Kemi runs a Lagos fabric import business. Her Ghanaian supplier wants payment in cedis. Until February 2026, that naira left Nigeria, became dollars in New York, and arrived in Accra as cedis. The route cost 6 to 8 percent in correspondent bank fees.
This is not an edge case. This is how most intra-African cross-border payments still work.
Africa built the world's most advanced domestic mobile money system. Two trillion dollars moved through African mobile wallets in 2025. The continent hosts 52 percent of all mobile money accounts on Earth. When money tries to cross a border, it still routes through New York in most corridors.
The correspondent banking problem is structural, not technical. African currencies are not priced directly against each other in liquid markets. Converting naira to cedis requires pricing each against the US dollar separately. Two FX spreads, two transaction fees, and settlement delays measured in days. All for a transaction that never needed to leave the continent.
The cost shows up clearly. Sub-Saharan Africa remains the most expensive region for cross-border payments. The average cost is 8.78 percent per transaction, against the UN target of 3 percent. In the South Africa-Zimbabwe corridor - one of the continent's highest-volume remittance routes - costs average 12.7 percent. Southern Africa as a whole runs between 12 and 25 percent.
The aggregate loss is significant. Africa's cross-border payment market reached $329 billion in 2025. It is tracking toward $1 trillion by 2035. FX liquidity gaps cost African businesses an estimated $5 billion annually. That cost is not in fees. It is in trade that does not happen because the math does not work. A manufacturer in Lagos cannot price a payment to a supplier in Dar es Salaam with certainty. The order does not happen.
The domestic contrast makes the cross-border failure hard to explain away. Within Kenya, M-Pesa settled KSh 41.68 trillion in the year to March 2026. Within West Africa, MTN MoMo processed $500 billion in a year. Those numbers represent money moving cheaply, instantly, and reliably between accounts on the same operator network. The moment a payment crosses a border, it exits the operator's settlement system. It re-enters the correspondent banking infrastructure that Africa's mobile money revolution was supposed to bypass.
Two infrastructure layers are being built in parallel. They address different parts of the same problem.
The first is the interoperability layer. Onafriq connects over one billion mobile money wallets across 43 African countries. It also reaches 500 million bank accounts through more than 2,000 cross-border payment corridors. In February 2026, Onafriq and PAPSS activated the first wallet-based corridor between Nigeria and Ghana. A naira in an MTN Mobile Money wallet in Lagos now arrives as cedis in Accra. Settlement is near real-time. No dollar conversion step. The Central Bank of Nigeria approved a six-month pilot for individuals and SMEs.
Before this integration, routing between MTN Ghana and M-Pesa Kenya required multiple operator accounts and manual conversion steps. Onafriq's network layer handles that friction without exposing it to the user. The technical work of messaging, routing, and currency conversion is absorbed at the infrastructure level.
The second is the settlement layer. The Pan-African Payment and Settlement System, known as PAPSS, launched in January 2022. It is backed by the African Union and Afreximbank. PAPSS now connects 17 countries, 14 national switches, and over 150 commercial banks. It settles in local currencies against a multicurrency reserve pool. Where PAPSS is live, an intra-African payment can clear and settle entirely in local currencies. The dollar system is not in the chain.
The critical gap is that most active corridors have one layer but not both. Onafriq provides messaging and routing. PAPSS provides settlement. A corridor without both still needs dollar intermediation somewhere in the chain. The Nigeria-Ghana corridor matters because it is one of the first to run both layers together at commercial scale.
The operator positioning is hardening fast. The territorial logic is becoming clear.
MTN MoMo holds 69.5 million monthly active users. It processed $500 billion in transaction value in 2025. MTN has positioned itself as the corridor builder for West and Central Africa. Its wallet presence spans multiple countries within the same trade networks. M-Pesa controls East Africa at depth, with 40.99 million active customers and KSh 41.68 trillion in annual transaction volume. Safaricom's strategy has shifted toward service layers - credit, savings, merchant payments - rather than outbound corridor expansion. Airtel Money, with 54.1 million customers, fills corridors where MTN and Safaricom have gaps across East and Central Africa.
None of them has solved compliance at corridor level. Each new cross-border corridor requires a licensed entity in both jurisdictions. It requires AML monitoring across two regulatory regimes. It requires FX reporting to both central banks. This is not a technology problem. The technology to move money between wallets in near real-time exists and has existed for years. The compliance architecture to activate a new corridor legally, quickly, and at scale does not exist.
Every new corridor requires months of regulatory negotiation, licensing fees, and capital requirements in the destination market. The Nigeria-Ghana pilot is a six-month trial, not a permanent commercial product. The reason is compliance, not code.
For African builders working in payments infrastructure, the territorial map is already settled at the wallet layer. MTN, Safaricom, and Airtel have distribution that cannot be replicated from scratch. The structural gap is in the compliance and settlement middleware. This is the layer that would allow a licensed operator to activate a new corridor in 90 days, not 18 months. It does not exist at scale on the continent.
The corridor that costs 12 percent and three days today will eventually run at 1.5 percent and settle in minutes. That shift will not come because a better wallet is built. It will come when someone builds the compliance architecture that makes corridor activation cheap, fast, and repeatable. African builders who understand the regulatory layer - not just the technical layer - are the ones positioned to close that gap.
