A Banca d'Italia mystery-shopping study published July 30, 2026, executed 200 real USDC transfers across ten remittance corridors linking Italy with Argentina, Brazil, South Africa, the UAE, and Japan. Total costs ranged from 0.30% to 8.96% of the transferred amount. Blockchain fees accounted for...
"Stablecoin-based solutions do not seem to be systematically cheaper than traditional methods, as cost-effectiveness depends on on- and off-ramp fees." — Paolo Angelini, Deputy Governor, Banca d'Italia (April 20, 2026)
A Banca d'Italia mystery-shopping study published July 30, 2026, executed 200 real USDC transfers across ten remittance corridors linking Italy with Argentina, Brazil, South Africa, the UAE, and Japan. Total costs ranged from 0.30% to 8.96% of the transferred amount. Blockchain fees accounted for just 0.4% of total expense on average. The remaining cost — exchange spreads, card surcharges, fiat on-ramp and off-ramp fees — came from the same legacy infrastructure stablecoins are supposed to bypass.
The findings land at an inflection point. The stablecoin market now exceeds $313 billion in capitalization, with USDT and USDC settling over $7.2 trillion in February 2026 alone — surpassing the U.S. ACH network. Yet in the $212 billion global remittance market, stablecoins hold roughly 1% of flows, unchanged since 2023. This study quantifies why: the blockchain layer works; the fiat layer surrounding it does not.
Authors Alberto Di Iorio, Enrica Di Stefano, Michele Mascioli, and Giorgio Trebeschi — all from Banca d'Italia's Directorate General for Markets and Payment Systems — conducted the study under Paper No. 86 in the bank's Markets, Infrastructures and Payment Systems series. The research employed a "mystery shopping" approach: researchers posed as ordinary remittance senders and executed real transfers rather than simulating them.
Each transfer moved 200 USDC on the Ethereum blockchain between March 24 and 26, 2026. Ten corridors were tested, covering both directions (Italy-to-destination and destination-to-Italy) for five country pairs. The researchers used centralized exchanges — Binance and Kraken for European legs, Ripio (Argentina), Foxbit (Brazil), Valr (South Africa), BitOasis (UAE) — to handle fiat-to-crypto and crypto-to-fiat conversions.
The study tracked five distinct phases of each transfer: (1) fiat account funding, (2) USDC purchase, (3) on-chain transfer, (4) USDC sale, and (5) local currency withdrawal. This full-journey measurement is significant. Most industry estimates of stablecoin transfer costs isolate the blockchain leg — the cheapest phase — and ignore everything else.
The study did not measure peer-to-peer stablecoin markets, informal agent networks, or non-custodial wallet transfers where costs and counterparty risks differ substantially.
The ten corridors produced wide dispersion. A summary:
| Corridor | Total Cost (% of $200) | Primary Cost Driver | |---|---|---| | Italy → Argentina | 0.30% | Exchange-rate distortion (official vs. market rate) | | Italy → Japan | 1.30% | Low exchange spread | | Japan → Italy | 1.60% | Low exchange spread | | Brazil → Italy | 2.21% | Moderate exchange/withdrawal fees | | Italy → Brazil | 2.70% | Moderate exchange/withdrawal fees | | Italy → South Africa | 4.58% | Bank withdrawal delays, exchange fees | | South Africa → Italy | 5.44% | Bank withdrawal delays, exchange fees | | Italy → UAE | 7.20% | Card surcharges, exchange fees | | Argentina → Italy | 8.96% | 5.36% purchase cost; Dólar MEP distortion | | UAE → Italy | 8.95% | 3.8% card surcharge; 6.17% combined on-ramp |
The Italy-to-Argentina corridor's 0.30% cost is an outlier driven by Argentina's parallel exchange rate regime. The official peso rate creates an artificial discount that flatters the stablecoin pathway. The reverse corridor — Argentina to Italy — cost 8.96%, the study's most expensive route. The 30x cost differential between identical country pairs in opposite directions underscores a core finding: corridor economics, not blockchain technology, determine transfer cost.
The UAE corridors were the study's second-most expensive, driven primarily by BitOasis card surcharges of 3.8% and combined funding/purchase costs of 6.17%. Japan, with mature and liquid exchange infrastructure, produced the most consistent results in both directions.
The study's most consequential finding is its cost decomposition. Across all ten corridors, the blockchain transfer — the on-chain leg — averaged 0.4% of total cost. In the Brazil-to-Italy corridor, blockchain fees were as low as 0.01%.
The remaining 99.6% of cost, on average, came from:
The paper concluded: "Funding, conversion and withdrawal rather than the on-chain transfer" drove expenses. This means the value captured by fiat intermediaries — banks, card networks, exchanges — dwarfs the value captured by the blockchain itself. The on-chain rail is nearly free. The on-ramps and off-ramps extract the rent.
Settlement speed showed similar heterogeneity to cost, and for the same reason: local payment infrastructure.
The blockchain confirmation step completed in under 15 minutes across seven of eight tested corridors. But end-to-end speed — from initiating fiat transfer to receiving fiat in the destination — depended entirely on whether the origin and destination countries had instant payment rails.
Sub-20-minute corridors: Italy (connected via TIPS, the ECB's instant payment system), Brazil (Pix), and Argentina (Transferencias 3.0) enabled full round-trip completion in under 20 minutes. These three countries have invested heavily in domestic instant payment infrastructure.
1-2 business day corridors: South Africa, which lacked instant payment rails for this use case, required standard bank transfer settlement times of one to two business days — comparable to traditional remittance services.
The implication is direct: stablecoin speed advantages exist only where both endpoints have instant payment systems. Where they do not, stablecoins settle at the speed of the slowest fiat leg.
Wise (formerly TransferWise) serves as a relevant benchmark because it represents the best-in-class fintech alternative to traditional remittance services.
Of seven comparable corridors, USDC beat Wise on cost in three:
USDC lost to Wise in four corridors, including both UAE routes and Italy-to-Brazil (USDC 2.70% vs. Wise 2.20%).
The 3-of-7 win rate against a single fintech competitor suggests stablecoins have not yet achieved the cost structure required for broad remittance dominance. Where USDC wins, it wins on corridors with specific structural advantages (Argentina's parallel rate, Brazil's liquid crypto market, South Africa's high traditional costs). Where it loses, it loses on corridors where Wise has optimized its own payment infrastructure.
Against the World Bank's global average remittance cost of 6.65% (Q2 2024, the most recent available figure), stablecoins performed better in most corridors tested:
| Corridor | Stablecoin Cost | World Bank Average | |---|---|---| | Brazil | 2.21% | 9.96% | | South Africa | 5.44% | 15.23% | | Japan | 1.60% | Not specified | | UAE | 8.95% | 2.65% |
The Brazil and South Africa results are notable: stablecoin costs ran 78% and 64% below World Bank corridor averages, respectively. These are high-cost corridors where traditional remittance operators extract substantial margins. Stablecoins offer meaningful savings here.
The UAE result tells the opposite story: at 8.95%, stablecoin cost exceeded the World Bank corridor average of 2.65% by 238%. The UAE's well-developed traditional remittance infrastructure, built around its large migrant worker population, already delivers low-cost transfers that stablecoins — burdened by card surcharges and exchange fees — cannot match.
The stablecoin market capitalization reached approximately $313 billion by mid-2026, a 23% year-over-year increase. Approximately 269 million addresses hold stablecoin balances globally. Monthly settlement volumes surpassed the U.S. ACH network in February 2026. Yet actual real-economy payments — as opposed to trading, arbitrage, and DeFi activity — account for only $350-550 billion of the $28-62 trillion in gross annual stablecoin transfers, according to Reap Global's analysis.
Cross-border B2B stablecoin payments are projected to reach $5 trillion by 2035, up from approximately $13.4 billion in 2026. The consumer remittance use case — the subject of this study — remains a fraction of that, with stablecoins accounting for an estimated 5-10% of flows in the U.S.-Mexico corridor (the largest single bilateral remittance route) and roughly 1% of global flows overall.
The Banca d'Italia data suggests the bottleneck is not the blockchain. On-chain costs are negligible. The bottleneck is the regulated financial infrastructure that connects blockchains to bank accounts. Until fiat on-ramps and off-ramps become cheaper, faster, and more competitive, stablecoin remittances will remain corridor-specific rather than universally superior.
This aligns with Governor Fabio Panetta's broader stance. Elected Chair of the BIS Board in June 2026, Panetta has consistently argued that improving domestic fast-payment systems and interlinking them across borders may solve the remittance cost problem more effectively than stablecoins alone.
The Banca d'Italia study does not conclude that stablecoins are ineffective for remittances. It concludes something more precise: that stablecoin efficiency is a function of fiat infrastructure quality, not blockchain performance. The on-chain layer is fast and cheap. The off-chain layer — where regulated entities extract fees for compliance, conversion, and cash handling — determines whether a stablecoin transfer saves money or costs more than a Wise transfer.
For the stablecoin industry, this presents a structural challenge. The technology works. The economics depend on variables that stablecoin issuers do not control: exchange fee competition, card network surcharges, local banking integration, and national instant payment system availability. The corridors where stablecoins deliver genuine savings — high-cost routes in Latin America and Africa — are precisely the corridors where traditional infrastructure is weakest and most extractive.
Whether stablecoins can move beyond their current 1% global remittance share will depend less on blockchain throughput or regulatory frameworks and more on whether fiat on-ramps and off-ramps converge toward competitive pricing. As of August 2026, that convergence has not occurred.