Report contents
Industry landscape
Where the company competes: two issuers, thin switching costs, powerful distributors
Circle Internet Group issues USDC, a dollar-denominated stablecoin, and earns most of its revenue from interest on the cash and short-term Treasury bills that back it. Strip away the crypto vocabulary and the business is a licensed money issuer running a very large, very short-duration bond portfolio, funded by holders who expect to redeem at par on demand. The profit pool is the spread between what the reserves earn and what Circle pays to put USDC into circulation.
Upstream, the suppliers are ordinary and replaceable: custody at BNY Mellon, reserve management through a BlackRock-managed money market fund, monthly attestation by Deloitte. None of them captures Circle's margin, and none is a bottleneck in the way a sole-source component supplier would be. Downstream is where the economics get decided. Exchanges, custodians, wallets, payment processors and merchants sit between Circle and the end holder, and the largest of them have negotiated a share of reserve income on the USDC they hold. Coinbase's arrangement is the big one. That is why distribution expense, not reserve yield, is the main swing factor in Circle's results: the company books the issuance margin but keeps only part of it. Circle's non-reserve revenue, from minting fees, cross-chain transfer tools and tokenized fund products, is real but secondary.
The competitive arena is narrower than "crypto." It is dollar stablecoin issuance for payments, settlement and dollar savings, sold to exchanges, fintechs, remittance corridors and, increasingly, merchants. USDC circulates globally, but demand clusters in three places: offshore dollar users in countries with weak currencies who want a dollar claim they can hold in a phone; trading venues that need a dollar leg; and payment corridors in developed markets where settlement speed matters.
Geography cuts two ways. The network is global, which means scale in liquidity and integrations compounds. Permission is national. The GENIUS Act, signed in July 2025, created a federal framework for payment stablecoins in the United States; the EU runs a separate regime under MiCA; and issuers still piece together state money-transmitter licenses. A stablecoin has to be admitted market by market, so the addressable footprint expands jurisdiction by jurisdiction, not all at once.
| Issuer | Token | Backing | Reporting | Main demand |
|---|---|---|---|---|
| Circle | USDC | Cash and short-dated Treasuries | Monthly attestation; SEC reporting | Payments, settlement, offshore dollar savings |
| Tether | USDT | Treasuries plus gold, secured loans and other assets | Quarterly attestation | Offshore trading and store of value |
| PayPal | PYUSD | Cash and Treasuries | Attestation; trust charter | PayPal and Venmo flows |
| Ethena | USDe | Crypto collateral hedged with futures | Public dashboard | Yield-seeking crypto users |
| Banks (e.g. JPMorgan) | Tokenized deposits | Commercial bank deposits | Bank disclosure | Institutional settlement |
Dollar stablecoin issuance is concentrated at the top: USDT and USDC are the two large issuers, and the rest of the field is a long tail of small tokens. Applying the practical test, the leaders can be counted on one hand, which points to meaningful barriers. Those barriers are trust and licensing, not technology. A stablecoin lives on a promise to redeem at $1, and one visible break can cost an issuer its franchise: when Silicon Valley Bank failed in March 2023, USDC traded below $0.90 because part of its reserve cash was stuck at the bank. Reserve quality and custody arrangements are the product. Licensing adds a second wall, since compliance, audits and exchange integrations cost money a startup cannot easily front.
The catch is on the demand side. Switching costs for a holder are close to zero: redeeming USDC and buying USDT takes minutes, and a merchant that accepts one can accept the other at no extra cost. Network effects are therefore moderate, not binding. That is why the distribution layer, not the token, extracts the rent.
Regulation cuts both ways, and this is the most important open question. The GENIUS Act legitimizes the product and shuts out outright frauds, but it also builds a clear compliance path for well-capitalized entrants that already own customer relationships: PayPal, bank consortia, payment processors. The barriers are high enough to keep the field short and low enough that the newcomers are large. Circle's position is defensible on reserve quality and regulatory standing, and it is not protected by lock-in.
Competitive advantage
Does the company have a moat? Yes, though distributors collect much of it
Circle Internet Group is one of two large issuers of dollar stablecoins, and the barriers that keep that field short are genuine: a dollar token is a promise to redeem at par, backed by reserves a holder can inspect and licenses a regulator can verify. What a reader may miss is where the rent lands. Circle books the reserve income and then pays a large share of it to the exchanges, wallets and payment processors that put USDC in front of users. The moat protects the franchise. The margin gets renegotiated with a handful of counterparties.
Concentration passes the first test. Two issuers dominate dollar stablecoin supply, and the compliance, attestation and integration costs of entry are high enough that the tail of small tokens has not closed the gap. Sustained excess returns are a messier test. ROE ran 43.9% in 2023 and 34.2% in 2024, turned negative in 2025, and works out to roughly 9.5% for the first nine months of 2026. A barrier that lets the kept spread swing from 23.6% to 8.7% and back is weaker than the concentration alone implies.
Market position: second of two large issuers, and few others count
Market position: Fairly Strong. USDC ranks second to Tether's USDT and sits with it in the top tier; every other dollar token is small by comparison. Rank and tier carry the argument here, because two credible issuers rather than twenty is what an entry barrier looks like when trust is the product. Circle's edge is narrower than leadership: it is the audited, compliance-first alternative, priced against an incumbent with more circulation.
Stability: the standing holds; the pricing does not
Stability: Moderate. No dated share series for 2021-2026 establishes whether USDC gained, held or lost ground, so five-year stability cannot be confirmed, and the call rests on structure. Redeeming USDC and buying another dollar token takes minutes at close to no cost, and a merchant that accepts one can accept the other with no added work. Reserve quality and licensing hold the franchise in place. Habit and integration do not.
Revenue growth is not evidence of share gain. Revenue went from $1.45 billion in 2023 to $2.75 billion in 2025 and $2.25 billion in the first nine months of fiscal 2026, per the company's statements, but reserve income rises with token balances and with short-term rates. A larger base and higher rates lift revenue even while a rival grows faster.
Profitability: high ROE on a thin capital base
Returns: Moderate. ROE of 43.9% in 2023 and 34.2% in 2024 looks like clear evidence of advantage until you check the denominator. Equity was $570 million at the end of 2024 and $3.86 billion at September 30, 2026 after the June 2025 IPO, so the same dollar profit now supports a far larger base. ROA, which fits a business whose balance sheet is its factory, was 0.74% in 2023, 0.43% in 2024 and roughly 0.4% for nine months of 2026.
A standard after-tax ROIC does not apply. Invested capital as equity plus debt minus cash came to about negative $140 million at the end of 2024 ($570 million of equity plus $40 million of debt less $750 million of cash), because the earning assets, the Treasury bills and cash backing USDC, are funded by holders' redeemable claims. The 95% liabilities-to-assets ratio at September 30, 2026 is those claims. Read the business as a spread: reserve yield minus what Circle pays to distribute the token.
Margins: the kept spread moved 15 points in a year
Margins: Moderate. Gross margin was 23.6% in 2024, 8.7% in 2025 and 24.0% in the first nine months of 2026; net margin went from 9.3% to -2.5% to about 15%. The 2025 net loss was $70 million against operating cash flow of $540 million, so the swing was more accounting than cash. Even so, a company that surrenders 15 percentage points of gross margin in one year and recovers most of it over the next three quarters is negotiating its margin, not setting it. The September 2026 quarter, at 34.1% gross margin against 17.8% in June, shows how much rides on that split: at a revenue run rate near $3 billion, one point of gross margin is worth about $30 million a year, an estimate from the reported figures.
Overall rating: Moderate
Overall rating: Moderate. Circle holds a real position: a two-issuer top tier, licenses and a reserve record a new token cannot buy quickly, and revenue that grew from $80 million in 2021 to $2.75 billion in 2025. Profitability is where the moat thins out. Holders switch at almost no cost, the largest distributors take a negotiated share of reserve income, and the 2025 margin shows how little Circle keeps when that split moves the wrong way. Market position rates Fairly Strong and profitability Moderate; the overall rating follows profitability, because a barrier that cannot hold a margin does not hold much.
Where the advantage comes from and how long it can last
The advantage has three sources: trust in the reserve, the license perimeter, and liquidity scale. The single metric that tests all three is the share of reserve income Circle keeps after paying distributors, which shows up as gross margin. In the first nine months of 2026, gross profit equaled about 0.9% of average total assets annualized against revenue of about 3.8%, an estimate from reported figures that overstates the reserve yield slightly because some revenue is fees. Distribution absorbs the difference.
Reserve quality: the trust barrier is real, and it breaks first
Source rating: Fairly Strong. A dollar token is diligence made liquid. The holder is buying an attestation, a custody arrangement and a redemption promise, and moving to a rival for a few basis points of yield is not worth re-underwriting any of that. That is customer captivity of a narrow kind: it binds the holder to the token without binding the holder's operations to it. The asset side is a commodity, with Treasury bills held at BNY Mellon and managed through a BlackRock fund. Nothing there is proprietary, and no supplier captures Circle's margin. The barrier is a record of par redemption plus the reporting that supports it, which a new issuer earns by surviving stress.
What is that worth at the point of pricing? Less than the trust suggests. The split with distributors is the largest cost line, and in 2025 it ate nearly the whole spread: revenue rose about 64% while gross profit fell to $240 million. A trust barrier that lets the kept spread fall that far protects the token while the economics leak away. What breaks it: a reserve shortfall or a second depeg. The March 2023 episode, when USDC traded below $0.90 on a bank failure, showed how fast that asset reprices. The test is whether a repeat would now be cushioned by the stronger reserve mix.
Licenses: the wall is high and the gate is wide
Source rating: Fairly Strong, with a catch. Entry runs through the GENIUS Act framework signed in July 2025, the EU's MiCA regime and a patchwork of state money-transmitter licenses. Compliance is a fixed cost that a startup cannot front and a large company absorbs without noticing. Circle gains from having built the apparatus early and from a reserve mix that fits what the rules permit: cash and short Treasuries. Tether's backing, which includes gold and secured loans, is costlier to conform if it wants the US market, an inference from the reserve rules rather than a reported win.
The catch is that the same statute that legitimizes USDC hands a clear path to firms that already own the customer. PayPal issues its own token to its own user base, and bank consortia can do the same inside closed networks. While the rules bar issuers from paying yield to holders, the token stays a commodity and the advantage goes to whoever controls placement. If the regime ever permits interest-bearing stablecoins, the economics shift decisively toward the distributors. What breaks the license moat: a large distributor issuing its own token and defaulting its users to it, at which point Circle's compliance standing keeps it in business without keeping its margin.
Liquidity and integrations: scale that helps without locking anyone in
Source rating: Moderate. Deeper liquidity means tighter on-chain spreads and more trading pairs, which pulls in the next integration, and the fixed costs of compliance and market coverage spread across a bigger base. The problem is that a network effect needs captivity to become a barrier. Here the holder and the merchant are indifferent to which dollar token clears, so the loop never closes. Circle's distribution scale is largely rented: the reach belongs to the exchange or the processor, and the price of that reach is the split that moved gross margin 15 points. If a top distributor launched a competing token, depth would not save USDC, because the integration was never exclusive to Circle.
Durability: the scarce asset is the license and the record
Substitutes arrive from outside the token market. A bank deposit, a card rail and a tokenized deposit from a commercial bank all meet the settlement need, and JPMorgan's tokenized deposits already serve institutional settlement inside its own network. None of them needs to look like a stablecoin to take the use case. Against those alternatives, Circle's durable assets are its regulatory standing and a multi-year record of redemption at par; USDC's brand recognition is real and is not captivity, because the wallet owner decides.
The open question is whether Circle can hold a larger share of the reserve spread as the license perimeter widens and well-capitalized distributors enter. The record since 2024 does not settle it: the kept spread has swung by 15 points within eighteen months, a range that looks negotiated, and the September 2026 recovery to 34.1% is one quarter, not a trend. If gross margin holds above 30% for four quarters while reserves keep growing, the moat would look stronger than Moderate. If it slides back toward the 2025 level with revenue still rising, the barrier protects the token and someone else collects the profit.
Generated by DeepVal AI. Data and models may contain errors or be outdated. For research only; not personalized investment advice.
Continue reading the full report
- Comparison and summaryHow Circle Internet Group, Inc. compares with its closest rivals, and our moat rating
Free to sign up · 20 research points · unlocking this report uses 10
20 points cover 2 full reports. No card needed.
Research reflects its generation date. Preview the first two modules.
More on Circle Internet Group, Inc.
Circle Internet Group, Inc.
Business, moat and key questions
Open company page