Ripple's Trust Bank Is a Permission Slip, Not a Moat: $2.5B of Reserves, $11.7M of Capital, and a Crowded Template

The approval is narrow and specific, and the narrowness is the news. Ripple National Trust Bank operates as a trust company and activities related to it, including fiduciary activities, under 12 USC 92a and 12 CFR 5.26. It manages RLUSD's segregated reserve on a directed basis, performs collateral trustee services for RLUSD holders in a fiduciary capacity, and can hold crypto for affiliates and unaffiliated institutional clients. RLUSD issuance stays where it was: Standard Custody & Trust Company, a New York limited-purpose trust company supervised by the state regulator.

Three conditions do the real work. First, capital: $11.7 million of tier 1, half of it or $5.85 million parked in eligible liquid assets, plus 180 days of operating expenses, all enforced for the bank's first three years. Second, perimeter: the bank must not meet the Bank Holding Company Act definition of a bank, so no deposits and no lending. Third, and least discussed, the OCC reserved the right to require Ripple to conform or divest the collateral trustee structure to comply with the GENIUS Act and whatever implementing rules follow. the federal licence can still be rewritten to fit the stablecoin rulebook.

Run the arithmetic before anyone calls this a bank. The minimum tier 1 capital is $11.7 million. The reserves the bank administers are $2.5 billion, and they do not even sit on Ripple's balance sheet - they sit at BNY, the primary reserve custodian, as dollar deposits, U.S. government securities and cash equivalents, with monthly independent attestations. The bank's exposure is fiduciary and administrative. Its capital is 0.5% of the float it supervises, and it earns custody and trustee fees on that float rather than lending against it.

RLUSD launched in December 2024 and reached $1.7 billion by June 2026 and $2.5 billion by October 2026, so the float is compounding at a genuinely fast rate. It is still 3.4% of the $73.3 billion of USDC outstanding at the end of June. Read that gap as the honest valuation of the charter: it removes friction from a $2.5 billion float, not yet from a $73 billion one.

Nobody needs a deposit franchise to make money on a stablecoin float. They need the reserve yield, and they need to keep it. Circle's second-quarter 2026 results show what happens when an issuer does not own its distribution: $701 million of revenue and reserve income, $668 million of reserve income, $412 million of distribution, transaction and other costs, leaving $289 million of revenue less distribution costs and a 41% margin. Fifty-nine cents of every revenue dollar left the building before shareholders saw any of it.

Ripple's structure keeps the other side of that trade. The reserves sit inside Ripple's own issuance chain, and the distribution rails were bought rather than rented: Ripple acquired Rail for $200 million to build an end-to-end stablecoin payments platform, alongside GTreasury and Hidden Road. Apply Circle's own disclosed 3.5% reserve return rate to RLUSD's $2.5 billion and the gross reserve income is about $88 million a year, before custody fees and operating costs - and none of it has to be paid away to a distribution partner. Ripple's advantage is not a charter, it is owning the pipe between the float and the customer.

Ripple was in the first wave. On December 12, 2025 the OCC conditionally approved five national trust charters at once, including Ripple, Circle, Paxos, BitGo and Fidelity Digital Assets. Seven more applications were publicly cleared in the following ten months, including Bridge and Crypto.com in February 2026, Coinbase on April 2, 2026, World Liberty Trust Company in August, and three more on September 18, 2026. The capital template barely moves: Ripple needs $11.7 million of tier 1, Agora $10 million, Catena $10 million, Bastion $6 million.

The sharpest comparison is Agora, which received preliminary conditional approval on September 18, 2026 and intends to move AUSD issuance itself into its own bank. That is the step Ripple did not take, because RLUSD issuance sits with a New York state trust company. Bastion will white-label stablecoins for other firms. When the regulator is selling charters to issuers, white-labelers and AI-agent banking projects at the same capital bar, the charter stops being a differentiator. scarcity was the moat, and the OCC has been selling it since December.

A charter only matters if float moves through it, and float moves through custodians and distributors. Ripple has assembled both sides: BNY holds the reserves and Standard Chartered now offers institutional custody of RLUSD on the XRP Ledger from its Dubai branch, while SBI VC Trade puts RLUSD in front of Japanese institutions and Korean and Turkish partners extend the reach.

Note what this table says about where the money lands. For BNY, custody of $2.5 billion is about 0.005% of the $53.1 trillion of assets under custody and administration it oversaw at the end of March 2025 - a rounding error next to $10.2 billion of revenue in the second quarter of 2026. For Standard Chartered, which reported $11.6 billion of operating income and 17.6% return on tangible equity in the first half, digital-asset custody is a fee-growth option, not a franchise. the fee owners collect whether Ripple or Circle wins the float.

Ripple now runs a two-tier structure that the regulator has pre-approved for revision. A New York state trust company issues RLUSD and holds the reserves under state supervision with monthly attestations; a federally supervised trust bank administers that reserve and stands as collateral trustee for holders if the state issuer fails. The OCC's approval letter reserved the right to require Ripple to conform or divest that trustee structure to comply with the GENIUS Act and future implementing regulations, and that condition stays in force for three years. A federal licence wrapped around a state issuer is a structure the OCC has signalled it can rebuild.

And none of this creates a new buyer for XRP. The trust bank's business is RLUSD reserves and institutional custody; the token is not on the approved-activity list. The market read it accordingly: XRP was $1.32 and up 2.08% on October 27, 2026, a modest move for a final federal approval, and the token had slipped below $1 on August 18, 2026, the day a Korean regional bank agreed to adopt Ripple Payments. the charter monetises RLUSD, not the token.

The synthesis is straightforward. Ripple has won a licence, a federal supervisor for its reserves, a trustee role behind its float and a custody business it can sell into institutions - and none of that makes it a bank in the sense the banking industry means. What it makes Ripple is the only participant in this race that has stopped renting its distribution, which is why the charter is worth more to Ripple than to a rival with the same licence and a partner-dependent model. trade the fee flow, not the licence.