Only 0.4% Happens On-Chain: The Bank of Italy Measures the Real Cost Structure of Stablecoin Transfers

A central bank that sent the money itself

Published as No. 86 in the Bank of Italy's "Markets, Infrastructures, Payment Systems" series, the study applies mystery shopping, a method in which researchers pose as ordinary customers and execute real transactions, to stablecoin remittances for what is believed to be the first time in a systematic way. Using Italy as a hub, the team sent 200 USDC (roughly 200 US dollars, about 30,000 Japanese yen) in both directions across corridors linking Italy with Argentina, Brazil, South Africa, the United Arab Emirates and Japan. The 200 dollar amount was chosen deliberately to match the World Bank's Remittance Prices Worldwide benchmark. Transactions were executed on March 24 and 26, 2026, using Ethereum as the default network, with Binance and Kraken on the Italian side and local venues on the other: Ripio in Argentina, Foxbit in Brazil, BitOasis in the UAE and Valr in South Africa.

Taking apart the "stablecoin sandwich"

The study's key design choice was to decompose each transfer into five sequential phases: funding a local exchange account with fiat, purchasing USDC, transferring it on-chain, selling it for local currency, and withdrawing to a bank account. The industry describes this structure as a "stablecoin sandwich," in which a cheap blockchain settlement layer is wrapped between two expensive fiat conversion steps. This experiment measured the bread and the filling separately for the first time.

The verdict was unambiguous. The on-chain leg averaged 0.4% across all corridors and fell as low as 0.01% on the Brazil to Italy route. By contrast, the UAE to Italy corridor reached a total cost of 8.95%, largely because the only available funding method was a credit card carrying a 3.8% surcharge. Withdrawal fees also proved operator-specific, charged by some platforms and waived by others. Cost, in other words, is determined not by blockchain performance but by the intensely practical question of which exchange and which funding instrument a user happens to choose.

Domestic payment rails, not blockchains, determine speed

The same pattern appeared in execution times. The on-chain transfer completed within fifteen minutes in seven of eight corridors, confirming that the blockchain is not the bottleneck. The constraint lies on the fiat side. In jurisdictions with instant payment systems, Italy's TIPS, Brazil's PIX and Argentina's Transferencias 3.0, funding and withdrawal each completed in under a minute and the entire end-to-end transfer concluded within twenty minutes. In South Africa, which relies on standard bank transfers at both ends, settlement stretched to one or two business days, erasing any speed advantage. The authors conclude that stablecoin remittances and domestic instant payment infrastructure are complements rather than substitutes.

Japan's paradox: cheap on paper, unusable in practice

The Japanese case most vividly demonstrates that regulatory design shapes outcomes. According to the paper, retail access to dollar-pegged stablecoins in Japan is confined to a single domestic operator, and that operator does not permit direct outbound transfers to foreign exchanges, so the experiment had to route through an un-hosted wallet as an intermediate step. Stringent quantitative limits further forced the transfer to be fragmented into smaller pieces, producing a level of operational complexity incompatible with ordinary retail use. Nominal costs were competitive at 1.6% for Japan to Italy and 1.3% for Italy to Japan, the latter conducted on Solana with the on-chain component at just 0.15%, but these figures are not directly comparable because the transaction chain departed from the baseline methodology. Citing an estimate that Japan's share of global yen-denominated fiat-to-crypto exchange volume collapsed from roughly 60% in the mid-2010s to around 4% in 2024, the authors warn that excessively rigid rules do not extinguish demand but push it offshore and into decentralised finance.

Against incumbents, the result is a draw

Measured against World Bank average sending costs, stablecoins won in every corridor except the UAE. Brazil offered the sharpest contrast at 9.96% versus 2.21%, and South Africa at 15.23% versus 5.44%. Yet in a direct comparison against Wise, a leading money transfer operator, USDC was cheaper in three corridors and more expensive in four. With the global average remittance cost still hovering around 6.4%, well above the sub-3% target set by the G20 and the United Nations Sustainable Development Goals, the fair characterisation is that stablecoins undercut poor incumbent channels but do not reliably undercut good ones. A Federal Reserve note published in March 2026 reached a compatible diagnosis, observing that while the on-chain cost of moving stablecoins is near zero, the meaningful cost structure resides at the on-ramp and off-ramp.

Business Development Insights

1. Both the revenue and the friction sit in the conversion layer, not on-chain. With the blockchain leg compressed to 0.4%, business models premised on faster or cheaper settlement rails have little room to differentiate. The on-ramp and off-ramp, fiat deposits and withdrawals, local exchange rates and withdrawal fees, are simultaneously the profit pool and the principal source of customer friction. Middleware that optimises venue selection and routing, aggregators that bundle multiple local off-ramps, and tooling that makes effective all-in cost visible to corporate treasurers all represent clear commercial openings.
2. Depth of integration with domestic instant payment infrastructure translates directly into competitive advantage. Brazil's PIX, which settles in roughly three seconds on average and reaches more than 90% of the adult population, is what made that corridor both fast and cheap. Japan possesses high-quality instant rails of its own in the Zengin system and Cotra transfers, so operators able to wire stablecoin deposits and withdrawals directly into them hold a structural edge. Partnerships with banks and funds transfer service providers should be treated as core architecture rather than a peripheral concern.
3. The largest value creation lies in designing the off-ramp out of existence. The paper's strongest policy observation is that if recipients could spend stablecoins directly on goods, services, rent or school fees without reconverting to local fiat, the economic advantage would be substantially greater. This is the model the industry calls the "toast" variant, and it points to B2B settlement, cross-border payroll and merchant acceptance, use cases where funds received can be pushed straight into the next payment, as the real target. In Japan, yen-denominated stablecoins such as JPYC and JPYSC have emerged, and the amended Payment Services Act took effect on June 1, 2026, loosening constraints on how trust-type reserve assets may be managed. Whether an economy can be built in which value is received in yen-denominated tokens and spent in the same form will be decisive for commercial success in the Japanese market. At the same time, policymakers are increasingly alert to the possibility that wider use of dollar-pegged stablecoins accelerates currency substitution, a consideration that must be built into any regulatory strategy.

Sources

• CoinPost, "Bank of Italy concludes stablecoin transfers offer no cost advantage" https://coinpost.jp/?p=729183
• Banca d'Italia, "Are Stablecoins Efficient for Remittances? Evidence from a Mystery Shopping Exercise by Banca d'Italia", Markets, Infrastructures, Payment Systems No.86, July 2026 https://www.bancaditalia.it/pubblicazioni/mercati-infrastrutture-e-sistemi-di-pagamento/approfondimenti/2026-086/N.86-MISP.pdf?language_id=1
• CoinDesk, "Bank of Italy research suggests stablecoins aren't necessarily cheaper for remittances" https://www.coindesk.com/business/2026/08/01/bank-of-italy-research-suggests-stablecoins-aren-t-necessarily-cheaper-for-remittances
• Federal Reserve Board, "Payment Stablecoins and Cross Border Payments: Benefits and Implications for Monetary Policy Implementation", FEDS Notes, March 30, 2026 https://www.federalreserve.gov/econres/notes/feds-notes/payment-stablecoins-and-cross-border-payments-benefits-and-implications-for-monetary-policy-20260330.html
• SBI VC Trade, USDC service page https://www.sbivc.co.jp/services/crypto/usdc
• Financial Services Agency of Japan, explanatory materials on the bill amending the FIEA and the Payment Services Act, April 2026 https://www.fsa.go.jp/common/diet/221/02/03.pdf