How AfriPay stops competing on price and starts owning the category
A path to the only app in our markets where a customer holds money in the currencies they actually use, funds a send from a bank account or their own crypto wallet, and spends from a card, on infrastructure Cashpot owns. Bvndle sits in it as a wallet of its own.
How we got here
Who signs this memorandum
Kelvin Orogun
Chief Executive, Cashpot
Kabelo Lehlongwane
Head of Group Technology, Products and Platforms, VFD Group
Nonso Okpala
Group Managing Director, VFD Group
What VFD gets out of this
Bvndle becomes a wallet inside this app, in four markets, on rails we build once.
The instrument model AfriPay needs in order to hold several currencies is the same model Bvndle needs in order to be funded at all. A tag is only useful if something can pay into it. Every wallet added here is one a tag can draw on, with no second integration and no separate build.
This is not a new line of spend. The linked-instrument work is inside Stage Now, already co-funded, already on the critical path for the corridor. Bvndle inherits it.
- Distribution in four marketsEvery AfriPay customer in Nigeria, the UK, South Africa and Ghana becomes reachable by a Bvndle tag.
- A funding layer, finishedBank account, card, crypto wallet and balance, all linked to one verified identity.
- The Tag Payments substrate, liveThis is the group identity and payments layer proved on a real corridor rather than in a pilot.
Product roadmap
Three stages, each one earning the right to the next.
AfriPay today moves money into Nigeria on a rented engine, under a licence the CBN no longer issues to companies like ours. That scarcity is an advantage with a ceiling on it: rented technology caps what the licence can be worth. Every stage below is a decision on its own, and nothing in the later stages is committed by approving the first.
| Stage | What ships | What it runs on | What it needs from a regulator |
|---|---|---|---|
| Now0 to 9 months | The UK to Nigeria path, live and earningThe business as it stands, on an engine we own | The existing app, re-pointed to a Cashpot-built engine as the rebuild lands | Nothing new. The IMTO licence Cashpot holds already covers it |
| Next9 to 24 months | A new app: crypto enabled, multi-wallet readyOne identity, many balances, many funding sources | One ledger, a stablecoin settlement core, a named counterparty in each market | Nigerian SEC registration and a UK cryptoasset permission |
| Target24 months and beyond | Payments in local currency both ways, with virtual cardsThe category position nobody else holds | The same engine, with a licensed partner per country and a card programme on top | A licence per market per capability, and card issuing |
The hard word in the target state
Both ways. Nigeria's IMTO licence permits inbound transactions only, and requires beneficiaries to be paid in Naira. The right to send from a market is something this plan acquires over three years, never something it assumes, and no amount of technology substitutes for it.
App architecture
One app over many wallets, with a settlement core that does not grow with the map.
Every balance in the app is a claim on a named entity in a named country. The app is a surface, not a licence, which is exactly what lets one product span four regulators.
One app
One verified identity, many linked instruments: bank, card, crypto wallet, Bvndle tag
Cashpot engine
One ledger, one set of checks, one operations surface
Settlement core
USDT in a Fireblocks vault, policy bound, one pooled buffer for every corridor
Local leg, per market
Kraken in the UK. A licensed counterparty in Nigeria, South Africa and Ghana
The settlement core is what makes many markets affordable. On correspondent banking each corridor needs its own pre-funded account, sized for its worst day and stranded where it sits. A pooled stablecoin buffer serves every corridor and moves in minutes, so working capital stops scaling with the number of countries. That is the argument for the crypto layer, and it is a treasury argument, not a product one.
Three constraints shape the build, and all three are regulatory rather than technical. A Nigerian wallet is a V Bank account surfaced in the app, because holding a customer balance is not international money transfer. A Nigerian crypto entity may register in only one SEC category and may not combine two. And Kraken settles the UK leg only: it supports no Naira, no Rand and no Cedi, so each African leg needs its own contracted counterparty.
Timeline
Two regulatory dates are fixed. Everything else sequences behind them.
Nigerian VASP registration ahead of 30 June 2027, and the UK cryptoasset regime commencing 25 October 2027. Both are outside our control, so the plan works backwards from them and puts the unblocking work first.
| Workstream | Starts | The date it must clear | Why it sits here in the order |
|---|---|---|---|
| Engine rebuild, both rails | Now | None | Needs nobody's permission and sits under every market |
| Naira counterparty | Now | None | Unblocks the corridor economics and the SEC category choice |
| Linked instruments | Now | None | The multi-wallet foundation, and Bvndle's funding layer |
| Managed corridor test | Q4 2026 | None | Proves the numbers before any capital is committed |
| Nigerian SEC registration | Q4 2026 | 30 June 2027 | Category decision first. Approval in principle is available through the incubation route |
| UK cryptoasset permission | Q1 2027 | 25 October 2027 | Registration under the money laundering rules applies in the meantime |
| South Africa and Ghana | Q3 2027 | Partner licensed | Receive-first markets, added once the core is proven |
| Virtual card issuing | 2028 | Sponsor bank per region | The last layer, and the least scoped |
Working weeks from agreement in principle, not calendar commitments. Rebuild timings assume no third-party dependency stretches them.
Cost
About one to two million dollars of spend over three years, not twenty.
The figures below are best estimates, marked as such, so the order of magnitude is decidable now rather than after three more quotes. The distinction that matters is spend against capital: regulatory capital is locked on the balance sheet, not consumed.
What you are approving now
$180k – $315k
Stage Now in full. The engine rebuild, infrastructure and tooling. No new licence, no locked capital.
Whole programme, spend
$1.0m – $2.2m
All three stages over roughly three years, gated stage by stage.
Whole programme, capital locked
$0.5m – $2.2m
Regulatory capital and collateral. Held, not spent, and returnable if a licence is surrendered.
| Line | Best estimate in US dollars, and what the amount buys |
How the figure was arrived at |
|---|---|---|
| Stage Now · spend | ||
| Engine rebuildengineers and delivery | $150k – $250k | EstimateThree to four people over eight months at fully loaded rates, on the AI-assisted cadence already in use |
| Infrastructureservers, hosting, failover | $20k – $40k | EstimateYear one, for a regulated payments workload |
| AI and agent tooling | $10k – $25k | EstimateModel hosting and ops tooling, year one |
| Stage Now total | $180k – $315k | No capital locked. The IMTO licence is already held and already paid for |
| Stage Next · spend | ||
| New multi-wallet appmobile, web, back end | $300k – $500k | EstimateFour to six people over nine to twelve months, including design and security testing |
| Fireblocksorchestration, year one | $40k – $120k | EstimateSubscription rather than per transaction. Public plans start near $36k; payments use cases price above that |
| Censeprovenance and AML, year one | $25k – $75k | EstimatePer check or subscription, to be quoted |
| Nigerian SECfees and counsel | $50k – $85k | EstimateRegistration and filing fees of roughly ₦30m, plus Nigerian counsel |
| UK authorisationcounsel and compliance build | $125k – $350k | EstimatePayments or e-money permission plus the cryptoasset permission. Range reflects which permissions are sought |
| Treasury controls and operations | $50k – $100k | EstimatePolicy, reconciliation and exception handling for an owned vault |
| Stage Next total | $590k – $1,230k | Before capital, which is set out below |
| Stage Target · spend | ||
| South Africa and Ghanatwo partner integrations | $60k – $160k | EstimateIntegration and onboarding per market, partner terms on top |
| Virtual card programmeset-up | $100k – $300k | EstimateSponsor bank, programme manager and scheme onboarding. The least scoped line here |
| Additional compliance and operations | $75k – $150k | EstimateHeadcount and tooling as markets are added |
| Stage Target total | $235k – $610k | Partner-dependent, and re-estimated once Stage Next lands |
| Regulatory capital · locked, not spent | ||
| Nigerian IMTO | Already held | Held today$1m operating capital plus ₦10m annual renewal. Published, and already on Cashpot's balance sheet |
| Nigerian SEC registrationthe biggest single variable | $225k – $1.50m | Estimate₦300m at the ancillary tier, ₦2bn if the activity is classed as an exchange, converted at ₦1,331 to the dollar on 18 September 2026 |
| UK initial capital | $150k – $400k | EstimateScales with the permission sought and with transaction volume |
| Card collateral and partner floats | $100k – $300k | EstimateScheme collateral, Stage Target only |
| Capital total | $475k – $2.20m | On the balance sheet, not through the profit and loss |
Where the range actually comes from
One decision moves the number more than everything else combined. Registering in Nigeria at the ancillary tier locks ₦300m, about $225,000. Being classed as an exchange locks ₦2bn, about $1.5m. That single classification is a $1.28m swing, and settling it early is worth more than sharpening any other line on this page. Kraken, for what it is worth, charges no licence or account fee at all: the venue costs a trading fee of 0.20% falling to 0.04% at volume, and nothing else.
The ask
Approve Stage Now, and preserve every option after it.
Agreement in principle, so the work that needs nobody's permission can start while the licensing runs alongside. This commits $180,000 to $315,000 and no regulatory capital.
- One rebuild carrying both rails. Cashpot builds, owns and operates. VFD contributes architecture, shared build costs, project management and governance. No share of the licence, no change to the regulatory perimeter, no new filing.
- Three workstreams start now. The engine rebuild, selection of a Naira counterparty, and the linked-instrument model that carries AfriPay's multi-wallet and Bvndle's funding layer in a single build.
- The baseline pack within four weeks. Working capital held today, settlement timing and all-in cost per send, so the estimates on this page are replaced by our own numbers.
- A decision on the Nigerian SEC category by year end. It is the largest variable in the plan and the cheapest one to settle.
The Nigerian SEC registration and the UK cryptoasset permission run on their own clocks and are not conditions of any of the above.
Agreed in principle
“Be fearful when others are greedy, and greedy when others are fearful.”
Nonso Okpala, Group Managing Director, VFD Group
Kelvin Orogun
Chief Executive, Cashpot
Kabelo Lehlongwane
Head of Group Technology, Products and Platforms, VFD Group
Nonso Okpala
Group Managing Director, VFD Group
Signature blocks are for wet signing unless a signatory is signed in.