Stablecoin Earn on Client-Owned Vault Positions

Clients hold stablecoin balances that earn nothing, while the same tokens earn a return on the public lending markets that regulated European institutions already use. Every month the product stays switched off, the clients who want that income move their balances to a venue that offers it. A curated lending vault keeps those balances, and those clients, on the institution's own rails: the client's tokens leave the balance and enter an audited vault, the client holds a vault position against them, and earns what borrowers pay, using custody and a licence the institution already holds.

What makes the product possible is a distinction in the regulation. Paying a rate on a balance is prohibited outright under MiCA Art. 50, which is why one major venue switched off stablecoin rewards for European clients on 1 December 2024. Lending is not prohibited, and the model is already live: Deblock, a MiCA-licensed French neobank, reported more than USD 100 million of client funds earning within one year of launch, Societe Generale FORGE became the first regulated bank on this infrastructure in July 2025, and six named institutions are in production across the EU and Czechia. The institution's expensive parts, the authorisation, the custody, the fiat ramps and the years of client due diligence, are already built, and what remains is the connection and the discipline to build it in the shape the regulation permits.

Check one thing before anything else. Does your authorisation cover reception and transmission of orders, and portfolio management? A curated vault menu can be read as both. On the engagement this blueprint comes from, counsel identified exactly those two characterisations as fatal for a CASP licensed only for custody, exchange and transfers - so those permissions decide whether the product exists at all.

• Yes if you hold those permissions, already custody client stablecoins that earn nothing, and want a disclosed fee line on balances you already hold.
• Not if the permissions are missing - a different blueprint fits those licences. Not if your clients' priority is a holding-period tax exemption: the vault exchange is a disposal, and in some jurisdictions it costs the client more in tax than it pays in yield. Not if you want to pay a rate on balances, which is the one thing Art. 50 actually prohibits.

• Tokens leave, a position comes back. The statement shows that vault position, and it has to: the visible exchange is what the regulatory reading rests on.
• The counsel test is one sentence: would the client receive this benefit by simply keeping the balance? If no, the client is being paid for lending the tokens.
• You already hold the expensive parts: the licence, the custody, the ramps and the KYC. The rail provides audited contracts, independent curation and APIs, and holds no wallet, no keys, no KYC and no licence by design.
• Two rails, one structure. A direct protocol integration pays for itself above roughly EUR 15-25M of client funds in vaults; below that, a yield API delivers the identical client product in weeks instead of months. The rail choice changes nothing about the regulatory analysis, and the later migration is ordinary engineering - clients notice nothing.
• Positions are never hostage. They sit on-chain, client-owned, reachable directly on the underlying protocol. If a vendor disappeared tomorrow, the funds would not.