Circle's Forty Markets and the Missing Denominator

Circle said forty. It did not say which forty.

That is the entire hard content of the disclosure. A number. A direction. A verb — targets. And no list. No corridor names. No licensing timeline. No float targets. No partner banks. No redemption latency figures. No date by which any of it becomes measurable, and no baseline against which "forty" is an increment.

I have spent a lot of my working life reading disclosures that look like this. In 2017 I sat with forty-five smart contracts from early-stage projects and read them line by line, because the whitepapers had stopped being useful somewhere around page two. Three of those contracts carried reentrancy paths that would have drained their pools. The marketing pages said "audited." The code said something else. The code does not lie, but it can be misunderstood — and the misunderstanding almost always begins with a number that has no denominator.

Forty markets is that number. Forty out of how many? Forty of what unit — sovereign jurisdictions, payment corridors, currency pairs, licensed entities, banking partners? Forty by when, at what cost, against what float?

This is not a piece about whether Circle is a good company. It is a piece about how to read an announcement whose only verifiable content is a count. In a sideways tape, where direction is scarce and narrative is abundant, the discipline that pays is the discipline of asking what a number is denominated in.

What Was Actually Said

The substance is thin, and I want to be precise about how thin, because precision here is the whole exercise.

Circle is targeting expansion into forty additional markets. Global demand for stablecoins is rising. That expansion could accelerate blockchain adoption broadly. And there is a financial-inclusion framing layered on top.

That is it. Two factual claims and two narrative claims. There is no protocol upgrade in this announcement. No code change. No new architecture. No supply data. No revenue guidance. No disclosure of which chains, if any, are involved. Anyone who tells you this announcement is a technology story is telling you about their own imagination, not about Circle's press release.

So the honest first move is to strip the narrative and ask what a market expansion actually requires at the operating level.

Circle is a regulated issuer of a fiat-backed payment token. USDC is not a governance asset and not a utility asset. It is a claim on a pool of short-dated US-dollar instruments, redeemable one for one, issued by a company that files with the SEC and trades on the NYSE. That structure determines everything downstream. It determines what "expansion" means, what it costs, and where the revenue actually lands.

I want to be blunt about one thing before going further, because it is the most common category error I see in this sector. USDC has no token economy in the sense that word is normally used. There is no emission schedule. No inflation curve. No staking yield funded by dilution. No governance capture. Applying a tokenomics framework to USDC is like applying a tokenomics framework to a checking account. The value capture sits in the equity — the listed shares of the issuing company — not in the token. The token is the product. The equity is the claim on the product's economics.

Once you accept that, the forty-market announcement stops being a crypto story and becomes a distribution story with a very specific arithmetic behind it.

The Float Is the Product

Here is the revenue model, stripped of decoration.

Circle holds reserves against outstanding USDC. Those reserves sit predominantly in short-dated US Treasury instruments and overnight repo. The yield on those instruments accrues to Circle. Against that gross yield, Circle pays distribution costs to partners who bring float onto the platform — most visibly the large exchange relationships that hold customer dollar balances in USDC. What remains is the company's operating margin.

So the core equation is simple. Gross reserve income equals float multiplied by the short rate. Net revenue equals that, minus distribution. Everything else — API fees, cross-chain fees, subscription services — is real but secondary in the current P&L.

This has an uncomfortable implication that the market routinely glosses over. Circle does not earn on volume. Circle earns on standing float.

Read that twice. A corridor can move two billion dollars of USDC in a month and contribute almost nothing to revenue if the average balance held in that corridor is small and transient. Conversely, a corridor with modest transfer volume but deep stored balances — households keeping dollar savings, businesses holding working capital — contributes real money to the income statement even though it looks quiet on a block explorer.

This is why the forty-market number is so hard to value. Markets are not revenue units. Float is the revenue unit. And float per market is exactly what the announcement does not disclose.

I learned this distinction the hard way in 2022, when I personally worked through the reserve proofs of five major lending protocols after the Terra collapse. What I found was not fraud. It was something more mundane and more dangerous: assets that looked large in aggregate but were concentrated in a small number of counterparties, so that the headline number was stable right up until the moment it was not. The aggregate was a comfort. The composition was the truth. I advised the five hundred members of my copy-trading group to exit three days before the broader crash, and the aggregate saving was around one point two million dollars. That call did not come from a forecast. It came from reading composition instead of size.

The same discipline applies here. Forty markets is a size claim. It says nothing about composition. And composition — which corridors, what float, what latency, what licensing cost — is where the entire economic question lives.

Why Forty Markets Is a Banking Problem

Here is the part of stablecoin expansion that almost nobody writes about, because it is boring and because it does not fit the on-chain frame.

A new market does not require a new blockchain. It requires a bank account.

Think about what has to exist for a person in a new corridor to actually use USDC the way the inclusion narrative promises. They need an on-ramp where local currency converts to USDC at a spread that does not eat the use case. They need an off-ramp where USDC converts back to local currency, settled into a local bank account, within a time window that makes the transaction worth doing. They need a licensed entity or a partner entity permitted to hold and move customer funds in that jurisdiction. They need anti-money-laundering controls calibrated to local rules. They need a treasury operation that manages local liquidity so the off-ramp does not run dry on a Friday afternoon.

None of that is a cryptography problem. All of it is correspondent banking, licensing, and treasury.

I built a slippage-protection bot for a hundred and fifty users in 2020, during the gas spikes that made ordinary swaps behave like auctions. We hit a ninety-four percent success rate by ordering transactions defensively and routing around the worst of the extractable value. That was a genuinely hard engineering problem and I am proud of it. But I want to be honest about what it taught me: the engineering was the easy part of keeping that community solvent. The hard part was the fiat side — the timing of conversions, the settlement windows, the counterparty that said the transfer would clear Tuesday and cleared it Thursday. The chain was deterministic. The banking rails were not.

That asymmetry is the correct lens for the forty-market announcement. Each new market is a new set of banking relationships plus a licensing file, wearing a crypto costume. The delivery timeline is governed by regulators and partner banks, not by block times. And the cost is front-loaded while the float is back-loaded — you pay the licensing and integration cost years before you know whether the corridor will hold balances.

This is why I read the announcement as a sales pipeline rather than a product launch. A pipeline is real. A pipeline is also not revenue.

CCTP Solves the Bottleneck That Was Not Binding

I need to address the technical piece, because it is the part sophisticated readers will reach for, and reaching for it is a trap.

Circle's genuine technical contribution is cross-chain native issuance. The mechanism is straightforward in outline: USDC is burned on the source chain, an attestation is produced by a permissioned signing service operated by Circle, and an equivalent amount is minted natively on the destination chain. There is no pooled custody contract holding user funds on the source side, which is precisely the failure mode that destroyed so many lock-and-mint bridges. The pooled-custody risk is removed. What replaces it is attester liveness and correctness — a permissioned service whose availability and honesty you now trust instead of a smart contract's collateralization.

That is a real improvement. It is also a narrower improvement than the marketing suggests. The trust model did not become trustless. It became differently centralized. You moved from trusting a pool to trusting a signer.

More importantly for this discussion: the cross-chain problem was never the binding constraint in the corridors the forty-market expansion is aimed at.

In a remittance corridor, the friction is not moving value from one chain to another. The friction is the last mile — getting from a token balance to spendable local currency, at a rate that beats the incumbent remittance operator, with settlement certainty. CCTP does nothing about that. It is an elegant solution to a problem that sits upstream of the actual bottleneck.

I have watched this pattern repeat across the sector for years. Teams solve the problem they can see in the code, then discover the market was blocked somewhere else entirely. In 2021 I sold out of my Bored Ape positions near the mid-year peak for roughly a hundred and eighty thousand dollars in profit, and the reason was not price prediction. It was that I had started tracking community retention metrics and found that the projects with the loudest roadmaps had the worst retention curves — because their actual bottleneck was never art or mint mechanics. It was whether anyone wanted the thing after the mint closed. They kept shipping features at the wrong layer.

Cross-chain transfer is the wrong layer for this expansion. The right layer is fiat settlement depth. Watch whether Circle discloses off-ramp partners and settlement windows. If it does not, the forty markets are cross-chain announcements in a corridor costume.

Not All USDC Is Circle's Float

Here is a technical distinction that matters enormously for anyone trying to verify this expansion independently, and it is almost universally ignored.

Not every token labeled USDC on a given chain is native Circle issuance.

When a chain lacks native support, third parties historically wrapped or bridged USDC, producing a derivative that carries the label and a different risk profile. That derivative does not add to Circle's outstanding float. It is a claim on someone else's claim. Only native issuance — minted by Circle, redeemable by Circle — contributes to the reserve base and therefore to revenue.

This sounds like pedantry. It is not. It means that if you try to measure Circle's expansion by summing USDC across chains using a token-screener, you will overstate the float. You will be counting derivatives. And you will be counting them precisely in the newest, least liquid markets — which is where the expansion claims are being made.

The correct measurement is narrower and harder: native issuance per chain, plus a jurisdiction attribution that Circle does not publish. The second part is the real problem. A token does not carry a passport. Native USDC minted on a global chain can be held by anyone anywhere. So the very concept of a "market" in this announcement is an organizational construct — a sales territory — not an on-chain observable.

That is the quiet admission inside the number forty. A market here is a business unit, not a measurable on-chain quantity. Which means the announcement is not falsifiable by looking at a block explorer. It can only be falsified by watching the company's reported float and the composition of its distribution costs over subsequent quarters, at a lag of months.

Trust is earned in drops and lost in buckets. A disclosure that cannot be checked in real time asks for trust in advance. That is not disqualifying. It is, however, a reason to size any conclusion accordingly.

The Corridor Metric Nobody Publishes

If I were building the diligence file on this expansion — and structurally, that is what I am doing here — I would not start with market count. I would start with two ratios per corridor.

The first is float-to-volume. Take the average standing balance of USDC held by non-institutional wallets attributable to a corridor, and divide it by gross monthly transfer volume in that corridor. A high ratio means dollars are being stored there. Storage is float. Float is revenue. A low ratio means money is passing through — often exchange rebalancing, arbitrage, or market-maker inventory — and passing through generates volume without generating income.

This single ratio separates real adoption from round-tripping, and it is the reason I am skeptical of every "stablecoin volume hits record" headline. Volume can be manufactured by a handful of automated actors moving the same dollars in circles. Float cannot. Float requires someone to decide to hold.

The second ratio is redemption latency: the median time from a merchant or user receiving USDC to spendable local currency landing in a local bank account. This is the actual product in a payment corridor. If it is measured in days, the use case is savings and hedging. If it is measured in minutes, the use case is payments. Those are different businesses with different float characteristics, different regulatory exposure, and different competitive dynamics.

Neither ratio appears in the announcement. Neither appears in most research coverage. And without them, forty markets is a number that could mean almost anything — from forty licensed corridors with deep float potential, to forty partnership memoranda with a website landing page.

I want to be fair here. Circle is not obligated to publish corridor-level economics, and no serious company would. My point is not that the disclosure is dishonest. It is that the disclosure is insufficient for the conclusion the market will draw from it. There is a gap between what was said and what will be heard, and that gap is where capital gets mispriced.

The Competitive Map, Honestly Drawn

The stablecoin sector is effectively a duopoly with a long tail, and the shape of that duopoly determines whether expansion into forty markets is opportunity or attrition.

The incumbent by size and by reach holds the largest share, is deepest in exactly the corridors being targeted, and built its position through early penetration of exchange settlement and emerging-market peer-to-peer networks. The approximate scale gap is on the order of two to three times in outstanding supply, though any specific figure should be checked against current data rather than taken from an article.

The challenger's differentiation is compliance architecture, reserve transparency, and institutional trust. That is a real differentiation. It is also a differentiation that matters most to the customers who need it least — regulated institutions, funds, payment companies with audit obligations. The retail user in a high-inflation economy choosing between two dollar tokens does not price in attestation standards. They price in whether the local agent will take it, whether the spread is tight, and whether the app on their phone supports it.

That is the honest competitive read. USDC's advantage is strongest where the volume is smallest, and weakest where the float is largest. Institutional corridors have deep balances per participant but few participants. Retail corridors have shallow balances per participant but enormous aggregate potential. The forty-market expansion is a bet that the second category can be won with the tools built for the first.

I am not certain that bet fails. I am certain it is not free. Tether's position in these markets is a network effect, and network effects do not yield to better documentation. They yield to better liquidity at the point of use, which is a local, expensive, relationship-driven business.

Compliance Is a Cost Structure, Not a Moat

Here is where I part company with the consensus reading.

The prevailing view is that stablecoin legislation in the United States and the European framework have handed the compliant issuer a durable moat. Raise the compliance bar, the argument goes, and the non-compliant competitor gets squeezed out of regulated venues, and share migrates to the compliant one.

There is something to this. Regulated venues do prefer regulated instruments. But the conclusion is overdrawn, for a reason that gets almost no airtime.

Compliance is replicable. It is expensive, slow, and tedious, but it is not defensible in the way a network effect is defensible. Any bank consortium, any payment processor, any well-capitalized entrant can buy the same lawyers, hire the same compliance officers, and file the same licenses. What they cannot easily buy is a decade of accumulated liquidity depth in a hundred local markets. The moat is on the side the narrative treats as commodity, and the commodity is on the side the narrative treats as moat.

I say this as someone who spent 2024 building a compliance checklist for AI-driven trading agents alongside two legal experts. The work was valuable and I would do it again. But I came out of it with a clear-eyed view of what compliance frameworks actually do: they raise the fixed cost of participation. Raising fixed costs favors incumbents in the short run and favors capital in the long run. It does not, by itself, create a durable advantage for any particular incumbent — because the constraint it imposes is a checkbook constraint, not a capability constraint.

So when I read that forty new markets are being targeted on the strength of a compliance position, my reaction is not skepticism about the compliance. It is skepticism about the assumption that compliance travels. It does not travel for free. Entering forty jurisdictions means satisfying forty regulatory frameworks, or some grouped subset of them, each with its own licensing timeline, capital requirement, and supervisory temperament. The compliance advantage that helps you in one jurisdiction is a cost center in the next.

That is the inversion worth holding onto. The regulatory clarity that the market is pricing as an accelerant is, operationally, a lengthening of the delivery schedule. Every new market adds a queue, not a customer.

The Freeze Function and the Inclusion Story

The financial-inclusion framing deserves direct scrutiny, because it is doing real work in this announcement and it is not neutral.

USDC, like any regulated fiat-backed token, carries issuer-level controls. The issuer can freeze balances and blacklist addresses in response to legal process or sanctions requirements. This is not a bug smuggled in by careless engineers. It is a design requirement for operating inside the regulated financial system, and it is the same power a bank has when it freezes an account.

But it does change the character of the inclusion claim. A dollar token that can be frozen at the issuer's discretion is a dollar token where the holder's access depends on a permissioned counterparty. That is a perfectly reasonable product. It is not the same product as self-custodied, permissionless value transfer, and the two get conflated constantly in this sector's marketing.

For the corridors being targeted, this matters concretely. Consider a user in a country with capital controls. The appeal of a dollar token is partly that it sits outside the domestic banking system. But the token is issued by a US-regulated entity that will honor legal process from US authorities and increasingly cooperate with local ones. The user is not outside the system. They are at the edge of it, on a permissioned ramp.

I do not raise this as an accusation. I raise it because the gap between the inclusion narrative and the actual trust model is where users get surprised, and surprise in this domain is expensive. In the silence of the dip, the weak hands break — and so do the assumptions people never examined about what they actually held.

If the forty markets are real, and if they are concentrated in economies with unstable currencies, then Circle is taking on a policy exposure that does not appear on any risk dashboard: the host jurisdiction's interest in restricting dollar substitution. That is not a technical risk and it will not be caught by an audit. It is a political risk that arrives as a licensing decision, sometimes overnight.

The Arithmetic of a Narrowing Spread

Now the part I think is most underpriced.

Circle's revenue is float times the short rate, less distribution. The float is growing slowly, by licensing and integration. The short rate is not under Circle's control and is not growing.

Run the arithmetic illustratively. If the average yield on reserve assets moves from roughly five percent to roughly three percent, float must grow by about two thirds just to hold gross reserve income flat. Two thirds. Not forty percent. Not forty markets. Two thirds of the entire outstanding base, added, merely to stand still on the top line.

That is the frame the announcement should be read inside. The expansion is happening into a rate environment that is, on any reasonable multi-year view, less generous than the one that produced the recent earnings base. Volume and float have to grow faster than the spread compresses, or the revenue line declines even as the market count rises.

This is not a prediction about rates. I do not make those. It is an observation about which variable dominates. For a business whose gross revenue is float times spread, a change in spread is a larger lever than a change in distribution reach, and the two are being announced in opposite directions at the same time.

The distribution cost line compounds the issue. A meaningful portion of gross reserve income flows out to partners who supply float. As competition for float intensifies — and it will, because every regulated entrant needs the same partners — the share retained by the issuer comes under pressure. Gross float growth can coexist with net revenue stagnation. That is the scenario the market is least prepared to price, because it does not show up in any adoption metric.

What Would Change My Mind

I have been hard on this announcement. Let me be equally clear about what would move me to a constructive read.

A published corridor list with named licensed entities. Not partners — entities, with license numbers and jurisdictions. That would tell me the pipeline has cleared regulatory gates rather than marketing gates.

A float disclosure segmented by region, even coarsely. Three buckets would do it: North America, Europe, rest of world. If rest-of-world float is growing faster than the aggregate, the expansion thesis is working and I would say so.

Redemption latency figures for a handful of named corridors. Even a range. Even a median with no distribution. This is the single most informative number a stablecoin issuer could publish, and the fact that nobody publishes it is the sector's largest disclosure gap.

And a distribution cost ratio — the share of gross reserve income paid out to partners. That one number tells you whether the moat is widening or the toll road is getting more expensive.

Absent those, what we have is a count. Forty. A number without a denominator, a timeline, or a cost basis. It is not nothing — pipelines have value and Circle has earned the benefit of the doubt on execution more than most. But a pipeline is not float, and float is not revenue, and revenue in this business is a function of a rate the issuer does not set.

The question I would put to anyone pricing this announcement as a win: which of the forty, and how much float, and by when? If the answer is that we will know in a year, then the honest position today is that we are holding a narrative, not a measurement. The code does not lie, but it can be misunderstood. So can a press release. The difference is that the code will eventually tell you which one you were reading.