BCEAO Sets an 8,000 CFA Free-Transfer Threshold

From 2 November, West Africa's central bank will require covered interoperable electronic-money transfers to pass through PI-SPI, its shared instant-payment switch. Domestic transfers between individuals will be free while a user's cumulative outbound amount stays at or below 8,000 CFA francs per day with a given participant, according to the BCEAO's 2 October notice.

That changes routing and permitted prices for a defined class of transfers. It is not yet evidence that wallets have lost fee income or that government disbursements arrive faster. Those outcomes require transaction and cost data the announcements do not supply.

Across the eight states of the West African Economic and Monetary Union, a common rail can make different accounts easier to connect. It can shift competition toward the cost of serving an active customer. Neither effect follows automatically from connecting an institution to the switch.

The rule has a boundary

The BCEAO notice makes PI-SPI routing compulsory for the interoperable electronic-money transactions it covers. Its scope includes transfers between a bank or microfinance account and an account at an electronic-money issuer or payment institution. It also covers transfers between electronic-money issuers and payment institutions. It does not say that every transfer made inside a mobile-money network must migrate to PI-SPI.

The free domestic allowance is cumulative per user, per participant and per day. There is no limit on the number of transactions within it. Above 8,000 CFA francs, the notice allows an outbound fee from zero to 0.8% of the pre-tax transfer amount. Receipt is free without an amount or transaction limit. Equivalent tariff conditions for transfers between UEMOA countries start on 1 June 2027.

The wording leaves an operational question. The notice describes a cumulative free threshold but then refers to transfers above 8,000 CFA francs when setting the possible fee. It does not spell out how a provider must charge a transaction that crosses the day's threshold. The implemented schedule and confirmation screen will matter. A cap is not a promise that every provider will charge it.

The BCEAO says the free threshold covers 75% of electronic-money transactions. That is a share of transactions by count, not a share of value, revenue or users. Many small transfers could fit below the threshold while fewer large transfers carry most of the value. Without the value distribution and existing fees on covered routes, 75% cannot yield a credible lost-fee estimate.

Connection is the supply side

In its October notice, the BCEAO reports 175 connected participants and access for more than 38 million people. Its 24 June update counted 80 connected participants, while another 74 institutions were in live testing. The connected count has risen by 95. The access figure describes potential reach through participants, not people who sent or received a PI-SPI payment.

The June update also extended the deadline for effective public service to 30 September for banks and electronic-money and payment institutions, and to 30 June 2027 for supervised microfinance institutions. Those dates show why connection, public availability and use should be measured separately. An institution can pass integration tests before customers find the service in an app or use it repeatedly.

None of those two BCEAO notices gives a comparable series for completed PI-SPI transfers, active senders, failed transactions or the fees actually charged by providers on the covered routes. The 175-participant count therefore supports a claim about network supply. It cannot support one about adoption, and the 38 million access figure cannot be treated as an active-user count.

The economics could still move. Mandatory routing may reduce the need for separate bilateral links between participating providers. A free allowance may make small cross-provider transfers more attractive. But a provider's margin also depends on switch costs, fraud and support expense, any fee it chooses above the allowance, and whether customers shift from existing in-network transfers. Public announcements do not yet quantify those variables. A serious margin test compares like-for-like covered transfers before and after 2 November, including any change in cash-out pricing.

The treasury offers a use case, not a result

The public side of PI-SPI has a firmer operational signal. In an 8 October announcement, the BCEAO says Côte d'Ivoire's treasury regularly sends mass payments through the switch. It says the other public treasuries are being integrated progressively. That is a report of repeated use by one treasury, not evidence of a completed rollout across the union.

For an administration, the relevant outcome is whether a payment batch reaches the right accounts sooner and at lower all-in cost than its former channel. The central bank describes the ability to send and receive salaries, scholarships, subsidies and tax payments across account types and to follow them in real time. It does not identify which Côte d'Ivoire programmes used PI-SPI or publish batch volume, delivery time, rejection rate, reconciliation cost or a comparison with the former payment route. The Ivorian treasury's own PI-SPI page describes authorised participants, but supplies no such performance series.

The treasury's batch is distinct from an individual's person-to-person transfer. It shows the shared rail can carry a public payment flow. It does not prove the free-transfer rule pays for itself or that the same performance will hold across eight governments.

What November can prove

The first useful check after the rule starts is narrow: whether covered transfers actually route through PI-SPI, whether users see the free allowance and lawful fees in provider screens, and whether receipts remain free. The next check is behavioural: completed-transfer counts, active users, success rates and transaction values by route. Only then can a before-and-after comparison test whether the rule moved traffic away from private links or merely changed its path.

Côte d'Ivoire's treasury can supply a separate benchmark. Publish the size and value of its batches, the share completed on time, rejected instructions, recipient complaints and total cost per successful payment against the previous channel. Those figures would turn an encouraging deployment claim into a measurable delivery result.

PI-SPI's 2 November change is a real constraint on how covered transfers are routed and priced. The unresolved question is who uses that opening, at what cost and with what reliability. Until the providers and treasuries publish those measurements, the rule is a test of payment economics, not its verdict.