What Nigeria’s eNaira taught engineers about integrating a CBDC into real banking systems

Log in or create a free account to get started.

Read History Go to My Account Log Out Enhanced by Google Opinion What Nigeria’s eNaira taught engineers about integrating a CBDC into real banking systems The eNaira launched in October 2021. By March 2022, colleagues on my team at FCMB had taken the eNaira integration

Listen to this article Share The eNaira launched in October 2021. By March 2022, colleagues on my team at FCMB had taken the eNaira integration live in the mobile app. The plan was to extend coverage across our digital estate, and I built the integration for the web portal, intended for both retail and corporate banking. The web build reached the pilot phase. It was never promoted to production because usage was low. And when we later retested the integration, it failed: not on code, not on connectivity, but because the float account, the pool of eNaira a bank maintains to fund its customers’ wallets, held a balance of zero. There was nothing in the pipe to test with. The product’s failure had become the reason its own infrastructure couldn’t pass a test. We had built a bridge, watched nobody cross, and then discovered the river had dried up underneath it. Most analyses of the eNaira are written by economists from the outside. This one is written from inside the integration layer, where the gap between what a central bank designs and what a market wants becomes a number on a settlement ledger. That number, in our case, was zero.

The eNaira’s architecture is, on paper, respectable. It runs on Bitt Inc.’s Digital Currency Management System, with Hyperledger Fabric, a permissioned distributed ledger, as the underlying transaction network. The IMF’s review of the project’s first year noted that every permissioned node in the system was a CBN computer, which makes the arrangement function much like a centrally managed distributed database. The blockchain was real; the decentralisation was not. The distribution model is two-tier: the CBN mints and holds the liability, and commercial banks act as intermediaries, onboarding customers, funding wallets from their float accounts, and handling the customer relationship. This is the same prudent design the Bank of England sketched for the digital pound and the ECB for the digital euro. For a participating bank, the integration surface looks like this: wallet provisioning gated by the national identity stack (BVN and NIN, mapped to a tiered KYC structure), funding and defunding flows between customer deposit accounts and eNaira wallets, and a float account against which all of it settles. The wallet tiers were conservative by design, from a ₦120,000 balance cap at the bottom tier up to ₦5 million at full KYC, with zero interest at every level and no credit facility. The CBN was explicit that this was a payment instrument, not a savings product. From an engineering standpoint, none of this was exotic. The flows resembled any other wallet integration a Nigerian bank has done a dozen times. Which is precisely the problem worth sitting with: the technology was never the hard part.

Here is the uncomfortable shape of the eNaira when you lay its features side by side with the instrument it was implicitly competing against. USDT, which Nigerians were adopting at one of the highest rates in the world, is dollar-denominated, permissionless, uncapped, lightly identified, and fungible with a global financial ecosystem. The eNaira is naira-denominated, identity-gated, capped, fully surveilled, and fungible with nothing. A central bank’s mandate is risk containment, so every design decision is optimised for what could go wrong rather than why anyone would use it. Caps to prevent deposit flight from banks. Zero remuneration to protect the banking system’s funding base. Identity gates to satisfy financial intelligence reporting. Each decision is defensible in isolation. Stacked together, they engineered out every reason to adopt. The product was the residue left over after risk management finished.

The stated reasons for building the eNaira were financial inclusion, cheaper remittances, and payment efficiency. The sequencing tells a different story. In February 2021, the CBN banned banks from servicing cryptocurrency transactions. Eight months later, it launched the eNaira. That is not a coincidence; it is a substitution play. Nigerians were converting naira to USDT at scale, partly as an inflation hedge, partly to escape a rationed official FX window whose rate everyone knew was fiction. The peer-to-peer USDT market had become something far more dangerous to the CBN than a payments competitor: a public, liquid, real-time counter-quote on the naira that contradicted the official rate every minute of every day. The economy had quietly re-anchored to the parallel rate. The CBN was defending a price that governed nothing except those who received the subsidy.

The market’s revealed demand was a single thing: an exit from the naira into dollar exposure, at the honest market rate, without rationing. USDT was winning because it served that demand. The only digital currency product that could have beaten it was one that served the same demand better and that is the one product the CBN can never build, because preventing flight from the naira is the mandate. The eNaira was therefore not really competing for demand; it was instructing demand to reverse direction. It asked users to flee back into the depreciating asset, at the official fiction of a rate, with surveillance added, to swim against the only current that existed, as an act of policy compliance dressed up as a market product.

A key takeaway from that integration was that you cannot ship a product whose success would violate your mandate. For the eNaira to win, it would have had to do what the market wanted, facilitate movement out of the naira at the honest rate, which is precisely what the CBN exists to prevent. A product built to enforce policy against the market’s revealed preference is not a product; it is a compliance instrument with a UI.

The IMF’s review of the eNaira’s first year (Ree, 2023) found that in any given week, 9 per cent of issued wallets went unused, roughly 14,000 transactions per week across the entire country, and that total transactions since launch, around 802,000, were fewer than the wallets themselves, meaning the typical holder never used theirs even once after opening it. Our ledger and the IMF’s statistics were describing the same silence. Low adoption shelved the integration; the shelved adoption emptied the float; the empty float then broke the integration’s own verification. There is a particular kind of clarity that comes from building infrastructure for a product nobody uses.

The eNaira still technically exists. So does our work. Somewhere in FCMB’s infrastructure there is a mobile integration that went live in March 2022 and, judging by the float account it was meant to draw on, found almost nothing to carry, and a web integration stranded in pilot, built carefully, unable even to pass a retest, because the test requires money in a float that adoption never filled. A bridge to a permissioned blockchain, waiting for a river. A small monument to the difference between issuing money and earning demand for it.

Olukowi Oladimeji Ayodeji is a backend engineer specialising in fintech infrastructure, with experience spanning core banking integrations, CBDC implementation, and payments systems in Nigeria.