Three central bank research papers published within a ten-day window in July 2026 converge on the same conclusion: stablecoin infrastructure works, but not in the way its promoters claim. The Bank of Italy sent 200 real USDC transfers across ten corridors and found total costs ranging from 0.30% ...
"Capital controls cut bank dollarisation by up to 32 percentage points but showed no effect on stablecoin inflows." — Boris Hofmann, Aaron Mehrotra & Jan Paulick, BIS Working Paper No. 1270
Three central bank research papers published within a ten-day window in July 2026 converge on the same conclusion: stablecoin infrastructure works, but not in the way its promoters claim. The Bank of Italy sent 200 real USDC transfers across ten corridors and found total costs ranging from 0.30% to nearly 9% — with fiat on- and off-ramps, not blockchains, driving the price. The Bank for International Settlements analyzed stablecoin flows across 130+ economies and found dollar-pegged tokens bypass capital controls that suppress bank dollarization by up to 32 percentage points. BCG and Allium Labs pegged real-economy stablecoin payments at $350–550 billion in 2025 — less than 1% of the $62 trillion in gross stablecoin transaction volume that year.
Taken together, the three studies suggest stablecoins are economically significant but structurally different from their marketing pitch. Blockchain fees are negligible; the expensive part is the interface with the fiat banking system. Regulatory leakage is real but not because stablecoins are efficient — rather, because they route around controls that apply only to bank deposits. And headline transaction volumes overstate actual payment activity by roughly 100x.
The Bank of Italy's Markets, Infrastructures, Payment Systems division published a paper on July 30, 2026, titled "Are Stablecoins Efficient for Remittances? Evidence from a Mystery Shopping Exercise." The methodology was unusual for a central bank: researchers sent actual money. They executed 200 transfers of 200 USDC each across ten bidirectional corridors linking Italy to Argentina, Brazil, South Africa, the UAE, and Japan.
The results were mixed. Total transfer costs ranged from 0.30% to almost 9% of the amount sent. In corridors with well-developed instant payment rails — Japan, the UAE — transactions settled in under 20 minutes. In corridors without such infrastructure — parts of South America, Sub-Saharan Africa — settlement took one to two business days.
The stablecoin transfers beat the World Bank's reported global average remittance cost of 6.49% in most corridors. But they underperformed Wise, the fintech benchmark, in four of seven comparable corridors. The headline finding: stablecoins do not offer a "systematic cost advantage" over existing digital payment channels. Cost competitiveness depends almost entirely on local infrastructure quality.
The Bank of Italy study decomposed each transfer into its constituent costs. The on-chain transaction fee — the gas or network fee for moving USDC on a blockchain — was negligible in every corridor. It typically amounted to well under $0.01 for transfers on low-cost chains.
The cost was elsewhere. Exchange fees charged by platforms at the on-ramp (fiat-to-USDC conversion) and off-ramp (USDC-to-local-currency conversion) constituted the majority of total costs. Currency conversion spreads added further friction, particularly in corridors involving less-liquid currencies.
This finding aligns with broader industry data. According to a 2026 analysis of stablecoin off-ramp economics, FX conversion spreads range from 15 to 200 basis points depending on customer segment (consumer vs. institutional). Network gas, FX spread, payout fees, and KYC tier each move the all-in cost by 30 to 200 additional basis points. Card-based on-ramps charge 3–5.5% in fees; bank transfers cost 0–2% but require one to three business days.
The implication is structural: stablecoin cost efficiency is bounded by fiat infrastructure quality, not by blockchain throughput or fee markets. A stablecoin transfer between two Coinbase accounts with no currency conversion is effectively free. A stablecoin transfer between an Italian bank account and a South African mobile money wallet passes through multiple intermediaries, each extracting margin — precisely the same dynamic that makes traditional remittances expensive.
Ten days before the Bank of Italy paper, the BIS published Working Paper No. 1270 on July 21, 2026. Authors Boris Hofmann, Aaron Mehrotra, and Jan Paulick analyzed deposit dollarization data across 130+ economies from 1990 to 2019 and stablecoin flow data from Chainalysis covering 184 countries from 2017 to 2024.
The central finding: capital controls are effective against bank deposits but ineffective against stablecoins. Countries that require regulatory approval for residents to hold foreign-currency bank accounts showed deposit dollarization ratios 25 to 32 percentage points lower than countries without such rules. The same controls had no statistically significant effect on stablecoin inflows.
The mechanism is straightforward. Bank deposits sit inside regulated institutions subject to supervisory reach. Stablecoins move on public blockchains and can reside in self-hosted wallets outside institutional oversight. Exchanges, peer-to-peer markets, and decentralized protocols provide alternative entry and exit points that do not require licensed banking relationships.
The BIS estimated total stablecoin market capitalization at approximately $320 billion as of end-May 2026, with 99.4% of fiat-backed stablecoins denominated in US dollars. The paper noted this dynamic is "especially pronounced in emerging markets and developing economies" and flagged concerns about irreversibility — once stablecoin dollarization takes hold, reversing it may prove difficult.
The BIS's June 2026 annual report added a related observation: stablecoins fail to meet four foundational monetary system requirements — singleness, elasticity, interoperability, and integrity. This framing positions stablecoins as functional but architecturally incomplete relative to central bank money.
A BCG white paper published in January 2026, in collaboration with on-chain analytics firm Allium Labs, attempted to separate real economic activity from the noise in stablecoin transaction data.
The headline finding: real-economy stablecoin payments reached $350–550 billion in 2025, growing approximately 60% year-over-year. But overall stablecoin transaction volume was estimated at $62 trillion in the same period. The gap — roughly 100x — reflects the dominance of trading, arbitrage, treasury management, and internal routing in on-chain stablecoin flows.
B2B settlement accounted for approximately 60% of real-economy stablecoin payment volume. Among corporates already using stablecoins, 41% reported cost savings of at least 10% on cross-border B2B payments. On a $50 million annual payment volume, that translates to $5 million or more recovered annually.
The BCG data provides crucial context for interpreting headline stablecoin metrics. The $8.8 trillion figure widely cited for H1 2026 stablecoin flows includes the full spectrum of on-chain activity. Actual payment usage — the economic activity that matters for remittance and commerce comparisons — is orders of magnitude smaller.
Traditional remittance operators are not standing still. Western Union launched USDPT, a dollar-backed stablecoin on the Solana blockchain, on May 4, 2026. The token is issued by Anchorage Digital Bank, a federally regulated crypto custodian holding an OCC federal trust charter. USDPT is designed for agent settlements — replacing SWIFT-based interbank flows rather than serving retail customers directly. A consumer-facing service, Stable by Western Union, is planned for launch in over 40 countries later in 2026, with initial deployment in the Philippines and Bolivia.
The strategic logic is defensive. Western Union's traditional margin comes from FX spreads and transfer fees averaging 5–7% on small remittances. A stablecoin settlement layer compresses the back-end cost structure but risks cannibalizing fee income. As Forbes noted in a May 2026 analysis, Western Union's stablecoin "automates the end of its own margin."
Wise charges fees starting at 0.57% depending on currency pair. For a $500 transfer, stablecoins can save $5–20 compared to Western Union but the advantage narrows or disappears against digital-native competitors like Wise, as the Bank of Italy data confirmed.
The convergence of these three studies points to a specific infrastructure gap. The blockchain layer — the part the crypto industry has spent billions building — is not the bottleneck. Gas fees are sub-cent on most modern networks. Settlement finality ranges from seconds to minutes. The technology works.
The bottleneck is the fiat interface: KYC verification, bank account connectivity, local payment rail integration, FX conversion, and regulatory compliance at the on-ramp and off-ramp. These are the costs that make a stablecoin transfer cost 9% in one corridor and 0.30% in another.
For infrastructure builders, the implication is that value capture in stablecoin payments will accrue not to blockchain protocol developers but to entities that solve fiat connectivity at scale. This is consistent with the observed strategic behavior of incumbents. Coinbase reports USDC held in its products reached a $20 billion all-time high in Q2 2026, while stablecoin revenue — primarily from the fiat interface — generated $292 million in the quarter. Circle's revenue model depends on float income from reserves, not on-chain transaction fees.
The BIS capital-controls finding adds a regulatory dimension. As stablecoins grow, regulators in emerging markets face a choice between attempting to control on- and off-ramps (effective but restrictive) or accepting stablecoin flows and adapting monetary policy tools (pragmatic but disruptive to existing frameworks). The data suggests prohibitionist approaches redirect users to unregulated offshore platforms rather than suppressing demand.
The July 2026 research cluster from the Bank of Italy, BIS, and BCG provides the most empirically grounded assessment of stablecoin utility published to date. The findings do not support the claim that stablecoins are uniformly cheaper or faster than traditional payment rails. They do support the claim that stablecoins route around regulatory and institutional barriers that constrain bank-based systems.
The economic value in stablecoin infrastructure lies not in the blockchain layer — which is commoditized and cheap — but in the interfaces between digital and fiat systems. Entities that control on-ramps, off-ramps, FX conversion, and regulatory compliance will capture the economic surplus. The blockchain is the pipe; the value is in the faucets at either end.
For the $307 billion stablecoin market, this reframing matters. Growth will be driven not by lower gas fees or faster finality but by fiat infrastructure investment in underserved corridors — the same corridors where the Bank of Italy found stablecoin costs approaching 9%. The technology is solved. The plumbing is not.