Stablecoins have been pulling deposits out of community banks for years. Vantage Bank and Custodia just published a white paper arguing they built the architecture to stop it.
The paper, released June 18, 2026, details the Hazel Network, a token sub-ledger that sits alongside a bank’s existing core system. No rip-and-replace required.
The compliance officer at a mid-size Texas community bank reading this white paper would find something most blockchain proposals skip entirely: a layered enforcement mechanism that fires before any on-chain movement begins. That detail matters more than the token mechanics. It is what makes this proposal feel less like a crypto pitch and more like a banking infrastructure argument.
Three Screening Layers Nobody Else Is Talking About
Every token transfer inside Hazel triggers three distinct compliance checks before settlement can execute. The first is off-chain: AML, fraud, and sanctions screening on the related fiat movements. The second runs blockchain analytics automatically, flagging high-risk wallet addresses and holding the transaction until a compliance department clears the alert. Third is an on-chain sanctions oracle that queries in real time and blocks any transaction touching a sanctioned entity.
That is protocol-level enforcement. Not a policy document. Not a checkbox.
The system also ties reserve management directly into the smart contract itself. The contract automatically blocks any transfer out of the consortium that would exceed Custodia’s verified reserve balance. Banks don’t manage this manually.
The October 2025 announcement from Vantage and Custodia had already signaled a patent-protected framework for exactly this kind of integrated token. Patents 11,392,906, 12,450,578, and 12,579,525 cover the foundational design, with additional filings pending on the unified-token model.
Caitlin Long, CEO of Custodia, said the announcement was “six years in the making” and credited Vantage with helping the team “tailor our original design to solve real problems for banks.” American Banker placed Long at number two on its inaugural list of the most innovative people in finance this year.
The Token That Changes What It Is at the Border
Jeff Sinnott, President and CEO of Vantage Bank, described Hazel as “a compliant, secure ledger that extends your core and enables banks to operate at software speed while fully preserving your core customer relationships.”
The core of the technical design is a single smart contract issuing what the white paper calls a dual-character tokenized U.S. dollar. Inside the consortium, it is a bank deposit, with the issuing member bank as obligor and FDIC insurance attached. Outside the consortium, that same token becomes a stablecoin backed at minimum 1:1 by cash and short-duration Treasuries, designed to meet GENIUS Act requirements.
No conversion transaction happens at the boundary. The token’s legal character shifts programmatically as it crosses.
That is the piece most coverage has treated as a product feature. It isn’t. It’s a structural answer to the disintermediation problem that has been eating community bank balance sheets. When a customer moves dollars into Circle’s USDC or Tether, those funds leave the bank’s deposit base and land on a third party’s balance sheet. Hazel inverts that flow. The stablecoin is designed to find its way back to the originating bank as a tokenized deposit, without any manual conversion step, no crypto exchange in the middle.
Integration Without Replacing the Core
Three integration tiers exist. The basic model needs no core integration at all, operating “like a traditional correspondent banking delivery model,” according to the white paper. Implementation time: 4 to 6 weeks. The advanced model uses scheduled batch delivery of posting and compliance files alongside an online banking SDK. The enterprise tier builds full real-time API integration for near-immediate debits and credits.
A reference implementation has been live on Ethereum mainnet since March 2026. Vantage, Custodia, and integration partner Infinant completed the first of four required testing phases at scale. Phase two begins soon.
One pilot already announced: Participate, a bank-to-bank loan participation network. Tokenized deposits would handle the closing and servicing of loan participations, a use case that involves multiple banks, ongoing payment flows, and settlement reconciliation. More than 600 banks currently use Participate’s network.
Full availability to banks and their customers is targeted for Q4 2026.
What Banks Still Need to Verify
The compliance architecture is built in. The patent protection exists. Ethereum mainnet has been running the reference implementation since March. None of that removes the execution risk from a bank considering consortium membership.
Testing is still in phase one of four. The first formal group of member banks has not yet been announced. How the token’s legal characterization holds in a contested regulatory environment is still an open question; the GENIUS Act framework is new and its interpretation will develop over time.
The white paper does not address what happens if a sanctioned entity obtains a wallet that the on-chain oracle has not yet flagged. Blockchain analytics tools have coverage gaps. That is a real risk and one that compliance teams at potential member banks should pressure-test before signing on.
