A new front has opened in the war for on-chain money. Five U.S. regional banks — Huntington Bancshares, First Horizon, M&T Bank, KeyCorp, and Old National Bancorp — announced on March 17 that they are building the Cari Network, a tokenized deposit platform on ZKsync's Prividium infrastructure, ta...
A new front has opened in the war for on-chain money. Five U.S. regional banks — Huntington Bancshares, First Horizon, M&T Bank, KeyCorp, and Old National Bancorp — announced on March 17 that they are building the Cari Network, a tokenized deposit platform on ZKsync's Prividium infrastructure, targeting a Q4 2026 commercial launch. Days earlier, BNY activated its own tokenized deposit service with Citadel Securities, Galaxy, Circle, and Ripple Prime as early participants. JPMorgan's Kinexys is expanding JPM Coin to the Canton Network.
These are not pilot programs buried in innovation labs. They represent a coordinated strategic response by the banking system to the $318 billion stablecoin market — and the first serious attempt to bring FDIC-insured, fractional-reserve bank money natively onto blockchain rails. The Federal Reserve Bank of New York's February 2026 staff report frames the stakes: the choice between stablecoins and tokenized deposits is really a question about whether society wants money creation separated from lending or fused together. This is the most consequential monetary infrastructure debate since the 2008 financial crisis.
For years, traditional banks watched from the sidelines as stablecoins captured a function that used to be exclusively theirs: the movement of dollar-denominated value. Tether (USDT) controls $187 billion in market capitalization. Circle's USDC holds $75.7 billion. Together, stablecoins process volumes that rival Fedwire on some days. The total stablecoin supply crossed $318 billion by January 2026 and transaction volumes are forecast to approach $1 trillion monthly by year-end.
Banks are no longer watching. In the span of 90 days, three distinct tokenized deposit initiatives have reached production or late-stage development:
This is not incremental innovation. It is the banking system building parallel monetary infrastructure on blockchain rails.
Tokenized deposits are fundamentally different from stablecoins. Understanding this distinction is critical for anyone evaluating the future of on-chain money.
Stablecoins are bearer instruments issued by non-bank entities. When a user holds USDC, they own a claim on Circle's reserve of U.S. Treasuries and cash equivalents. The reserves sit in segregated accounts. The stablecoin issuer does not lend those reserves. This is, in economic terms, narrow banking — money creation without credit intermediation.
Tokenized deposits are digital representations of a deposit liability held at a federally insured bank. When BNY creates a tokenized deposit, the underlying funds remain on the bank's balance sheet. The bank continues to lend against those deposits through the fractional-reserve mechanism. The token holder has a claim on the bank, not on a segregated reserve pool.
The implications are profound:
| Feature | Stablecoins | Tokenized Deposits | |---------|------------|-------------------| | Issuer | Non-bank (Circle, Tether) | Chartered bank | | Backing model | 1:1 reserve (Treasuries, cash) | Fractional reserve | | Insurance | None (priority claim in insolvency) | FDIC insured | | Credit creation | No | Yes | | Yield to holder | Generally none (accrues to issuer) | Potentially interest-bearing | | Regulatory regime | GENIUS Act / OCC | Existing bank regulation | | Transfer model | Bearer instrument (peer-to-peer) | Account-based (bank-mediated) |
This structural difference has downstream consequences for monetary policy transmission, credit availability, and systemic risk — consequences that the Federal Reserve is now actively modeling.
BNY's launch is the most operationally significant. The bank created on-chain mirrored representations of client deposit balances, initially targeting collateral and margin workflows. Early participants — Citadel Securities, DRW Holdings, Galaxy, ICE, Circle, Ripple Prime, Baillie Gifford, Invesco, and WisdomTree — represent the core plumbing of institutional capital markets.
The strategic logic is clear: if collateral movements happen 24/7 on-chain using tokenized deposits, the entire settlement layer moves inside the banking system rather than relying on stablecoin rails. BNY is not competing for retail payments. It is capturing the institutional settlement layer.
The Cari Network represents something different: a collective response by mid-size banks to protect their deposit franchises. Led by Gene Ludwig — the former Comptroller of the Currency who subsequently built Promontory Financial Group — the consortium chose ZKsync's Prividium, a permissioned ZK-rollup anchored to Ethereum.
Prividium offers banks a critical feature: privacy with auditability. Transactions are shielded from public view but can be verified by regulators. As Matter Labs CEO Alex Gluchowski noted: "Financial infrastructure is undergoing the same shift computing went through decades ago, from siloed databases to shared, programmable infrastructure."
The Mid-Size Bank Coalition of America has backed the project, signaling that this is not a one-off experiment but an industry-wide strategic bet.
JPMorgan's Kinexys has processed over $1.5 trillion in tokenized transactions since inception. The next phase involves issuing JPM Coin (JPMD) natively on the Canton Network — a privacy-enabled blockchain designed for synchronized financial markets — in collaboration with Digital Asset. A cross-chain delivery-versus-payment test with Ondo Finance and Chainlink successfully settled tokenized U.S. Treasuries against JPM deposits across chains.
This is the most technically ambitious play: taking bank money multi-chain while maintaining institutional-grade controls.
The competitive dynamics between stablecoins and tokenized deposits go beyond technology. They represent fundamentally different visions for the monetary system.
Stablecoins extract deposits from the banking system. When a user converts $1,000 of bank deposits into USDC, that money leaves the bank's balance sheet. Circle parks it in Treasuries. The bank loses the ability to lend against those deposits. At scale, this is deposit disintermediation — precisely the scenario that has historically alarmed regulators.
Tokenized deposits keep money inside the banking system. The token is a new interface for an existing bank deposit. Lending capacity is preserved. FDIC insurance remains intact. The monetary policy transmission mechanism — the Fed's ability to influence credit conditions through interest rates — stays functional.
This is why the American Bankers Association has published practical guides urging banks to adopt tokenized deposits as a defensive measure. As the ABA noted in March 2026: stablecoins with a total supply exceeding $310 billion now exist "in the wild" — and regional banks face the question of how to protect their deposit franchises against 24/7, programmable, global competitors.
The New York Fed's February 2026 Staff Report No. 1179, "Stablecoins vs. Tokenized Deposits: The Narrow Banking Debate Revisited" by Xuesong Huang and Todd Keister, provides the most rigorous academic framework for this competition.
The paper's core finding: the optimal policy depends on the regulatory environment.
The practical implication: in the current U.S. regulatory environment — where the GENIUS Act provides a clear framework for stablecoins and the OCC is implementing supervisory standards — the likely equilibrium is coexistence. But the Fed's research makes clear that regulatory design, not market competition, will primarily determine which instrument dominates.
The GENIUS Act, signed into law on July 18, 2025, created the first federal framework for payment stablecoins. It explicitly excludes permitted stablecoins from securities and commodities regulation and establishes one-to-one reserve backing requirements. The OCC issued a 376-page proposed rulemaking in early 2026 to implement the law, with comments due May 1, 2026. The Act's effective date is January 18, 2027.
Here is the paradox: by legitimizing stablecoins, the GENIUS Act simultaneously accelerated bank adoption of tokenized deposits. The legislation made the competitive threat concrete and legal. Banks can no longer argue that stablecoins will be regulated out of existence. Instead, they must compete — and tokenized deposits are their competitive response.
The regulatory asymmetry, however, favors banks. Tokenized deposits inherit decades of existing bank regulation. There is no new licensing regime, no novel reserve requirements, no regulatory uncertainty. For risk-averse institutional clients, the regulatory clarity of a bank deposit — even a tokenized one — may be more attractive than a stablecoin governed by legislation that is not yet in effect.
The on-chain money wars have entered a new phase. For five years, stablecoins operated in an open field — no bank-native competitor existed on blockchain rails. That era is over. The banking system is deploying tokenized deposits across institutional settlement (BNY), regional payments (Cari Network), and cross-chain capital markets (JPMorgan Kinexys) simultaneously.
The economic stakes dwarf anything in DeFi. This is about who controls the plumbing of the dollar system: non-bank stablecoin issuers operating under the GENIUS Act, or chartered banks extending their existing deposit infrastructure onto programmable rails. The New York Fed's research suggests neither side wins completely — the optimal outcome involves both. But the terms of coexistence will be written by regulators, not by protocol governance votes.
For institutional allocators, the signal is clear: tokenized deposits are the banking system's native blockchain asset class. They preserve FDIC insurance, credit creation, and monetary policy transmission. Whether they can match the composability, speed, and global reach of stablecoins will determine the architecture of the next financial system.