A Bank of Italy mystery shopping exercise, published July 30, 2026, tested 200 USDC remittances across ten bidirectional corridors linking Italy with Argentina, Brazil, South Africa, the UAE, and Japan. Total transfer costs ranged from 0.30% to 8.96% of the amount sent. The blockchain leg of each...
"If stablecoins could be spent directly in the real economy — for goods and services, rents, or school fees — without reconversion into local fiat currency, the economic advantages of stablecoin-based transfers would be substantially higher." — Bank of Italy, Stablecoin Remittance Study (July 2026)
A Bank of Italy mystery shopping exercise, published July 30, 2026, tested 200 USDC remittances across ten bidirectional corridors linking Italy with Argentina, Brazil, South Africa, the UAE, and Japan. Total transfer costs ranged from 0.30% to 8.96% of the amount sent. The blockchain leg of each transaction averaged 0.4% of total cost. The remaining 60–99% of expenses came from fiat on-ramps, off-ramps, exchange spreads, and currency conversion — infrastructure that has nothing to do with distributed ledgers.
The study lands at a moment when the stablecoin remittance narrative is under active stress-testing by incumbents. Western Union launched its USDPT token on Solana in May 2026; MoneyGram deployed MGUSD on Stellar in June 2026. Both firms are integrating stablecoin rails into legacy networks spanning over one million combined retail locations. Yet the Bank of Italy data suggests the cost problem was never on-chain. It sits at the fiat boundary — the same boundary these incumbents already control.
Global remittance flows are projected to exceed $900 billion in 2026, according to World Bank estimates. The global average cost of sending $200 remains 6.35–6.49%, more than double the UN Sustainable Development Goal target of 3%. Stablecoins were marketed as the fix. The empirical evidence is more complicated.
The Bank of Italy deployed a mystery shopping methodology — researchers acted as retail users, executing actual USDC transfers through commercially available platforms. The sample: 200 transactions across ten corridors (five country pairs, bidirectional). The stablecoin tested was USDC, the second-largest dollar-pegged stablecoin by market capitalization.
The study measured total end-to-end cost: fiat-to-stablecoin conversion (on-ramp), blockchain transfer fees, and stablecoin-to-fiat conversion (off-ramp) including any foreign exchange spreads. This all-in approach distinguishes it from industry benchmarks that often cite only the on-chain transfer fee — a number that is functionally irrelevant to the end user.
The researchers acknowledged limitations: 200 transactions across ten corridors using a single stablecoin and a limited set of platforms. The findings, the study noted, cannot be "readily generalized" to other providers or corridors. That caveat matters. But so does the fact that this is one of the first empirical, controlled tests of stablecoin remittance costs conducted by a central bank.
The cost variance across corridors was extreme:
| Corridor | Total Cost (% of transfer) | |---|---| | Italy → Argentina | 0.30% | | Italy → Japan | 1.30% | | Japan → Italy | 1.60% | | Brazil → Italy | 2.21% | | Italy → Brazil | 2.70% | | Italy → South Africa | 4.58% | | South Africa → Italy | 5.44% | | Italy → UAE | 7.20% | | Argentina → Italy | 8.96% | | UAE → Italy | 8.95% |
The cheapest corridor (Italy to Argentina, 0.30%) was 30x less expensive than the most expensive (Argentina to Italy, 8.96%). The same country pair, reversed, produced a 29x cost differential. This asymmetry is not a blockchain phenomenon. It reflects differences in local exchange liquidity, regulatory constraints on fiat conversion, and the availability of instant payment infrastructure on each end.
Six of ten corridors came in below the World Bank's global average remittance cost of 6.35–6.49%. Four exceeded it. The corridors involving the UAE were consistently the most expensive, with both directions exceeding 7%.
The study's most consequential finding: blockchain transaction fees averaged 0.4% of total cost across all corridors. The remaining cost — in some cases exceeding 8.5 percentage points — came from:
This cost structure inverts the standard narrative. Proponents argue that blockchain removes intermediaries and cuts fees. The Bank of Italy data shows the blockchain layer is already functionally free. The expense sits entirely in the fiat infrastructure wrapping the on-chain transfer — infrastructure operated by exchanges, banks, and payment processors.
The 2026 stablecoin off-ramp market confirms this pattern at scale. Industry data shows off-ramp fees ranging from 0.5% to 3.0% depending on provider and corridor, with FX conversion spreads adding 15 to 200 basis points. Network gas fees, by contrast, are negligible on most chains — fractions of a cent on Solana, under $0.01 on Stellar, and typically under $0.50 on Ethereum Layer 2s.
Settlement speed was the study's most favorable metric for stablecoins. On-chain settlement completed in under 15 minutes in seven of eight directly comparable corridors. End-to-end transfer times — including fiat conversion on both ends — fell under 20 minutes where instant payment systems operated at both endpoints:
South Africa was the outlier: standard bank transfer delays pushed total settlement to one to two business days, negating the speed advantage of on-chain settlement.
The speed finding reinforces the study's central theme: stablecoin performance is a function of fiat infrastructure quality. Where instant payment rails exist on both ends, stablecoins deliver near-real-time settlement. Where they do not, the blockchain cannot compensate for legacy banking delays.
The study benchmarked USDC transfers against Wise (formerly TransferWise) in seven comparable corridors. The results:
USDC cheaper than Wise (3 corridors):
Wise cheaper than USDC (4 corridors):
The 3–4 split against Wise — a fintech that operates on traditional banking rails — is significant. Wise's cost advantage comes from optimized correspondent banking relationships and pooled liquidity, not from technological superiority. The fact that a bank-rail fintech beats stablecoins in a majority of tested corridors suggests the cost problem is operational and structural, not technological.
The Bank of Italy study arrives as legacy remittance operators are moving onto stablecoin rails themselves:
Western Union launched USDPT on Solana on May 4, 2026, issued by Anchorage Digital Bank N.A. (the first federally chartered crypto bank in the U.S.). USDPT is designed for agent settlement, not retail — an alternative to SWIFT for moving value between Western Union's 600,000+ locations. A consumer-facing product, Stable by Western Union, is planned for 40+ countries in 2026.
MoneyGram deployed MGUSD on Stellar on June 2, 2026, issued by Bridge (a Stripe subsidiary). MGUSD is embedded directly into the MoneyGram app, enabling self-custodial dollar-denominated balances for 60 million customers across nearly 500,000 retail locations.
The incumbent pivot changes the competitive calculus. Western Union and MoneyGram already operate the fiat on-ramp and off-ramp infrastructure that the Bank of Italy identified as the dominant cost driver. By issuing their own stablecoins, they can internalize what was previously exchange fee leakage — capturing the on-chain efficiency without ceding margin to third-party exchanges.
This is a vertical integration play. The incumbents are not being disrupted by stablecoins. They are absorbing stablecoins into existing distribution networks.
The Bank of Italy data quantifies what the industry calls the "fiat boundary tax." The 2026 off-ramp market has segmented into two tiers:
Retail off-ramps (personal USDC/USDT to bank account): fees of 0.5–3.0%, plus FX spreads of 50–200 basis points. Total all-in cost: 1.0–5.0%.
B2B/institutional off-ramps (treasury flows, payroll, merchant settlement): fees of 0.1–0.5%, plus FX spreads of 15–50 basis points. Total all-in cost: 0.25–1.0%.
The gap between retail and institutional pricing explains much of the variance in the Bank of Italy corridors. A retail user in Argentina faces exchange controls, limited platform competition, and wide FX spreads. An institutional user in Japan accesses tighter spreads and lower fees through regulated exchanges with deep liquidity.
Stablecoin transaction volumes reached approximately $33 trillion in 2025, according to industry tracking data — but this figure includes all transaction types, not only remittances. Actual stablecoin payment volume (P2P, B2B, cross-border) reached $390 billion in 2025, more than double 2024 levels, per McKinsey and Artemis Analytics. The growth is real. The question is whether the cost structure serves retail senders — the people remittances are supposed to help.
The study carries three policy signals:
1. Domestic instant payment infrastructure matters more than blockchain. Corridors with instant payment systems on both ends (Italy-Brazil, Italy-Argentina) showed the lowest costs and fastest settlement. Investment in domestic payment infrastructure — PIX-style systems — may deliver more remittance cost reduction than stablecoin adoption alone.
2. Prohibitionist regulation backfires. The study found that restrictive regulatory approaches in certain corridors pushed users toward offshore platforms with higher fees and lower consumer protections. The Argentina corridors illustrate this: the 0.30% Italy-to-Argentina cost reflected favorable on-ramp conditions in Italy, while the 8.96% reverse route reflected the friction of Argentine exchange controls.
3. Direct stablecoin spending would change the equation. The study's most forward-looking observation: if recipients could spend stablecoins directly — for rent, goods, school fees — without converting to local fiat, the off-ramp fee disappears entirely. This scenario requires merchant acceptance infrastructure that does not yet exist at scale in most remittance-receiving countries.
The Bank of Italy study is a calibration event for the stablecoin remittance narrative. It does not show that stablecoins are useless for remittances — the Italy-to-Argentina corridor at 0.30% is a genuine result. It shows that stablecoin performance is almost entirely determined by the quality and cost of fiat infrastructure at each end of the transfer. The blockchain in the middle is already close to free.
This finding has structural implications. It means the entities best positioned to capture stablecoin remittance value are not crypto-native startups but incumbents who already control fiat on-ramps and off-ramps — banks, money transfer operators, and regulated exchanges. Western Union and MoneyGram have arrived at the same conclusion and are acting accordingly.
The path to cheaper remittances runs through better fiat infrastructure, more competitive exchange markets, and regulatory frameworks that enable — rather than prohibit — stablecoin access. The blockchain is not the bottleneck. It never was.