Griffin's $20M Series B Puts a UK Bank Licence Behind Stablecoin Reserves

Griffin closed a $20 million Series B at a valuation north of $100 million, and the London bank is spending it on the least glamorous, most load-bearing part of the stablecoin stack: reserve services and onchain cross-border payments.

Small check. Big signal.

Here's the thing. Griffin isn't a fintech renting someone else's licence. It holds its own UK banking authorisation, and it's the institution sitting under Uber's UK payments plumbing. When a company that already moves real money for a real customer base decides fresh capital should go toward stablecoin reserves, that's not a narrative trade. That's a product roadmap.

The licence came first

Griffin started in 2017 with a thesis that was boring and correct: banking infrastructure is bad, and developers should be able to call an API instead of spending nine months begging a correspondent bank for a relationship. Banking-as-a-service. Compliance baked in. The unsexy plumbing that everything else depends on.

Then came the licence. Griffin worked through the Prudential Regulation Authority's authorisation process and came out the other side as a fully licensed UK bank, clearing mobilisation and operating under its own charter. That's the part people skip past. A banking licence isn't a badge. It's permission to hold customer money, and it's the difference between a company that can custody stablecoin reserves and a company that can only talk about it.

The Uber mandate followed. Uber's UK operation needed a bank partner that could handle payment flows without the usual legacy friction, and Griffin fit. Nobody writes blog posts about payout rails. Nobody cares about infrastructure until it breaks.

So the sequence matters here. Licence first. Distribution second. Capital third. The Series B is the last step, not the first, and that ordering tells you more about Griffin's prospects than the valuation does.

Now the $20 million goes to work on two things: holding reserves for stablecoin issuers, and moving money across borders onchain rather than through the correspondent banking chain.

Reserves are the bottleneck

Let's talk blob economics for a second, except the blobs aren't the problem. Throughput is table stakes now. Ethereum's rollups have cheap data availability, Solana keeps grinding through blocks, and every L2 with a sequencer and a dream can push thousands of transactions a second. The execution layer has never been less interesting.

The real bottleneck is what sits behind the token. Every stablecoin is a promise backed by something, and that something needs a custodian with a balance sheet, an audit trail, and a regulator who can walk in the door. For years the answer was a bank in a friendly jurisdiction, a monthly attestation, and a lot of hand-waving about segregation.

Ask the obvious question. Who actually wants a licensed UK bank holding their reserves? Issuers who want to sell into European payment corridors and need a counterparty their enterprise customers can underwrite. That's the buyer. And right now there aren't many banks willing to take that business, because stablecoin reserves sit awkwardly between deposits, custody, and securities depending on which regulator is asking.

Griffin is betting that awkwardness is a moat. I think that's right, and I think the $100 million valuation looks cheap. A licensed bank with an existing enterprise customer base and a stablecoin reserve product is a rare combination, and rare things get repriced when the market notices. Compare it to what crypto raises for a token with a whitepaper and a Discord, and the asymmetry is almost funny.

But here's the tradeoff. Banking moves at banking speed. Griffin's compliance committee will outrank its product team every single time, and that's a feature until it's a growth ceiling. The crypto-native reserve managers will beat Griffin to market on features and lose on trust. Somebody has to bridge that gap, and the bridge usually costs more than the founders expect.

On cross-border, the pitch writes itself. The global average cost of sending $200 is still around 6%, and the money that gets shaved off in the middle is the correspondent banking chain taking its cut for moving information between ledgers that don't talk to each other. Stablecoin rails collapse that chain into a mint, a transfer, and a redemption. The saving isn't a basis point story. It's a structural one.

Who loses? The correspondent banks clipping fees on corridors where they add nothing but latency. The offshore reserve shops that built a business on regulatory ambiguity, because ambiguity gets priced out once a licensed alternative exists. And any stablecoin issuer that treated reserves as a marketing line rather than a balance sheet problem.

What happens next

Watch the first corridor, not the press release. The reserve service is the easier build, because custody of T-bills and cash inside a licensed bank is fundamentally an operations problem. Cross-border settlement is harder, since it needs liquidity on both ends and a redemption path that works on a Sunday.

My guess is Griffin starts with UK-to-EU and UK-to-something-in-Asia, where enterprise demand is loudest and where stablecoin regulation is clearest. The UK has been slow-walking its stablecoin regime while the EU's MiCA framework has been live since December 2024. That gap is an opportunity for a bank that can serve both sides without repapering every contract.

The scaling roadmap just got more interesting, and not because of a throughput breakthrough. Because the boring layer, the one with licences and reserve accounts and regulators who file reports, is finally getting capital.

Two thresholds to track over the next twelve months. If Griffin announces a named stablecoin issuer as a reserve client, the round looks smart in hindsight. If it announces three corridors and no clients, the round looks like a bet on a thesis rather than a business.

Either way, the signal is clear. Stablecoin infrastructure is consolidating around regulated balance sheets, and the companies that spent 2021 through 2023 chasing yield on reserves are about to find out that a bank charter is the only yield that compounds.