The Stablecoin Banker, June 9

June 9th, 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.

ICYMI: Banking Stablecoins whitepaper

Klaros, Hunton, and Omnia just released a deep dive whitepaper on how banks can implement stablecoin payments. Banking Stablecoins: A Regulatory and Operational Framework covers stablecoin use cases, balance sheet models, and deep regulatory analysis relevant to any banker working on a stablecoin strategy. This is not another glossy vendor info-sales brochure, and will multiply your understanding of the space.

"I don't like that we're going to support stablecoins, but our customers want them."
— Jack Dorsey, CEO, Block (May 27)

In This Issue

• Four companies launched stablecoin products, but we're reminded that scale isn't distribution
• Paxos becomes the second US securities clearing company, and it runs on chain
• Mastercard finally catches up to Visa with stablecoin settlement

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

Lots of Scale - No Distribution

Four companies launched USD-backed stablecoins or stablecoin payment features into large existing user bases. SoFi, an OCC-chartered national bank, launched SoFiUSD to its 15 million members, with instant swaps into/out of deposit balances. Block's Cash App rolled out USDC payments to roughly 60 million users, with free conversion from each user's existing dollar balance. MoneyGram launched MGUSD across a network of 500,000 retail locations and 60 million customers. International-focused payroll platform Deel began its launch of DLUSD, starting in Argentina with plans to reach 1.5 million contractors across 150 countries. Deel's wallet allows contractors to hold DLUSD, spend it using a Deel Mastercard debit card, off-ramp it to local fiat accounts, and eventually earn rewards on the balance. Sources [CoinDesk, Yahoo Finance, PYMNTS]

My Take:

This week's news invites a distinction that gets lost in the excitement: scale and distribution are not the same thing.

Deel and MoneyGram both have scale across millions of users and global networks of payment service providers. Using a stablecoin for balances solves a real operational problem: offering dollar balances to users in dozens of countries through a US bank would require a partner willing to ledger millions of foreign retail accounts and absorb the BSA/AML exposure, for less yield than a stablecoin can pay. The further elegance of the model is that it disconnects pay-in from pay-out so that the companies can scale faster by no longer needing to be in the middle of every transaction. They can use local on-ramp providers to convert fiat to stablecoin, the balance sits in self-custody so the provider doesn't need to be a custodian, and the off-ramp runs through the card networks and a long tail of local stablecoin-to-fiat providers. A bank never needs to enter the picture directly.

The companies clearly have the scale where issuing can matter, but should they use a branded coin? The issuers of DLUSD and MGUSD (Bridge and M0 respectively) share ~90% of reserve yield with the brand sponsor. But, so do many existing third-party stablecoins. The only reasons to brand a stablecoin are either for the marketing boost, or because the brand believes it has a shot at actually building distribution.

I think the rationale for both MoneyGram and Deel is heavily tilted towards marketing, and it's through this analysis that we see the difference between scale and distribution. Notice that neither company is working to create off-platform utility for its token. Deel shows users a USD balance and forces conversion into USDC or USDT the moment they want to move money out. MoneyGram has said nothing about distribution beyond its existing network, which already serves MoneyGram customers without stablecoins. These are captive stablecoins, which are sensible for their business models to profitably hold balances, but not for driving general acceptance.

Their scale is in their user base and how many access points they have for their private network. The market excitement seems to confuse this scale for distribution. This is where even the most powerful institutions get stuck. JPMorgan could get every client to hold and transact in a JPM stablecoin, and the volumes would be enormous. But if those transactions all clear on JPMorgan's balance sheet, the token is window dressing. The Holy Grail is what Circle and Tether have achieved: unrelated third parties, with no prior relationship to the issuer, accepting and using the token because it's genuinely useful to them. That is fundamentally different from any closed-loop model, regardless of scale.

CashApp and SoFi are playing a different game. They are trying to add payments value to their balance-first product by seamlessly connecting to stablecoin rails. Their par-redeemability model in which the user sees a dollar balance and the app handles stablecoin conversion silently (and for free at the moment) makes it easy for users to transact with any stablecoin counterparty. That is exactly the instant par-redeemability model I describe in the Banking Stablecoins paper, and it's the one most natural for a bank. CashApp is the cleanest example because they are using USDC, whereas SoFi is promoting SoFiUSD, presumably thinking they will be able to crack the distribution nut.

My hypothesis on how the issuance landscape may evolve: While the game is certainly not over, I have yet to see a credible strategy to crack USDC/USDT distribution. New issuers like SoFi face the tradeoff of hobbling their balance product by forcing the use of their native stablecoin with no distribution, or connecting directly to USDC/USDT and losing whatever advantage their scale gives them to attract distribution (which is overstated anyway). Meanwhile, swapping between stables will only get cheaper and easier as liquidity increases, balance products use their float income to offset costs, and issuers start offering in-kind mint and redeem with tokenized debt products (which I wrote about on May 26). I think that the balance products will all use self-issued or third-party stablecoins for balances. They will offer the Deel/CashApp-like experience at low, or no, cost. Value will flow between balances using USDC/USDT. This implies slower growth rates for those tokens vs. all of the others.

So for the bankers following along, the advice remains the same. Use stablecoin payments as a way to make your balance product better than anyone else's. Don't bother issuing a token, unless the performative act for marketing is worth the legal bills. Read our new paper to understand the implementation, regulatory positioning, and business model.

First Banks, Now Securities Clearing Gets Unbundled

On May 28, the SEC granted stablecoin issuer Paxos a temporary registration as a clearing agency, authorizing it to operate as a central securities depository and securities settlement system — the first blockchain-native firm to hold the registration. It will clear and settle eligible securities through the Paxos Settlement Service, a ledger that tokenizes both the security and the cash leg and settles them atomically: a bilateral delivery-versus-payment model with no central counterparty, no guarantee fund, and no multilateral netting. The registration is temporary, lasting up to 18 months, and carries operating conditions including a limited number of participants in its first year. It makes Paxos only the second central securities depository ever registered in the United States, ending a five-decade stretch in which DTC was the sole securities depository. In parallel, the SEC delayed its "innovation exemption" for tokenized stocks after exchanges objected to a draft provision that would have allowed third-party synthetic tokens — digital representations of shares issued by intermediaries without the company's involvement. The delay is indefinite pending resolution of token rights, shareholder protections, and dividend administration. Sources [Paxos, Unchained]

My Take:

For months I've argued that banking is being unbundled. Trust charters are decoupling digital asset services from banks, and with the advent of tokenized everything, they will be able to provide more bank-like services. Federal Reserve payments accounts will enable non-depository institutions to offer payment services. The through-line, as I put it in October, is that tokenization lets us break a bundled financial product into its constituent attributes and price each one on its own. The Paxos CA license is that story for the securities market.

First, let's define the bundle. DTCC fuses five things into one utility: custody, clearing, netting, settlement, and risk mutualization. Anyone wanting to clear trades in the US must be a DTCC member. It was built to solve a genuine crisis: in the late 1960s, trading volume buried Wall Street's back offices in paper, and the NYSE had to close on Wednesdays just to let them catch up. The fix was to lock the physical certificates in a single depository and settle on a net basis. Everyone trades throughout the day and squares up net differences in cash and securities only at the close of business. Because members don't entirely trust each other, they post margin and pay into a shared guarantee fund.

Fast forward, and the system built to cure slowness has become an engine for leverage. Net settlement lets a member trade against money it doesn't have yet with the guarantee fund backstopping the gap. Currently, each dollar of backstop guarantees $50 of trades. But that fund isn't a public good like a library or a road. It's insurance, and the fully-funded traders who bring no risk are subsidizing the premium for the leveraged firms that do. Separate those two use cases and the funded trades could settle on far simpler, cheaper rails.

Trading on-chain works because it solves for the problems that originally drove the creation of net settlement. Smart contracts guarantee atomic settlement - either each party gets their asset or the trade fails completely, guaranteed. The cash leg moves at the same speed as the securities leg thanks to tokenization. The cost is mere pennies. We see these benefits in the shape of the atomic gross settlement capability for which Paxos received approval. That means no netting, no insurance fund, and no margin.

This unbundles the market because it allows us to separate the benefits of trade settlement from the benefits of netting and margin. Traders can still get margin from lenders, just not from the settlement system. High-frequency traders can use an off-chain book with net settlement and pay directly for that benefit, while settling the net trades on-chain. The benefits go further. Tokenized assets settling on-chain means that traders actually hold their assets in bearer form, making them much more portable. This enables the creation of cross-asset-class trading, e.g. direct trading of stocks for bonds, portfolio margin across assets, and more.

So why should a banker who never touches DTCC care? For fifty years there was one securities depository in this country; there are now two, and the second one runs on a blockchain. This is another example of a regulator licensing a blockchain-native firm to peel one slice off an incumbent's bundled monopoly and sell it standalone. This is similar to how trust charters are offering custody, trading, and prime services outside of banks, or how Fed payment accounts will pull payment execution outside of banks. DTCC had the lead, a regulatory moat, and the comfort of being indispensable, and it still ended up with a competitor inside the perimeter. "We're regulated and they're tiny" is exactly what every incumbent says right before the unbundling reaches it. The exercise worth doing at your bank is to name which component of your bundled service is most exposed to being priced and sold on its own, and decide how you might offer that on a standalone basis, or how you make the bundle so good that customers don't seek the standalone.

Mastercard Catches Up to Visa

On June 3, Mastercard moved 24/7 on-chain card settlement from pilot to production, letting card issuers and acquirers settle in regulated stablecoins USDC, PYUSD, RLUSD, SoFiUSD and others across eight blockchains, with CBW Bank, Cross River, Lead Bank, ARQ, and Nuvei as early participants; the same capability adds intraday, weekend, and holiday fiat settlement. The launch followed Mastercard's receipt of a NYDFS BitLicense on May 27. Also on June 3, CoinDesk reported, citing three sources, that Stripe, Visa, and Mastercard are close to launching a joint stablecoin platform, with Coinbase in talks to join; no name, structure, or token details were disclosed. Sources [Mastercard, CoinDesk]

My Take:

Mastercard is finally catching up to Visa. Mastercard actually held the lead in 2022, when it launched the Mastercard Token Network and then fell behind, and this week shows why. As a New York company, Mastercard couldn't touch stablecoin settlement without a NYDFS BitLicense, a restriction that California-based Visa never faced.

Realistically, these announcements just restore parity to what Visa has done for more than a year: let card issuers settle daily in stablecoins. That matters for two reasons. First, daily settlement cuts the network's risk and frees the collateral an issuer must post to guarantee settlement. For a bank whose opportunity cost on that collateral is a business or consumer loan, that's a real unlock. However, funding weekend settlement requires holding stablecoin, which quietly pushes the idle-capital problem back onto the bank's own balance sheet.

Second, and more threatening to banks, when the network accepts stablecoin settlement, the sponsor of a stablecoin-backed card can skip the bank and the fiat system entirely, bypassing both the weekend funding gap and the bank's costs. This is how issuers like Rain work: when a card transaction is authorized, the issuer takes the backing stablecoin into its possession and can send it directly to the network on the next settlement cycle, no prefunding required except by the user. That's most compelling for issuers like Rain that hold non-bank card issuing licenses. Uptake of those products has been minimal in the US so far, but it's growing fast everywhere else, as evidenced by Rain's most recent funding valuation of $2 billion in January. The card networks are building the rails that make a bankless card issuer viable.

This all sounds similar to the discussion on unbundling of the DTCC. The 5-day settlement system was built in a time when money only moved during banking hours. To bridge that system to a 24x7 card payment network, a daisychain of liabilities forms, with regular net settlement, and collateral to reduce risk. Stablecoins allow that to unbundle, with immediate gross settlement now viable and separable from funding mechanisms for delayed settlement, whether that's the bank, the card program, or the cardholder choosing to borrow funds to make a settlement down the chain.

We seem to be moving beyond the surface level excitement of stablecoins as merely faster money, and starting to pursue the deeper value of using tokenization to decompose calcified financial systems built for a different era. The faster money lens is not inaccurate, but it belies the transformational potential of the technology and leaves many, especially those in the US, scratching our heads about the domestic utility of stablecoins. Even here, stablecoin cards can make sense in the unbundled system. For example, the cardholder could more easily switch lenders without changing their card number, lower their APR with instantly-issued lines backed by tokenized assets, or pool credit across cards.

Decomposition is as much an opportunity as it is a threat for banks. While stablecoin settlement enables companies like Rain to issue without ever touching a bank, it will also give banks more flexibility to create differentiated card products. The Visa-Mastercard-Stripe deal, should it prove to be accurate, reflects the degree to which the card networks want stablecoins to happen. Issuing and acquiring banks have a unique opportunity now to lean in and leverage their existing distribution to shape how the market evolves.

Coupon Clippings

Falcon Finance and Anchorage Solve the GENIUS Act Yield Puzzle

Anchorage Digital Bank issued fUSD, a GENIUS Act-compliant stablecoin that offers institutional holders 3% annual rewards. The coin remains compliant because the underlying asset manager, Falcon Finance, pays the rewards - not Anchorage. Yield finds a way, and it does so quickly. fUSD is institutional-only today, so it's one particular route around GENIUS rather than the universal retail model, but it's a clean illustration of the pattern. It likely threads CLARITY too: the entity paying the reward is an asset manager that isn't an "affiliate" of the issuer, so a yield prohibition aimed at issuers and their affiliates doesn't obviously reach it. [Full Story]

Zero Hash Files for an OCC Trust Charter

Zero Hash applied to the OCC for a national trust bank charter, proposing stablecoin issuance, custody, trade execution, and transfer-agent services under federal supervision; it already powers crypto and stablecoin rails for Morgan Stanley and roughly 60 fintechs. The Morgan Stanley relationship is what makes this more than another entry in the charter race — a federally chartered Zero Hash would compete directly with bank trust departments for the digital-asset mandates those departments assumed were theirs. Ten crypto firms now hold conditional or operational national trust charters with another fourteen pending. Sources [Banking Dive]

Who Owns the Phone Book for AI Agents?

The barriers to an AI-agent economy are collapsing fast. The Linux Foundation launched DNS-AID, an open-source project that lets AI agents and MCP servers discover and verify one another through the existing DNS system, which is already used by every internet-connected device in the world. If domain owners adopt the standard, any user can instantly discover the organization's services from their AI frontend (Claude, ChatGPT, etc.). This is potentially great for banks. Today, most bank frontends are monolithic, treating every user exactly the same despite obvious differences that could be used to create tailored experiences. Since banking services are highly commoditized, the bank is probably no better than the user at building the banking UI. We're moving into an era where the frontend for a bank can be an AI chatbot, or a custom UI built by an AI chatbot. Not only that, the bank gets to dynamically interact, asking questions to gather rich intelligence in every session. The bank that owns discoverable service endpoints may end up mattering more than the bank that owns the prettiest app. [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.