September 23, 2026
Three stablecoin developments bankers should know about

The Stablecoin Banker is a periodic newsletter keeping bankers on top of the stablecoin industry. I highlight top stories that are relevant to banks, with my insights and commentary to draw out the most important conclusions.

If you find this content useful, please feel free to forward it to your friends and colleagues. They can also subscribe directly or read back issues at thestablecoinbanker.com. If you're interested in stablecoin services for your bank, feel free to reach out to us at Omnia.

"I think we're done. It's over."
— Sen. Cynthia Lummis, after the CLARITY cloture vote failed 49–50 (Sept. 15, 2026)

In This Issue

• Stablecoin meets fiat, and somebody gets the bill
• A Stablecoin and a Deposit Token Walk Across the Border
• Core or Filler: Stablecoin's Two Jobs in the Global Account

Plus, tidbits you may have missed in our Coupon Clippings section.

Stablecoin Meets Fiat, and Somebody Gets the Bill

Three deals in three days moved stablecoin companies deeper into fiat payments. Circle agreed to buy Tazapay for $400 million in stock. Tazapay is a Singapore-based cross-border payments company with local payout rails in more than 100 markets. MVB Financial said it will join a Visa Direct pilot, delivered through stablecoin infrastructure firm Velocity, that lets the bank fund and settle push payouts in stablecoin. Velocity's licensed partners perform the conversion and hold the wallets. A day later, Velocity extended its Series A to $48 million at a $200 million valuation, adding Visa, Circle and Ripple as investors. Finally, Coinbase and Moov announced a partnership to offer stablecoin acceptance, settlement and payouts to the more than 1,000 community banks and credit unions on Moov's platform. Under the arrangement, Coinbase's custodial wallets hold the funds and its Payments API moves the stablecoin, while Moov connects those functions to the institutions' existing payment systems. Sources [Banking Dive, Velocity, Coinbase]

My Take:

As I've written before, stablecoins are becoming money and these three deals reflect the process of merging the fiat and tokenized money systems. They also expose the messy interface between the two where a 24x7 rail carrying a new asset has to meet a 9-to-5 rail carrying only fiat. That interface requires new intermediaries, new funds flows and prefunding, and somebody to pay for the convenience of moving more seamlessly across the interface. As companies roll out fiat-to-stablecoin capabilities, the open question is who will pay the new cost, since there's no market default.

Circle is coming at this from the stablecoin side. Until now, it has largely left the edges of the blockchain, the on-ramps and off-ramps, to partners. Circle Payments Network admits only regulated financial institutions who provide ramp functions on their own terms, yet in my experience the vast majority of stablecoin-to-fiat flow runs through liquidity providers that aren't regulated banks. Tazapay gives Circle its own ramps, much as Rail did for Ripple and RLUSD last year. The difference between partnering and owning is incentive. A partner routes customers to whichever coin works best for them, increasingly using a yield-sharing coin for balances and converting to/from USDC only when a transaction demands it. A ramp that Circle owns has every reason to make USDC the only option. I suspect Circle will offset ramp fees with backend reserve yield, meaning USDC holders and distributors would pay for the services while ramp users are less likely to swap into a competing coin.

Coinbase is making the same move toward community banks, and the press release isn't encouraging. Coinbase holds the wallets and moves the stablecoin, not Moov and not the bank. I wrote at length about this arm's-length model in our paper, Banking Stablecoins. It typically means sending customer PII to the stablecoin provider, prefunding that drags on the balance sheet, and handing the customer experience to a third party that I doubt will look after the bank's customers the way the bank would. MVB's arrangement shares the flaw, with conversion and wallets performed by Velocity's licensed partners. With Moov, who pays is undecided. Coinbase could absorb the cost in exchange for customer access, Moov could treat it as a loss leader to win more banks, the banks could eat it in NIM, or customers could foot the bill.

Velocity looks like most other stablecoin payments companies, which makes the checks from Visa, Circle and Ripple less interesting than the pilot itself. Here the bank may come out ahead. Visa requires senders to fund cross-border payouts in advance, and those balances do not earn interest. Using stablecoins enables just-in-time funding to Visa, which doesn't eliminate pre-funding as much as shifting the problem elsewhere: now the bank will need to pre-fund its stablecoin balance on its balance sheet or through an arrangement with Velocity or a third party, drawing on the respective party's cost of capital. Depending on the coin it chooses, though, the bank can earn some return on that balance, which Visa never paid. The benefit is narrow. Domestic push-to-card payouts already clear through the normal card settlement cycle, so there is little capital to free up there. And because instant payouts are priced at a premium, I expect the bank to keep whatever it saves.

Stablecoin is being woven into the financial system around existing institutions, whether through Visa Direct funding, stablecoin payments or some other route. Implementations create new cost structures that need to be apportioned. Bankers considering a stablecoin product should map the actual flow of funds to work out the costs, including the cost of capital, and to understand the customer experience. Then, figure out who will pay for it all.

A Stablecoin and a Deposit Token Walk Across the Border

U.S. Bank said on that it completed a live pilot cross-border payment using USBDC, its own dollar-backed stablecoin, on the public blockchain Stellar. The payment moved between U.S. Bank's own entities in North America and Europe and involved no client. It tested minting, redemption, freezing and clawback, and the bank did not disclose the amount or how the coin's reserves are held. Separately, DBS and Citi's New York office completed a Singapore-to-US dollar payment using tokenized deposits over Swift's blockchain-based ledger, which 17 banks are piloting. DBS said the payment completed in minutes, compared with up to two business days today. Sources [U.S. Bank, DBS]

My Take:

Last issue I floated the idea that a bank-issued stablecoin could stand in for both nostro/vostro arrangements and tokenized deposits. Two weeks later we have a live example of each approach to cross-border bank payments. Set side by side they are the same transaction, with one difference that matters to any bank deciding which to build on.

U.S. Bank didn't describe its funds flow, so here is how I expect it works. A customer's deposit is used to mint USBDC, and the reserve behind the coin sits on U.S. Bank's own balance sheet as a deposit, identical from an accounting perspective to the customer deposit it replaced. The coin goes to the European entity, which holds it as a bank-owned asset and credits the recipient's deposit account. The European-held coin, paired with the deposit liability recognized in the US, is a nostro/vostro relationship. When Europe redeems the coin, the US bank sends a Swift wire. (Note that the OCC's proposed diversification rules cap stablecoin reserves held as deposits at a single bank at 40%, so in reality U.S. Bank would need to hold some reserves in US Treasury assets or on deposit with another bank. In a consortium model like the Big Bank Coin, reserve federation across multiple banks solves this challenge.)

DBS and Citi reach the same outcome with a different product. The sending bank creates a deposit token and passes it to the receiving bank, which credits the beneficiary. (In reality the Swift system lets banks avoid ever holding each other's tokens, so it looks more like a faster database that tracks netting obligations.) Later the position is cleared with a conventional interbank money movement, the same one the Swift wire triggers in the stablecoin version.

For bankers: focus on the fact that the same payment can be made with a deposit token or a stablecoin, on the same netting and nostro/vostro mechanics. Then note the difference. With the deposit token, banks are limited to transacting among themselves. With the stablecoin, a bank can transact directly with anyone it chooses to approve, no network membership required.

Core or Filler: Stablecoin's Two Jobs in the Global Account

Nu, the largest digital bank in Latin America with more than 140 million customers, introduced Nu Global, a multi-currency account available in more than 35 countries, starting with Europe and Latin America. All deposits are converted into stablecoins USDC or EURC, which earn 3.50% and 2.20% APY respectively, and the account comes with a virtual Mastercard that spends worldwide with no exchange-rate markup. Nu's home markets of Brazil, Mexico and Colombia are not yet connected, nor is the US, where Nu launched a separate bank-partnered account the same day. Column, the nationally chartered bank founded by Plaid co-founder William Hockey, launched a set of international products together: US dollar accounts for non-US individuals and businesses, multicurrency accounts with their own IBANs and sort codes on SEPA Instant and the UK's Faster Payments, card issuing on its own processor, and 24/7 conversion between dollars and USDC or USDT. Column's cards can be backed by stablecoin balances and settled with Visa and Mastercard in stablecoin. Sources [Nu, Column, Column]

My Take:

The product format for the global account is being cemented, and stablecoin is finding its place in it. Money movers like MoneyGram and Deel got there first, with stablecoin balances that have cards attached (or are in flight, as described for MoneyGram in the clippings below). Nu and Column show the two models now emerging, and banks that serve fintechs or cross-border customers will be measured against both.

Nu runs the fintech model, which is striking coming from the largest digital bank in Latin America. Historically, offering accounts across countries meant finding a sponsor bank in each one, or partnering with a licensed payments operator and complying with local fintech rules. Stablecoin unlocked a different approach: give the user a stablecoin wallet, custodial or not, and connect it to local providers that convert fiat into stablecoin. Decoupling the on-ramp from the account lets a firm plug into many jurisdictions at once. Payments work the same way: local providers for local bank payments, a stablecoin-backed card for spending anywhere, and a single sponsor bank or payments company as the fallback for Swift. The business looks more like asset management plus payments than banking. The provider earns a spread on the stablecoin yield plus FX markups or flat fees. That Nu chose this route over extending its own deposits across borders says something about what each costs to build at global scale.

Where stablecoins are stateless (a non-custodial balance is held nowhere and everywhere at once), a bank's balances are tied to a specific jurisdiction. A fintech builds a global product starting from stablecoin. A bank has to build one starting from a deposit. Column did the expensive work: dollar and foreign-currency accounts for US and non-US holders, local payments access in Europe and the UK, and global card issuing. Stablecoin fills the gaps that are tricky to fill on the banking side. Column can issue cards globally against a traditional account, yet it also offers stablecoin-backed cards, because they require little or no prefunding with the network.

So there are two models: stablecoin-core accounts that use banks to reach payments, and deposit-core accounts that use stablecoin to route around legacy-rail friction. Stablecoin-core is much cheaper to build but can cost more to run, since stablecoins are often more expensive than bank payments in high-volume currency corridors. Deposit-core is far more expensive to build but comes with a lower cost of capital and the flexibility of a full bank charter. I expect the fintech model to get cheaper to run as liquidity deepens and spreads compress. Meanwhile the appetite in banking for a Column-style build apparently remains muted. Just compare Column's API with the API at any other fintech sponsor bank.

Either way, stablecoins are finding their role, as a core balance primitive or as the filler between existing financial system primitives. Sponsor banks and banks with international customers should compare their feature set to Column's, and recognize that the gaps are addressable today by stablecoins at a fraction of the build costs. The gaps may not be visible in customer attrition rates today, but the development of both stablecoin models, and the visibility of entrants like Nu, will only raise the bar.

Coupon Clippings

CLARITY Stalls, and the Regulators Don't Wait

The Senate rejected a motion to advance the CLARITY Act on September 15 by a vote of 49–50, well short of the 60 needed. Democrats voted no unanimously, with ethics provisions covering the president's family the main sticking point, and Republicans Susan Collins, Josh Hawley and Jerry Moran also voted no. Thom Tillis switched his vote to no to preserve a motion to reconsider, so the bill isn't formally dead, though Republican backer Cynthia Lummis said, "I think we're done." The failure leaves crypto in a regulatory grey zone, and the SEC and CFTC are already moving to fill it: within two days both chairs promised rules with or without legislation, though agency rules are easier to unmake than statute. For stablecoins the failure is good news, because the route by which issuers remit reserve yield to distributors, who pass it on to holders as "rewards," stays open for now. The bill already restricted those workarounds, and the banking groups that stood to benefit opposed it for not going far enough. Their next shot is the OCC's final GENIUS rule, due by November, which as proposed would presume such arrangements violate the yield ban. [Full Story]

MoneyGram Swaps the Bank Account for a Stablecoin Balance

MoneyGram launched the MoneyGram Card in Colombia, a digital Visa card built with Rain and Crossmint inside its app, that spends from a stablecoin balance, with expansion across Latin America planned. The card supports USDC at launch, and MoneyGram says it will add MGUSD, the stablecoin it launched in the US in June, in the near future. The company won't call it a debit card, saying it "is stablecoin-backed and is not a traditional bank account or fiat-currency balance," and it replaces the bank-issued account and debit card MoneyGram closed in December 2025. MoneyGram is building its global banking product on the fintech model I describe above, with stablecoin at the core, and it traded away a bank product to do it. That the card runs on USDC and not MGUSD speaks to the flexibility of stablecoins: MoneyGram didn't have to wait for its own coin to reach Colombia, and the eventual move to MGUSD will probably be a swap rather than an account migration. It has a reason to make that move, since its own coin brings it reserve yield. [Full Story]

Thanks for reading. If you don't hear from us for a bit, don't worry—we'll be back when there's something important to share.

Omnia is a provider of stablecoin infrastructure for banks that want to capture growing demand for stablecoins. If you're interested in learning more about us, please get in touch.

Getting value? Forward to a colleague and subscribe.