Stablecoins can cross a blockchain in minutes, but a transfer between wallets does not establish that a supplier was paid or a migrant sent money home. New classifications put observable payment activity in the hundreds of billions of dollars, far below headline transfer volume, while live card and bank programs show why the distinction matters.
Summary
- Allium estimates $401 billion to $527 billion of stablecoin payments in the first eight months of 2026, compared with the much larger total of transfers recorded on public chains.
- A separate Allium and FXC Intelligence analysis put stablecoins at roughly 0.31% of the $44.3 trillion retail cross border payments market in 2025.
- Visa says more than 160 stablecoin linked card programs operate on its network, with program payment volume up nearly 200% year over year.
- The classification can change: Visa’s September dashboard refresh removed additional exchange, bot and routing activity from adjusted volume after greatly expanding its labeled address set.
- The amount reaching a recipient, the conversion cost and the location of both parties remain difficult to infer from a public token transfer alone.
A dollar token sent from one blockchain address to another leaves a clear transaction record. It does not leave an invoice, the residence of the sender, the identity of the ultimate beneficiary or the exchange rate paid to turn the token into spendable local money. Those missing facts separate an impressive onchain total from the narrower question of commerce.
Allium’s September study estimates at least $401 billion of payment volume through August 2026 and a broader upper estimate of $527 billion. It says businesses received 58% to 64% of classified payments. The range reflects analytical uncertainty, not two competing official ledgers. This is substantial activity, but it is not an assertion that every stablecoin transfer settles a good or service.
What happens between a transfer and a payment?
A customer may give dollars to an intermediary, which buys USDC, sends it to a receiving partner and sells it for pesos. That journey can create multiple blockchain transfers for one underlying obligation. The customer might never hold a token, and the merchant might receive only bank money. Conversely, a business can settle a cross border invoice directly from its token wallet to a supplier’s wallet and leave just one visible transfer, while compliance and accounting happen elsewhere.
The same ledger also records exchange deposits, liquidity moves, lending collateral, decentralized exchange swaps, bridges between networks, issuer mints and redemptions, and treasury sweeps between wallets owned by one firm. Some are economically important; most are not a completed payment to an independent recipient. A wallet address is not a country. Inferring an international transaction requires reliable information about counterparties and their locations, often held by payment companies rather than published onchain.
In its 2026 working paper on stablecoin transactions, the Bank for International Settlements examined 593 million token transfer events within 141 million Ethereum transactions in 2025 for USDT, USDC and PYUSD. Nearly 60% of transfer events occurred inside complex transactions, the researchers found. Counting each event as a standalone payment would misclassify a large part of the activity. That result applies to the sampled Ethereum assets and period; it is not a universal percentage for all stablecoins or chains.
Even a clean classification of a transfer as payment leaves a second question: Was it cross border? An American company can pay an American contractor through a globally accessible blockchain. A transfer between two domestic exchange addresses could ultimately serve an international remittance offchain. Public records alone cannot settle either case. The economic unit is the obligation discharged between real parties, not the number of token movements.
The hundreds of billions have boundaries
Allium’s $401 billion to $527 billion estimate covers the first eight months of 2026. Its earlier work with FXC Intelligence estimated that stablecoins handled roughly 0.31% of $44.3 trillion in retail cross border payments during 2025, up from 0.2% in 2024. The first figure is a payment estimate across use cases and a different period; the second is a share of a defined cross border market. They should not be divided as if they describe the same denominator.
Another Allium and BCG report offers a useful warning about definitions. It compares more than $62 trillion of annual stablecoin transfers with about $4.2 trillion of activity it describes as real economy payments, around 7%. Yet its conservative count of observable bilateral payments for goods and services in 2025 is only about $350 billion to $550 billion. The $4.2 trillion category is broader than that bilateral subset. The report also says its conservative method excludes some offchain use, including internal exchange settlement and stablecoin card payments. Quoting its 7% and its bilateral estimate as interchangeable payment totals would hide that difference.
The Visa Onchain Analytics methodology sorts transfers into payments, decentralized finance, exchange flows, investment and trading, store of value, issuance, routing and infrastructure. Its adjusted volume removes labeled exchanges, contracts, bots, bridges and other activity, using heuristics for unlabeled wallets. A consumer payment, a payroll transfer and a person to person transfer can all enter its payments category; a dollar sent to an exchange normally does not. The model is an inference from address labels and behavior, not a global receipt database.
That inference was materially revised on September 18. Visa’s dashboard changelog says a refreshed Allium identity set expanded from about 15 million to roughly 600 million labeled addresses. Adjusted volume decreased because more exchange, contract and automated addresses could be excluded. Visa cited an automated Solana program moving tokens through thousands of throwaway wallets that had inflated volume. The definition stayed nominally the same while the data used to apply it improved. Historical comparisons should use a consistent dashboard vintage.
Where the live payment evidence appears
There are independently observable examples beyond classification models. In October, Visa said more than 160 stablecoin linked card programs were running, with their payment volume up nearly 200% year over year. It said about 17% of that card volume in fiscal 2026 to date came from business and commercial programs. These are network figures for Visa programs, not the share of all global commerce and not proof that each merchant accepted a token. Cardholders can spend through existing rails while issuers and settlement partners use stablecoins behind the scenes.
Visa has separately reported an annualized stablecoin settlement run rate above $20 billion. Annualized is a pace extrapolated from recent activity, not a calendar year total. Settlement between a card network and its partners is a different economic step from each retail purchase. Adding card purchases and the corresponding stablecoin settlement could count one commercial cycle twice. The distinction is visible in crypto.news coverage of Visa’s card expansion.
One discrete test is easier to bound. Lloyds and Visa’s seven day pilot settled $750,000 of cross border obligations using USDC, with the reported transfer arriving in under an hour. This demonstrates a working route between participants. It cannot establish that banks can repeat the economics at mass scale, across all currency corridors or with identical compliance obligations. The fiat entry, token movement and eventual local payout each have costs and operational dependencies.
In an earlier Hyundai Card remittance trial reported by crypto.news, a $20,000 intercompany payment moved between U.S. and Mexican entities using USDT. The seven minute reported settlement is a measure of that trial, while a corporate treasury team would also care about prefunding, conversion spread, beneficiary account availability, weekend liquidity and transaction monitoring. Speed at the blockchain layer is only one component of delivered payment quality.
Why business flows dominate the measured use
Allium estimates that businesses receive most classified payment value and that business to business settlement reached around $137 billion to $153 billion in the period covered by its analysis. A single supplier invoice can outweigh thousands of consumer purchases, so business dominance by value does not imply most people are paying with stablecoins in shops. Enterprises also have repetitive international obligations where a faster funding cycle or access to dollar liquidity may be valuable.
Stablecoins can move at any hour. Traditional bank transfers may be limited by business hours, intermediary banks and domestic payment windows. But a stablecoin still needs an issuer, reserve and redemption mechanism; a business often needs a regulated exchange or payment provider to convert it; the recipient needs local liquidity and sometimes a bank account. A fast transfer to a wallet with no economical off ramp can be a poor payment. The BIS 2026 annual report cautions that entry and exit costs can make the total cross border cost as high as or higher than a bank transfer in some cases. Crypto.news coverage of routing costs also examines how a quoted foreign exchange rate differs from the delivered route.
The value proposition varies by corridor. Where dollar banking is scarce and local transfers slow, access to a liquid dollar token can shorten a treasury cycle. Where domestic instant payments and competitive foreign exchange already work well, the gains may be smaller. A service charging a visible blockchain fee of cents could still earn a wide exchange spread or require collateral that ties up capital. A credible comparison prices the whole route, from the sender’s funding instrument to the beneficiary’s usable balance.
The crypto.news report on stablecoin card payment growth notes the Allium payment range alongside Visa’s program expansion. These are complementary observations from different measurement systems. Allium infers transactions from blockchain data; Visa sees activity inside its card network. Neither can simply be summed with the other without mapping overlap.
The final mile decides whether speed has value
A remittance sender cares about the amount the recipient can actually use. The chain can confirm a token transfer while an off ramp delays identity checks, awaits banking hours or finds insufficient local currency liquidity. An intermediary can quote a favorable token transfer fee and recover its cost in the exchange rate. A business recipient may prefer a dollar balance and retain the stablecoin; another may need local currency for payroll or tax. Those are different services even if their onchain transactions look alike.
This is why market share based on visible wallets is fragile. Some payment providers net thousands of customer obligations internally and put one settlement movement onchain. Others route the same obligation through more than one address. Card payments can use stablecoin funding without a merchant wallet. Allium’s conservative bilateral measure omits some offchain processes, while broad gross transfer measures count infrastructure several times. The actual commerce probably lies between measures with different coverage, and no honest estimate can erase the uncertainty by quoting more decimal places.
Payment growth also has a concentration question. If a handful of high value business corridors account for most dollars, headline growth can coexist with limited adoption among ordinary consumers. A large treasury transfer is not a million separate remittances. Conversely, a corridor with many small payments may have modest dollar volume yet a significant effect on the people using it. Count of completed beneficiary payments, median amount and total value together would give a fuller picture than value alone.
The practical control points are often outside a public blockchain. An issuer decides who can mint and redeem its token; a bank provides reserves and fiat settlement; exchanges and payment firms manage liquidity, screening and account access. A recipient may have recourse against a service provider for a mistaken payout, while an irreversible token transfer itself cannot resolve the dispute. Financial institutions considering the rail must test fraud handling and operational recovery alongside processing speed. A payment system is judged by whether the right party receives usable money under reliable rules, not just whether a hash appears in a block explorer.
For future comparisons, researchers should disclose which chains, tokens and offchain routes are included; how wallets are labeled; whether transfers involving exchanges and smart contracts are excluded; and whether a classification change restates prior periods. The September Visa refresh demonstrates why this matters: more labels lowered adjusted volume even though the underlying historical transactions did not disappear. A decline after a methodology change cannot be treated automatically as customers leaving the market. Likewise, rapid reported growth can reflect better identification of payment processors rather than entirely new commerce.
Four numbers that answer four different questions
The global supply of stablecoins measures the outstanding token balance at a point in time. Allium put that stock at about $303 billion in August 2026. It can support many payments during a year because a token can change hands repeatedly. Comparing annual payment value with supply therefore measures turnover, not the fraction of coins assigned permanently to commerce. A token held on an exchange as trading collateral can also circulate among traders without paying a supplier once.
Gross transfer volume is a flow across addresses. It is closer to counting every movement of a banknote between tills, vaults, intermediaries and customers than to counting only purchases. The $62 trillion annual figure in the Allium and BCG study is a broad record of such transfers. Its approximately $4.2 trillion classification of real economy activity uses filtering, and its $350 billion to $550 billion conservative bilateral goods and services estimate narrows the definition further. All are flows, but their scopes differ materially.
The cross border share introduces a new denominator. FXC Intelligence and Allium compared classified stablecoin use with $44.3 trillion in retail international payments for 2025. Its 0.31% estimate says little about wholesale treasury settlement, exchange transfers or domestic payroll. A higher value in a later year may reflect more genuine usage, improved identification or a changed market size. A comparable time series needs the same definition on both sides of the fraction.
Visa’s card volume is a network measure. It is based on transactions through card programs, some of which use a stablecoin for funding while the merchant is paid through conventional acquiring and settlement. The 200% growth refers to those programs, not a 200% rise in all stablecoin payments or the fraction of every Visa purchase. The $20 billion annualized stablecoin settlement run rate is another measure of obligations among network partners. One customer purchase can generate both a card transaction and a later settlement step.
Keeping these four categories separate prevents a common error in market stories: adding supply, transfers, completed purchases and wholesale settlement into one supposed adoption total. The values are related but not mutually exclusive. A token can be issued, sent to an exchange, passed through a payment provider, redeemed and used to settle a card obligation in a series of entries around one commercial event. The useful question for a cross border feature is where the independently completed obligation sits in that chain.
How a payment processor might prove its claim
A processor can report the number and value of customer instructions it received, the number actually completed, the origin and destination corridors, and the local currency credited to beneficiaries. It can disclose median end to end time and the distribution of delays, rather than the fastest blockchain confirmation. If it quotes costs, those should include funding, foreign exchange spread, blockchain fees, compliance charges, redemption and payout. The quote should specify who bears each cost. Such disclosure would permit a comparison with bank and card alternatives without exposing personal transaction details.
The difference between a test and recurring business is visible in the Lloyds and Visa pilot. A $750,000 settlement demonstrates that the route can work under the conditions of that seven day trial. It does not report a quarterly count of independent customers or a completed end user remittance program. A bank’s internal obligation to another institution is a legitimate cross border settlement, but it answers a different question from whether a family can send a small amount abroad more cheaply and reliably.
Merchant acceptance is similarly easy to overstate. A stablecoin linked card can let a consumer spend a token at a conventional terminal while the retailer receives its local currency from an acquirer. That is meaningful utility for the cardholder and an issuer, but it is not direct stablecoin acceptance by the merchant. Direct acceptance requires the seller to receive and manage the token or contract with a processor to do it on the seller’s behalf. The settlement chain determines where the stablecoin actually changes hands.
Public researchers face limits that processors do not. A processor can match an invoice, verified counterparties and a payout record. An independent analyst sees token movements and may infer intent from labels and behavior. Better labeled addresses improve the model but do not make an unlabeled customer wallet self describing. The Visa dashboard’s September revision is evidence of how much improved identity information can change adjusted volume after the transactions have already happened.
Regulation and redemption shape the route
A payment token’s usefulness depends on its issuer’s promise to redeem and the firms willing to accept it. A dollar stablecoin may remain close to $1 across major exchanges, yet a recipient in a thin local market can face a wide spread when cashing out. Banking relationships can halt deposits or withdrawals even as the blockchain runs continuously. Providers must satisfy rules for identity, sanctions, travel information and consumer protection in their operating jurisdictions. The blockchain transfer is one part of that regulated service.
For a company, an international payment has an accounting life after settlement. It needs to record the invoice, the exchange rate, any fees, the identity of the payee and whether tax or reporting obligations arise. A recipient who keeps a stablecoin on a balance sheet takes issuer and custody exposure until redemption. A firm that converts immediately needs reliable market depth and a payout partner. These constraints explain why a fast pilot may grow slowly despite sound technical performance.
The institutional route can still have advantages. A company that routinely pays suppliers in several time zones may reduce prefunded balances and weekend delays if stablecoin liquidity and local payout are reliable. A recipient without practical access to dollars through banks may value a dollar denominated token. The advantage is corridor specific. It must be demonstrated through delivered amounts and repeat transactions rather than extrapolated from total transfers on the issuing chain.
The data available in October establishes that classified payments exist at scale and that some card and bank programs are live. It does not establish the fraction of every public transfer that represents a cross border purchase. The question can be answered more precisely as providers share completed transaction data and researchers maintain stable classifications across time.
What to watch
The next useful disclosure is a consistent series of completed payments by purpose and corridor, with the sender’s total cost and the beneficiary’s received amount. Payment processors can publish those operational measures without exposing private customer identities. Updated Allium classifications and Visa’s methodology log will show how much a changed label set alters the historical totals. More bank and card programs can demonstrate repeat use, though a pilot should remain a pilot until it reports volume at commercial scale.
The direction of travel is clearer than an exact global market share. Real payment cases exist and classified volume has grown. Yet the trillion dollar transfer headlines mostly describe the motion of tokens through trading and financial plumbing. The estimate of commerce depends on classification, while the best cross border test includes both ends of the payment and all the costs between them.
FAQs
Is every stablecoin transfer a payment?
No. Transfers can reflect trading, collateral, bridging, issuance, internal treasury moves or a payment. A blockchain event alone does not reveal the underlying obligation.
How much payment volume did Allium estimate for 2026?
Its September study estimated $401 billion to $527 billion for the first eight months of 2026 across classified stablecoin payments, with methodological uncertainty.
What share of cross border payments used stablecoins?
Allium and FXC Intelligence estimated about 0.31% of a $44.3 trillion retail cross border market in 2025. It is a defined estimate for that year, not the share of every international financial transfer.
Why are the $4.2 trillion and $350 billion to $550 billion figures different?
The former is a broader real economy activity category in the Allium and BCG study; the latter is its conservative estimate of observable bilateral payments for goods and services in 2025.
Do Visa stablecoin cards mean merchants receive stablecoins?
Generally the programs use familiar card acceptance, and the merchant may receive conventional currency. Stablecoins can be used in funding or settlement behind the transaction.
Does a cheap blockchain fee make a cheap remittance?
Not necessarily. Funding, foreign exchange, redemption, compliance and local payout costs determine what the sender pays and recipient receives.
Can blockchain data tell whether a transfer crosses a border?
Only with reliable information about counterparties and their location. Addresses alone have no geographic nationality.
What would improve measurement?
Consistent labels, published methodology revisions and processor data on completed payments, corridor, cost and beneficiary amount would narrow the uncertainty.
Disclaimer: This article is for information and educational purposes only and does not constitute financial or investment advice. Figures reflect regulatory filings and reporting available at the time of writing and change with each disclosure. Nothing here is a recommendation to buy, sell, or hold any security or asset. Always do your own research. Information is accurate as of October 7, 2026.
Be the first to comment