OCC Part 15 Liquidity Rules: How a Permitted Coin Is Supposed to Survive a Run
Part 15’s reserve list answers what you may hold. The liquidity rules answer how fast that book must turn into dollars. That is the run chapter. It is still a proposal (OCC NPRM, February 25 / March 2, 2026 — not final). Treat the numbers as the working supervisory standard, not as a signed regulation.
GENIUS itself does not write a cash sleeve or a redemption clock. The Comptroller does.
The Liquidity Stack
Part 15 builds four layers. Mix them up and you mis-price the product.
| Layer | What it is | What it is for |
|---|---|---|
| 1. 1:1 fair-value reserves | Eligible assets ≥ coins outstanding, at all times | Solvency of the peg |
| 2. Daily / weekly liquidity sleeves | 10% same-day cash; 30% inside five business days | Monetization speed |
| 3. Concentration caps | 40% any one institution; 50% of the cash sleeve at one bank | Custodian / deposit-bank failure |
| 4. Redemption SLA + operating backstop | Two business days (seven calendar days if a 10% / 24-hour run); 6–12 months of opex in a separate buffer | Customer exit and issuer survival |
Layer 1 can be all T-bills and still fail a run if Layers 2–4 are thin. Layer 4’s opex buffer is not part of the 1:1 reserve. It is capital-like liquidity for the operating company when the website, the transfer agent, or the lawyers are the constraint.
Daily Liquidity — the 10% Sleeve
Each business day, at least 10% of required reserves must be:
- demand deposits or insured shares payable on request, or
- money standing to the credit of an account at a Federal Reserve Bank.
That sleeve is same-day cash. T-bills, even T+0 on-the-run bills, do not count until they are cash or a Fed balance.
Concentration inside the sleeve: no more than 50% of that 10% may sit at any single eligible institution. On a $40 billion coin, $4 billion must be cash-like, and no one bank may hold more than $2 billion of that cash. The rule is written for a regional-bank failure, not for a T-bill market freeze.
Large issuers ($25 billion+ outstanding) add a symbolic insured sliver: 0.5% of reserves in insured deposits, capped at $500 million. On a $200 billion coin that is $500 million — useful optics, not a run backstop.
Weekly Liquidity — the 30% / Five-Day Sleeve
At least 30% of required reserves must be daily-liquidity assets or amounts unconditionally due within five business days from:
- pending sales of reserve assets,
- maturing reserve assets, or
- other maturing transactions.
A one-week T-bill and a trade that has already sold and is awaiting settlement can live here. An 80-day bill cannot, unless it has already been sold.
On the $40 billion coin, $12 billion must be five-day money. Combined with the 10% cash sleeve, the incremental “weekly-only” bucket is 20 points — $8 billion of very short paper or receivables.
This is the OCC’s answer to a Monday–Friday redemption wave that is larger than overnight cash but smaller than a full reserve liquidation.
Weighted Average Maturity — the Rate-Risk Gate
The entire reserve book must have WAM ≤ 20 days. Eligible T-bills may have residual maturity up to 93 days. The 20-day cap stops a book that is legally all “short T-bills” but clustered at 70–90 days from taking a mark-to-market hit that breaches the continuous 1:1 fair-value test.
Liquidity and rate risk are the same problem at the short end. A 20-day WAM means the book rolls constantly. It also means the issuer is a perpetual bid for 4-week and 8-week bills and a light user of 13-week paper.
OCC asked whether large issuers should get a tighter WAM and small issuers a looser one. That question is still open.
Redemption Clock — the Customer-Facing Liquidity Rule
Proposed SLA:
- Ordinary redemptions: two business days.
- If redemption requests exceed 10% of outstanding issuance in any 24 hours: the issuer may take up to seven calendar days.
That is a gated run, not a bank-deposit promise and not an MMF floating-NAV plus liquidity fee (though it rhymes). A $40 billion coin can be asked for $4 billion in a day and legally take a week. Holders who model “always $1 T+0” are modeling a product Part 15 does not require.
Compare: Bank of England systemic sterling coins aim at end-of-day convertibility in steady state. Part 15 is slower on purpose. The OCC is optimizing for T-bill monetization time, not for Fedwire instant payments.
What Happens When Liquidity Fails
If fair-value reserves slip under 1:1 — including because the cash sleeve was spent and bills have not settled — the issuer must notify the OCC and stop issuing new coins immediately. After 15 consecutive business days under-reserved, it must begin liquidation and redemption.
There is no PCA-style “undercapitalized but still open” middle. Liquidity failure becomes an issuance ban, then a wind-down. That is why the 10/30 sleeves exist: to keep the 15-day clock from starting in a garden-variety bill selloff.
Option A vs Option B
The 10 / 30 / 40 / 50 / 20-day grid is identical under both options.
- Option A: principles-based diversification, with the grid as a safe harbor.
- Option B: the grid is mandatory every business day.
Liquidity officers at large issuers will run the grid either way. Option A only matters if someone wants to argue a non-conforming book is still “sufficiently liquid.” Do not underwrite that argument for a $20 billion+ coin.
What Part 15 Liquidity Is Not
- It is not a central-bank standing facility. The Fed is not the buyer of last resort for the coin.
- It is not FDIC insurance on the token. The 0.5% insured-deposit sliver insures a bank deposit inside the reserve, not the holder’s coin.
- It is not overcollateralization. The OCC said so. Marks and sleeves replace a 102% requirement.
- It is not a solution to platform yield. Idle coins earning 4% on an exchange can still run to the exchange, not off it. Liquidity rules protect par redemption, not deposit stability at banks.
Worked Book: $40 Billion Permitted Coin
| Bucket | Minimum | Form |
|---|---|---|
| Same-day cash | $4.0 bn | Fed balances / demand deposits, split across ≥2 names |
| Five-day money (inclusive of cash) | $12.0 bn | Cash + bills/receivables due in ≤5 business days |
| Rest of book | $28.0 bn | T-bills and T-bill repo, WAM of the whole book ≤20 days |
| Single custodian / bank / repo name | ≤ $16.0 bn | Combined exposure |
| Insured deposits if applicable | $200 mn | 0.5% of $40 bn |
| Opex backstop | Separate | ~6–12 months operating costs, reset quarterly |
A competent treasurer will hold more than $4 billion cash. The 10% is a floor written by lawyers, not a treasury target written by people who have seen a three-day bill fail.
Investor Implications
- A Part 15 coin is T+2 cash with a T+7 gate in a 10% run. Price it between a government MMF and a bank deposit, closer to the MMF.
- Bill-market stress that cheapens 8-week paper is a mark problem (1:1 fair value) before it is a cash problem (10% sleeve). The WAM cap is there so those two problems arrive together and stay small.
- Multi-custodian architecture is now a liquidity rule, not an operational preference. Single-bank reserve designs will not clear Option B.
- In a real run, the binding constraint may be dealer capacity to take T-bills, not the legal sleeve. Part 15 assumes the bill market works. Size positions as if that assumption can fail for a week.
- Platform APY, if it survives CLARITY, can increase redemption risk at the issuer (users leave the chain for the exchange, then the exchange redeems). Liquidity rules do not model that second hop.
Bottom Line
Part 15 liquidity is a short government-money ladder plus a two-day promise. Ten percent cash, thirty percent inside a week, twenty-day WAM, two-day redeem, seven-day gate at 10%. That is enough to make a permitted coin a serious cash instrument. It is not enough to make it Fed funds. Until the rule is final, assume those five numbers are how the OCC will judge whether a dollar token can get out of its own way.
