The OCC’s Stablecoin Rulebook: A Banker’s Guide to What Actually Matters

The Stablecoin Banker — Special Edition

The OCC released its Notice of Proposed Rulemaking for GENIUS Act stablecoin issuers — the most detailed regulatory blueprint we've seen for how federally chartered stablecoin issuers will actually operate. Most of the coverage has focused on the headline provisions: the eligible reserve assets, the two-day redemption window, the yield prohibition. That's all in there. But the 376-page document reveals far more about the OCC's vision for this industry than the summaries suggest, and I think the most consequential design choices have been largely overlooked.

I approached this document through the lens of how banking actually works. Capital adequacy, liquidity management, operational risk, the economics of running a regulated financial institution. When you apply those concepts to the OCC's framework, the implications become clearer.

Four things jumped out.

First, the economics of issuance. The capital and operational backstop requirements create a cost structure that structurally favors certain kinds of issuers over others.

Second, the bank run protections that the OCC has engineered with a sophisticated liquidity architecture, borrowing directly from bank liquidity management principles.

Third, a threat to stablecoin-as-a-service created by siloing rules that would upend the business model of these issuers.

Fourth, broad yield prohibition, which is definitely not settled. The comment period questions reveal that the OCC itself is uncertain about whether the ban protects bank deposits or merely reroutes the competitive threat through a more complex channel.

This is a proposed rule. Nothing is final. But the design choices in this document will shape the stablecoin industry for a decade, and bankers should understand them now — not after the comment period closes.

The Economics of Issuance: Who Actually Makes Money?

There's a persistent narrative in banking circles that stablecoin issuers are rate-sensitive businesses — that if interest rates drop, the model breaks. Having worked directly on stablecoin issuance at Paxos and built my own startup around it, I can tell you the narrative is incomplete. The OCC's proposed framework makes this even clearer.

The capital structure for stablecoin issuers is currently anchored to operational risk, not reserve size. There's an absolute minimum capital floor of $5 million, with de novo issuers facing individualized floors based on their business plans — the December conditional approvals ranged from $6 million to $25 million. Capital instruments are limited to CET1 and AT1 equity only. No subordinated debt, no hybrid instruments. This is simpler and more restrictive than what banks face, but the total capital required is modest relative to the reserve base. However, the Office has included questions on whether variable capital charges should be included in various forms - on uninsured deposits, based on total reserves, or based on credit/price/interest rate risk.

On top of capital, the OCC introduces an operational backstop: issuers must hold 12 months of trailing operating expenses in ultra-liquid assets — Fed cash, FDIC-insured demand deposits, or qualifying T-bills — completely separate from reserves. If the backstop is breached, issuance freezes. Two consecutive quarterly breaches trigger mandatory full redemption. This may help address the concerns raised in the December 15 newsletter about tokenholder priority in bankruptcy by ensuring the issuer has funds on hand to pay for bankruptcy trustee and debtor in possession lenders.

Let's look at the implications for issuer economics through a live example with Circle, holding roughly $75 billion in reserves. While we don't know where OCC will land on their reserve capital, the Office presented a potential variable capital component that would put Circle's required reserves at 0.2% of assets, or $150 million. At a 15% cost of equity, that's about $22.5 million per year. That's only 3bps on reserves — plenty of room in a low rate environment.

On the surface, one might think that stablecoin issuance is wildly expensive. Circle spent $1.7 billion on distribution costs in 2025, and another $1.2 billion on operating expenses. However, the actual work of minting, burning, redeeming, and managing reserves is not operationally expensive. A purpose-built stablecoin issuer runs a lean technology stack with a focused compliance function. Tether runs a $200 billion stablecoin with 60% fewer staff than Circle.

The operational backstop reinforces this lean design. Every dollar of operating expense inside the issuing entity requires a dollar of low-yield assets sitting idle in the backstop. That creates a direct financial incentive to push expensive functions — marketing, distribution, retail customer support — outside the issuer into a parent company, affiliates, or third-party partners. The OCC's vendor management rules prevent hollow-shell issuers, so core technology, compliance, and risk management must stay internal. But everything else can be externalized, run with zero capital requirement, and flexed up and down with rates.

This has a profound structural implication: direct-to-consumer issuance is penalized at the issuer level. Consumer-facing functions are operationally expensive — more headcount, more support infrastructure, more compliance complexity. Each function added to the issuing entity raises operating expenses, which raises the backstop, which locks up more low-yield assets. The rational design is a lean B2B issuing entity with the consumer relationship living in a separate distribution layer — a bank, a fintech platform, or both.

So who benefits? Standalone issuers are best positioned at scale. They're purpose-built, can be run operationally lean, and their fixed regulatory costs become negligible as the reserve base grows. Bank subsidiaries have a theoretical advantage through shared infrastructure, but bank operational costs are orders of magnitude higher than modern fintechs. The real advantage for banks is distribution and customer relationships, not issuance efficiency. Issuer-as-a-service platforms like Bridge and Brale survive, but with a narrower value proposition — the OCC's rigid reserve framework largely eliminates custom reserve construction per branded stablecoin, so what they sell becomes operational efficiency across multiple clients, not product differentiation.

The minimum viable scale for a standalone issuer, accounting for backstop, capital, compliance, and OCC assessment fees, is approximately $2–3 billion in outstanding stablecoins. Below that, fixed costs don't amortize. Above it, the economics work in almost any realistic rate environment.

The Liquidity Architecture: Engineering a Bank-Run Buffer

The GENIUS Act specified eligible reserve assets but said almost nothing about how issuers should manage liquidity. The OCC has filled that gap with a reserve management framework that borrows heavily from bank liquidity regulation, with rules on both the sources of liquidity and the demand for liquidity.

On the asset side, the eligible reserve list is unchanged from the GENIUS Act: Fed reserve balances, demand deposits at US banks, T-bills of 93 days or less, repos, reverse repos, money market funds, and tokenized versions of each. But the OCC layers on new constraints: 10% of reserves held as deposits split across at least two banks or the Fed (of which 0.5% must be insured deposits for large issuers); a 20-day weighted average maturity cap on the entire portfolio; and 30% of reserves must be receivable and due unconditionally within 5 business days.

That last requirement deserves attention. This is a contractual maturity test, not a marketability test. Assets that could be sold within 5 days don't qualify unless they're already contractually maturing or due within that window. This means the 30% tranche is effectively limited to Fed balances, demand deposits, overnight repos, money market funds with next-day redemption, and T-bills in their final days before maturity.

On the liability side, the OCC sets a 2-business-day maximum redemption window. The rules go even further, mandating an automatic 7-day redemption deferment if more than 10% of outstanding stablecoins are redeemed in any 24-hour period. This is a subtle but important piece of market psychology design. If an issuer had discretion to delay redemptions and exercised it, the market would read it as distress — the way a money market fund "breaking the buck" triggers panic. By making the extension mandatory and threshold-triggered, the OCC has removed that signal. "We're in the automatic 7-day window" becomes a neutral regulatory status, not a management decision. The OCC has effectively given issuers a bank-run buffer without the reputational cost of appearing to invoke one.

There is one unresolved tension worth flagging. The framework works cleanly as long as the liquidity waterfall provides sufficient runway to avoid forced sales of longer-dated assets. In a simultaneous rate shock and large redemption wave — the SVB dynamic — the tail portfolio could face mark-to-market losses that exceed the operational risk capital floor. The OCC has substituted portfolio construction rules for capital requirements as the primary interest rate risk mitigant. That's a legitimate philosophical choice, but it's not uncontested — and the variable capital component still under consideration would partly address this gap.

Issuer-as-a-Service: Under Threat

Companies like Paxos, Bridge and Brale are built on the assumption that a stablecoin is a stablecoin is a stablecoin. Stablecoin issuance is entirely brand-independent, from minting and redemption to reserve management to compliance. The material unique element is some amount of branding and go-to-market. At the very large end, an issuer might derive a material economic benefit from customizing these functions, but for everyone else, any differences lack distinction. Thus, IaaS providers market their ability to deliver scale economics to smaller tokens by reusing the exact same systems, people, and processes across brands.

The OCC is calling this model into question with concerns that reserve asset pooling across brands could create contagion risk. The theory is that market concerns about the liquidity or capital cushion on one branded stablecoin could spill over to the other brands powered by the same IaaS provider. Historical dislocations in the digital assets markets suggest that participants are quite sophisticated in their understanding of how things are interconnected, and in some cases using that knowledge to trigger panics.

While the rules don't specifically address this yet, the OCC has raised Question 172 of whether IaaS providers should be restricted in some way. Most draconian: require each issuer to issue only one stablecoin brand. Less draconian: require legally distinct reserve assets but otherwise allow the IaaS provider to pool all operational resources. Without fully distinct operational resources and capital reserves, it's impossible to eliminate a contagion risk. If the rules ultimately land on the draconian side, the IaaS business model will become much less efficient, and the minimum scale for a successful stablecoin will rise commensurately.

Yield: The Battle Isn’t Over

The yield prohibition is more broadly drafted than I expected: any contract, agreement, or arrangement to pay interest or yield — whether in cash, tokens, or other consideration — is banned. The operative test is proportionality to holding size. Anything that scales with how much stablecoin you hold is functionally yield, regardless of what you call it. This mirrors MiCA's approach in Europe.

But the prohibition contains a carve-out that may matter more than the ban itself. The OCC explicitly permits issuers to share reserve yield with distribution partners. If a bank distributes a stablecoin to its customers, the issuer can share reserve income with the bank. The bank can then pay yield to the customer through their deposit account — not as stablecoin yield, but as deposit interest funded by the distribution relationship.

This is the single most important commercial clarification in the entire 376-page document. It is a regulatory green light for the model I've been advocating: banks as the distribution layer for institutional stablecoin issuers, capturing economics without the operational burden of issuance. If you've been waiting for clarity on whether this model is permissible, the OCC just gave it to you.

The most viable workaround for delivering yield directly to stablecoin holders — outside the banking relationship — will almost certainly be tokenized money market funds. A KYC'd distribution platform registered as a broker-dealer markets tokenized MMF shares alongside the stablecoin, enabling instant swaps between the two. The stablecoin remains yield-free; the MMF pays yield to its shareholders as it always has. Legally distinct instruments, no yield flowing from issuer to holder. Banks with existing broker-dealer capabilities are structurally advantaged here over pure fintech platforms — they already have the securities infrastructure to offer this.

But the answers are far from final: the comment period questions reveal that the OCC itself isn't sure the prohibition is correctly calibrated. Questions 35 and 36 ask whether there should be a de minimis exception and whether the OCC should propose formal safe harbors for compliant arrangements. Question 37 goes the other direction — asking whether the prohibition should be expanded to cover indirect payments through affiliates or third parties. If this prevails in the final rule, it could tighten or eliminate the distribution profit-sharing carve-out. And Question 39 is the most revealing: the OCC explicitly asks what impact the prohibition would have on bank deposits — telling you the regulator itself is uncertain whether banning yield protects bank deposit franchises or merely redirects the competitive threat through a more complex channel.

What Bankers Should Do Now

My biggest takeaway is that banks are best positioned as distribution partners for existing issuers, not as issuers themselves. The economics of issuance favor purpose-built, operationally lean entities at massive scale — that's not most banks. But distribution is where the real value accrues. The yield prohibition ensures that reserve income flows to distribution partners rather than token holders' wallets. Banks have a massive distribution advantage that remains wildly underutilized. If the breadth of the OCC's yield prohibition sticks, banks can stop worrying about deposit competition and start capturing the economics of being the channel through which stablecoins reach customers.

Banks should be submitting comments during this period, particularly on Question 37. The distribution profit-sharing model is the most strategically valuable provision in these rules for community and regional banks. If it gets broadened into oblivion, the economics of bank participation change materially.