A CBDC is a direct liability of the central bank, issued digitally. That one fact — public money, not a commercial-bank promise — is the whole point and the whole controversy. The honest question is rarely “can we build it?” but “what problem does it solve that existing rails do not?”
A CBDC is sovereign currency issued digitally as a direct central-bank liability — public money on a device, distinct from both bank deposits and crypto.
A central bank digital currency is a digital form of a country’s sovereign currency that is a direct liability of the central bank — the digital equivalent of a banknote, not of a bank deposit. That distinction is everything: the money you hold in a bank app is the bank’s promise to pay; a CBDC would be the central bank’s own money on your device.
It is not crypto. There is no need for a public blockchain, no floating value, no anonymity by default. It is sovereign currency, centrally issued, with whatever ledger technology the issuer chooses — which may or may not be a distributed ledger.
The motivations vary by country: defending the role of public money as cash declines, improving payment efficiency and resilience, advancing financial inclusion, enabling programmability, or simply not ceding the digital-money future to private stablecoins and foreign systems. Whether a CBDC is the best tool for any of these is exactly the live debate.
Wholesale CBDC is a quiet upgrade to interbank plumbing and is where most credible progress sits. Retail CBDC is the loud, contested one — and the one struggling to find adoption.
| Wholesale CBDC | Retail CBDC | |
|---|---|---|
| Who holds it | Banks and financial institutions | The general public — people and businesses |
| What it replaces | Existing central-bank reserves / RTGS settlement | Cash and, potentially, some bank deposits |
| Main goal | Faster, cheaper, programmable interbank & securities settlement | Public access to digital public money; inclusion; payment choice |
| Controversy | Low — an upgrade to plumbing institutions already use | High — bank disintermediation, privacy, surveillance, state reach |
| Maturity | Many live trials; closest to real deployment | Few live; most retail projects remain pilots or are paused |
Account-based: you prove who you are to transact (like a bank account). Token-based: you prove you hold a valid token (like cash, value-bearing on the device). Token designs are more cash-like and more offline-capable; account designs are easier to integrate and police.
Almost every serious retail design is two-tier: the central bank issues, but commercial banks and PSPs handle onboarding, wallets, KYC and customer service. The central bank does not want to run 50 million retail accounts — and intermediation softens the bank-disintermediation problem.
A CBDC that only works online cannot replace cash where power and connectivity are unreliable — a first-order requirement across much of Africa, and a hard engineering and double-spend problem.
The defining tension. Cash is private; a naively designed CBDC could give the state a record of every transaction. Designs reach for tiered privacy (cash-like anonymity for small amounts, full KYC for large), but the trust question is political, not technical.
Caps on how much CBDC a person can hold, and whether it pays interest, are deliberate brakes to stop deposits fleeing banks for risk-free central-bank money — especially in a crisis.
CBDC can carry conditional logic (expiry, restricted use). Powerful for targeted disbursement; alarming if it becomes control over how citizens spend. Most central banks are careful to disclaim the latter.
A retail CBDC in preparation. The ECB closed its multi-year preparation phase and in October 2025 moved to the next stage — building technical readiness. No issuance decision has been taken; that needs EU legislation. On current timelines a pilot could begin around mid-2027 and issuance, if approved, around 2029.
One of the first live retail CBDCs (2021) — and the cautionary tale. Adoption stayed very low; a CBN official publicly conceded in late 2025 it was “not a rosy story.” Nigeria is reportedly re-evaluating and tilting toward a wholesale focus, with offline / USSD / feature-phone efforts to revive retail use. The clearest evidence that issuing is the easy part and adoption is the hard part.
SARB’s experimentation has centred on wholesale: Khokha 1 (interbank settlement on DLT) and Khokha 2 (DLT for debt issuance/settlement and tokenisation), with further phases exploring interoperability and tokenised assets alongside SA’s banks. The pragmatic, plumbing-first path — no rush to a retail digital rand.
Wholesale and cross-border CBDC experiments are advancing; large-economy retail CBDCs are mostly stuck in pilots, paused, or facing political resistance. Issuance is solved; demand, privacy trust, and bank-stability concerns are not.
In markets that already have fast, cheap instant payments, the marginal benefit of a retail CBDC over what exists is thin — which is exactly why several advanced economies are hesitating. “Because others are building one” is not a use case.
The eNaira proves you can launch a retail CBDC and watch almost no one use it. Without a sharp reason to prefer it over cash, cards and mobile money — and without merchant acceptance — it sits idle.
If risk-free central-bank money is too attractive, deposits leave commercial banks, raising funding costs and, in a crisis, enabling instant digital bank runs. Holding limits and non-remuneration exist to blunt this — and complicate the product.
You can design tiered privacy, but citizens have to believe the state will not surveil or programmatically control spending. In low-trust environments that belief is hard-won and easily lost — and it determines adoption more than any technical control.
Preventing double-spend without a live connection — the property that makes cash work — is an unsolved-at-scale engineering challenge, and the one that matters most for inclusion in low-connectivity regions.
Lead with the problem, then ask whether a CBDC beats the alternatives. If the goal is faster retail payments, a well-run instant-payment system usually gets you most of the way at a fraction of the political and technical cost. Wholesale settlement and cross-border are where CBDC’s marginal advantage is clearest.
Design for adoption, not just issuance. Merchant acceptance, a compelling reason to hold it over existing options, offline capability where connectivity is poor, and credible privacy are what separate a used CBDC from an eNaira-style museum piece.
Treat the two-tier model and holding limits as non-negotiable guardrails. Keeping banks in the loop and capping balances are how you get the benefits without triggering disintermediation or digital bank runs.
The wholesale case (settlement efficiency, cross-border, tokenisation) is the stronger near-term bet — which is why SARB’s Khokha work and Nigeria’s pivot both lean that way. Retail CBDC competes with already-dominant mobile money and must clear a high adoption bar, with offline support as a precondition for genuine inclusion.
A CBDC is not a shortcut around the unglamorous work — agent liquidity, acceptance, trust — that any digital money needs.
Cost of getting it wrong: spending political and technical capital on a retail CBDC that no one adopts (the eNaira lesson), or designing privacy and holding rules so poorly that you either trigger bank instability or forfeit public trust — either of which can poison the well for a decade.