CBDC Licensing & Intermediary Frameworks

Executive Summary

This document provides a regulator-grade comparative analysis of CBDC operator and intermediary authorization — who may distribute the CBDC, and under what licence — across all 20 jurisdictions in this domain:

United States,

Bahamas,

Brazil,

Canada, Eastern Caribbean,

Jamaica,

Australia,

China,

India,

Japan,

Singapore,

South Korea,

Thailand, the

EU,

Russia,

Sweden, the

UK,

Ghana,

Nigeria, and

South Africa. It feeds the cross-domain License Map.

Of the 20 jurisdictions, 6 have a live or launched retail CBDC whose distribution model is operationally tested: the Bahamas (Sand Dollar), Nigeria (eNaira), Jamaica (JAM-DEX), the Eastern Caribbean (DCash — pilot concluded, successor DCash 2.0 suspended Feb 2026), China (e-CNY, large-scale pilot operating at near-deployment scale), and India (e₹, dual retail/wholesale pilot). The remaining 14 are at pilot, research, design, or — in the United States' case — anti-CBDC stages, and their intermediary models are proposed, prospective, or "not applicable / not verified."

The central finding is unchanged across this refresh: no jurisdiction has created a standalone, free-standing CBDC intermediary licence. Every distribution model — launched or proposed — rests on existing financial-sector authorization (banking licences, payment-service-provider authorisations, e-money permits, money-transmission licences) sometimes extended with CBDC-specific obligations imposed via scheme rulebooks, bilateral agreements, or regulatory directives. The UK's proposed PIP/ESIP framework comes closest to a dedicated intermediary category but remains entirely at the design stage.

Two refreshed facts reshape the picture this cycle. First, China now lists 22 authorised operators (the PBOC doubled the roster from 10 to 22 on 2 Apr 2026), the broadest designated-operator base of any launched programme. Second, the Eastern Caribbean's licensed-FI distribution model is now dormant — DCash 2.0 was suspended by the ECCB Monetary Council on 13 Feb 2026 in favour of a conventional Fast Payment System + CAPS, so its 21-FI network no longer operates.

Adoption, wallet, and transaction figures cited below are self-reported by the issuing central banks or their trackers and should be read as such; they are flagged accordingly.

1. Operator / Intermediary Authorization Taxonomy

Across the 20 jurisdictions, every intermediated CBDC design (and all 20 are intermediated or would be) channels distribution through pre-existing licensees. Within that, six distinct authorization models emerge — distinguished by who is eligible, how they are admitted (designation vs. open application vs. mandate), and whether non-banks have a primary-distributor path.

Model A: Existing Bank Licence + Central-Bank Designation (China, India, Jamaica)

Description: Only institutions already holding a banking/deposit-taking licence may distribute the CBDC, and admission is by central-bank designation or bilateral authorisation — there is no open application route and no new licence category.

• China (e-CNY): The PBOC designates "authorised operators." As of 2 Apr 2026 there are 22 authorised operators (the roster was doubled from 10 with 12 new banks added; the full new list is NOT FULLY VERIFIED). The original ten were six state-owned commercial banks (ICBC, ABC, BOC, CCB, BoCom, PSBC), two joint-stock banks (CMB, CIB), and two internet banks (WeBank/Tencent, MYBank/Ant Group). Non-bank payment institutions (Alipay, WeChat Pay) act as front-ends/exchange services but are not designated operators. From 1 Jan 2026 the e-CNY framework was reclassified toward a bank-deposit-money model (verified wallets earn 0.05% interest), tightening the bank-only character. Admission is by PBOC designation, not application.
• India (e₹): ~19 scheduled commercial banks participate through bilateral RBI authorisation (still ~19 banks in late 2025). Eligibility requires an existing banking licence plus technology readiness (a compliant e₹ wallet app), KYC/AML infrastructure, and reporting; no additional capital beyond existing banking norms. RBI has additionally authorised non-bank operators (PayNearby, Bankit) as distribution agents — a limited non-bank path layered onto the bank-only core. The process is invitation-based with no public application pathway.
• Jamaica (JAM-DEX): Only deposit-taking institutions (DTIs) licensed by the Bank of Jamaica may serve as wallet providers; admission is by BOJ designation. As of 2026, three DTIs are authorised (NCB/Lynk, JN Bank/JN Pay, and Sagicor Bank), but only two have a live public wallet (Lynk and JN Pay — JN Pay launched July 2025; Sagicor had not launched a public consumer wallet as of end-2025).

Outcomes: Distributor count is structurally small (22 in China, ~19 in India, 2–3 in Jamaica); entry barriers are maximal (existing bank licence required); fintech/non-bank entities are largely excluded from primary distribution (India's agent layer is the partial exception).

Model B: Extended Payment-Service-Provider Framework (EU — Proposed)

Description: Distribution obligations attach across the full spectrum of already-supervised PSPs — credit institutions, payment institutions, and e-money institutions — without any new licence.

EU (Digital Euro): Under the proposed Digital Euro Regulation (COM(2023) 369, not yet enacted), the following would be required to distribute the digital euro:

1. Credit institutions authorised under CRD/CRR
2. Payment institutions authorised under PSD2 (and forthcoming PSD3/PSR)
3. Electronic money institutions authorised under EMD2

Additionally, public entities (post offices, local authorities) could distribute to users without bank accounts. This is the broadest potential intermediary base of any jurisdiction — potentially thousands of supervised entities across the 20 euro-area states. PSPs would not pay scheme fees. The digital euro builds on the existing legal authorization framework under PSD2 by extending distribution obligations to supervised PSPs. The Eurosystem's preparation phase concluded in October 2025 and the Governing Council moved the project to its next phase (29–30 Oct 2025); a PSP pilot call opened 5 Mar 2026, but issuance still requires the Regulation's adoption (ECON committee vote scheduled 23 Jun 2026).

Outcomes: Broadest potential distributor base; no new licence category; mandatory PSP participation reduces single-provider bottleneck risk; competition across banks, fintechs, and EMIs.

Model C: New Dedicated CBDC Intermediary Categories (UK — Proposed)

Description: The jurisdiction proposes new, CBDC-specific intermediary roles distinct from existing banking or payment licences.

UK (Digital Pound): The Bank of England's October 2025 design note on intermediary roles proposes two new categories — within a public-private platform model where users interact with the intermediary, not the Bank:

Payment Interface Providers (PIPs):

• Provide essential digital-pound services to end users (account provisioning, onboarding, payment initiation, wallet management, balance/transaction queries)
• Responsible for KYC/AML and data protection
• Direct API connection to the BoE core ledger
• Subject to both legislative/regulatory requirements and the BoE scheme rulebook
• Expected participants: banks, building societies, fintechs, PSPs
• Capital requirements: NOT YET SPECIFIED

External Service Interface Providers (ESIPs):

• Deliver specialised value-added services without directly managing funds (e.g. Payment ESIPs for third-party initiation; Data ESIPs for spending analysis/budgeting)
• Indirect ledger access — connect through PIPs, not to the core ledger
• Lighter rules/obligations than PIPs; must obtain user consent

No decision to launch has been taken; primary legislation would be required, and the proceed/don't-proceed decision is expected later in 2026.

Outcomes: Most innovative intermediary framework but entirely at design stage; the PIP/ESIP split mirrors PSD2's PI/TPP distinction; creates an explicit non-bank primary-distributor path; capital and authorisation criteria remain undefined.

Model D: State-Directed Mandatory Bank Participation (Russia)

Description: No new licence; existing bank licences carry a compulsory CBDC-integration obligation, phased by systemic importance.

Russia (Digital Ruble): The CBR (sole issuer and platform operator) mandates phased integration. Users interact only via their existing bank's mobile app; one wallet per person sits on the CBR platform. Government/budget use went live narrow-scope on 1 Jan 2026, with the full large-scale launch legally mandated for 1 Sep 2026.

Banks must provide wallet opening, transfers, and payment for goods/services; the CBR publishes message-album specifications (e.g. ES 2026.01) that banks must implement.

Outcomes: Guarantees comprehensive bank coverage by compulsion; no choice for banks (non-compliance is a regulatory breach); banks bear implementation costs ("hundreds of millions of rubles" each); no non-bank primary-distributor path.

Model E: Central-Bank Platform with Broad Licensed-FI Distribution (Bahamas, Nigeria, Eastern Caribbean)

Description: The central bank operates the core platform and admits a broad range of licensed financial institutions — including non-bank FIs and, in some cases, agents — as distribution channels. No CBDC-specific licence; existing FI licences govern.

• Bahamas (Sand Dollar): The CBOB operates the platform (Cortex DLT, built by NZIA/Movmint). Wallet providers ("Authorised Financial Institutions," AFIs) must hold an existing CBOB licence as a commercial bank (Banks & Trust Companies Regulation Act), credit union (Co-operative Credit Unions Act), money-transmission business, or payment service provider. No bank account is required to open a wallet. AFIs increased to 10 firms (2025). In July 2024 the CBOB announced a two-year plan to mandate commercial-bank Sand Dollar integration (effective ~mid-2026), and a 2025 agency-banking consultation now requires Sand Dollar operability.
• Nigeria (eNaira): The CBN is sole issuer and validator; distribution is two-tier through Deposit Money Banks (DMBs), Payment Service Banks (PSBs), Microfinance Banks (MFBs), and Primary Mortgage Banks (PMBs), governed by the Regulatory Guidelines on the eNaira (eff. 25 Oct 2021) plus BOFIA 2020 — no standalone eNaira Act, no CBDC-specific licence. Despite the broad institution set, the programme is largely dormant and is pivoting to a wholesale CBDC focus (eNaira v2.0), with the CBN tilting toward a private stablecoin (cNGN).
• Eastern Caribbean (DCash) — now dormant: The ECCB distributed DCash through 21 participating licensed FIs (including 11 credit unions) plus 10 government agencies (Jan 2024), with Bitt Inc. operating the technical infrastructure — the broadest and only multi-national licensed-FI model. The DCash pilot concluded January 2024 and the DCash 2.0 successor was suspended by the ECCB Monetary Council on 13 Feb 2026, in favour of a Fast Payment System + CARICOM (CAPS) integration, so this distribution network is no longer operational.

Outcomes: Widest range of intermediary types (incl. non-bank FIs, credit unions, and — Nigeria — agents); central bank carries heavy operational burden; broad distribution has not produced adoption (Bahamas circulation ~0.39% of cash; Nigeria 98.5% wallet inactivity after one year; DCash now suspended).

Model F: No Distribution Model (Research / Design / Anti-CBDC) — 9 jurisdictions

Nine jurisdictions have no operational distribution model because no CBDC is issued. Their intermediary arrangements are prospective (would leverage existing bank/PSP licensing) or — for the US — barred outright.

• United States (anti-CBDC): No CBDC and the strongest anti-issuance stance of any major economy. Executive Order 14178 (23 Jan 2025) bars all federal CBDC activity; a statutory CBDC moratorium passed both chambers in differing forms but is NOT enacted (H.R. 1919 passed the House 219–210; the 21st Century ROAD to Housing Act's CBDC moratorium passed Senate 89–10 and an amended House version 396–13, but the versions remain unreconciled). New Fed Chair Kevin Warsh (sworn in 22 May 2026) reaffirms opposition to a retail CBDC. The operative strategy is stablecoin-first under the GENIUS Act (S.1582, Public Law 119-27). Any hypothetical US CBDC was envisaged (2022 "Money and Payments" paper) as intermediated through existing financial institutions — but none exists. No distributor model.
• Brazil (Drex): Pilot restarted — the Hyperledger Besu platform was decommissioned 4 Nov 2025 and the project is rebuilding "from scratch." Distribution would be via BCB-authorised financial institutions (banking licence + Resolution 315/2023 pilot authorisation) in consortia, on a tokenized-deposit model (banks move client deposits into Drex wallets). The earlier pilot involved 16 consortia (50+ companies); admission was pilot-invitation only, not open. No live distributor model post-decommissioning.
• Canada: Retail CBDC scaled down (Sep 2024); no issuing authority designated. Any future distribution would leverage RPAA supervision of ~1,600 PSPs. The Stablecoin Act (Bill C-15) received Royal Assent 26 Mar 2026 but is not yet in force and expressly excludes CBDCs. No distributor model.
• Australia: Wholesale-focused; retail shelved. Project Acacia's pilot wholesale CBDC was distributed via Exchange Settlement Account (ESA) holders acting as distributors (ANZ, Banking Circle Australia, Cuscal) after KYC. The Acacia final report (18 May 2026) covered 20 use cases (12 pilot + 8 PoC) with A$4.4m wCBDC issued — research/pilot issuance, not a live launch. No retail distributor model.
• Japan: No CBDC; 2026 decision pending. A two-tier intermediated model is assumed (BOJ core ledger; commercial banks and PSPs manage wallets/KYC under existing banking/payment licensing). No distributor model yet.
• Singapore: Wholesale-first; first live wholesale CBDC issuance on the SGD Testnet (first live transaction Nov 2025) via institutional participants (DBS, OCBC, UOB, Standard Chartered) under existing MAS regulation — no new licence. Retail PBM runs through existing commercial-bank infrastructure. No retail CBDC distributor model.
• South Korea (Project Hangang): Hybrid deposit-token model; commercial banks issue won-pegged deposit tokens to users via bank apps, BOK runs the settlement layer. Phase 2 restarted March 2026 with 9 banks; not open to non-bank/fintech providers. Explicitly not a full retail CBDC.
• Thailand: Retail pilot completed (2024) with no full-scale plans; pilot wallets were run by financial-service providers (Krungsri, SCB, 2C2P). Focus pivoted to a Programmable Payment Sandbox for stablecoins. No retail distributor model going forward.
• Sweden (e-Krona): Research/analysis only; no issuance decision and no licensing framework. An anticipated two-tier model (Riksbank issuer/operator; commercial banks and PSPs as intermediaries under existing licences) exists only on paper, with Finansinspektionen supervision of intermediaries. Trajectory now tied to the EU digital euro. No distributor model.
• Ghana (eCedi): Pilot declared "completed" by the BoG Governor on 22 May 2026, now reframed toward wholesale/cross-border design; retail eCedi unlaunched, enabling legislation unpassed. The pilot used existing licensed institutions (CalBank, Vodafone Cash) on a two-tier model, leveraging bank/e-money/mobile-money-agent licensing. Separately, the Virtual Asset Service Providers Bill, 2025 passed Parliament Dec 2025 (crypto/VASP licensing — distinct from the eCedi). No live distributor model.
• South Africa: No CBDC; all work is PoC/sandbox. Project Khokha participants were licensed banks + the JSE; the SARB Position Paper (27 Nov 2025) found no compelling immediate need for a retail CBDC; Project Khokha 2x continues wholesale work. No retail distributor model.

(Note: Models C, D, and parts of B/E overlap with the "not-launched" status above where the design is documented but unissued; the taxonomy classifies by the intended authorization architecture, while §2 marks live vs. proposed.)

2. Operator / Intermediary Authorization — All 20 Jurisdictions

Aggregates (N = 20), derived from the matrix:

• 6 of 20 have a live/launched retail CBDC distribution model (Bahamas, Nigeria, Jamaica, Eastern Caribbean [now suspended], China, India). Counting only currently operational networks gives 5 of 20 (Eastern Caribbean's is dormant).
• 2 of 20 run a live wholesale CBDC distribution model (Singapore — live issuance; Australia — completed pilot issuance). South Korea operates a live hybrid deposit-token pilot.
• 20 of 20 intermediate (or would intermediate) distribution through existing licensees; 0 of 20 have created a standalone CBDC intermediary licence.
• 8 of 20 provide a non-bank primary-distribution path (live or proposed): EU, UK, Bahamas, Eastern Caribbean (design), Sweden (anticipated), Ghana (mobile-money agents), Canada (hypothetical PSPs), plus partial paths in India and Nigeria (agents) — bank-only models dominate the launched set.
• 1 of 20 is anti-CBDC (United States) — no authorisation is possible.

3. Arbitrage Analysis — Where Are New Entrants Welcome vs. Bank-Only?

Lowest barrier to entry (intermediary access)

1. EU Digital Euro (proposed) — any entity already authorised as a PSP (payment institution, EMI, or credit institution) under PSD2/EMD2 would be required to participate; no additional licence. The most accessible regime for established fintechs.
2. UK Digital Pound (proposed) — the PIP/ESIP model envisions non-bank entities (fintechs, building societies) as primary distributors; capital requirements not yet specified, but the design intent is broad participation.
3. Bahamas Sand Dollar — MTBs and PSPs participate alongside banks; the broadest live non-bank path among launched programmes.
4. Ghana eCedi (design) and Sweden e-Krona (anticipated) — both envisage mobile-money agents / PSPs, though neither has launched.

Highest barrier to entry

1. China e-CNY — closed roster of 22 PBOC-designated operators; no application process; invitation-only.
2. Russia Digital Ruble — only existing credit institutions; participation is mandatory for banks, no non-bank path.
3. India e₹ — invitation-based bilateral RBI authorisation; scheduled commercial banks (agents a narrow exception).
4. Jamaica JAM-DEX — DTIs only; three authorised, two live, after nearly four years.
5. South Korea — banks only; not open to non-bank/fintech providers.

Barred entirely

• United States — federal CBDC activity is prohibited by EO 14178; no intermediary can be authorised because no CBDC may be issued.

Where new entrants are welcome vs. bank-only

Cost of participation (where documented)

4. Compensation Model Analysis

Who pays for CBDC distribution rails remains among the most consequential unresolved CBDC design issues. Each model allocates costs differently.

EU Digital Euro: Multi-Stakeholder Fee Model (Most Developed)

• Inter-PSP fee (merchant-side): two methods proposed — cost-based (regulator sets limits on the most efficient merchant PSPs' costs plus reasonable profit) or market-based (aligned with existing interchange). Spain proposed a 0.2% cap (aligned with the IFR 2015 debit-card interchange cap) for a 5-year transition; EuroCommerce merchants proposed 0.05% with a 2-cent/transaction cap and a 4-cent total merchant ceiling.
• PSP scheme fees: zero.
• Consumer charges: basic digital-euro services free for individuals; PSPs may charge for value-added services.
• Eurosystem costs: development ~EUR 1.3bn; annual operations ~EUR 320m from 2029; funded by seigniorage. Framework agreements with EU-controlled providers were announced 2 Oct 2025 (est. EUR 432.1m, max EUR 1.164bn).

China e-CNY: State-Subsidised Model (with a 2026 shift)

• The PBOC does not charge authorised operators or individuals; operators do not charge individuals for exchange — a zero-fee design to drive adoption.
• No published merchant-fee schedule; merchants register through the e-CNY app ecosystem.
• From 1 Jan 2026 the reclassification toward bank-deposit liabilities (verified wallets earn 0.05% interest, first quarterly payout 20 Mar 2026) means commercial banks now treat e-CNY within their deposit base subject to reserve requirements — an implicit compensation/cost mechanism.

Russia Digital Ruble: Free-for-Individuals, Phased Commercial Fees

The CBR has published the most detailed fee schedule of any jurisdiction:

Banks bear infrastructure cost; the CBR operates the platform at state expense; market players risk losing up to RUB 50bn/year in reduced transaction fees.

UK Digital Pound: Undefined

No compensation model has been published. UK Finance has warned the digital pound risks becoming an "unfunded obligation for banks and payments providers."

India e₹, South Korea, Singapore, Australia, Brazil, Bahamas, Jamaica, Nigeria, Eastern Caribbean, Ghana, Japan, Sweden, Thailand, South Africa, Canada, US: NOT VERIFIED / not applicable

No publicly documented intermediary compensation models were found for these jurisdictions as of the source matrix. (South Korea's pilot recorded ~5bn won per-bank infrastructure spend against only 692.46m won in Phase-1 transaction value — a cost-to-activity imbalance, not a compensation model.)

Who bears the cost of CBDC distribution?

5. Merchant Acceptance Analysis

Mandatory acceptance

EU Digital Euro (strongest mandatory framework, proposed): Legal-tender status would require acceptance by all merchants who currently accept digital payment means (limited, well-framed exceptions); merchant-fee caps apply. The Council's December 2025 position endorsed mandatory acceptance with proportionate fees, and the European Parliament adopted amendments endorsing both online and offline use (Feb 2026).

Russia Digital Ruble (phased mandatory by revenue):

• 1 Sep 2026: retailers with revenue >RUB 120M must accept.
• 1 Sep 2027: revenue >RUB 30M.
• 1 Sep 2028: remaining retailers >RUB 30M. Exempt: outlets <RUB 5M annual revenue.

Legal tender but voluntary acceptance

• Jamaica JAM-DEX: legal tender (BOJ Amendment Act 2022) but no mandatory-acceptance requirement; ~2,379 merchants registered (Sept 2024, self-reported). Governor Byles is pushing banks to retrofit POS (JN Bank expected first by July 2026).
• Bahamas Sand Dollar: legal tender (CBOB Act 2020) but acceptance effectively voluntary.
• Nigeria eNaira: legal tender; voluntary acceptance; programme dormant.

Voluntary acceptance

• China e-CNY: no mandatory acceptance; ~5.6m registered merchants (mid-2023, self-reported); zero merchant fees; Alipay/WeChat QR integration.
• India e₹: no mandatory acceptance; ~300,000 merchants (end 2024, self-reported); UPI-QR interoperability lets existing UPI merchants accept e₹.
• UK Digital Pound (proposed): voluntary, phased (online first, then in-store QR, then NFC/POS).

Not applicable (no live retail CBDC)

US (anti-CBDC), Canada, Australia (wholesale), Singapore (wholesale), South Korea (deposit-token pilot), Thailand (pilot ended), Japan, Sweden, Brazil (pilot restarted), Ghana (pilot completed), South Africa, and Eastern Caribbean (suspended) have no operative merchant-acceptance regime.

6. Banking & Settlement Access for CBDC Intermediaries

Can non-bank entities participate as primary distributors?

Settlement / central-bank ledger access

Direct API to central-bank ledger: UK (proposed — PIPs connect directly to the BoE core ledger; ESIPs indirectly through PIPs); EU (PSPs connect to the Eurosystem settlement platform built on TIPS).

Through existing settlement systems: Jamaica (wallet providers integrate with JamClear RTGS + eCurrency Mint); Russia (banks connect to the CBR platform; users access via bank apps); India (banks provide wallet apps over the RBI core platform).

Central bank manages all infrastructure: China (PBOC centralised ledger; operators provide the user layer); Bahamas (CBOB Cortex DLT; AFIs connect through CBOB); Eastern Caribbean (ECCB platform — now dormant); Nigeria (CBN Hyperledger Fabric, sole validator).

7. What Works and What Doesn't

What works

1. Leveraging existing licensing avoids regulatory delay. Every launched CBDC (China, India, Bahamas, Jamaica, Nigeria, Eastern Caribbean) used existing bank/PSP licensing rather than a new category — enabling faster launch than design-stage jurisdictions (EU, UK). India launched its pilot within months of amending the RBI Act; China designated existing banks with no new instrument.

2. Broader intermediary bases correlate with broader adoption potential. China's inclusion of internet banks (WeBank, MYBank) tied to Alipay/WeChat drove the largest CBDC merchant base (~5.6m, self-reported); India's UPI interoperability leverages 300M+ existing merchant touchpoints. Distribution restricted to a handful of traditional banks (Jamaica: 2–3 providers) yields the narrowest acceptance.

3. Mandatory bank participation addresses "bank reluctance." Both the Bahamas (July 2024 mandate, ~mid-2026 effective) and Russia (2026–2028 phased mandate) moved from voluntary to mandatory after banks proved reluctant to invest. Jamaica's years-long single-provider phase demonstrates the cost of relying on voluntary bank adoption.

4. Agent and non-bank networks extend reach — in principle. India's PayNearby/Bankit agents and Nigeria's super agents/MFBs extend access to the unbanked, though neither has translated into meaningful adoption.

What doesn't work

1. Invitation-only, bank-only distribution creates bottlenecks. Jamaica's single-provider phase and India's ~19 banks (out of thousands of FIs) constrain distribution. China's 22 operators work only because they include the platforms (Alipay, WeChat) that already dominate mobile payments.

2. Licensing breadth without compelling use cases fails to drive adoption. Nigeria authorised many intermediary types yet hit 98.5% wallet inactivity after one year; the Bahamas has 10 AFIs but only ~0.39% of cash in circulation. Breadth is necessary but insufficient.

3. Undefined compensation deters intermediary investment. The UK's missing compensation model led UK Finance to warn of an "unfunded obligation." Where compensation is clear (Russia, EU), intermediaries can make informed decisions.

4. Central-bank operational burden is unsustainable at scale. The Bahamas and Nigeria show that central banks acting as platform operator and regulator face capacity constraints — Nigeria is pivoting to wholesale.

5. A licensed-FI network can be abandoned. The Eastern Caribbean built the only multi-national licensed-FI distribution network (21 FIs across 8 states) and then suspended DCash 2.0 (Feb 2026) in favour of a conventional Fast Payment System — a reminder that a distribution model is only as durable as the central bank's commitment.

6. Some jurisdictions decline to build a distribution model at all — by design. South Africa (Position Paper 27 Nov 2025: no compelling immediate need), Canada (retail scaled down), and the United States (anti-CBDC, EO 14178) demonstrate that not authorising CBDC intermediaries can be a deliberate, defensible choice.

8. Recommended Framework

Based on the comparative evidence across all 20 jurisdictions:

8.1 Tiered intermediary licensing

Tier 1 — Full CBDC Distributors (analogous to UK PIPs / EU PSPs): direct connection to the central-bank platform; full KYC/AML; account provisioning, wallet management, payment initiation; settlement account required. Eligible: banks, authorised payment institutions, EMIs. Dual supervision (central bank + conduct authority). Capital calibrated to CBDC-specific operational risk (funds sit on the central-bank ledger, reducing prudential risk).

Tier 2 — Value-Added Service Providers (analogous to UK ESIPs): indirect access via Tier 1; payment initiation, analytics, budgeting; no direct fund management; lighter burden. Eligible: fintechs, technology and data firms.

Tier 3 — Physical Distribution Agents: agent networks for onboarding, cash-to-CBDC conversion, support, under contract with Tier 1. Essential for inclusion in unbanked populations. Eligible: post offices, retail chains, mobile-money agents (Ghana/India model).

8.2 Compensation model design

1. Basic individual use: free (consensus across all studied jurisdictions).
2. Merchant fees: capped below card interchange — Phase 1 (0–3 yrs) 0.1–0.15% with a 5-cent ceiling; Phase 2 cost-based.
3. Intermediary scheme fees: zero/minimal.
4. Platform costs: seigniorage (EU model) or government budget.
5. Value-added fees: market-determined.

8.3 Merchant-acceptance phasing

Prerequisites before mandatory acceptance: enforced fee caps; POS subsidies / zero-cost integration; a universal QR standard (Russia model); offline capability for network-disrupted areas.

8.4 Competition and market structure

• Avoid single-provider bottlenecks — require 3–5 Tier 1 intermediaries at launch (Jamaica is the cautionary case).
• Prevent incumbency lock-in — open API standards, interoperability mandates, portable wallet credentials.
• Enable non-bank participation — payment institutions and EMIs should have equal Tier 1 access; agent networks extend reach without full licensing.
• Commit, or decline cleanly — the Eastern Caribbean's suspension shows the cost of building a network without durable commitment; South Africa shows that a clearly-reasoned "no" can be the responsible decision.

9. Decision Matrix

Recommended calibration for most jurisdictions: Moderate across dimensions, with aggressive non-bank access to maximise competition and inclusion.

10. Sources and Verification

Primary sources by jurisdiction (operator/intermediary authorization)

China: PBOC White Paper (July 2021); PBOC Action Plan effective 1 Jan 2026; 2 Apr 2026 doubling of authorised-operator roster to 22 (12 new banks; full list NOT FULLY VERIFIED). India: RBI Act 1934 (as amended by Finance Act 2022); RBI Annual Report 2025-26 (released 29 May 2026); ~19 participating banks + non-bank agents (PayNearby, Bankit). Jamaica: Bank of Jamaica (Amendment) Act, 2022; BOJ DTI licensing; three authorised DTIs, two live public wallets (Lynk, JN Pay) as of end-2025. Bahamas: Central Bank of The Bahamas Act, 2020; AFI licensing (banks, credit unions, MTBs, PSPs); 10 AFIs (2025); July 2024 mandatory-integration plan. Nigeria: Regulatory Guidelines on the eNaira (eff. 25 Oct 2021) + BOFIA 2020; DMBs/PSBs/MFBs/PMBs + agents; Nov 2025 wholesale-pivot acknowledgement. Eastern Caribbean: 2020 ECCB Agreement amendments; 21 licensed FIs + 10 govt agencies (Jan 2024); DCash 2.0 suspended by Monetary Council 13 Feb 2026 (Communiqué 4 May 2026). EU: COM(2023) 369 Digital Euro Regulation (procedure 2023/0212(COD), not enacted); preparation phase concluded Oct 2025; PSP pilot call 5 Mar 2026; ECON vote scheduled 23 Jun 2026. UK: BoE Intermediary Roles & Scheme Rulebook design note (October 2025); PIP/ESIP categories; UK Finance "unfunded obligation" analysis. Russia: Federal Law No. 340-FZ (1 Aug 2023); CBR phased mandate (1 Sep 2026 / 2027 / 2028); government/budget use live 1 Jan 2026. Australia: Project Acacia final report (mr-26-13, 18 May 2026) — ESA-holder distributors (ANZ, Banking Circle Australia, Cuscal); A$4.4m wCBDC across 8 use cases. Singapore: PSA 2019 / FSMA 2022; SGD Testnet institutional participants; first live wholesale transaction Nov 2025. South Korea: Project Hangang Phase 2 (9 banks, from March 2026); commercial-bank deposit-token issuance; Bank of Korea Act. Thailand: retail pilot concluded 2024 (Krungsri, SCB, 2C2P); Programmable Payment Sandbox pivot. Brazil: BCB Resolution 315/2023; consortia model; Hyperledger Besu decommissioned 4 Nov 2025; rebuild from scratch. Canada: retail scaled down (Sep 2024); RPAA supervision of ~1,600 PSPs; Stablecoin Act (Bill C-15) Royal Assent 26 Mar 2026 (not in force; excludes CBDCs). Japan: assumed two-tier (banks + PSPs); 2026 retail-CBDC decision pending. Sweden: no issuance decision; anticipated two-tier (banks/PSPs); trajectory tied to the EU digital euro. Ghana: Bank of Ghana Act, 2002; pilot "completed" per Governor 22 May 2026; VASP Bill 2025 passed Parliament Dec 2025 (distinct from eCedi). South Africa: SARB Act 1989; Project Khokha (banks + JSE); Position Paper 27 Nov 2025; Project Khokha 2x. United States: EO 14178 (23 Jan 2025) bars federal CBDC activity; statutory moratorium passed both chambers in differing forms (H.R. 1919; ROAD to Housing Act) but NOT enacted; GENIUS Act (S.1582 / PL 119-27); Fed Chair Kevin Warsh sworn in 22 May 2026.

Items marked NOT VERIFIED / not applicable

1. India e₹ intermediary compensation — no public RBI model.
2. UK PIP/ESIP capital requirements — explicitly "NOT YET SPECIFIED" by the BoE.
3. UK compensation model — none published.
4. Brazil Drex distributor model post-decommissioning — rebuilding from scratch; merchant-acceptance framework not specified.
5. Bahamas / Jamaica / Nigeria intermediary compensation — no public documentation.
6. South Korea / Singapore / Australia intermediary compensation — not documented (wholesale/pilot scope).
7. Japan distributor model — assumed two-tier; not yet specified (no issuance).
8. Sweden / Ghana / Canada distributor models — anticipated/hypothetical; no live framework.
9. South Africa retail distributor model — not applicable (no retail CBDC).
10. China full 22-operator roster — 12 new banks added 2 Apr 2026; full list NOT FULLY VERIFIED.
11. Eastern Caribbean — DCash 2.0 distribution network suspended (13 Feb 2026); no operational model.
12. United States — not applicable; federal CBDC activity barred (anti-CBDC).

Self-reported figures: adoption, wallet, merchant, and circulation figures (e.g. China ~5.6m merchants; India ~300,000 merchants; Bahamas circulation ~0.39% of cash; Jamaica ~2,379 merchants; Nigeria 98.5% wallet inactivity; Russia pilot stats) are self-reported by the issuing central banks or their trackers and have not been independently audited; they are flagged as such throughout.

---

Last updated: 2026-06-04