Dollar access without dollar accounts

Executive summary

The price a dollar token has to beat is a bank price. In the World Bank's Remittance Prices Worldwide for the third quarter of 2025, the global average cost of sending $200 was 6.36%, down from 6.49% in the first quarter, and the average for Sub-Saharan Africa as a receiving region was 8.46%, the most expensive of any region. Banks charged 14.99% on the same measure, against 4.72% for money transfer operators and 4.59% for digital services. Of the thirteen corridors costing more than 20%, nine originated in Sub-Saharan Africa, and remitting from South Africa cost 15.65% on average.

That price has a supply-side cause the BIS has documented. Bulletin No 87 of 30 May 2024, by Garratt, Koo Wilkens and Shin, shows active correspondents and corridors falling between 2011 and 2022, with Africa recording the largest regional decline on its chart, at roughly 40%, and concludes that emerging markets are particularly affected. The BIS press release of 13 December 2021 put the global fall in correspondent banking relationships at about 25% between 2011 and 2020, on SWIFT payment-message data covering more than 200 countries.

Stablecoins have grown into that gap, and in one currency. The BIS Annual Economic Report of 24 June 2025 notes that over 99% of stablecoins are dollar-denominated, that they offer foreign-currency access to those without dollar accounts, and that the result in vulnerable economies is what it calls stealth dollarisation. Its verdict is that stablecoins fall short on the three tests of singleness, elasticity and integrity and "may at best serve a subsidiary role".

The conclusion the evidence supports is narrower than either the promotional or the alarmist reading. Nigeria's currency appreciated and its reserves rose while onchain volumes grew, so the flows are not only a hedge against depreciation. And nobody, the IMF included, can yet size the payment use separately from the trading use.

Key findings

1. The cost gap the tokens exploit is a size and channel gap, not a technology gap. In the third quarter of 2025 the global average cost of a $200 transfer was 6.36% but of a $500 transfer only 4.08%, and the bank channel cost 14.99% against 4.72% at money transfer operators. Sub-Saharan Africa's 8.46% receiving average is where flat costs bite hardest on small amounts.

1. The dominant instrument is one dollar token on one chain. On 25 July 2026, $89.05bn of stablecoin supply sat on Tron, about 28.5% of the $312.14bn DefiLlama tracked across 123 stablecoins. Tether reported $186.45bn of liabilities relating to issued tokens at 31 December 2025 and $183.44bn at 31 March 2026, the first quarter-on-quarter fall in that series, of about $3.0bn.

1. Regional volumes are large and are not payments statistics. Chainalysis reported on 10 September 2025 that Sub-Saharan Africa received over $205bn of onchain value between July 2024 and June 2025, up 52% year on year, with Nigeria at $92.1bn and South Africa at about $30.7bn, and that over 8% of value transferred in the region was in amounts below $10,000, against 6% globally. Monthly volume reached nearly $25bn in March 2025.

1. Local-currency tokens have not appeared at any scale. Nigeria's regulated naira token cNGN had about NGN 2.3bn outstanding across roughly 4,805 wallets on 12 June 2026, more than eighteen months after launch. Against $92.1bn of value received in the twelve months to June 2025, a naira-denominated alternative does not yet exist in any economically meaningful size.

1. The measured remittance corridor is a fraction of the onchain number and moving differently. World Bank data put Nigeria's personal remittance receipts at $22.13bn in 2024, up from $19.55bn in 2023 but still below the 2018 peak of $24.31bn. The $92.1bn Chainalysis figure is not the same object, and no public dataset separates the payment part from the trading part.

Analysis

Set the payment economics out plainly and the source of the saving is obvious. A stablecoin corridor has three legs: the sender converts local currency into a dollar token at a bid-offer spread, the token moves onchain for a network fee measured in cents on the chains these corridors use, and the recipient converts back at another spread. The middle leg is where cost collapses towards zero. The two ends are where cost survives, and because they are priced as proportional spreads rather than as the flat correspondent-banking and compliance charges embedded in a bank wire, the saving is largest precisely where Remittance Prices Worldwide records the highest cost: small amounts, thin corridors, bank-originated transfers. The report's own numbers make the point without reference to tokens: the global average falls from 6.36% at $200 to 4.08% at $500 because much of the cost is fixed per transaction.

The economics therefore differ for business payments. A six-figure supplier payment pays the same two spreads and the same trivial network fee, so the proportional saving is smaller and what is bought is settlement time. Thunes and Yellow Card announced on 28 October 2025 an arrangement connecting Yellow Card's infrastructure, credited in the release with about $6bn of historical volume of which 99% is attributed to stablecoins across 34 countries including 20 in Africa, to Thunes' payout network across more than 130 countries and 80 currencies. The stablecoin leg carries value between two payout networks; it replaces neither.

The demand story is usually told as dollar scarcity, and the Nigerian evidence complicates it in a useful way. The IMF's 2026 Article IV staff report, completed on 7 May 2026, records gross international reserves rising to $46bn in 2025 from $40bn at end-2024, the naira appreciating 10% year on year against the dollar in March 2026, inflation nudging up to 15.4% year on year in the same month, and greater exchange rate predictability having reduced risk premia and strengthened investor confidence. That is a currency stabilising, not one in flight. Yet onchain volumes rose 52% across the region in the year to June 2025. The staff report addresses the point directly, noting that growing stablecoin use for cross-border transactions brings both opportunities and risks and that increasing use of dollar-denominated stablecoins raises risks to monetary sovereignty, capital flow management, financial stability and financial integrity.

Read together, the most defensible interpretation is that dollar scarcity opened the corridor and payment utility keeps it open. Once an importer or a freelancer has learned to settle in a dollar token that clears in minutes at any hour, a stabilising exchange rate removes one reason to hold it and none of the reasons to use it. The BIS framing of stealth dollarisation captures the risk correctly, but the mechanism is now settlement convenience as much as store of value, and the policy answer to the two is not the same.

That distinction is visible in what the regulators are doing. Kenya's approach is licensing. The Virtual Asset Service Providers Act, assented to by President Ruto in October 2025, was completed by regulations gazetted on 24 July 2026 as Legal Notice No 134 in Gazette Supplement No 185. The Central Bank of Kenya takes virtual-asset-to-fiat conversion and stablecoin issuers; the Capital Markets Authority takes exchanges, token issuance platforms, initial coin offerings and tokenisation. Obligations include governance and capital requirements, customer due diligence, asset safeguarding, seven-year record retention and periodic reporting, and the rules reach foreign providers targeting Kenyan customers without a local presence.

Nigeria's approach is to treat the tokens as monetary instruments. The Central Bank of Nigeria's Payments System Vision 2028 refers to stablecoins repeatedly, and its stated intention is to license fully fiat-collateralised stablecoins with 100% high-quality reserve backing, daily attestations, monthly audits and continuous supervisory visibility through observer nodes on approved networks, with a minimum share of the reserves behind foreign-currency tokens held domestically at licensed banks. A supervisory pilot for virtual-asset service providers began on 31 March 2026. That domestic-reserve condition is the clause to watch: it converts an offshore dollar liability into a partly onshore one, and issuers whose reserves sit in US Treasury bills and US banks will resist it.

What the evidence cannot settle is whether the corridor is net stabilising. The clearest formal work is IMF Working Paper WP/26/144 of July 2026 by Brandon Joel Tan, a calibrated model rather than an estimate from African data. It finds stablecoin access raising welfare by about 1.2% in consumption-equivalent terms when exchange rate misalignment is low and reducing it by 6.3% when misalignment is high, the sign reversing at a threshold of about 0.59, and average crisis exposure rising from 3.9% in a cash-only baseline to 7.4% in a full stablecoin economy and 12.9% at maximum misalignment. Its motivating case is Bolivia, where virtual-asset transactions multiplied twelvefold between July 2024 and May 2025 and the USDT-boliviano rate became the reference parallel price. The conditional result matters more than the magnitudes: the same instrument helps when the official rate is roughly right and accelerates the run when it is not.

Methodology and data

Cost figures are from Remittance Prices Worldwide, Issue 54, and refer to the third quarter of 2025 unless stated; the database prices $200 and $500 transfers through named providers and averages them by receiving region, sending country and instrument type. Correspondent banking figures are from BIS Bulletin No 87 and the CPMI data release of 13 December 2021, which draws on SWIFT payment messages from more than 200 countries and jurisdictions.

Stablecoin supply figures are DefiLlama aggregates read on 25 July 2026: $312.14bn across 123 tracked tokens, of which $89.05bn was issued on or bridged to Tron. Tether's quarterly figures are liabilities relating to issued digital tokens as reported in its attestations, not audited financial statements. Nigeria's remittance receipts are the World Bank personal-remittances-received indicator, annual, latest populated year 2024.

Regional volume figures are Chainalysis estimates of onchain value received, geolocated by service attribution, for July 2024 to June 2025. Macroeconomic figures are from Nigeria's 2026 Article IV staff report. Regulatory detail for Kenya and Nigeria comes from reporting of the gazetted regulations and the CBN blueprint, dated 24 July 2026 and 15 June 2026.

Limitations

Corridor-level stablecoin volume does not exist in public data, and that is the largest gap here. Chainalysis's value received attributes onchain activity to services and services to countries by a proprietary method; it counts exchange-internal and trading transfers alongside payments, cannot be netted, and is not a balance-of-payments figure. No claim that stablecoins have displaced a measurable share of a named corridor is supportable from published data, and this piece makes none.

The two prices in the comparison are not measured the same way. Remittance Prices Worldwide surveys named providers on a fixed methodology and does not cover crypto channels. Providers, for their part, do not publish price lists: Yellow Card's site directs pricing enquiries to a sales contact rather than a fee schedule, so the round-trip cost of a stablecoin corridor, meaning both conversion spreads plus the network fee, cannot be computed from public sources. The $6bn volume figure is self-reported, cumulative, and given without a start date.

The regional correspondent-banking declines in BIS Bulletin No 87, including the roughly 40% figure for Africa, are read from a chart rather than a table and are approximate. The 25% global figure from the 13 December 2021 release is published but stops at 2020.

The causal question is open. Nigerian volumes grew while the naira appreciated and reserves rose, which cuts against a pure dollar-scarcity explanation without establishing the payment-utility one, since the same period saw a general rise in global crypto activity. Distinguishing the two would require transaction-level data on purpose and counterparty that no provider or regulator publishes. The CBN's own priorities include building blockchain analytics for naira-to-stablecoin transactions, an admission that the supervisor lacks this data too.

Finally, WP/26/144 is theory. Its welfare and crisis-exposure numbers are simulation outputs under one calibration and its case study is Bolivia; nothing in it estimates Nigeria, Kenya or Ghana, and citing its magnitudes as measured effects for African economies would misuse it.