Circle's Guidance Increase Is Almost Entirely an Arc Presale Illusion

Total revenue and reserve income reached $701.3mn, up 7% year-on-year, while adjusted earnings before interest, taxes, depreciation and amortization (adjusted EBITDA) reached $143mn, up 8%. USD Coin (USDC) in circulation declined sequentially to $73.3bn at period end, but average circulation, a more representative measure of the base generating reserve income, rose 25% year-on-year to $76.5bn.

The increase in other revenue guidance does not reflect organic acceleration. It is entirely driven by the expected recognition of revenue from the ARC token presale and masks a modest downward revision to the business excluding Arc. This change reflects both a less supportive crypto environment and Circle's decision to reallocate part of its resources towards Arc development.

The higher-for-longer macro backdrop supports reserve income in the near term but extends the model's structural dependence on an interest-rate-sensitive revenue stream. At the same time, the announcement of Open USD, a rival stablecoin-distribution consortium backed by Visa, Stripe and BlackRock among others, in late June 2026 introduces new competitive pressure through a model that shares reserve income with distributors, although the project is not yet operational.

Why the guidance increase is misleading

Circle's Q2 results delivered solid year-on-year growth but more mixed sequential trends. Reserve income increased by 5%, other revenue by 41% and adjusted EBITDA by 8%. Revenue less distribution costs margin (RLDC margin) reached 41.2%, compared with 38.2% a year earlier, but declined slightly from the first quarter. USDC in circulation fell from $77.0bn to $73.3bn at period end, while average circulation increased from $75.2bn to $76.5bn. Onchain volume declined from $21.5tn to $14.8tn, although the comparison was distorted by approximately $9tn of exceptional repricing activity on Aerodrome, a decentralized exchange on the Base network, in the previous quarter.

The main takeaway concerns the increase in other revenue guidance, from a range of $150mn to $170mn to a range of $310mn to $330mn. Circle now includes $180mn of expected revenue from the ARC token presale, while guidance excluding Arc has been reduced to a range of $130mn to $150mn. The increase is therefore more than fully explained by Arc, as presale revenue offsets a $20mn reduction in the business excluding ARC across the entire range. RLDC margin excluding Arc is expected to remain close to 39%.

This revision mainly reflects the accounting recognition of non-recurring revenue rather than organic acceleration. Circle attributes the reduction in guidance excluding Arc to a less favourable crypto environment and to the priority given to Arc development, which limited certain new blockchain integrations. The presale covered 807.5mn tokens priced at $0.30 each and generated approximately $242.2mn in gross proceeds. These proceeds were recorded as deferred revenue and are recognised as the relevant performance obligations are satisfied.

Diversification remains limited. Subscription and services revenue increased by 58% year-on-year to $28.2mn but declined sequentially, while transaction revenue fell to $5.3mn. CPN reached $14.7bn in annualised volume without any contribution yet separately identifiable in the financial statements. At the same time, adjusted operating expenses increased by 23% to $146mn, highlighting the persistent gap between investment in new growth engines and financial materiality.

Reserve dependence, rates and distribution costs

The model remains highly concentrated. Reserve income represents $667.7mn of the $701.3mn in total revenue and reserve income, equivalent to 95.2%. It consists of interest and dividends generated by reserve assets and depends primarily on two variables: average USDC in circulation and the reserve return rate.

The reserve return rate declined by 66 basis points year-on-year to 3.48%. Circle estimates that this reduction lowered quarterly reserve income by $113.9mn, an effect offset by 25.2% growth in average circulation. Based on the sensitivity disclosed in Form 10-Q, a US Securities and Exchange Commission quarterly filing, a 100 basis point decline in interest rates would reduce reserve income over the following 12 months by $737mn and distribution and transaction costs by $360mn, implying a mechanical net impact of approximately $377mn.

The higher-for-longer environment therefore supports results in the near term without reducing the model's structural dependence. The main risk would be monetary easing before the new businesses generate a sufficiently meaningful level of recurring revenue.

The distribution mix is moving in different directions. The share of USDC held on Coinbase increased from 21% to 30%, while the share held directly on Circle rose from 10% to 17%. The off-platform share declined from 69% to 53%. The net economic impact remains difficult to isolate because it depends on average balances and contractual terms. Hyperliquid, a high-performance decentralized derivatives exchange, illustrates this complexity. At the end of June, 90% of its associated USDC was held on Coinbase and 10% on Circle, with no published detail on the precise revenue-sharing arrangements.

Operational growth engines

Subscription and services revenue increased from $17.8mn to $28.2mn year-on-year, up 58%, and represents the largest component of other revenue. This category includes integrations and maintenance, fund management, time-based access and user licences, without a more detailed breakdown. It therefore does not consist entirely of recurring revenue. Sequentially, revenue declined by $7mn, which management attributed to fewer blockchain integrations as resources were prioritized towards Arc. Among the three growth engines considered, this is the only one whose contribution is already directly visible in the income statement, although it remains modest at group level.

CPN reached $14.7bn in annualized volume at the end of the quarter, calculated using the previous 30 days, up 76% sequentially. The network had enrolled 175 institutions, an increase of 29%. By 31 Jul, annualized volume had reached approximately $23bn, while CPN and the associated payment products covered more than 58 countries. Management expects to begin monetising the network in the second half of 2026 but has not specified the timing, the expected size of the contribution or how it will be presented in the financial statements.

Circle National Trust opened on 24 Jul, initially providing fiduciary digital asset custody services to Circle and its affiliates. On 31 Jul, Circle also received a limited purpose trust company charter from the New York Department of Financial Services (NYDFS) for Circle Internet Trust Company. USDC reserve management remains a future capability. The quarterly documents do not yet disclose third-party custody contracts, a commercial rollout timetable or pricing. The regulatory advantage is therefore tangible, but its conversion into revenue remains to be demonstrated.

These three growth engines are at different stages of maturity. Subscription and services already generate revenue, although the contribution remains modest at group level. CPN is entering its monetisation phase. Circle National Trust remains primarily a regulatory and operational capability. The next calibration point will be the announced start of CPN monetisation in the second half of 2026.

Arc's optionality moves into operational testing

Arc represents the main source of long-term growth optionality presented by Circle. Chairman and CEO Jeremy Allaire believes the network could represent a larger opportunity than USDC itself. Its public mainnet is scheduled to launch on 16 Sept.

Pre-launch traction is significant. As of 30 Jun, the testnet had processed 502mn transactions across 2.8mn transacting wallets, with uptime of 99.99%. More than 100 partners were also participating in the private mainnet as of 20 Jul. The validator cohort includes BlackRock, Depository Trust & Clearing Corporation (DTCC), Mastercard, Visa and Standard Chartered. BlackRock plans to deploy BUIDL, its tokenized money-market fund, on Arc, while DTCC is expected to enable the tokenization of assets held by The Depository Trust Company.

The ARC presale provides a one-off contribution to 2026 results but does not indicate the recurring economics of the network. These will depend on effective mainnet adoption and the development of revenue from staking, network transactions and partnerships, for which neither timing nor materiality has yet been quantified.

The 16 Sept launch will therefore represent the first operational test. The main indicators to monitor will be the number of applications actually deployed, economic volumes, fees generated, validator activity and the integration timetable for BUIDL and assets tokenized by DTCC.

Indicators to monitor

Reserve income. Movement in the reserve return rate, meaning the average return generated by reserve assets, relative to the path of the Federal Reserve, average USDC circulation rather than the period-end balance, and the distribution of balances across Circle Platform, Coinbase and off-platform holdings.

Subscription and services. Sequential and year-on-year growth, any recovery in blockchain integrations following the launch of Arc, and any future breakdown between licences, integrations and fund management fees. Disclosure of customer retention and concentration metrics would also improve visibility into the quality and recurring nature of these revenues.

Circle Payments Network. Initial revenue from the monetisation expected in the second half of 2026, growth in Total Payment Volume (TPV) and the number of institutions enrolled, together with any disclosure of a take rate or an identifiable contribution in the income statement.

Circle National Trust. Initial digital asset custody contracts with third-party customers, the timetable for any internalization of reserve management and disclosure of a pricing model.

Arc. Following the mainnet launch, the number of active applications, economic volumes and fees generated, the deployment of BlackRock's tokenized BUIDL fund, integration with DTCC, effective validator participation and the emergence of recurring revenue sources identified by management, including staking, transaction fees and partnerships. The accounting recognition of the ARC token presale should be monitored separately from the operational economics of the network.

Three developments could alter the pace of the transition. Sustained high reserve yields would extend the monetisation window. The emergence of an identifiable CPN take rate would accelerate diversification. Wider adoption of reserve income sharing models by major distributors would weigh on the economics of the current model.

The second quarter doesn't settle the structural investment case. It confirms the resilience of the reserve income engine and the reality of operational traction, but also the persistent gap between adoption and recurring financial contribution. The launch of Arc and the announced start of CPN monetisation will represent the next meaningful calibration points.