Seventeen U.S. commercial banks, led by JPMorgan Chase, Citigroup, Bank of America, and Wells Fargo, are building a shared tokenized deposit network through The Clearing House, targeting a first-half 2027 launch. A parallel effort — the Cari Network — unites five regional lenders on a ZKsync-base...
"The industry is facing a radically different future around on-chain payments and finance." — David Watson, CEO, The Clearing House
Seventeen U.S. commercial banks, led by JPMorgan Chase, Citigroup, Bank of America, and Wells Fargo, are building a shared tokenized deposit network through The Clearing House, targeting a first-half 2027 launch. A parallel effort — the Cari Network — unites five regional lenders on a ZKsync-based rollup with a Q4 2026 production target. BNY, the world's largest custodian with $57.8 trillion in assets under custody, already went live with tokenized deposits in January 2026.
The initiatives represent the banking sector's operational counter-offensive against the $300 billion stablecoin market. The U.S. Treasury Borrowing Advisory Committee (TBAC) has flagged $6.6 trillion in transactional bank deposits as potentially "at risk" from stablecoin migration. Citigroup's revised forecast projects stablecoin supply reaching $1.9 trillion (base case) to $4.0 trillion (bull case) by 2030. The banks' response: wrap deposits in blockchain programmability while keeping funds inside the regulated, FDIC-insured perimeter.
This report examines the architecture, participants, economic logic, and structural limitations of these competing tokenized deposit networks, and assesses whether they can neutralize the stablecoin threat before deposit erosion becomes material.
On June 5, 2026, The Clearing House — the bank-owned payments company that operates the RTP and CHIPS networks — announced a "bank-led on-chain money initiative" designed to enable clearing and settlement of tokenized commercial bank deposits at scale.
Core participants: JPMorgan Chase, Citigroup, Bank of America, and Wells Fargo.
Extended roster: BNY, BMO, Citizens Financial, Fifth Third, HSBC, Huntington, KeyBank, PNC, Regions, Santander, TD Bank, Truist, and U.S. Bank.
That makes 17 institutions, collectively holding the majority of U.S. commercial bank deposits.
What it does: The network will convert traditional bank deposits into blockchain-based tokens that move 24/7 with instant settlement between participating institutions. It connects on-chain activity with established fiat rails — specifically the RTP and CHIPS networks — enabling automated workflows and richer transaction data.
What it doesn't have yet: No blockchain vendor has been selected. No network name, pricing model, or rulebook has been published. The target launch is H1 2027, which means the initiative is approximately 12 months from production with critical infrastructure decisions still unmade.
The Clearing House processes roughly $2 trillion in payments daily through its existing networks. Adding a tokenized deposit layer represents a significant expansion of its mandate, but the operational scaffolding — KYC/AML integration, interbank liquidity rules, governance structures — remains unspecified.
While the major banks coordinate through The Clearing House, a separate consortium targets a different segment of the market.
The Cari Network, founded by Gene Ludwig — former Comptroller of the Currency under the Clinton administration and founder of Promontory Financial Group — brings together five regional lenders: Huntington Bancshares, First Horizon, M&T Bank, KeyCorp, and Old National Bancorp.
Technology stack: Cari selected Matter Labs' Prividium, a ZKsync-based ZK rollup anchored to Ethereum, as its blockchain infrastructure. This makes it the first major bank consortium to deploy on an Ethereum Layer 2 for deposit tokenization.
Timeline: Q3 2026 pilot with participating banks testing creation, transfer, and redemption of tokenized deposits. Q4 2026 target for full production availability across all five founding banks.
Design principle: Deposits remain FDIC-insured. Tokens represent actual customer deposits at chartered banks, not a new asset class. Funds never leave the banking system — they simply acquire blockchain-native programmability.
In April 2026, Cari announced a strategic partnership with Tassat to accelerate network development. Ludwig has stated publicly: "Banks should be leading the next phase of digital money, not reacting to it."
The Cari approach differs from The Clearing House in two respects: it targets retail-scale use cases, and it has already selected its blockchain infrastructure. Whether a ZK-rollup on Ethereum can meet the throughput, privacy, and regulatory requirements of multi-bank retail deposit settlement at scale remains to be validated.
These networks are not purely hypothetical. Two of the largest financial institutions are already operating tokenized deposit infrastructure.
BNY launched tokenized deposit capabilities on January 9, 2026. The service creates on-chain representations of deposits held in client accounts, enabling near real-time transfers while keeping underlying balances within the traditional banking environment. Early adopters include the Intercontinental Exchange (ICE), Citadel Securities, DRW, and Circle. Since 2023, BNY has participated in Singapore's Project Guardian, testing interoperability of tokenized deposits and FX with OCBC Bank.
JPMorgan's Kinexys platform has processed more than $4 trillion in cumulative transactions since launch, with average daily volume exceeding $7 billion. Kinexys operates JPM Coin (JPMD) and has expanded to support eight currencies: USD, EUR, GBP, AUD, HKD, JPY, CNY, and SGD. The platform has also deployed JPM Coin on Coinbase's Base Layer 2 for institutional clients and announced integration with the Canton Network for issuance, transfer, and redemption.
These deployments provide live proof-of-concept data, but they serve institutional clients only. Retail-facing tokenized deposits — the segment most directly competitive with consumer stablecoin adoption — do not yet exist in production.
The banking sector's urgency is driven by specific quantitative assessments of stablecoin-driven deposit migration.
Total U.S. commercial bank deposits: $19.14 trillion as of May 2026, according to Federal Reserve H.8 data.
Total stablecoin market cap: Approximately $300 billion as of June 2026, with Tether (USDT) at $186.8 billion and USDC at $75.8 billion. Together they control approximately 88.6% of the market.
TBAC assessment: The U.S. Treasury Borrowing Advisory Committee identified $6.6 trillion in non-interest-bearing transactional deposits as "at risk" of migrating to stablecoins. This is the first time TBAC formally examined stablecoins as a material threat to bank funding.
Citigroup projections (revised September 2025): Base case of $1.9 trillion in stablecoin supply by 2030 (up from $1.6 trillion initial estimate). Bull case of $4.0 trillion (up from $3.7 trillion). Citi estimates stablecoin-driven deposit displacement of $182 billion to $908 billion over the same period.
Treasury lending impact: Treasury estimates that stablecoin deposit migration could reduce U.S. lending capacity by $1.26 trillion, affecting mortgages, student loans, and small business credit.
At current levels, the stablecoin market represents approximately 1.6% of U.S. bank deposits. The threat is not current scale — it is trajectory. Stablecoin supply grew from $131 billion in January 2024 to $300 billion by June 2026, a 129% increase in 30 months. If Citigroup's base case holds, the stablecoin market will be 6.5x larger in four years.
Stablecoin reserve composition matters. Issuers now hold more than $120 billion in T-bills to back their tokens. TBAC projects that continued organic stablecoin expansion could create up to $900 billion in additional T-bill demand versus the current $6.4 trillion bill market. This creates a secondary pressure: stablecoins don't just drain deposits — they compete with banks for safe-asset collateral.
The terms are often conflated. They describe fundamentally different instruments.
| Dimension | Tokenized Deposit | Stablecoin | |-----------|-------------------|------------| | Issuer | Chartered bank | Private company (Tether, Circle) | | Regulatory status | Bank deposit, subject to banking regulation | Payment instrument under GENIUS Act | | Insurance | FDIC-insured (up to $250K) | No deposit insurance | | Reserve backing | Bank's balance sheet (fractional reserve) | 1:1 reserves in T-bills, cash, repos | | Credit intermediation | Yes — bank lends against deposits | No — issuer holds reserves, does not lend | | Network | Permissioned (bank consortium) | Public blockchain (Ethereum, Tron, Solana) | | KYC/AML | Embedded at bank level | Varies; on-chain is pseudonymous | | Programmability | Smart contract-enabled within network | Fully programmable, composable with DeFi | | Availability | 24/7 (new capability) | 24/7 (native) | | Interoperability | Between participating banks only | Any wallet, any chain, any protocol |
The core structural difference: a tokenized deposit preserves fractional-reserve banking. The bank can lend against the deposit. A stablecoin does not — the issuer holds reserves idle. This is what the New York Fed calls the "narrow banking" distinction.
In February 2026, New York Fed economists Xuesong Huang and Todd Keister published Staff Report No. 1179: "Stablecoins vs. Tokenized Deposits: The Narrow Banking Debate Revisited."
The paper reframes the tokenized deposits vs. stablecoins debate as a modern version of a question that has recurred since the 1930s: should the institution that creates money also be the one that lends it?
Core argument: Stablecoins function as narrow banks — they take deposits (in the form of stablecoin purchases) and hold safe assets (T-bills, cash). They do not lend. Traditional banks do both: they hold deposits and extend credit.
Three policy scenarios identified:
The implication: there is no universally correct answer. The optimal policy depends on the regulatory environment and the banking sector's risk appetite. The current U.S. approach — allowing both stablecoins (under GENIUS Act) and encouraging tokenized deposits — fits the third scenario.
A separate June 2026 New York Fed staff report (SR 1185, "Stablecoin Disintermediation") found that banks holding stablecoin-issuer deposits are lending less, providing early empirical evidence that stablecoin reserves do affect credit intermediation.
The GENIUS Act (P.L. 119-27), signed into law in July 2025, created a federal regulatory framework for payment stablecoins. Key provisions relevant to the deposit competition:
The yield restriction is significant. Without direct interest payments, stablecoins compete with bank deposits on speed and programmability, not on yield. If that restriction holds, tokenized deposits may not need to match stablecoin functionality — they just need to match speed while retaining the yield advantage of a bank account.
The banking industry's concern: the yield restriction may erode over time, or be circumvented through exchange-level reward programs. If stablecoins eventually offer competitive yields, the deposit drain accelerates significantly.
The Clearing House network is pre-product. No blockchain vendor, no network name, no pricing model, no rulebook. A 12-month timeline to H1 2027 launch is aggressive for an initiative of this scale involving 17 institutions with competing internal blockchain strategies (JPMorgan runs Kinexys; Citi has its own digital asset platform).
Interoperability between networks is undefined. The Clearing House network, Cari Network, BNY's system, and JPMorgan's Kinexys all operate on different blockchain infrastructures. Whether tokenized deposits on one network can settle against deposits on another is an open question. Stablecoins, by contrast, are natively interoperable across public chains.
Retail readiness is years away. All current tokenized deposit deployments are institutional. The Cari Network's Q4 2026 retail target would be the first consumer-facing implementation — but it covers only five banks. Stablecoins already serve millions of retail users globally.
The composability gap. Tokenized deposits operate on permissioned networks. They cannot interact with DeFi protocols, decentralized exchanges, or cross-chain bridges. For users who value programmable money that works across open ecosystems, tokenized deposits are a constrained substitute.
International coordination is absent. Singapore's Project Guardian and the EU's blockchain settlement experiments suggest parallel interest, but no interoperability standards exist between U.S. bank tokenized deposits and international equivalents.
The U.S. banking system is responding to the stablecoin threat with coordinated infrastructure rather than purely legislative defense. The Clearing House initiative, Cari Network, BNY's live deployment, and Kinexys collectively represent the most significant blockchain adoption effort by regulated banks in history.
The question is timing. Stablecoin supply has grown 129% in 30 months to $300 billion. The Clearing House network won't launch until mid-2027 at the earliest. If Citigroup's base case holds, stablecoins will reach $1.9 trillion by 2030 — a period during which tokenized deposit infrastructure is still scaling.
The New York Fed's framework suggests the optimal outcome is coexistence: tokenized deposits for wholesale settlement and credit intermediation, stablecoins for retail payments and open-network programmability. The market may ultimately arrive there. The risk for banks is that by the time their networks reach production scale, the deposit migration they were designed to prevent has already occurred.
The $19.14 trillion U.S. deposit base is not threatened today. The $6.6 trillion in transactional deposits flagged by TBAC represents the battle line. Whether banks can make tokenized deposits competitive before stablecoins breach that perimeter will determine the structure of American banking for the next decade.