explains
Tokenized Deposits vs Stablecoins vs CBDC
Three instruments. Three issuers. Three risk profiles. Three regulatory regimes. Trad-fin readers consistently conflate them.
Published
Tokenized deposits are commercial bank money on a ledger. Stablecoins are issuer backed instruments.
Reader Brief
Tokenized deposits are commercial bank money on a ledger. Stablecoins are issuer-backed instruments. CBDCs are central bank money. Each does something different.
Reading Guide
Four moves that distinguish three instruments trad-fin readers consistently conflate.
Three instruments, three issuers, three risk profiles, three regulatory regimes.
**Tokenized deposit:** commercial bank money on a ledger. Liability of the issuing bank. FDIC/DGS-insured to threshold. Endogenous money creation via lending. **Stablecoin:** licensed non-bank issuer money. Liability of the issuer. 1:1 reserve, generally uninsured. Exogenous, reserve-bounded supply. **CBDC:** central bank money. Liability of the central bank. Sovereign-credit backing. The instruments are not substitutes - each is structurally suited to a specific use case. Operators that bet on "stablecoins beat tokenized deposits" or "CBDC kills stablecoins" misunderstand the architecture. The instruments are layered, not competitive.
Tokenized deposits scale through the bank-correspondent network; stablecoins scale through open issuance and exchange networks.
Production and near-production tokenized bank-money systems include J.P. Morgan Kinexys Digital Payments, formerly JPM Coin System, which J.P. Morgan said was averaging more than $5B daily in 2026; Citi Token Services, which Citi said moved from pilot to a live commercial solution in 2024; and bank/central-bank experiments such as Project Agora [1][3][7]. Common characteristics: permissioned ledger, not public chain; institutional users; value transfers within or between participating banks. The architecture extends the existing correspondent-banking network; it does not replace it. Stablecoins extend differently: open public chains, any wallet can hold, no bank relationship required, 24/7 transferability, geographic neutrality. The two architectures coexist because they serve different access patterns - tokenized deposits dominate where banks are; stablecoins dominate where they are not.
Wholesale CBDC matters more than retail CBDC for stablecoin operators - mBridge is the most credible institutional challenger.
Retail CBDCs, including e-CNY, eNaira, and Sand Dollar, compete with stablecoins for retail digital-money demand in some markets. They do not directly replace institutional cross-border clearing networks. Wholesale CBDC, including mBridge, Project Agora, and Project Helvetia, directly competes with stablecoin clearing for wholesale cross-border settlement. The current status is narrower than "commercial production": BIS says mBridge reached MVP stage in 2024, Saudi Central Bank joined that year, and the MVP platform is enabled to undertake real-value transactions subject to jurisdictional preparedness [2]. Project Agora is still a prototype-stage public-private research project [1]. mBridge-style settlement can move cross-border trade finance and FX in central bank money directly - no correspondent banking and no commercial issuer credit risk. This is the architecture that wholesale stablecoin clearing aspires to but cannot reach without sovereign infrastructure.
The mature cross-border stack uses all three.
**Customer-facing legs:** tokenized deposits at the originating and beneficiary bank. Insured, programmable, integrated with existing customer accounts. **Transit:** stablecoin between licensed operators in different banking networks. Open, 24/7, geographic-neutral. **Sovereign settlement:** wholesale CBDC central-bank-to-central-bank where corridors are CBDC-enabled, mBridge-class. This is layered architecture, not winner-take-all. Each instrument occupies the layer it is structurally suited to. Operators that build clearing networks should expect to integrate all three over time - tokenized deposits for bank counterparties, stablecoin for non-bank licensed counterparties, wholesale CBDC where central banks have built corridor infrastructure.
The Three Instruments at a Glance
The single table that resolves most of the confusion.
The instruments differ by issuer, liability, backing, insurance, yield, money creation, production status, and cross-border reach. Those differences are structural, not branding distinctions.
| Property | Tokenized deposit | Stablecoin | CBDC |
|---|---|---|---|
| Issuer | Commercial bank | Licensed non-bank issuer | Central bank |
| Liability of | The issuing bank | The stablecoin issuer | The central bank |
| Backing | Bank balance sheet, fractional reserve | 1:1 reserves, cash plus T-bills | Sovereign credit |
| Insured? | FDIC / DGS up to threshold | Generally not | Sovereign by definition |
| Yield possible? | Yes, bank may pay interest | Generally no, regulatory prohibition | Design-dependent, most retail CBDCs no |
| Money creation | Endogenous, lending creates deposits | Exogenous, 1:1 reserve | Direct central bank issuance |
| Status | Production/early commercial, Kinexys and Citi Token Services | Production, USDC, USDT | Retail live cases plus wholesale pilots/MVPs, including mBridge |
| Cross-border use | Limited, within bank or partnership | Broad, open networks | Bilateral CBDC corridors only |
Tokenized Deposits
Bank money digitally re-issued onto a ledger: same legal claim, different operational substrate.
A tokenized deposit is bank money digitally re-issued onto a ledger. It keeps the same legal claim as a deposit at the issuing bank, but moves on a different operational substrate.
Production tokenized deposit systems (2024-2025)
**J.P. Morgan Kinexys, formerly Onyx / JPM Coin System:** permissioned institutional digital payments and deposit-token infrastructure. J.P. Morgan reported in April 2026 that Kinexys had processed more than $3T since inception and averaged more than $5B daily [3]. **Citi Token Services:** Citi said Token Services for Cash moved from pilot to a live commercial solution in 2024, facilitating multimillion-dollar transactions for institutional clients [7]. **Project Agora, BIS plus 7 central banks plus commercial banks:** cross-border tokenized deposit and central-bank-reserve prototype, not a finished product [1]. **USDF Consortium and other bank-led networks:** bank-issued tokenized deposit experiments with permissioned access. Common characteristics: permissioned ledger, not public chain; institutional users; value transfers within or between participating banks.
Why tokenized deposits are not stablecoins
Tokenized deposits are bank deposits on a different ledger. They keep the deposit insurance, the bank balance sheet exposure, and the regulatory regime of bank deposits. Stablecoins are a separate instrument with separate issuer, separate reserve rules, separate license. A tokenized deposit holder is a creditor of the bank. A stablecoin holder is a creditor of the stablecoin issuer. The practical implication: tokenized deposits scale through the existing bank-correspondent network. Stablecoins scale through open issuance and exchange networks. The two architectures coexist but address different segments.
Stablecoins
Commercial issuer money: not a bank, not a central bank, and increasingly a regulated payment category.
Stablecoins are commercial issuer money. They are not banks and not central banks. They emerged from the crypto market and acquired regulatory legitimacy through MiCA, the GENIUS Act, and parallel stablecoin frameworks [6][8][9].
What stablecoins do that the other two cannot
**Open issuance:** any wallet on a public chain can hold the token without a bank relationship. Tokenized deposits require a bank account; CBDCs require a central bank wallet or proxy through a bank. **Cross-issuer interoperability:** USDC and USDT trade against each other on open markets. Two banks tokenized deposits do not natively interoperate. **24/7 transferability:** stablecoins move continuously. Tokenized deposits move during ledger operating hours; CBDCs depend on central bank operating model. **Geographic neutrality:** stablecoins serve corridors where the originating or beneficiary bank has no relationship. The trade-off: stablecoin holders bear issuer credit risk that bank deposit holders do not bear, and CBDC holders never bear.
CBDC
Central bank money in digital form, with retail and wholesale forms that are often confused.
CBDC is central bank money in digital form. Retail CBDCs serve households and firms; wholesale CBDCs serve banks and financial institutions. The wholesale version matters most for institutional stablecoin clearing.
| Property | Retail CBDC | Wholesale CBDC |
|---|---|---|
| Users | Households, firms | Banks, financial institutions |
| Live examples | e-CNY, China; Sand Dollar, Bahamas; eNaira, Nigeria | mBridge MVP, China + UAE + Hong Kong + Thailand + Saudi participation; Project Helvetia, Swiss SNB |
| Cross-border use | Almost none | Active pilots and MVP-stage wholesale corridors |
| Threat to stablecoins? | Indirect, substitute for retail USD demand | Direct, substitute for wholesale settlement use case |
Why wholesale CBDC matters more than retail CBDC for stablecoin operators
Most stablecoin volume by value is institutional or platform-mediated settlement. Retail CBDCs, including e-CNY and eNaira, can compete with some retail digital-money behavior but do not directly replace the institutional cross-border use case. Wholesale CBDC, including mBridge, Project Agora, and Project Helvetia, directly competes with stablecoin clearing for the wholesale cross-border settlement use case. mBridge reached MVP stage in 2024, and BIS says the MVP platform is enabled to undertake real-value transactions subject to jurisdictional preparedness [2]. This is the most credible sovereign challenger to the stablecoin clearing model, but it is not yet a broad commercial network.
The mBridge case study
mBridge is a multi-CBDC bridge platform built by BIS Innovation Hub with the Digital Currency Institute of the People Bank of China, Hong Kong Monetary Authority, Bank of Thailand, Central Bank of UAE, and Saudi Central Bank. BIS says it reached MVP stage in mid-2024 and that BIS handed the project over to the partners in October 2024 [2]. It is designed to settle cross-border payments and FX transactions in central bank money, directly. No correspondent banking. No commercial issuer credit risk. This is the architecture that wholesale stablecoin clearing aspires to but cannot reach without sovereign infrastructure.
When Each Is the Right Tool
The instruments are not substitutes. Each is structurally suited to a specific use case.
A mature cross-border payment stack will combine all three: tokenized deposits at bank-facing legs, stablecoin in transit between licensed operators, and wholesale CBDC where corridors are enabled.
| Use case | Best instrument | Why |
|---|---|---|
| Intra-bank treasury, single global bank, internal liquidity | Tokenized deposit | Stays within the bank; insured; programmable |
| Cross-border B2B between licensed FIs | Stablecoin, or wholesale CBDC where available | Open network, 24/7, no correspondent banking |
| Wholesale FX between G10 central banks | Wholesale CBDC | Sovereign settlement, no commercial credit risk |
| EM USD demand, households | Stablecoin | No bank relationship needed; available where local USD access is restricted |
| Domestic retail payments in advanced economies | Tokenized deposit or retail CBDC | Insured; sovereign trust; no need for crypto rails |
Why this matters for cross-border architecture
A mature cross-border payment stack will combine all three: - Tokenized deposits at the originating and beneficiary bank for the customer-facing legs. - Stablecoin in the middle for transit between licensed operators in different banking networks. - Wholesale CBDC settlement between central banks where corridors are CBDC-enabled, mBridge-class. The instruments are layered, not competitive. Operators that bet on "stablecoins beat tokenized deposits" or "CBDC kills stablecoins" misunderstand the architecture. Related reading: Six Pathways maps the full stack.
Counter-Arguments & Limitations
Where this analysis can be challenged, and the counter-counter.
Wholesale CBDC will absorb the institutional B2B flow once mBridge expands.
**The argument:** mBridge has demonstrated that wholesale CBDC infrastructure can support real-value cross-border transactions in central bank money, with no commercial issuer credit risk. As mBridge expands and similar consortia, including Project Agora, JURA, and Helvetia successors, come online, the institutional segment that stablecoin clearing serves today shifts to wholesale CBDC. Stablecoin clearing networks built for that segment are betting on a transitional window. **Counter-counter:** mBridge is corridor-consortium architecture by design - each new corridor requires central-bank agreement, technical integration, and political alignment. BIS describes the platform as MVP-stage and enabled for real-value transactions subject to jurisdictional preparedness, not as a broad commercial rail [2]. Expansion to dozens of jurisdictions is therefore a governance question as much as a technology question. Stablecoins serve open networks where no CBDC corridor exists or will exist on a relevant timeline. Both architectures will scale; the question is corridor coverage. Operators with cross-border footprints in EM, US-EU-EM triangulation, or non-G20 corridors will need stablecoin rails for the foreseeable future. Capture of the corridors that mBridge actually overlaps is the real competitive question - not whether wholesale CBDC absorbs the institutional flow wholesale.
Tokenized deposits will absorb stablecoin use cases as banks add programmability.
**The argument:** Kinexys, Citi Token Services, and Project Agora are adding 24/7 settlement, programmable money features, including atomic DvP and conditional payments, and cross-bank interoperability. As tokenized deposits acquire the operational properties currently distinctive to stablecoins, the stablecoin advantage compresses. The open-network framing is marketing for a use case that bank-issued tokenized deposits will serve better with insurance and supervisory backing. **Counter-counter:** Tokenized deposits are structurally bound to permissioned ledgers and bank account access. They cannot serve self-custody, retail-EM USD demand, or open-network use cases. The user in Lagos who holds USDT does not have a J.P. Morgan account and will never get one. The Argentine SME that hedges peso depreciation in stablecoin cannot do so via tokenized deposit at any bank. Programmability is not the binding constraint; account access is. Tokenized deposits dominate the segment where banks are present and willing; stablecoins dominate the segment where they are not. The market is larger than the bank-served segment, and stablecoin coverage of the non-bank-served segment is structural, not transitional. The two instruments coexist for the same reason that bank money and physical cash coexist: different access requirements at different price points.
Evidence And Sources
This raw HTML export preserves source visibility for crawler and contractor review. Indexing decision: index, follow.
- Project Agora - BIS Innovation Hub
- Project mBridge - BIS Innovation Hub
- Kinexys 2026 Milestones - J.P. Morgan
- Tokenisation in the Context of Money and Other Assets - BIS
- CBDC Tracker - Atlantic Council
- High-level Recommendations for Stablecoin Arrangements - Financial Stability Board
- Citi Token Services Marks New Milestone - Citi
- Crypto-assets - Markets in Crypto-Assets Regulation - European Commission
- S.1582 - GENIUS Act - US Congress