If you’ve were somehow offline last week then you may not be aware that on Monday 14th September the Senate failed to advance the CLARITY Act, 49 votes to 50, eleven short of cloture. There are plenty of post mortems that effectively wrote themselves, crypto’s market structure push is dead for 2026, the industry’s year in Washington ends with nothing, its all about the midterms etc. Whatever else died with the bill, the airtight yield prohibition the bank lobby spent the spring asking for, the one I wrote about in Caught Inside, died too. I read a lot of the coverage and couldn’t help think that a lot of it was missing the point.
Because a month ago, on 18th August, the Treasury Department published a proposed rule that most community bankers will never read, and unlike the bill that failed on Monday, it doesn’t need actually need a single vote to become the law. The rule implements section 3 of the GENIUS Act, which has been law for fourteen months. It governs who may issue payment stablecoins in the United States and what digital asset exchanges may offer for sale, and it is addressed to a cast of characters that doesn’t include a single community bank. So I’m thinking about stablecoin issuers and crypto exchanges, or even offshore token operators. As a banker, it kinda feels like reading someone else’s mail.
I think reading it that way may be a mistake, and I want to try and run through why.
The GENIUS Act created a new legal category of money. The payment stablecoin, a digital dollar claim backed 1:1 by T-bills and bank deposits, redeemable at par, issued only by federally or state supervised entities. Since then the implementing machinery has been grinding forward in two separate workstreams, and I think the distinction between them does matter. In February the OCC published its proposal, 376 pages of reserve composition rules, redemption timelines, capital floors and examination schedules. I think of it like that rule is the cage. It describes how a permitted issuer must live. The August Treasury rule does something different. It draws the fence line around the country. It defines what it means to issue a stablecoin “in the United States,” who may lawfully do it, what an exchange may put in front of an American customer, and how far US law reaches when the issuer sits in El Salvador or Singapore.
So in my mind, cage rules determine how your competitors operate and fence rules determine who your competitors are. For a community bank whose entire franchise is built on gathering local dollar deposits, a federal rule that settles who may compete for dollar balances, from where, and starting on what date, is not someone else’s mail. It’s really like the zoning decision for your street.
What Treasury actually decided
If we strip away the 43 comment questions and the definitional scaffolding then I think the proposal makes three moves.
First, it settles the geography. A stablecoin is issued “in the United States” if either the issuer or the recipient is located here, and an entity is located here if it’s organized under US law or runs its principal place of business here. Issuance by anyone other than a permitted issuer becomes a federal crime carrying up to $1 million in fines and five years in prison per violation, and Treasury takes care to say the rule is intended to have extraterritorial effect whenever a US person is on the receiving end.
Second, it opens a door that I think most people assumed was closed. The statute lets foreign issuers keep serving the US market if their home regime is determined comparable and they register with the OCC. I know that some read it as that this covered secondary trading only, and the coins could circulate on US exchanges but the issuers themselves stayed outside. Treasury rejects that reading. A registered foreign issuer from a comparable regime may issue directly into the United States, full parity with domestic permitted issuers. No regime has been determined comparable yet, and the determination requires unanimity from Treasury, the Fed and the FDIC. But the door is there.
Third, and I think probably most consequentially, it writes the offshore world into law. Proposed section 1523.2(c) gives a non-US issuer a safe harbor if it reasonably believes its coins are going to people outside the United States, runs controls designed to keep it that way, and doesn’t market to Americans, then it has done nothing unlawful. Securities lawyers will recognize the architecture immediately, it’s Regulation S adapted for tokens. What it means in practice is that Tether, with roughly $183 billion outstanding and a bit over 60 percent of the global stablecoin market, can keep serving Latin America, Africa and Southeast Asia indefinitely, lawfully, so long as it stays on its side of the fence.
Then there is the clock. One prohibition runs now where no exchange may offer a foreign issued stablecoin in the US unless the issuer can technologically comply with lawful orders, meaning it can freeze and seize on demand. The other runs later, from July 18th 2028, no digital asset service provider may offer or sell any payment stablecoin to a person in the United States unless it comes from a permitted or passported issuer. That is the delisting date for every non compliant coin on every US venue, backed by penalties of up to $100,000 per day for exchanges that keep listing designated foreign coins.
A perimeter, a passport, a safe harbor, and a date. I went looking and yep, we have run this exact experiment before.
The last time Congress fenced private money
In 1863 and 1864 the National Bank Acts created a new class of federally chartered banks authorized to issue national bank notes, private currency backed by specified collateral, federal bonds deposited at the Treasury, redeemable at par. Sound familiar? It should. The GENIUS Act’s permitted issuer, with its segregated T-bill reserves and par redemption obligation, is the same design with a distributed ledger where the engraved plates used to be.
But the charter alone didn’t reorder the system. In 1865 Congress added the clock with a 10% tax on state bank notes, effective the following year. Notice what Congress did not do. It didn’t ban state banknotes, it didn’t seize them, it didn’t declare them worthless. Rather it made them uneconomic to circulate after a date. And guess what? The state banks understood the message. Hundreds converted to national charters within three years, and the number of state banks collapsed from around 1,500 to roughly 250.
Section 3(b)(1) is the 10% tax. Non permitted stablecoins aren’t banned in 2028, they lose US distribution. To me this is the same mechanism and same design philosophy. Don’t outlaw the incumbent instrument, fence it out of the domestic market on a schedule and let the issuers choose between conversion and exile.
So the funny thing is the state banks that lost the note issuing business didn’t disappear. Shut out of currency, they rebuilt their franchises around something the tax couldn’t reach. Deposit accounts and the checks drawn on them. By the 1890s state banks outnumbered national banks again, and deposit banking, the business lots of readers of this newsletter are in, was substantially their invention. The fenced out party built the next system. I’d keep that in mind when assessing what Tether, and the wallet providers and exchanges living around the yield prohibition, do with the next two years. Institutions excluded from a charter don’t stop competing. They move to whatever the fence doesn’t cover.
The Eurodollar rerun
Of course, there’s a second precedent buried in the safe harbor In the 1960s, Regulation Q deposit ceilings and the Interest Equalization Act fenced the domestic dollar market. The dollars didn’t disappear. They moved to London, beyond the reach of both rules, and became the Eurodollar market, which grew from a curiosity of a few billion dollars into the multi trillion offshore funding pool that ended up setting the marginal price of dollars everywhere, including for the US banks the rules were meant to protect!
Treasury’s safe harbor is a deliberate Eurodollar design, and I doubt very much the drafters are naive about it. The rule constructs a two tier dollar. An onshore tier of permitted issuers, OCC supervised, freeze capable, holding their reserves in T-bills and US bank deposits, and an offshore tier that may serve the rest of the world provided it doesn’t solicit Americans. The offshore tier is really not a loophole. It’s the design. Offshore stablecoins extend dollar usage into markets American banks will never reach, they generate structural demand for Treasuries, and the enforcement lever, the issuer’s ability to freeze at government request, stays attached the whole time.
For a community banker the Eurodollar lesson is about prices, not geography. London didn’t take Main Street’s deposits in 1965. It set the rate expectations and funding structures that Main Street eventually had to live with. An offshore stablecoin tier serving your commercial customers’ suppliers in Mexico and Vietnam does the same thing. Customers will learn what instant, weekend, near free dollar settlement feels like from their counterparties, and then they will ask you why their own money moves on your schedule.
What the fence means for your funding
Now lets put the perimeter next to a banks balance sheet.
Today the entire stablecoin market is roughly $300 billion against more than $17 trillion of deposits in US insured institutions, which is why it’s still possible to dismiss all of this as just noise. I’d be careful with that comfort. The OCC’s own economic analysis of its companion rule works from forecasts of $500 billion in issuance this year and cites Standard Chartered’s projection of $2 trillion by 2028, the same year the distribution fence closes. And Treasury’s borrowing advisory committee has sketched scenarios in which as much as $6.6 trillion of deposit value could migrate into stablecoins over time. Lets be honest, nobody knows which of those numbers is right. If someone claims to, they’re selling something. But the perimeter rule converts the question from whether this competition is legal to merely how big it gets, and it stamps a date on the answer.
The New York Fed has already measured what happens to a bank that touches this flow the wrong way. Its staff research on stablecoin disintermediation found that partner banks holding stablecoin reserve deposits behave like narrow banks against them. One studied bank held roughly $1.5 billion in additional reserve balances and contracted its loan to asset ratio by 14 percentage points. Reserve deposits are real funding, and they may be the most accessible way for a community bank to participate in this system, the OCC is openly asking whether issuers should be required to place 20% of reserves at banks under $30 billion in assets, and whether insured stablecoin deposits should be spread across the country’s 4,380 insured banks rather than concentrated at a handful of custody giants. That’s a genuine opportunity, and community bank comment letters should fight hard for those allocations. But reserve deposits are hot, wholesale, and analytically nothing like an operating account of a local business. They fund a bigger balance sheet, not necessarily a bigger loan book.
The deeper point here is about what the fence protects and what it doesn’t. The GENIUS Act deliberately excluded tokenized deposits from the stablecoin regime entirely. A deposit recorded on a distributed ledger is still a deposit, still insured, still the banks to pay interest on. Congress fenced the stablecoin, then left the depository institution a product category with every advantage the stablecoin is denied. The permitted issuers can’t pay yield, and where the yield goes instead is a story I walked through in Money at Rest, Money in Motion. The bank can pay it. Their instrument can’t count as a settlement asset between banks unless it’s inside the perimeter. Banks deposits already are the settlement asset. I’m increasingly convinced the strategic question for regional and community banks isn’t how to fight stablecoins, it’s whether they’ll use the protected lane Congress drew for them before the 2028 reshuffle concentrates the market around whoever moved first.
A quick word about the trap door
Treasury’s rule gives examples of “knowing participation” in an unlawful issuance, conduct that attracts the criminal penalties. So things like taking on an obligation to redeem or guarantee someone else’s stablecoin, coordinating the machinery of an issuance such as customer onboarding or minting, and distributing newly issued coins to their first purchasers. Sponsor banks have spent three years learning that their fintech programs compliance failures become their consent orders. This feels like the same lesson with a criminal statute behind it. If a program partner touches token issuance anywhere in its stack, someone at the bank should be able to say which side of the permitted issuer line that partner sits on.
The comment window
In 1865 the state banks found out where the fence was when the tax arrived. Nobody asked them. Treasury’s comment window closes October 19, a month from now, and from what I can see, as of this week the docket has drawn twenty six comments. Twenty six, on the rule that decides who may compete for dollar deposits in the United States, from where, and starting on what date. The OCC’s parallel rulemaking is openly soliciting views on exactly the questions that determine whether community banks are participants in this system or bystanders to it. Where reserve deposits must sit. Whether they’re spread across four thousand banks or five. How the perimeter treats the products you already offer.
I expect the mega banks have their comment letters half written. The crypto industry has no doubt been drafting since the statute passed. The trade associations will file something suitably general. What’s usually missing from these dockets is the specific, operational voice of the banks that hold the country’s local deposit relationships, which is a pity, because this time the rule reaches their core product from three directions at once.
The Senate may have just spent its summer on the bill that failed. The fence is going up anyway, on schedule, in a docket almost nobody has written to. I think the only question is whether banks help decide where the gates are.
References
Legislation and regulatory sources
OCC Notice of Proposed Rulemaking implementing the GENIUS Act, Bulletin 2026-3 (Feb. 25, 2026), including the accompanying economic analysis and issuance forecasts ($500B 2026; $2T 2028, citing Standard Chartered)
Senate cloture vote on the CLARITY Act fails 49-50 (Sept. 15, 2026), CNBC; CoinDesk
Market data and deposit impact
Stablecoin market capitalization data, DefiLlama (total ≈ $300B; USDT ≈ $183B, September 2026)
Federal Reserve Bank of New York, Staff Report No. 1185, Stablecoin Disintermediation (partner-bank reserve behavior; loan-to-asset contraction)
Historical precedents
National Banking Acts of 1863 and 1864, Federal Reserve History, and the Act of March 3, 1865 (10 percent tax on state bank notes)
Interest Rate Controls (Regulation Q), Federal Reserve History
Catherine R. Schenk, The Origins of the Eurodollar Market in London: 1955-1963