The Redemption Clause: Washington Split the Stablecoin Industry With a Verb"

"article": "For eighteen months, the stablecoin debate in Washington was framed as a question of size — how large a reserve, how frequent the attestations, how deep the haircuts on short-term Treasury bills. That framing was a decoy, and most of the industry bought it because a decoy dressed as a compliance checklist is the easiest kind of narrative to sell. The line that actually matters in the Treasury's new proposed rulemaking is not a number. It is a verb: redeem.\n\nOn July 10, the Office of the Comptroller of the Currency handed Circle the only unconditional trust bank approval in the sector — First National Digital Currency Bank — while Ripple, BitGo, Fidelity, and Paxos sat in a conditional queue that had already been open since December 2025. Four firms waiting, one firm through. If you were reading headlines, you filed that under regulatory tempo problems, the natural sluggishness of agencies that move at the speed of comment periods. If you were reading §1523.1(c), you filed it under something colder. The OCC is not slow. It is deliberately rationing the only license that grants a private entity the legal right to owe the public money at par.\n\nThat distinction — between a firm that moves dollars and a firm that promises to redeem them — is the entire architecture of the GENIUS Act's notice of proposed rulemaking, and almost nobody is trading on it yet.\n\nTracing the fractal logic beneath the chaos: the same line that separates a bank from a payment processor is now being redrawn around stablecoins, and the side of the line you land on determines whether you are regulated like a money fund or audited like a money transmitter.\n\nContext: what the GENIUS Act actually built\n\nThe Guiding and Establishing National Innovation for U.S. Stablecoins Act — GENIUS, for the acronym-starved — is not a technology law. It is a classification law. Congress defined a category of asset called a \"payment stablecoin,\" attached a set of obligations to whoever issues one, and then handed the Treasury and five other agencies the job of writing the fine print. That fine print is the NPRM, and its comment period stays open until October 19, 2026. Everything contested lives inside that window.\n\nTo understand how unusual the American design is, you have to hold it against its peers. Europe's MiCA moved first and moved bluntly: it treats stablecoin issuers as a single regulated class, imposes reserve and redemption rules on all of them, and is already in its implementation phase. Hong Kong's stablecoin regime, which I have watched from the inside of the city's licensing consultations, is narrower and more self-interested than the press releases admit — a bid to capture the Asian book of business that Singapore's MAS has been quietly consolidating. I have argued for two years that Hong Kong's framework is less an embrace of innovation than a turf claim dressed as prudence, and the NPRM makes that clearer by contrast: Washington is not claiming turf. It is building a hierarchy.\n\nWhere MiCA flattens, GENIUS stratifies. The American rule splits the stablecoin value chain into two regulated layers — issuers on top, distributors below — and then decides which layer a given firm occupies not by what it calls itself, but by what economic function it performs. That is the innovation, and it is a regulatory innovation rather than a technical one. There is no new cryptography here, no new consensus mechanism, no new scaling primitive. There is a sorting machine.\n\nThe sorting machine has a clock attached. The Treasury's staggered compliance calendar runs from conditional approvals in late 2025, through a rulemaking deadline in July 2026 that at least some of the six responsible agencies have already missed, to the GENIUS Act's effective date of January 18, 2027 — the same day foreign stablecoin restrictions begin — and finally to a full restriction date of July 18, 2028, after which registered digital-asset service providers will be barred from offering payment stablecoins issued by non-licensed entities. Read that last date slowly. It is the most consequential sentence in the entire framework, and it names no one.\n\nCore: the economy of a verb\n\nThe mechanical heart of the NPRM sits in §1523.1(c), and it defines an issuer through economic substance rather than a list of permitted activities. Strip the legal language and you get a two-branch test:\n\nIf an entity takes on an obligation to redeem a stable value at face value, and holds that value on behalf of the public, it is a Payment Stablecoin Issuer — a PPSI — and it inherits the heaviest regulatory burden in the system: capital, licensing, supervision, the works.\n\nIf an entity merely transfers, custodies, or distributes stablecoins without ever assuming the redemption obligation, it is a Digital Asset Service Provider — a DASP — and its burden is lighter, centered on anti-money-laundering controls and sanctions screening rather than balance-sheet integrity.\n\nThe elegance — and the danger — is that the two identities are not mutually exclusive. A single firm can be both. The rule does not hand you a badge that says what you are; it audits what you owe. This is a genuine departure from the license-as-identity model that dominates financial regulation, where a charter defines a firm's universe of permitted conduct from the moment it is granted. Under §1523.1(c), the classification is fluid and fact-dependent. I spent six weeks in 2017 auditing early state-channel designs like Raiden Network, and the lesson that stuck with me was that whoever holds the liability is the party that defines the system — the routing layer is decoration, the settlement layer is power. The Treasury has now written that lesson into federal law. It has made liability-holding, not market-making, the thing that triggers the heaviest hand.\n\nThat is why the OCC approval queue is not a bottleneck. It is a chokepoint with intent. Circle's exclusive final approval means exactly one entity currently holds an unconditional federal charter built for digital-currency issuance. Ripple, BitGo, Fidelity, and Paxos hold conditional approvals — real, but reversible, contingent on meeting conditions that can be tightened at any review. Conservation of a scarce resource like this is not an accident of staffing. Scarcity is a narrative we agreed to believe, and in this case the Treasury and the OCC are the ones who agreed to believe it. The permission to owe the public money has been rationed, and the rationing is the policy.\n\nThe second layer of the architecture is where the growth is, and where the fight is. Circle's issuance rights mean little without distribution, and distribution is the part of the stack that does not require a bank charter. Coinbase, through its ties to Stablecore and Moov, has stitched together coverage spanning more than 3,000 community banks and credit unions on one side and more than 4,000 distribution endpoints on the other. Read that again as a sociologist rather than a trader. The stablecoin industry is not replacing the American banking system. It is threading itself through it, using the long tail of small banks as the last mile of dollar digitization. The community bank that cannot afford its own token infrastructure becomes a distribution node for someone else's.\n\nThis is the inverse of the story the industry tells about itself. The narrative is \"stablecoins disintermediate banks.\" The mechanism is \"stablecoins subcontract for banks.\" A three-thousand-branch dealer network is a moat that cannot be replicated with a whitepaper, and it belongs to whoever arrived with the integration playbook first. The DASP layer competes on coverage, not on charter, and coverage compounds. Following the signal through the noise floor, the loudest thing in the room is not Circle's approval. It is the quiet accumulation of last-mile rails by a small number of distributors who understood that the issuer license is a commodity and the distribution graph is not.\n\nNow the money. Here is where the NPRM does something genuinely clever, and where the incumbent banking lobby has correctly identified the soft spot.\n\nThe GENIUS Act forbids issuers from paying interest or yield to stablecoin holders. On the surface this looks like consumer protection — a rule to stop stablecoin issuers from being run like uninsured deposit accounts. It is not. It is a de-securitization device, and it is the single most important piece of legal engineering in the entire framework.\n\nRun the asset through a Howey test without the yield ban. A purchaser hands over money, that money is pooled into a common reserve, and the holder might reasonably expect a return derived from the efforts of the issuer managing that reserve. That is an uncomfortably close fit to an investment contract. The yield ban severs the \"expectation of profits\" prong at the root. By making it illegal to distribute reserve income to holders, the statute removes the very feature that would make a payment stablecoin look like a security. The prohibition is not protecting consumers from yield; it is protecting the asset from securities law. Remove the yield, and you remove the investment contract. The bug is the feature they didn't document: the ban on interest is what keeps the whole category out of the SEC's inbox.\n\nWhich reframes every argument about yield bans. Yields are merely attention taxes in disguise — the mechanism by which a financial product buys user loyalty without buying utility — and the Treasury has decided that payment stablecoins are not permitted to levy that tax. That is a policy choice with a clear beneficiary: the issuer, who keeps the reserve income outright. Circle does not pay you the interest on the Treasury bills backing your dollar. Circle keeps it. The yield ban is not a restriction on the industry; it is a subsidy to the issuance layer, paid by holders, enforced by statute.\n\nThis is a classic money-market-fund economy wearing crypto clothing. There is no Ponzi structure here — I want to be precise, because I have spent years dismantling the opposite claim. The reserve is real, it is backed by short-duration government paper, and the return comes from interest on that paper rather than from the deposits of later entrants. The economics are sound in the way a Treasury money market fund is sound, and fragile in exactly the same way: both are highly sensitive to the rate cycle. When rates fall, reserve income compresses, and issuers whose entire revenue line is the spread between zero and the policy rate discover they own a business with no pricing power. A Circle is not a growth company in a low-rate world. It is a spread business, and spreads mean-revert.\n\nNow watch the lobbying. The Bank Policy Institute, joined by the American Bankers Association, the Consumer Bankers Association, the Financial Services Forum, and The Clearing House, has argued in its comment that the interest and yield prohibition should be extended beyond issuers to DA SPs and exchanges as well. On its face this is a safety argument — \"Safety Across the Ecosystem,\" per the coalition's framing. Underneath, it is a market-structure argument. If distributors are allowed to pay interest-like rewards to attract users, they can compete for deposits without holding a charter, and that is a competitive threat to banks. Extending the yield ban to DA SPs would strip distributors of their primary customer-acquisition lever and force them into a fee-only operating model indistinguishable in economics from a bank's non-interest income line. The coalition is not asking for a safer ecosystem. It is asking the Treasury to close a competitive loophole before the non-banks learn how to use it.\n\nTruth emerges from the collision of opposites: on one side, a Treasury that wants a broad dollar footprint abroad; on the other, a banking lobby that wants a narrow, chartered footprint at home. Between them sits the definitional boundary of §1523.1(c), and neither side can win without redrawing it.\n\nThe vert