What happens when the same token that moves trillions in transfers suddenly becomes the currency used to pay network fees? That question sits at the heart of Circle’s Arc launch. USDC transfers climbed 49 percent in the third quarter to $13.9 trillion, yet average circulation only rose 8 percent to $73.3 billion. The gap between those two numbers is not just a curiosity. It shows how much activity can grow without a matching rise in the dollars people actually keep on hand. Arc opened on September 16 with USDC as its gas token, giving payment firms a single unit for both settlement and transaction costs. The real test is whether that convenience turns into lasting demand or simply reshuffles existing balances.
USDC Transfer Growth Meets Arc’s Gas Design
The numbers look impressive at first glance. Quarterly on-chain transfers jumped nearly half while the average amount of USDC in circulation edged up only modestly. Circle laid out the figures in its latest quarterly filing. A payment provider that receives USDC steadily throughout the day can recycle those inflows into new payouts. Another firm that must send money before customer funds arrive has to keep a larger buffer ready. The difference in working capital can be substantial even when the total payment volume looks identical.
Arc gives those firms a network where both the payment itself and the fee can be denominated in USDC. That removes one friction point. Yet the bigger variable for demand remains how long each dollar stays parked between settlements. Speed helps. So does the ability to net opposing flows. Neither factor shows up cleanly in a single gross-volume figure.
How Working Balances Shape Real Demand
Imagine a provider handling $100,000 in supplier payments every business day. If enough USDC arrives each morning and the entire day’s batch can be funded from those receipts, the starting balance might stay close to one day’s volume plus a modest operating cushion. If the firm must pre-fund five business days of activity, the same schedule suddenly requires $500,000 before any buffer is added. Same payment volume. Very different inventory needs.
I’ve found that this timing mismatch is where many discussions about stablecoin demand go sideways. People see large transfer numbers and assume the token is being held in equal measure. In practice the opposite can be true. Faster settlement can shrink the buffer some firms need. Recurring commitments can push others to keep more dollars on-chain simply because they cannot afford to miss a payout window. Both effects can run at the same time for different users on the same network.
A provider might also hold extra inventory because it cannot always top up liquidity the moment a customer wants to send money. The gap between a digital-dollar transfer and the availability of a local bank payout system is real. Sending the token is one step. Arranging conversion and meeting local banking cutoffs are others. Those operational realities determine how much USDC actually stays parked.
Turnover Ratios Tell Only Part of the Story
Divide $13.9 trillion in quarterly transfers by $73.3 billion in average circulation and you get a rough turnover of about 190 times. Using the reported growth rates to reconstruct the prior-year figures produces something closer to 137 times. Those ratios are approximate because the company reports rounded numbers, yet they still highlight a clear trend. Measured activity accelerated far faster than the amount of USDC held across the network.
The ratios cannot tell us that a typical unit of USDC changed hands 190 times or that every transfer represented a payment for goods and services. The same inventory can cycle through exchanges, trading pools, or settlement services repeatedly while other balances sit untouched. What the calculation does show is that gross activity pulled away from the stock of dollars available.
Circle Payments Network’s annualised volume sits in a different category altogether. The network reached $41.3 billion in annualised payment volume. That figure is a run-rate, not a twelve-month total, and it should not be added to the quarterly on-chain transfer number as if the two cover the same activity. Earlier in the year the annualised rate stood at $14.7 billion based on a thirty-day window. The acceleration is real. Whether it translates into higher balances held specifically on Arc is a separate question.
What Actually Needs Measuring on Arc
To understand demand on Arc itself, the useful data would identify addresses by role. Fee sponsors, market-making inventory, and firms paying customers or suppliers all look similar in raw transaction data. Treating every address with a stablecoin balance as a payment user risks counting trading inventory or internal transfers between accounts under common control.
With proper classification, two comparisons become possible. First, settled payment volume against the average balances of participating providers over the same period. Second, the change in those balances against net inflows onto Arc and net USDC issuance across all networks. Together those measures can separate growth in activity, a shift of existing funds from other chains, and genuine new demand for newly issued digital dollars.
Arc was public for only the final weeks of the quarter. Company-wide figures therefore cover far more activity than the new network alone. The next full quarter will give the first clean three-month window after launch. That period should start to clarify whether payment firms are parking meaningful working balances on Arc or simply routing existing flows through a new rail.
Gas Fees in the Same Currency Change the Budget Conversation
Using USDC for gas means businesses can budget network costs in the same unit they use for dollar settlement. That is a practical advantage. A firm can still manage the two balances separately. The account funding a large batch of transfers may hold only enough for fees, while a treasury or payments account carries the much larger sum owed to customers. Platforms that sponsor gas for users can maintain their own balance and fold the cost into pricing. In that setup the end customer rarely sees a separate gas line, yet the sponsor still needs USDC on hand.
The dollar denomination makes comparison with conventional service charges easier. It does not fix the fee for every transaction. The computational work required and network conditions at the moment of submission still matter. A simple transfer, a currency exchange, and a multi-contract operation will carry different costs. Batching also changes the relationship between the number of customer payments and the number of network operations. Ten individual payouts can become fewer settlement transactions between intermediaries. Counting network operations alone can therefore understate the number of real customer payments.
For demand purposes, the balance held by the fee payer and the amount spent over time measure different things. A provider that spends $1,000 in fees over a month might keep far less than that amount available at any moment if it tops up frequently. Another service that pre-funds for months can maintain a larger standing balance relative to observed spending. Both approaches are rational. Only the second adds more persistent USDC demand.
StableFX Introduces Its Own Inventory Question
Circle has brought stablecoin foreign exchange to Arc through StableFX. The request-for-quote model lets participants seek a price for exchanging one currency token for another. Payment-versus-payment settlement links the two sides so that delivery of one asset can be conditional on the other. That design reduces certain settlement risks, yet it still requires someone to hold inventory.
A firm converting USDC into a local-currency token cares about three prices at once: the exchange rate itself, the fee to settle, and the cost of turning the received token into funds that can be paid out locally. A competitive quote addresses only the first leg. Liquidity providers need assets ready to deliver on the quotes they make. That can mean holding several currencies and moving balances as demand shifts. Inventory that supports a regular two-way market can be reused. A series of large one-way trades forces rebalancing or wider spreads.
A dollar balance on Arc can therefore serve a purpose even if it never belongs to the person making the final payment. A liquidity firm keeps USDC to quote. A payments business holds it to fund settlements. A customer holds it to pay suppliers directly. If all three interact in the same transaction chain, an observer who treats every movement as new demand will overstate the picture. The standing demand comes from whoever retains the inventory needed to keep those routes open.
Perhaps the most interesting aspect is how routes serving the same customer can require different working balances. Where the recipient accepts USDC, the provider can deliver the asset directly. Where the recipient wants a local bank deposit, the provider needs conversion and a payout partner. The same dollar amount can therefore involve different inventory levels even before fees enter the calculation.
Moving Existing USDC Is Not the Same as Creating New Supply
Circle’s interoperability tools move supported assets among networks. In a burn-and-mint transfer the amount destroyed on the source chain and recreated on the destination represents a relocation of the same economic exposure. Arc can show an increase in local USDC while the source network shows a matching decrease. An observer looking only at Arc might interpret the mint as new supply. An observer tracking the token across networks sees an offsetting transfer.
The same distinction applies when one business buys USDC from another holder. The buyer’s balance rises and the seller’s falls. The purchase supports demand at the transaction level, yet it does not require Circle to issue a new unit unless someone further along the chain ultimately needs freshly minted coins. Net growth across the entire system still depends on issuance and redemption taken together.
Circle disclosed roughly $104 billion in gross minting and a similar amount of redemption during the quarter. The totals are rounded, so subtracting them cannot produce a precise net change. Large gross flows on both sides can coexist with a much smaller shift in the amount circulating. An analyst comparing Arc balances over time with activity on the networks from which funds arrive can begin to separate temporary settlement routes from balances that stay with identifiable providers for weeks.
Earnings Tell a Parallel but Different Story
Circle reported $865 million in third-quarter revenue and reserve income, up 17 percent. Within that total, reserve income came in at $658 million, down 8 percent from a year earlier. Higher average circulation helped, but lower returns on the reserve portfolio more than offset the volume growth. The company also recognised roughly $182 million in other revenue from the ARC token presale. That line is separate from income earned on the reserves backing USDC.
The distinction matters for how the quarter is read. Total revenue can rise while reserve income falls because commercial activity beyond the reserve portfolio is included. Presale proceeds and the amount recognised as revenue in a given period can also differ under accounting rules. Treating the two figures as independent sales would double-count the activity.
For a business evaluating Arc as a payment rail the distinction is more practical. USDC is the asset used for fees and can be the asset sent to counterparties. The ARC presale relates to a different token and different commercial arrangements. Its presence in the results does not quantify merchant balances or tell a supplier how much USDC it needs for its own settlement activity.
Payout Routes Can Generate Activity Without Leaving Merchant Balances
Circle’s planned acquisition of Tazapay is relevant precisely because the final recipient of a payment often wants ordinary bank funds even when the sender or payment provider prefers a stablecoin. Payout networks connect those preferences by handling conversion and distribution. The stablecoin can be central to the provider’s settlement process without ever appearing in the recipient’s account.
A local payout provider needs the right currency when it makes the payment. It may pre-fund a bank account, arrange credit, or convert incoming funds as they arrive. A dollar stablecoin can help move value between partners, yet the demand for USDC depends on which arrangement the provider chooses and how long the funds remain in token form. One provider might receive USDC, convert it immediately, and retain only a small gas balance. Another might keep a substantial portion of its dollar liquidity as USDC because it expects to fund several transactions later in the day. Both can report stablecoin-based activity. Only the second leaves a meaningful ongoing balance at the provider.
Circle expects the Tazapay transaction to close in 2027 subject to regulatory approvals. The company notes that the provider covers more than 100 markets. Those local routes can support higher transfer volumes while the amount of USDC actually held still depends on each provider’s treasury choices. An integration announcement shows that software and counterparties are lining up. The operational details that matter are which corridors go live, what share of settlements actually uses Arc, and whether providers hold balances overnight.
A payment provider’s inventory needs can also differ across the currencies on a single route. It may be able to replenish its USDC account around the clock while depending on a local banking partner that processes outgoing payments only at set times. An instruction that arrives after the local cutoff forces a choice: wait, pre-position additional currency, or route through another partner. Those decisions affect how much balance is needed, in which currency, and for how long. Faster on-chain settlement changes only one part of the sequence.
Practical Signals Worth Tracking Going Forward
Several concrete measures would help clarify whether Arc is generating lasting payment demand. Identifiable operating wallets that hold USDC between settlement cycles give a direct view of working capital. Recurring payments by the same providers across multiple weeks show whether activity is sticky. Net global issuance measured separately from USDC simply relocated onto Arc distinguishes new dollars from relocated ones. Persistent StableFX balances across the currencies used on active routes reveal liquidity committed to keeping markets open. Finally, the commercial rollout of Tazapay—launch dates, approved markets, and any reporting on live volumes—will show how local payout connectivity translates into on-chain activity.
Daily balance snapshots can miss important detail. A provider might hold substantially more USDC while processing an afternoon batch and return excess inventory to a treasury account before the day ends. A midnight snapshot would omit that intraday demand. Conversely, a large balance observed just before a payout batch can exaggerate what the firm usually carries. Tracking comparable intervals over several cycles helps separate a recurring need for funds from a balance assembled for one unusually large day.
The public chain records transfers between identified accounts, yet some of the steps that determine the customer outcome take place elsewhere. A payout to a bank account produces a different record from a transaction on Arc. An intermediary can group many customer payments into one settlement or use several transfers to complete a single instruction. Interpreting counts as completed payments requires knowing how the provider uses its wallets and how those movements map to instructions in its own system.
Putting the Pieces Together
USDC’s role as both settlement asset and gas token on Arc is a clean design choice. It simplifies budgeting for payment firms and removes one source of friction. The growth in transfer volume relative to circulation shows that activity can expand without a proportional rise in balances held. Whether that pattern continues on Arc depends less on the network’s technical design and more on the treasury practices of the firms that use it.
In my experience the most revealing questions are rarely the loudest ones. Gross transfer numbers attract attention. The quieter metrics—working balances retained between settlement cycles, the share of activity that stays on Arc rather than bridging away, and the inventory committed by liquidity providers—will ultimately determine how much lasting demand the network creates. Those figures are only beginning to become available.
The next few quarters should provide clearer data. Arc will have a full reporting period behind it. Payment providers will have more time to settle into routines. The Tazapay integration, once closed and operational, will add local payout options that could either increase turnover or encourage larger standing balances, depending on how firms manage the conversion step. Until those details emerge, the most accurate statement is that USDC is moving more, circulating only modestly higher, and now serving as the fee token on a new institutional network. How much of that activity turns into durable demand remains an open and interesting question.
One practical way to frame the outlook is to separate three distinct effects. First, higher payment volume that simply moves existing USDC through the system faster. Second, a shift of balances from other networks onto Arc without any change in global supply. Third, genuine new issuance driven by firms that decide they need larger permanent working capital in digital dollars. Only the third effect expands the overall stablecoin base. Arc’s design makes the first two easier. Whether it also encourages the third is the development that will matter most for long-term demand.
Payment firms themselves will decide that outcome through ordinary operational choices. How large a buffer they keep, how quickly they recycle inflows, whether they pre-fund corridors or rely on just-in-time liquidity, and how they manage the conversion to local bank funds will shape the inventory they hold. Those decisions are not glamorous, yet they determine the size of the balances that stay on-chain between transactions. Watching those balances over successive quarters will give a clearer picture than any single headline transfer number.
The design of Arc removes one historical friction by letting firms pay gas and settle in the same token. That convenience is real. It does not automatically create new demand. Demand still comes from the need to keep dollars available for future commitments. Faster settlement can reduce that need for some users. Greater reliability and broader local payout options can increase it for others. The net effect will become visible only as identifiable providers report or reveal the balances they actually retain.
Until then the story remains one of rising activity measured against relatively stable inventory. USDC transfers are higher. Circulation is modestly higher. Arc now offers a network where both the payment and the fee can live in the same unit of account. The question of how much lasting demand that combination creates is still being answered in real time by the firms moving the money.
