Stablecoins make perfect sense on paper. It goes without saying that if you replace a chain of correspondent banks with a blockchain transaction, then cross-border payments should become faster, cheaper and available around the clock due to the fundamental nature of blockchain versus the fundamental nature of the global commerce landscape’s correspondent banking network.
But new experiments by Banca d’Italia, the Italian central bank, and Project Agorá, a public-private collaboration, show that the economics of a stablecoin cross-border payment are considerably less straightforward.
The Italian central bank researchers conducted a “mystery shopping” exercise involving transfers of 200 USDC across 10 corridors connecting Italy with Argentina, Brazil, South Africa, the United Arab Emirates and Japan. Rather than modeling hypothetical transactions, they acted as ordinary customers, moving money from a bank account into USDC, transferring the tokens internationally, converting them back into local currency and withdrawing the proceeds.
The conclusion was less revolutionary than stablecoin advocates might expect. Stablecoins provided no systematic cost advantage over conventional remittance services. Total expenses ranged from 0.30% to almost 9% of the amount transferred, while completion times varied from less than 20 minutes to two business days.
Meanwhile, Project Agora’s cross-border settlement pilot using the tokenization of wholesale cross-border payments demonstrated an average time from payment initiation to settlement of about 80 seconds.
Taken together, the two central bank findings do not show that stablecoins are inherently inefficient. They reveal something more consequential for the payments industry: a faster settlement asset cannot, by itself, repair the fragmented financial systems surrounding it.
See more: The Stablecoin Sandwich Is Missing the Trust Layer
The Blockchain Is No Longer the Main Cross-Border Bottleneck
Cross-border B2B payments already account for the majority of global stablecoin payment volume, according to findings in the July 2026 edition of the Payments Innovation Tracker® Series, a PYMNTS Intelligence collaboration with Paymentology. But while a stablecoin transfer is often presented as a single digital movement, economically, it is closer to a five-stage supply chain.
The sender must fund an exchange account, purchase the stablecoin, transfer it over a blockchain, sell it at the destination and withdraw the resulting fiat currency. Banca d’Italia describes this structure as a “stablecoin sandwich,” with the blockchain transaction positioned between the fiat on-ramp and off-ramp.
In effect, stablecoins can replace one group of intermediaries while creating another. Cryptocurrency exchanges begin to function like digital correspondent banks, controlling access, liquidity, pricing and conversion. The efficiency of the payment therefore depends as much on exchange governance and market competition as it does on the underlying network
“We think of stablecoins as rails,” Mastercard Executive Vice President of Blockchain and Digital Assets Raj Dhamodharan told PYMNTS. “Each stablecoin can be thought of as a global ACH (automated clearing house), where the consumer doesn’t see the complexity.”
“The technology underneath this is quite powerful,” Dhamodharan said. “But that alone is not sufficient. To unlock the full value, really that orchestration needs to be provided.”
A near-instant transaction that remains stranded at an exchange, subject to an unfavorable spread or delayed by a bank withdrawal, is not an instant business payment.
See also: Nobody Told the ERP That Blockchain Won
Stablecoins Are Only as Fast as the Fiat Systems Around Them
In corridors supported by strong domestic instant-payment infrastructure, stablecoins performed well. Transfers involving Italy, Brazil and Argentina could be completed in less than 20 minutes because the fiat funding and withdrawal stages moved quickly through systems including TIPS, Pix and Transferencias 3.0.
In South Africa, however, the same stablecoin architecture required one to two business days. The delay came almost entirely from standard bank transfers used at the beginning or end of the transaction. The implication is almost paradoxical: stablecoins need efficient conventional payment systems to deliver their promised efficiency.
Those endpoints dominated the economics. Across the exercise, the blockchain portion contributed roughly 0.4% to the total cost, while exchange spreads, card charges, conversion fees and withdrawal expenses produced most of the variation. The study found total costs ranging from 0.30% for a transfer from Italy to Argentina to 8.96% in the opposite direction.
This complicates the argument that stablecoins allow countries to leapfrog legacy infrastructure. The study indicates that the best-performing stablecoin corridors are frequently those already supported by strong domestic payment systems. Markets with weak banking connectivity may have the greatest need for an alternative, yet their limited on- and off-ramp infrastructure can prevent stablecoins from realizing their theoretical advantage.
Read more: Crypto Experts Tell PYMNTS Where Digital Assets Go Next
For global trade, this changes where banks and payment providers should focus. Optimizing the ledger while ignoring local clearing access, bank-account connectivity and exchange liquidity will produce impressive settlement metrics without necessarily improving the customer experience.
The winning infrastructure may therefore look less like a standalone stablecoin network and more like an orchestration layer connecting tokenized money, real-time domestic rails, compliant exchanges and automated foreign exchange.
Data in “Waiting for Certainty: Why Most CFOs Are Holding Back on Crypto and Stablecoins”, a recent installment of PYMNTS Intelligence’s 2026 Certainty Project, shows that most middle market companies remain cautious about digital assets: 13% of firms use stablecoins and just 5% use other cryptocurrencies.
See more: Stablecoins Are Just Wildcat Banking With Better Wi-Fi
Project Agorá provides a useful counterpoint because it is testing tokenization without requiring money to leave the regulated banking system. Rather than placing a stablecoin between two fiat endpoints, Agorá seeks to make commercial bank deposits programmable and interoperable with central bank money. That could reduce reconciliation, funding and settlement risk without imposing a separate purchase-and-redemption cycle on users.
The contrast is not simply stablecoins versus banks. It is between tokenizing one leg of a payment and redesigning the complete transaction.
