pay.2nth.ai Tree modes mobile-money
modes · Mobile Money · Leaf

Africa skipped the bank account and went straight to the phone.

Mobile money is e-value held against a phone number, issued by a mobile operator (or licensed e-money issuer), cashed in and out through a network of agents. It is the single most consequential payments innovation the continent has produced — and the reference model the rest of the world is now studying.

Mobile Money M-Pesa E-money float Agents Interop

E-money on a phone number

Mobile money is e-value tied to a phone number and redeemable for cash through agents. It is the wallet Africa actually uses.

Mobile money is a stored-value account identified by a mobile phone number, not a bank account number. You load value by handing cash to an agent (cash-in), send it to any other number, pay merchants and bills, and withdraw it as cash (cash-out) at another agent. The phone — often a basic feature phone over USSD — is the interface; the operator or a licensed e-money issuer is the custodian.

It is deliberately not a bank account. The balance is e-money — a digital claim, fully backed, that the issuer must redeem for cash on demand. The point was to reach people banks could not serve profitably, using rails (the GSM network) and a footprint (airtime resellers) that already existed.

In 2025, Sub-Saharan Africa and North Africa together held around 1.2 billion registered mobile-money accounts — more than half the global total — and Sub-Saharan Africa had the highest rate of account ownership of any region. Global transaction value reached roughly $2 trillion, with East Africa alone processing around $806bn.

The GSMA model and the float

ComponentWhat it doesWhy it matters
E-money issuerRuns the ledger, issues e-value, sets pricing, manages KYC tiersUsually a telco subsidiary (e.g. Safaricom, MTN, Airtel) or a licensed EMI — regulated, but not always as a bank
Trust / float accountPooled bank account(s) holding 1:1 backing for all e-money in circulationThe safeguard: customer funds are ring-fenced from the issuer’s own balance sheet. Who earns the interest on it is a recurring policy fight
Agent networkCash-in / cash-out, registration, sometimes merchant servicesThe physical footprint. ~30m agents globally in 2025; liquidity is the operational core
USSD / app / SIMThe customer interfaceUSSD on feature phones is what makes it universal — no smartphone, no data bundle required

The model the GSMA codified has three moving parts: the issuer (operator / e-money licensee) who runs the platform and holds the float, the agents who do cash-in and cash-out, and the trust account that backs every unit of e-money in circulation.

M-Pesa, MoMo, Airtel Money

M-Pesa (Safaricom, Kenya — 2007)

The archetype. Launched as a way to send money home, it became the rails for an economy. By 2025/26 Safaricom reported on the order of 40m+ one-month-active M-Pesa users and cumulative transaction value past $450bn, with M-Pesa now central to the company’s revenue. The case study everyone cites.

MTN MoMo

MTN’s mobile-money arm, spread across West and Southern Africa (Ghana, Uganda, Cameroon, Nigeria and more). One of the two giants whose scale spans many markets and currencies.

Airtel Money

Airtel Africa’s wallet, strong across East, Central and West Africa. With MTN, anchors the pan-African duopoly outside the Kenyan M-Pesa stronghold.

Why telcos, not banks

Operators already had distribution (airtime agents), a customer identifier (the SIM), and reach into rural areas banks had written off. Mobile money is what happens when distribution beats balance sheet.

From walled gardens to open rails

It breaks the incumbent moat

Once value moves freely across networks, the dominant operator’s “everyone I pay is on my network” advantage erodes. Incumbents resist; regulators push.

It needs a settlement layer

Real-time messaging between wallets is the easy part. Net settlement between issuers — who owes whom, when, in central-bank money — is what makes it actually work and is where instant-payment switches earn their keep.

Bank ↔ wallet is the real prize

Linking mobile money to the formal banking and instant-payment system (and to cards) is what turns a closed wallet into a national payment instrument.

The first decade of mobile money was a set of walled gardens: you could only send easily to people on your own network. That is great for the dominant operator and bad for everyone else — it entrenches incumbents and fragments the market.

Interoperability — the ability to send from one wallet or bank to any other — is the policy battleground of the current decade. It arrives by two routes: scheme-level rails (mobile-money switches, instant-payment systems mandated or built by regulators) and bilateral wallet-to-wallet agreements.

Two different animals

Mobile money optimised for reach and simplicity; banks optimise for credit and balance-sheet services. The interesting products now blur the line — wallets that lend, banks that issue wallets.

Mobile moneyBank account
IdentifierPhone numberAccount / IBAN
BackingE-money, 1:1 in a trust accountBank’s balance sheet (fractional)
InterestUsually none to the user; debatedTypically pays (some) interest
Deposit protectionTrust-account safeguarding, not always deposit insuranceDeposit insurance where it exists
OnboardingMinutes, tiered KYC, often a feature phoneBranch / documents / minimum balance
Credit / lendingIncreasingly, via data-driven micro-loansCore business, but excludes the thin-file majority
ReachWherever there is an agent and a signalWherever there is a branch or smartphone + KYC

Gotchas

It is not a bank deposit

E-money is safeguarded in a trust account, but that is not the same as deposit insurance, and the user usually earns no interest on their balance. Conflating the two oversells the protection and undersells the risk if an issuer fails or the trust account is mismanaged.

Float interest is contested

A billion accounts holding balances generates serious interest income on the trust account. Who gets it — issuer, customer, or a fund — is a live regulatory fight (and a quiet revenue line). Watch the rules in your market.

Dominance becomes systemic risk

When one operator processes a large share of a country’s payments, an outage is a national event and the firm becomes too-important-to-fail. Regulators are waking up to mobile money as critical infrastructure, not a telco add-on.

Registered ≠ active ≠ value retained

Account counts flatter the picture. Many accounts are dormant or pure cash-in/cash-out conduits. The honest metrics are active users and value that stays digital.

Agent fraud and social-engineering

Fake reversal SMS, “send it back, I paid you twice” scams, and agent over-charging are endemic. The trust that makes the network work is also its biggest attack surface.

How to think about mobile money

If you are building on or competing with mobile money

Distribution beats features. M-Pesa won on agent density and a trusted identifier, not on UX. If you cannot match or partner into the agent footprint, you are not really in the market — you are an app on top of someone else’s rails.

Decide early: ride the rails or build your own. In most markets the rational move is to integrate with the dominant wallet(s) and the instant-payment switch, not to recreate a CICO network from scratch. Interop mandates increasingly make this possible — and increasingly mandatory.

Treat the float and KYC tiers as the regulated core. Where the trust account sits, who audits it, who earns its yield, and what each KYC tier permits are the questions that decide whether your product is licensable and survivable.

When mobile money is the right rail

For low-value, high-frequency P2P, bill pay, disbursements, and merchant payments to a population that is phone-rich and bank-poor, mobile money is usually the correct answer — reach and simplicity win.

It is the wrong primary rail when you need rich credit services, large-value B2B settlement, or strong cross-border reach without a corridor partner — there, banks, instant-payment systems, or stablecoin/cross-border rails carry more.

Cost of getting it wrong: building a closed wallet in an interop-mandated market (stranded), or assuming bank-grade protections and interest that the e-money model does not provide (mis-selling, regulatory exposure).

Where this sits in the tree

Primary sources