Central Bank Digital Currencies (CBDCs) Explained: Retail, Wholesale, and the Race to Digital Cash

What a CBDC actually is

A central bank digital currency is exactly what the name says: a currency issued by a central bank, in digital form. That one-sentence definition conceals a distinction that underpins the entire policy debate around CBDCs, and it is worth stating clearly before moving on. A CBDC is a direct liability of the central bank. When you hold a CBDC, you hold a claim on the central bank itself — not on a commercial bank, not on a private issuer, not on a blockchain protocol.

This is what separates CBDCs from every other form of digital money. A bank deposit is a liability of the commercial bank that holds it, protected up to statutory limits (£120,000 under FSCS in the UK — increased from £85,000 in December 2025 — and €100,000 under EU deposit guarantee schemes) but exposed to the insolvency of that institution beyond the threshold. A stablecoin is a liability of whatever private entity issues it — subject to the terms of its reserve management and its regulatory position in any given jurisdiction. Physical banknotes and coins are direct liabilities of the central bank. A CBDC extends that sovereign money property into digital form.

This distinction is not technical pedantry. It is the reason governments and central banks are interested in CBDCs in the first place, and it is the reason critics are alarmed by them. The same property that makes a CBDC robust (sovereign backing, no counterparty risk from a commercial bank) is what makes it a surveillance instrument of a fundamentally different nature from any private payment method.

Retail versus wholesale: two very different animals

All CBDCs fit into one of two categories, and the policy implications of each are entirely different.

Retail CBDCs are intended for use by the general public — businesses and individuals — for everyday transactions. They are the digital equivalent of physical cash. A retail CBDC wallet held by a citizen represents a direct claim on the central bank, bypassing commercial bank intermediation entirely. The policy questions retail CBDCs raise are correspondingly fundamental: What happens to commercial bank deposit bases? How is privacy managed when every transaction is logged on a sovereign ledger? What design choices — holding limits, interest rates, programmability — does the issuing authority build in?

Wholesale CBDCs operate between financial institutions rather than in the hands of the public. They are, in essence, a tokenised form of the reserves that commercial banks already hold at the central bank — but issued on a shared ledger that enables faster, more efficient settlement between banks, and potentially across borders. Wholesale CBDCs do not replace consumer payment infrastructure; they upgrade the plumbing underneath it. The privacy concerns that surround retail CBDCs are largely absent from the wholesale variant, since the counterparties are regulated financial institutions already subject to transaction reporting requirements.

Most of the political controversy around CBDCs — the talk of government surveillance, programmable spending restrictions, the abolition of financial privacy — concerns retail CBDCs specifically. Most of the commercially interesting near-term development, particularly in cross-border settlement, concerns wholesale CBDCs. The two categories tend to be conflated in public discourse, which generates more heat than light.

Where retail CBDCs are already live

Several retail CBDCs have launched, though their adoption rates vary widely.

The Bahamas Sand Dollar was the first retail CBDC to launch anywhere in the world, going live in October 2020. Issued by the Central Bank of the Bahamas, it was designed to extend financial services access to residents of the Bahamas' more remote islands, where banking infrastructure is limited. Adoption has been modest — the Sand Dollar remains a small fraction of overall payments — but it has established a proof of concept for island-economy CBDCs in general.

The Eastern Caribbean DCash was launched across the Eastern Caribbean Currency Union (ECCU) in 2021. A major outage in early 2022 lasted several weeks, and the project has since undergone significant revision. Its operational scope and status across ECCU member states has evolved considerably from the original pilot; readers should consult the Eastern Caribbean Central Bank's current publications for the up-to-date position.

Jamaica's JAM-DEX launched in 2022. Jamaica formally designated JAM-DEX as legal tender — as did the Bahamas with the Sand Dollar, which is Bahamian legal tender as the digital form of the Bahamian dollar, and the ECCB with DCash, which is legal tender across ECCU member states.

Nigeria's eNaira, which launched in October 2021, is the largest retail CBDC launched by an African economy. Adoption has been sluggish despite government promotion, and the Central Bank of Nigeria has iterated on the design several times. The eNaira experience is widely cited in CBDC research literature as a case study in the challenges of driving adoption for a new payment instrument when existing alternatives — mobile money, bank transfers — are already functional.

Then there is China's e-CNY (digital yuan), which exists in a category of its own. The People's Bank of China began trials in 2020 and has since expanded the e-CNY pilot to dozens of cities, distributed billions of yuan through lottery programmes and government subsidies, and integrated the system with major Chinese payment platforms. By scale of pilot and institutional infrastructure, the e-CNY is by far the most developed retail CBDC on earth. It is not yet considered a full national launch — the People's Bank has continued to describe it as a pilot — but the breadth and depth of the deployment is well beyond anything happening elsewhere. The e-CNY operates on a two-tier model: the central bank issues to commercial banks, which distribute to the public. Crucially, it is designed to coexist with WeChat Pay and Alipay rather than replace them, though analysts note it gives the People's Bank of China visibility into transaction flows that the private platforms previously kept proprietary.

The digital euro: technical work moves to the next phase

The European Central Bank's digital euro project is the most closely watched CBDC development in the Western world. Its progress has been careful and deliberate — which is another way of saying slow, depending on your perspective.

The ECB's preparation phase ran from November 2023 to October 2025, after a two-year investigation phase. In October 2025 the Governing Council moved the project into a further phase focused on technical readiness, market engagement and support for the legislative process. The ECB has selected potential platform providers and is preparing a controlled pilot, but it has not decided to issue a digital euro.

The proposed Digital Euro Regulation is progressing through the EU legislative process in parallel. Legislation must be in place before an issuance decision. On the ECB's working assumptions, a pilot could begin in 2027 and the Eurosystem aims to be technically ready for a potential first issuance in 2029; those are planning assumptions, not a launch commitment.

The ECB has been explicit about several design features it intends to build in. A digital euro would be a direct ECB liability, accessible to all eurozone residents. It would likely carry a holding limit — a cap on how much digital euro any individual can hold — specifically to address the disintermediation risk discussed below. It would be programmable in the sense of supporting conditional payments in commercial contexts (paying a contractor when a smart contract confirms job completion, for instance), but the ECB has stated it does not intend to programme the digital euro with spending restrictions. Critics note that stated intentions and technical capabilities are different things.

The MiCA carve-out is relevant here: the EU's Markets in Crypto-Assets Regulation explicitly excludes CBDCs issued by the ECB or EU member state central banks from its scope (see MiCA Recital 8 and Article 2). A digital euro would be sovereign monetary infrastructure, not a regulated crypto-asset. For a full treatment of what MiCA covers — and what it does not — see our article on MiCA, the EU's crypto regulation.

The digital pound: consultation complete, decision pending

The Bank of England and HM Treasury have pursued the digital pound — informal nickname "Britcoin" — through a joint consultation process. A consultation paper was published in February 2023, the consultation period closed in June 2023, and the Bank and HM Treasury published their joint response in January 2024, setting out their thinking on design and outstanding questions.

The Bank and HM Treasury are now completing a design phase that ends in 2026. They expect to publish their assessment and decision on next steps later in 2026. No decision has been made to build or issue a digital pound, and any later build phase would still require primary legislation and a further public consultation.

The design questions the UK is grappling with mirror those facing the ECB: holding limits to protect commercial bank deposit funding, offline capability so the digital pound can function without internet connectivity (important for financial resilience and for genuine cash-like access), privacy protections that are meaningful rather than purely rhetorical, and the relationship between a potential digital pound and the existing open banking infrastructure. For background on that infrastructure, our article on open banking and API-driven finance covers how the UK's current account data and payment rails work.

The United States: federal agencies are barred from pursuing a CBDC

The US position on a retail CBDC is the most resistant of any major economy, and understanding why requires understanding the political context as much as the technology.

The Federal Reserve has conducted research into CBDC technology, most notably through Project Hamilton — a collaborative research effort with MIT's Digital Currency Initiative that explored technical architectures for a potential US CBDC. Project Hamilton was a technology research project, not a policy decision: it examined how a high-throughput, privacy-preserving digital currency system could be built, without committing to build one.

The policy position hardened in January 2025. Executive Order 14178 prohibited federal agencies, except where required by law, from taking action to establish, issue or promote a CBDC and directed them to terminate ongoing creation initiatives. An executive order is not the same as a permanent statutory ban and can be changed by a later administration, but under the current federal policy there is no US retail CBDC in prospect.

It is important not to conflate the US CBDC picture with the launch of FedNow in July 2023. FedNow is the Federal Reserve's instant payment rail, enabling 24/7 bank-to-bank transfers settling in seconds. It is emphatically not a CBDC. FedNow operates between commercial banks and their account holders, just as existing payment rails do. The Fed has been explicit on this point, and it matters: FedNow modernises the interbank plumbing without touching the liability structure of the money flowing through it. Consumers still hold deposits at commercial banks; the Fed is not issuing a direct claim to the public.

The BIS and cross-border wholesale experiments

While retail CBDC development has been largely national, the most sophisticated wholesale CBDC work has been international — driven by the Bank for International Settlements Innovation Hub and its national partners.

The most prominent example was Project mBridge, a multi-CBDC platform for cross-border settlements involving the central banks of China, Hong Kong, Thailand, and the UAE, plus the BIS. mBridge built and tested a shared ledger on which central banks could conduct direct bilateral settlements in their respective CBDCs, eliminating correspondent banking intermediaries. The results were technically promising — settlement times fell from days to seconds for participating currencies.

In 2024, the BIS announced it was stepping back from mBridge, stating that the project had reached a stage of maturity to be run by the participating central banks directly. Analysts and commentators widely connected the decision to the project's geopolitical dimensions — the participating central banks span jurisdictions with significantly divergent geopolitical relationships, which created complications for a shared settlement platform at that scale — though the BIS did not publicly frame its departure in those terms.

Other BIS Innovation Hub experiments — Project Jura (cross-border settlement between France and Switzerland), Project Mariana (automated market maker-based FX settlement between France, Switzerland, and Singapore) — have explored different aspects of wholesale CBDC mechanics. These projects are research initiatives, not production infrastructure, but they are building the intellectual and technical foundations for what wholesale CBDC settlement could eventually look like at scale.

Stablecoins versus CBDCs: why the distinction matters

The comparison between private stablecoins and CBDCs surfaces constantly in policy discussions, and it is important to be precise about what the comparison actually involves.

Stablecoins — Tether's USDT, Circle's USDC, PayPal's PYUSD, and the many others competing for institutional and retail market share — are liabilities of their private issuers. They are regulated (or in some cases, not yet regulated) as private financial instruments. In the EU, dollar-pegged and euro-pegged stablecoins fall under MiCA as E-Money Tokens; their issuers must hold EMI licences and maintain reserves. The MiCA framework for stablecoins is detailed in our article on MiCA. In the UK, stablecoin regulation is being developed by the FCA under the Financial Services and Markets Act 2023.

A CBDC is none of those things. It is not a private instrument, not subject to private issuer risk, not covered by the deposit guarantee schemes that protect bank deposits, and not within MiCA's scope. It is sovereign money. The practical implication is that a CBDC carries zero credit risk from the issuer (the central bank cannot default on its own currency in the way a private company can default on its obligations) but carries a different kind of risk: the risk of the state having direct visibility of, and potentially control over, every transaction.

Whether that trade-off is acceptable — lower counterparty risk in exchange for reduced financial privacy — is the central policy question in every CBDC debate.

Programmable money: capability versus intent

One of the more unsettling features of CBDC architecture is programmability. Unlike physical cash, which can be spent anywhere without any technical constraint, a CBDC exists on a digital ledger that could in principle enforce conditions: spending restricted to certain categories of goods, amounts that expire if unused by a certain date, geographic limitations on where the currency can be spent.

The technical capacity for programmability is not a design accident — it is inherent in any software-based currency system. The question is whether central banks and governments choose to use it, and whether they can credibly commit not to use it in the future even if they do not use it today.

Most central banks designing retail CBDCs have stated clearly that they do not intend to programme their currencies with spending restrictions. The ECB has said the digital euro will not have programmable spending conditions attached by the issuer. The Bank of England has made similar statements about the digital pound. These statements are genuine in the sense that the institutions making them do not currently plan to restrict how CBDC holders spend their money.

Critics make a structural argument: that the relevant question is not the current intention of the current institutions, but what the technology enables if institutions or political conditions change. Physical cash offers a privacy guarantee that is embedded in its physical nature — there is no ledger of banknote transactions, no central authority that could reconstruct your spending history, and no mechanism to add spending conditions retroactively. A CBDC does not offer that guarantee by design, only by policy, and policies change.

This is not a fringe concern. It has been raised by civil liberties organisations, by legislators across the political spectrum, and by some economists who are otherwise supportive of digital currency innovation. It is the principal reason that retail CBDC proposals attract more organised opposition than wholesale CBDC work.

Disintermediation: the commercial bank problem

If consumers can hold their savings directly in a central bank-issued CBDC, why would they keep money in a commercial bank account? The answer, for many people, might be "they wouldn't" — particularly in a stress scenario when a commercial bank looks shaky. The ability to move instantly into a CBDC wallet that carries zero bank counterparty risk would make digital bank runs faster and more severe than anything that preceded them.

Even in normal conditions, widespread retail CBDC adoption could shrink commercial bank deposit bases — which are the primary funding source for bank lending. Banks that lose deposits to CBDC wallets have less to lend, or must fund themselves more expensively from wholesale markets. The knock-on effect on credit availability and the cost of borrowing is a serious macroprudential concern, not a theoretical one.

The standard response to this disintermediation risk is to build holding limits into retail CBDC design — capping how much digital currency any individual or business can hold. The ECB has discussed limits in the range of €1,000 to €3,000 per person as a design parameter (without committing to a specific figure). Holding limits constrain the CBDC to a payment instrument rather than a savings instrument, which preserves the commercial bank deposit base. They also limit the financial inclusion benefits of the CBDC, since individuals who want to hold more than the cap in government-backed digital money cannot do so.

The relationship between CBDCs and the broader financial infrastructure — including the Banking as a Service layer through which many fintechs access payment rails and deposit accounts — is an active area of regulatory discussion. Our article on Banking as a Service covers how that layer currently works and the regulatory pressures it faces.

Privacy: the foundational objection

The privacy concern with retail CBDCs is structural, not incidental. Physical cash is anonymous. When you hand over a banknote in a shop, no record of that transaction goes to any central authority. The central bank that issued the note does not know you spent it, where you spent it, or what you bought. This is not a bug in the cash system; it is a feature that has been considered integral to financial liberty in most liberal democracies.

Most proposed retail CBDC architectures, including the digital euro and digital pound designs, do not envisage the central bank having direct access to individual transaction records. The ECB's digital euro framework proposes that payment service providers — supervised private sector intermediaries — handle individual payments, with the central bank's visibility limited to aggregate data necessary for monetary oversight. The Bank of England has similarly stated that the digital pound would be designed so that neither the Bank nor the government would have access to personal transaction data. Critics accept this as the stated intent but make a structural argument: privacy protections built into a CBDC as a matter of policy are different in kind from the anonymity of physical cash, which is a property of the medium itself. Policy can change with legislation or government direction; the physical properties of cash cannot. The architecture creates a surveillance capability that does not exist with banknotes, even if that capability is not currently exercised.

Most central banks have proposed pseudonymity rather than anonymity as their privacy model — identifying CBDC wallets by an ID that is not directly linked to an individual's name, with a legal process required to de-pseudonymise. This offers some protection against routine surveillance while preserving the ability to investigate serious crime. Whether this is sufficient — and whether the legal processes required to de-pseudonymise will remain robust under political pressure — is contested.

Sources and methodology: This article draws on the ECB's October 2025 closing report and 2026 digital euro pilot material; the Bank of England's March 2026 digital pound progress update; Executive Order 14178 and Federal Register publication; BIS Innovation Hub project reports including mBridge, Jura and Mariana; and the respective central banks' public communications on live retail CBDCs. The MiCA carve-out is drawn from Regulation (EU) 2023/1114. No proprietary data is cited.