Bank of Italy: stablecoin remittances cost 0.3% to 9% once you count the off-ramp

The cheap part of a stablecoin remittance is the part almost nobody pays for on its own.

That is the finding of a Bank of Italy research paper, published in the central bank's Markets, Infrastructures and Payment Systems series as Paper No. 86, and reported by CoinDesk on August 1. Researchers ran a mystery-shopping exercise across 10 international corridors, sending USDC out of Italy to destinations including Argentina, Brazil, South Africa, the UAE and Japan, then measured the full journey — bank account to exchange to wallet to on-chain transfer to local fiat in the recipient's hands.

End-to-end cost came in between roughly 0.3% and almost 9% of the amount sent, depending on corridor and provider, per CoinDesk's account of the paper. Settlement ranged from about 20 minutes, where a domestic instant-payment system supported the withdrawal, to as long as two business days where the recipient fell back on a conventional bank transfer.

Network gas fees were a negligible share of the total. The money went to euro-to-USDC conversion, foreign exchange spreads, exchange fees, withdrawal into local currency, and domestic banking charges — the first and last mile, not the middle. The conclusion the researchers draw, as CoinDesk relays it, is not that stablecoins failed but that they have not been shown to be systematically cheaper than conventional money transfer operators once the whole chain is priced. The paper also credits stablecoins with genuine advantages: cost reductions in specific corridors, always-on settlement, and programmability.

Key facts

• 10 remittance corridors tested, originating in Italy; named destinations include Argentina, Brazil, South Africa, the UAE, Japan — Bank of Italy Paper No. 86, via CoinDesk (Aug 1, 2026).
• Total end-to-end cost: ~0.3% to nearly 9% of value sent — same.
• Settlement time: ~20 minutes to two business days — same.
• Blockchain gas fees: negligible fraction of total cost; conversion, FX spread and banking charges dominate — same.

The real-world read

The number "200 USDC" is doing two different jobs in CoinDesk's write-up: its summary calls it 200 remittances tested, its body calls it transfers of 200 USDC. Those are not the same claim, and the sample size versus the ticket size materially changes how much weight the range carries. The primary paper would settle it; the secondary account doesn't.

CoinDesk also pauses to note that central banks "may have a vested interest" in undermining stablecoins. Fair to flag — but the study's own finding cuts against the insinuation, since it exonerates the blockchain layer and blames the intermediaries sitting on either side of it. Discount marketing on the other side too: per-transfer costs of "less than a cent" describe an on-chain hop, not a remittance.

What isn't public here: which exchanges and off-ramp providers were used, the corridor-by-corridor breakdown, and when the shopping was done. Without those, the 9% top end can't be pinned to a named provider.

Opinion, and whose

The Bank of Italy's researchers argue that proliferating regulated off-ramps under MiCA, plus tighter integration between instant-payment systems and digital-asset infrastructure, could compress conversion fees — while FX spreads stay unavoidable. That is their forecast, not an observed result.

Sources

• CoinDesk (Aug 1, 2026), "Bank of Italy research suggests stablecoins aren't necessarily cheaper for remittances" — the sole secondary account here; it cites Bank of Italy Markets, Infrastructures and Payment Systems Paper No. 86, which is the primary document and has not been read directly for this item.
• Carried alongside that article on CoinDesk's site: a sponsored Binance "CEX landscape case study." That is marketing, unrelated to the research, and not used.

Not financial advice.