A Central Bank Sent 200 Real Remittances in USDC. The Off-Ramp Ate the Savings.

Researchers at Banca d'Italia did something the stablecoin debate badly needed: they became customers. In a mystery-shopping study published in late July as Paper No. 86 in the bank's Markets, Infrastructures and Payment Systems series, they sent $200 of USDC across ten corridors and measured what actually arrived on the other side.

The corridors connected Italy with Argentina, Brazil, South Africa, the United Arab Emirates and Japan, running in both directions for most pairs. Transfers were executed on 24 and 26 March 2026. The $200 figure was not arbitrary — it matches the World Bank's Remittance Prices Worldwide standard, which makes the results directly comparable with conventional providers.

The transfer was free. The doors were not.

Here is the finding that should reshape how these products are marketed: the blockchain leg — the part crypto marketing loves to quote — accounted for almost none of the total cost. What drove the spread was the on-ramp and the off-ramp: converting local currency into a token at the start, converting the token back into spendable money at the end, and the exchange-rate spread applied at each door.

So “stablecoins are cheaper” turns out to be corridor-dependent rather than structurally true. USDC won decisively on Brazil to Italy, at 2.21% against Wise's 4.68%–4.89%. It lost in four corridors: Argentina to Italy, Italy to UAE, UAE to Italy, and Italy to Brazil. South Africa carried the highest World Bank average in the sample at 15.23%, and USDC came in at 5.44% — a genuine improvement, and a reminder that the biggest gains appear where the incumbent system is worst.

The conclusion buried in the paper

The authors note that the single largest efficiency gain available would come from recipients being able to spend stablecoins directly in the real economy. That would remove the fiat conversion step entirely and eliminate the off-ramp cost which currently drives the bulk of total transfer costs.

Read that again with a payments hat on. The paper is not arguing that stablecoins fail. It is arguing that a transfer ending at a cash-out counter has an expensive final metre, and that the fix is to delete the counter — to let the received balance be spent as it is, at a shop, through a card or a QR code, rather than converted first.

Regulation shows up in the price

The study identifies regulatory design as a primary factor in whether stablecoin transfers are practical and affordable for ordinary users. Highly restrictive frameworks raise costs, reduce the number of available operators, and push users toward offshore platforms. Outright prohibitions do not remove demand; they redirect it into unregulated channels where it is harder to supervise and more expensive to access.

Governor Fabio Panetta reached a compatible verdict in May, arguing that stablecoins can help in selected corridors but are no universal cure for expensive remittances, and that the real work lies in upgrading domestic instant-payment rails and interlinking fast-payment systems across borders.

Context worth keeping

The World Bank's Q1 2026 data puts the global average cost of sending $200 at about 6.0%, down from 6.2% a year earlier — still double the 3% target set by the G20 and UN Sustainable Development Goal 10.c for 2030. That average also flatters reality: it is pulled down by high-volume corridors such as US to India at roughly 2.5%, while parts of Sub-Saharan Africa remain in double digits.

The likely outcome the paper sketches is not one system winning but two coexisting: regulated corridor traffic drifting toward bank-interoperable multi-currency ledgers, while peer-to-peer stablecoin use holds firm in high-inflation economies where families will accept a high offshore cost rather than watch savings erode in local currency.