A Bank of Italy field test has challenged the idea that putting a remittance on a blockchain automatically makes the full transaction cheaper or faster. Across the routes it tested, the central bank’s researchers found no systematic cost advantage for transfers using the USDC stablecoin. Total costs ranged from 0.30% to nearly 9% of the amount sent, while the on-chain leg contributed only a small share.
The findings put the accountability focus on the businesses and payment systems surrounding the blockchain. A customer still has to fund an exchange account, buy USDC, transfer it, sell it and withdraw local currency. Exchange pricing, spreads, payment-card charges and the quality of domestic bank-transfer rails can therefore outweigh inexpensive blockchain settlement.
The Bank of Italy published the study on July 30. Researchers Alberto Di Iorio, Enrica Di Stefano, Michele Mascioli and Giorgio Trebeschi used a mystery-shopping exercise involving transfers of 200 USDC between Italy and Argentina, Brazil, South Africa and the United Arab Emirates. Japan was examined separately because local restrictions prevented a directly comparable transfer chain. The transactions in the main exercise were conducted in March 2026; comparisons with Wise were simulated in April.
Low network fees did not guarantee a low customer price
The study separated each transfer into funding, stablecoin purchase, blockchain transfer, sale and fiat withdrawal. It found that the blockchain leg was not the main cost driver. In the researchers’ aggregate assessment, the on-chain transfer accounted for about 0.4% of total costs, leaving fiat conversion and intermediary pricing to determine most of what a sender paid.
Results varied sharply by route and direction. The Italy-to-Argentina transaction cost 0.30%, but the reverse route cost 8.96%. The researchers said the unusually cheap outbound result reflected Argentina’s exchange-rate conditions rather than an inherent technical efficiency. Transfers involving the UAE cost 7.20% from Italy and 8.95% in the reverse direction; the UAE-to-Italy transaction had to be funded by payment card because the mystery shopper did not have access to a bank account at the time.
A corridor-level comparison with Wise was mixed. USDC was cheaper in three of the seven routes for which the paper listed both usable figures, but more expensive in four. The stablecoin route cost 2.21% from Brazil to Italy, compared with a Wise simulation of 4.68% to 4.89%. In the opposite direction, USDC cost 2.70% against 2.20% for Wise. The authors cautioned that the USDC transactions and Wise simulations occurred on different dates, so market timing could have affected the differences.
This matters for providers advertising stablecoin payments on the strength of network fees alone. The relevant customer metric is the end-to-end amount delivered, not the gas fee or blockchain confirmation charge. Platforms also need to disclose which funding and withdrawal methods are available, the spread used at each conversion, fixed charges and any route-specific limits. Without those details, a low-cost settlement layer can obscure a substantially more expensive retail transaction.
Domestic payment rails determined whether the transfer was fast
The speed results followed the same pattern. The blockchain transfer took less than 15 minutes in seven of the eight comparable routes and about 30 minutes from South Africa to Italy. But the full process ranged from less than 20 minutes to one or two business days.
Where instant domestic systems supported the fiat legs—such as PIX in Brazil, Transferencias 3.0 in Argentina and instant euro payments in Italy—funding and withdrawal could complete quickly enough for an end-to-end time below 20 minutes. Where standard bank transfers were needed in South Africa, the full process took one to two business days. An Italy-to-UAE transfer was estimated at about one day because of the withdrawal step, but the paper notes that this estimate came from exchange documentation rather than a completed withdrawal.
The operational lesson is that stablecoin remittances and conventional payment infrastructure are complements. A fast token transfer cannot compensate for a slow or unavailable local payout rail. For payment firms, service-level claims should therefore cover the complete route and distinguish observed completion times from estimates based on provider documentation.
Regulation changed access, complexity and risk
Japan illustrates another control problem. The paper says retail access to dollar-pegged stablecoins was limited to one domestic operator, which did not permit direct outbound transfers to foreign exchanges. Researchers had to use an unhosted wallet and split the process because of quantitative limits. They concluded that the route was too operationally complex for a meaningful like-for-like comparison with the other corridors.
The authors argue that restrictive or prohibitionist regimes can shift demand toward offshore or unregulated channels, increasing anti-money-laundering and counter-terrorist-financing risks while reducing transparency. That is a policy conclusion from the study, not a finding that any provider in the experiment violated the law. It nevertheless highlights the trade-off regulators face: controls that block supervised access without eliminating demand may move activity to channels that are harder to monitor.
The study is evidence, not a universal verdict
The paper’s scope is narrow. It tested one stablecoin, a small number of countries and transactions of 200 USDC. Individual exchange choices, market conditions, customer eligibility and local rules can produce different results. The researchers explicitly say the findings should not be generalized to every stablecoin remittance arrangement.
Even with that limitation, the exercise identifies the controls that should be tested before a provider claims an efficiency advantage: end-to-end price by corridor and direction, exchange and card charges, payout availability, observed completion time, customer-support arrangements, transaction limits and the regulatory status of every intermediary. The blockchain may be the fastest and cheapest part of the chain, but the customer experiences the whole chain.