Remittances connect money earned abroad with spending at home. Mobile money changed how that connection works: a recipient can receive value in a phone account rather than collect every payment at a bank or money transfer counter. Where the payment involves different currencies, foreign exchange remains part of the transaction, even if the recipient sees only a local currency balance.
This is a distinct chapter in the development of foreign exchange trading in Africa. The central question is not how households speculate on currencies. It is how they obtain, convert and use money sent across borders, and how much disappears between the amount paid and the amount received.
Before Mobile Money: Access Was a Physical Problem
Before mobile wallets became established, formal remittance services depended heavily on banks, money transfer outlets and postal networks. A sender’s access to a bank did not mean the recipient had comparable access. Distance, paperwork and charges could make a small family transfer difficult to justify.
Informal arrangements filled some of those gaps. Migrants used friends, transport operators, traders and money dealers to deliver value home. A June 2007 IMF analysis of African remittance channels documented these alternatives and the high costs that discouraged small transfers through formal providers.
The practical problem had several parts. Money needed to reach the right person, arrive when needed and retain enough value to be useful. A cheap transfer requiring a costly collection trip was not necessarily cheap for the household.
Consider a worker sending money for a family expense due the following morning. A lower advertised fee would offer little comfort if the recipient could not reach the collection point in time. This distinction between a service being available and being usable helps explain the appeal of distribution through local agents.
The 2007 Turning Point: Domestic Transfers Before International Exchange
M-Pesa launched in Kenya in March 2007. Its early distribution model used existing businesses rather than requiring a new bank branch wherever customers needed service. Safaricom initially selected 400 of its largest airtime distributors as agents, a rollout documented in CGAP’s research on mobile money agent networks.
The historical distinction matters: moving money between two people within Kenya did not automatically involve foreign exchange. A domestic transfer in Kenyan shillings remained a shilling transaction. Its relevance to international remittances was the receiving network it helped establish.
Once a household could receive and use an electronic balance locally, an international transfer service had another possible destination for a payment. The international service still needed to handle the border crossing and any currency conversion. The wallet addressed the recipient’s end of the process.
What the agent network changed
Mobile money is not simply a banking app. In the model covered by GSMA’s mobile money service definitions, customers can use the service without a traditional bank account, while physical agents connect cash with electronic balances. At deposit, an agent accepts cash and credits electronic value. At withdrawal, the customer exchanges electronic value for cash.
This makes the agent’s role easy to underestimate. The phone carries instructions; the agent handles the physical notes. A wallet balance and cash in hand are related, but they are not interchangeable until someone can complete the withdrawal.
For a hypothetical recipient who needs to pay a cash-only supplier, that distinction is decisive. If the nearest agent cannot meet the withdrawal, the payment has reached the account but has not yet met the household’s need. Digital delivery does not make physical distribution irrelevant.
How Mobile Wallets Became International Remittance Channels
By the 2020s, mobile money had become an established channel for international transfers. Worldwide mobile money enabled remittances increased from $12 billion in 2020 to $34 billion in 2024. Sub-Saharan Africa accounted for more than 70% of the 2024 value, based on the GSMA State of the Industry Report 2025.
Those figures describe international remittance transactions involving mobile money. They do not mean that mobile wallets handled 70% of all remittances into Africa. The denominator matters: the regional share concerns the mobile money channel, not every bank transfer, cash collection or informal payment.
The report also recorded more favorable conditions for inbound than outbound international transfers among surveyed providers. Receiving money from abroad and sending money abroad therefore should not be treated as identical capabilities.
Where the currency conversion happens
A cross-border payment involves people or institutions in different jurisdictions. A cross-currency payment involves different currencies. These often overlap, but not always: a payment across a border within a monetary union may require no customer currency conversion. The distinction, and the separate problems of conversion and settlement, are set out in the BIS analysis of cross-border payment systems.
Consider a simplified transfer funded in US dollars and paid into a Kenyan shilling wallet. The sender accepts a quote, the service arranges the conversion, and the recipient receives the agreed shilling amount. Banks or other payment partners settle the obligations behind that payment.
The recipient does not need to buy shillings directly in the wholesale currency market. Nor does the appearance of shillings in a wallet mean that dollar banknotes have traveled to the local agent.
This example shows why the customer experience and the financial plumbing should be assessed separately. A short phone notification can represent a longer chain of funding, conversion and settlement arrangements. Simplifying the screen does not remove those arrangements.
Transfer Fees Are Only Part of the Foreign Exchange Cost
In the third quarter of 2025, the average measured cost of sending $200 to Sub-Saharan Africa was 8.46%, equivalent to $16.92. That regional benchmark is not a quote for any individual route. The World Bank’s September 2025 Remittance Prices Worldwide report separates transfer costs into the stated fee and the foreign exchange margin.
The exchange margin is the cost embedded in the conversion rate relative to a reference rate. A service can advertise a low fee while offering fewer units of the receiving currency. “Zero fee” is a pricing description, not a promise that conversion costs nothing.
For a useful comparison, hold the sender’s total spending constant. Comparing one provider’s fee-inclusive budget with another provider’s transfer principal gives a misleading result before the exchange rates even enter the calculation.
A worked comparison using fictional rates
Suppose a sender has exactly $200 to spend. The following quotes are invented solely to show the arithmetic; they are not current exchange rates or offers from real providers. Each fee is deducted from the $200 budget before conversion.
| Comparison item | Service A | Service B |
|---|---|---|
| Total sender spending | $200 | $200 |
| Transfer fee | $4 | $0 |
| Amount converted | $196 | $200 |
| Fictional exchange rate | KSh130 per dollar | KSh126 per dollar |
| Wallet credit before withdrawal charges | KSh25,480 | KSh25,200 |
Service A delivers KSh280 more despite charging a visible fee. Its better conversion rate outweighs that charge. Choosing Service B because the fee reads zero would leave the recipient with less money.
Now assume the recipient withdraws the entire payment and either service involves a KSh200 withdrawal charge. Spendable cash becomes KSh25,280 through Service A and KSh25,000 through Service B. The advantage remains, but neither wallet credit was the final cash amount.
For this comparison, the useful measure is:
Effective delivered rate = local currency available after relevant charges ÷ total amount paid by the sender.
If the recipient intends to spend directly from the wallet, use the applicable payment charges instead of assuming a cash withdrawal. Compare the actual intended use, not a cheaper transaction the recipient cannot use.
Formal Remittances Still Depend on Exchange Rate Incentives
A better delivery channel does not automatically persuade people to use it. The conversion rate also matters. Where official and parallel market rates differ sharply, senders may receive very different local currency amounts depending on the channel.
Egypt provides a historical example. Officially recorded remittances fell in 2023, while the gap between official and parallel exchange rates likely encouraged transfers outside recorded channels. Reported official flows rebounded after exchange rates were unified in March 2024, a development documented in the World Bank’s June 2024 remittance assessment.
The analytical lesson is that changes in recorded remittances can reflect a change of channel as well as a change in the amount families send. A rise in official receipts should not automatically be interpreted as an equivalent increase in household support.
This also separates payment technology from currency policy. A well-designed wallet can make an authorized payment easier to receive. It cannot, by itself, make an unattractive conversion rate attractive or settle a shortage of foreign currency.
The broader history of exchange controls and parallel currency markets explains why these incentives persist beyond any particular payment technology. For this article’s purpose, the distinction is straightforward: improving delivery and improving access to currency are connected tasks, but they are not the same task.
What Broader Access Still Requires
Mobile money’s next stage is not simply a larger number of registered accounts. Useful access requires reliable infrastructure, services that can exchange payments with one another, and arrangements that protect users. These are central policy priorities in the IMF’s 2025 paper on digital payment innovations in Sub-Saharan Africa.
Interoperability means that separate payment services can work together. For a recipient, its value is practical: money should not become difficult to use simply because the sender, employer or merchant uses a different network. International interoperability adds another layer because domestic systems must connect across jurisdictions.
Several questions therefore provide a better test of access than account registration alone. Can the intended recipient receive this transfer? Can they use the funds for the required expense? What happens if the payment is delayed or sent to the wrong account? Which organization handles the complaint?
Consider a family receiving school fees. A successful notification is only one checkpoint. The amount must be correct, the money must be available before the deadline, and the school must accept the proposed payment method. If an additional transfer or withdrawal is necessary, its cost belongs in the household’s assessment.
For a recurring payment, repeat that assessment periodically rather than treating the first quote as permanent. The useful comparison remains the same: equal sender spending, the same recipient, the same intended use and a comparable delivery deadline.
Access to Foreign Exchange Is Not Access to Forex Speculation
A remittance conversion meets a payment need. Speculative currency trading takes a position in the expectation that an exchange rate will move favorably. A household receiving money from abroad should not confuse the two simply because both involve foreign exchange.
A separate history, the development of Kenya’s retail forex industry, concerns brokerage access and trading services. That is different from the payment and distribution systems examined here.
The lasting importance of mobile money is better judged through ordinary transactions: whether a family can receive support, convert it on clear terms and use it without an expensive collection process. The strongest measure is not how modern the interface looks. It is how much usable money reaches the recipient, when it arrives, and whether there is a dependable remedy when something goes wrong.