of what counts as a loan sent through a protocol, so the number cannot be reproduced from the address Visa cites. Credit Coop itself is undescribed by anyone but Visa: its site renders client-side only, its legal name appears once in a footnote as CMBT Labs Inc., and there is no public account of which chains the Spigot runs on, who the participating lenders are, what the facilities cost, or what happens to the collateral if a programme fails between a draw and a repayment. Visa is also the interested party on both sides here. It supplies the settlement data that makes the credit assessable, it benefits when more programmes launch on its network, and the case study ends by inviting banks and institutional lenders to partner with it. Nothing in either document says whether Visa has capital at risk, takes a fee, or holds any interest in Credit Coop. And Sheffield's word for the arrangement is 'pilot', which is not how either written document describes it; neither says how many programmes are actually financed this way today, as against the 160 that exist.