Three Ways Banks Can Issue Tokenized Money

Stablecoins moved more than $300 billion in payment volume over the past year. Visa settles over $7 billion in stablecoin card volume annually. Trade.XYZ, an institutional perpetual venue, has cleared $100 billion on the S&P 500 perpetual since March 2026, all settled in stablecoins.

Banks see this volume. They want a piece of it. But the product they build to capture it determines what it costs their balance sheet, and that cost varies by a factor of four depending on which of three options they choose.

The downstream effects reach further than institutional trading desks. The architecture banks pick now shapes whether retail payments settle instantly, whether money moves on a Sunday, and whether cross-border transfers cost $40 or fractions of a cent. Each product family carries a different balance sheet penalty, and that penalty decides which use cases banks will actually build for.

The regulatory path is now open. The GENIUS Act created a federal framework for bank-issued stablecoins in July 2025. MiCA established the EU's stablecoin regime in June 2024. And Basel SCO60 — the cryptoasset chapter of the international framework that sets how much capital banks must hold against different risks — took effect January 2026. For the first time, banks have a clear rulebook for issuing tokenized money on public networks.

The question is which product to build.

Tokenized deposits

A tokenized deposit is the simplest option. The bank issues a token that represents a customer's existing deposit. The money stays on the bank's books. The token moves between wallets, but the underlying claim stays a deposit, with the same guarantee, the same regulatory capital treatment, and the same customer relationship.

JPMorgan's JPMD on Base is the live example. It's a permissioned deposit token on an Ethereum Layer 2 blockchain. Only vetted institutional wallets can hold it. The token is yield-bearing because the underlying instrument is a deposit, something a stablecoin issued under the GENIUS Act cannot offer.

The balance sheet impact is minimal. An institutional operating deposit moving into a tokenized deposit retains the 25% LCR runoff factor and the 50% ASF category for NSFR. No new liability is created. No reserve pool needs to be funded. The journal entry is a reclassification, not a new line item.

But the token can't leave the garden. Tokenized deposits only move within the bank's permissioned network or across participating consortium members. A token trapped inside that perimeter can't settle a trade on Hyperliquid or fund a margin call on a permissionless derivatives venue. For institutional clients whose settlement needs stay inside the bank's perimeter, this works. For clients who trade on public venues over the weekend, it doesn't.

First-party stablecoins

A first-party stablecoin is more expensive to operate and more powerful. The bank issues its own coin, backed dollar-for-dollar by a segregated reserve pool. The customer's claim converts from a deposit into a redemption contract against that reserve.

That conversion changes the liability profile. Before issuance, the bank funded itself with the customer's deposit, a stable wholesale liability under Basel rules. After issuance, the bank holds a redemption obligation that Basel SCO60 treats as a 100% LCR outflow and a 0% ASF for NSFR. The funding base shifts from the cheapest Basel category to the most expensive.

The reserves must be 1:1 in cash, central bank reserves, short-dated Treasuries, or Treasury-backed reverse repos. The bank can issue on its own consolidated balance sheet, through a separately incorporated subsidiary, or through a consortium with peer banks. Each structure carries different capital treatment.

What the bank gets in return is a 24/7 settlement asset it controls. A first-party stablecoin can settle trades, fund collateral calls, and move margin on weekends and holidays when traditional payment rails are closed. JPMD integrated with the Mastercard Token Network, which lets institutional clients settle alongside card-network flows in real time.

The trade-off is direct. Tokenized deposits protect the funding base but limit reach. First-party stablecoins extend reach but restructure the bank's liability profile from the inside.

Third-party stablecoins

A bank in this family doesn't issue anything. It helps customers convert funds into a stablecoin issued by someone else, typically USDC or a regulated equivalent. The customer's money leaves the bank's balance sheet entirely.

The broadest reach comes at the highest franchise cost. The customer can now settle on any public venue that accepts the stablecoin, including permissionless crypto derivatives markets where most institutional weekend flow clears. But the deposit relationship is gone. The bank becomes an onramp, not a counterparty.

Two structures are available. The bank can prefund by holding stablecoin inventory on its own books, converting central bank reserves into a Group 1b stablecoin asset. Under SCO60, that stablecoin attracts an 85% required stable funding factor in the NSFR and does not qualify as HQLA in the LCR. The bank's liquidity position weakens twice: the HQLA stock falls by the amount of the conversion, and the funding requirement rises by 85 cents on the dollar of inventory held.

Alternatively, the bank can extend a short-term secured loan against a blocked customer deposit, with the stablecoin delivered to the customer's whitelisted venue directly from the issuer. The bank never holds the stablecoin. The prudential cost is calibrated to the loan, not the coin, which puts it in a category bank treasury teams already manage.

No single product serves every client

A hedge fund may post initial margin on the bank's tokenized-deposit network, hedge weekend exposure with a third-party stablecoin, and rebalance through a consortium-issued first-party coin. A bank offering only one of these cannot fully serve that client.

Tokenized deposits defend the deposit franchise but reach a narrow set of venues. First-party stablecoins build a bank-controlled settlement rail but carry the heaviest Basel penalty. Third-party stablecoins access the widest liquidity but move the customer claim off the balance sheet.

Each choice carries a specific cost across LCR, NSFR, and risk-weighted assets. Most banks serving institutional clients on digital-asset venues will end up needing all three, priced and governed separately, running on different infrastructure stacks underneath.