There's a peculiar thing happening in Brussels right now. A stablecoin issuer — one of the most regulated, most transparent, most institutionally acceptable issuers in the entire industry — is quietly arguing against being forced to hold its reserves in banks.
Read that again.
The company that built its brand on regulatory friendliness is pushing back on the most conservative reserve rule European lawmakers can write. Circle, the issuer of USDC and EURC, wants the EU to swap a mandatory bank-deposit floor for what it calls "flexible liquidity rules." On its face, this looks like a firm asking to be held to a lower standard.
It isn't. Or rather — it isn't only that. And the gap between "looks like" and "is" is where the entire stablecoin regulatory war of 2026 is being fought.
I've been auditing token contracts since the EtheriumGold days in Prague, back when integer overflows were a feature of the landscape rather than a warning label. I've watched the industry learn, repeatedly, that the safest-sounding structure is often the one with the most concentrated hidden risk. This Circle story is the same lesson wearing a suit.
Let me explain what's actually on the table — and why almost everyone covering it is describing the wrong conflict.
Context: How We Got a Stablecoin Lobbying War
To understand why Circle is spending political capital on a reserve-composition rule, you have to understand what MiCA actually is — and what it was designed to prevent.
The Markets in Crypto-Assets regulation is the European Union's attempt to do something no major jurisdiction had done before: build a single, comprehensive legal framework for crypto assets that functions across all 27 member states. It entered application in stages, with the stablecoin provisions — the ones governing EMTs (electronic money tokens) and ARTs (asset-referenced tokens) — arriving first because they were deemed the most systemically urgent.
The logic was straightforward. Stablecoins are, functionally, money. They are used as settlement rails, as collateral, as the base unit of account for a huge share of on-chain activity. If a stablecoin breaks, the damage propagates instantly through DeFi, exchanges, and now — increasingly — traditional payment corridors. So the EU decided to treat issuers less like crypto startups and more like financial institutions.
That decision carried a structural consequence. Under MiCA, the largest issuers — those designated "significant" EMTs — face enhanced requirements. And one of those requirements, as generally understood, pushes a portion of reserves toward bank deposits. The reasoning is intuitive: bank deposits are insured, they're supervised, they sit inside the regulated banking perimeter. A stablecoin backed partly by insured deposits looks, to a regulator, like a safer stablecoin.
Circle operates in Europe through Circle France, which holds an EMI (electronic money institution) license. That license is its passport. Under MiCA's single-market design, a firm authorized in one member state should be able to serve the entire EU without seeking separate approval in each country. This is the "passport" principle, and it is the beating heart of the whole framework — the thing that makes MiCA worth more than the sum of its national parts.
So Circle has two live concerns. First: what assets count as acceptable reserves, and at what proportions. Second: whether the passport actually works in practice, or whether member states quietly erect local-presence requirements that gut the single market in all but name.
Both concerns surfaced in the context of MiCA's ongoing review. And both, I'd argue, are being misread by the market.
Core: The Reserve Philosophy War
Here's the thing nobody says plainly. The fight over stablecoin reserves is not a fight about safety. It's a fight about what kind of risk you're allowed to pretend doesn't exist.
When a regulator mandates that a stablecoin hold a slice of its reserves in bank deposits, they are making an implicit claim: bank deposits are the safe asset. When Circle argues instead for short-term government bonds, money market funds, and reverse repos, it is making a competing claim: liquidity, not insurance, is what makes a reserve safe.
Neither claim is wrong. Both are incomplete. And the incompleteness is where the actual economics live.
A bank deposit concentrates risk in a single counterparty. A Treasury bill spreads risk across a market.
That's the sentence the entire debate hinges on, and it almost never appears in the coverage.
Think about what a bank deposit actually is. You hand money to a bank. The bank owes you that money back, plus (historically) a small interest rate. The safety of your claim depends on the solvency of that specific institution. If the bank fails — and banks fail with depressing regularity, 2023 being the most recent reminder — your claim depends on deposit insurance, which itself has limits, and on the resolution regime, which is a political process as much as a financial one.
Now think about a short-term Treasury bill. You hold a claim on a sovereign. The risk is not counterparty solvency in the same concentrated sense — it's duration and interest-rate risk. If rates move against you and you're forced to sell before maturity, you take a loss. If you can hold to maturity, you don't. So the risk profile is fundamentally different: market risk instead of credit risk, spread instead of concentrated.
Which is safer? Wrong question. The right question is: safer against what?
Against a single bank failure, the T-bill wins. Against a systemic rate shock that forces fire-sales, the insured deposit wins — assuming the insurance holds. Against inflation, the T-bill wins by a mile, because bank deposits have historically paid rates that barely keep pace, while short-term government paper tracks the policy rate more closely.
And that last point — the yield gap — is the part Circle's public narrative carefully avoids.
I learned this lesson the hard way during the DeFi Summer of 2020, when I was pulling apart Aave's governance mechanics and Compound's collateral factors in parallel and realizing that the interest — who captures it, who bears the risk of it — is almost always the real story underneath the technical story. When someone tells you a change makes something "safer," the first question should always be: safer for whom, and who pays for that safety?
Apply that lens to Circle.
Circle's revenue model is not complicated. It holds reserves against USDC and EURC. Those reserves earn yield. Circle keeps the yield. Users of USDC get zero interest — the token is a payment instrument, not a savings account. So the spread between what the reserves earn and what Circle pays out (nothing) is, functionally, the company's gross margin.
This is not a scandal. It's the disclosed business model. But it means that every basis point of reserve yield matters enormously at Circle's scale. Reserves in the tens of billions of dollars, multiplied by a yield differential of even one or two percentage points, is a number with a lot of zeros.
Now layer on the rule in question. If MiCA forces a meaningful portion of reserves into bank deposits, and if bank deposit rates in the EU sit materially below short-term government paper rates — which, historically and structurally, they tend to — then the rule doesn't just change Circle's risk profile. It directly taxes Circle's revenue.
So when Circle argues for "flexible liquidity rules," it is simultaneously making a defensible prudential argument and protecting its income statement. Those two things are not in conflict. They point the same direction. And that alignment is exactly why you should be suspicious of anyone who presents the argument as purely about safety.
Let me be precise, because this is where a lot of commentary goes wrong. I am not saying Circle is lying. I am saying that a self-interested party making a technically correct argument is still a self-interested party, and the correct response is to evaluate the argument on its merits while pricing in the incentive. This is basic. It's the same discipline I applied back in 2017 when I found the integer overflow in that copycat ERC-20 contract and chose to publish rather than sell — because the incentive to stay quiet was real, and the only way to trust my own analysis was to expose it to scrutiny.
So let's scrutinize.
What "Flexible Liquidity Rules" Actually Means
When Circle says "flexible liquidity rules," the operative word is liquidity, and it's doing a lot of work.
A liquidity-based reserve framework asks: can this asset be converted to cash, quickly, at a predictable price, under stress? Short-term Treasuries, money market funds, and reverse repos generally pass this test. They trade in deep markets. They mature soon. Their price volatility is low over short horizons.
A deposit-based framework asks a different question: is this asset insured and supervised? Bank deposits pass that test — up to insurance limits, within a supervised institution.
Notice that these are two different definitions of safety that can, in specific scenarios, contradict each other. During the March 2023 banking stress, deposits at certain institutions were less liquid and less certain than short-term government paper — the opposite of the naive assumption. Depositors lined up; the paper kept trading. That episode is seared into the institutional memory of every serious reserve manager, and it's almost certainly part of why Circle is comfortable making this argument now.
But there's a counter-consideration the liquidity framework has to answer for: duration risk. If a reserve is parked in a money market fund or a short-dated bond and rates spike, the mark-to-market value falls. If redemptions force sales into that dip, the issuer realizes losses. This is not hypothetical — it's the exact mechanism that broke a certain well-known stablecoin's peg in 2022, when a reserve composition that looked "safe" on paper met a redemption wave it couldn't absorb.
So the honest framing is this: Circle is not arguing for a safer reserve. It is arguing for a differently-risky reserve — one whose risks it believes it can manage, and whose yields it prefers to keep.
Both the EU's instinct and Circle's instinct are defensible. The EU is optimizing for a world where a stablecoin failure doesn't become a banking failure. Circle is optimizing for a world where its reserves earn market rates. Neither is obviously wrong. The question is which failure mode the EU fears more.
And that, finally, is the real variable: not the technical merits, but the regulators' memory. Did the EU learn more from 2023's bank failures, or from the stablecoin de-pegs? The answer to that question will decide the rule, and no amount of technical argument from Circle will override it.
The Passport Problem
The second half of Circle's push is quieter but arguably more consequential: preserving cross-border issuance.
On paper, MiCA already grants this. Authorize in France, serve the union. The passport is the whole point.
In practice, the passport has an implementation problem, and it's the kind of problem that never makes headlines because it's boring. Member states, for a variety of reasons — domestic financial politics, supervisory caution, the desire to keep regulated activity on their own soil — have a tendency to want localized arrangements. Local custody. Local entities. Local reporting. Each requirement, in isolation, sounds reasonable. Stacked together, they erode the single market into a patchwork.
For Circle, this is existential in a way the reserve rule is not. If the passport works, Circle serves 27 countries with one legal entity, one compliance stack, one balance sheet. Marginal compliance cost per new market approaches zero. That is the entire economic case for a pan-European stablecoin strategy.
If the passport doesn't work — if Circle needs a licensed entity, a local custodian, and a separate compliance apparatus in every meaningful market — then the marginal cost per market explodes, and the economics of serving, say, smaller member states collapse entirely. You end up with a stablecoin that serves Germany and France and ignores Latvia and Malta, which is precisely the fragmentation MiCA was built to prevent.
So Circle's cross-border demand is not a grab for special treatment. It's a defense of the framework's central promise. And here's the subtle part: Circle's interest and the EU's stated interest are aligned on this one. The EU wants the passport to work — that's the political achievement it wants to claim. Circle is effectively volunteering to be the test case that proves the framework functions. That's a much stronger position than the reserve argument, and I suspect Circle knows it.
The Euro Question Nobody Is Asking
Now the part I keep coming back to, the part that makes the reserve debate look almost quaint.
MiCA does not treat all stablecoins identically. It distinguishes between euro-denominated tokens and non-euro tokens, and for "significant" non-euro tokens, it imposes restrictions on their use for payments within the EU. The intent is clear and, from a monetary-sovereignty standpoint, entirely rational: the EU does not want a dollar-denominated instrument to become the default payment rail inside its own borders, displacing the euro in digital commerce.
This is where USDC — not EURC — becomes the sensitive asset.
Circle's flagship product is a dollar stablecoin. Its European expansion is built on a euro token, EURC, which is a genuine and useful product but occupies a much smaller market. If MiCA's review tightens the treatment of significant non-euro tokens, Circle's European business could be structurally confined to the euro-denominated lane — a lane where it competes against bank-issued euro stablecoins that carry the implicit backing of the banking system, and where the addressable market is a fraction of the dollar stablecoin market.
That is a far bigger risk to Circle than any reserve-composition rule. A reserve rule affects margin. A payment restriction affects whether the product can exist at all in its most important form.
And notice: Circle's public framing — "preserve cross-border issuance" — conveniently bundles the euro-sovereignty concern into a broader, more sympathetic request. It's a smart piece of framing. It presents a narrow, self-interested worry as a universal principle about market access. I'd do the same thing in their position. But a reader should be able to see the seams.
Where Circle Sits in the Ecosystem
Step back and look at the structure, because the positioning explains the leverage.
USDC is not just a token. It's infrastructure. It's the settlement asset that a huge swath of DeFi treats as the safe leg of a trade. It's collateral on lending protocols. It's a trading pair on every major exchange. It's increasingly a rail for payments and remittances and tokenized real-world assets. Downstream, an enormous number of systems depend on it being liquid, stable, and present on the chains they use.
The dependency runs one direction more than the other. A DeFi protocol can, in principle, swap USDC for another stablecoin. In practice, that's costly, disruptive, and risky — you're changing the assumptions of every integration, every oracle, every collateral parameter. Migration costs are high, and high migration costs translate directly into negotiating power. Circle can push for favorable rules precisely because its product is hard to replace. This is a network effect wearing a regulatory strategy as a costume.
But — and this is the tension — that same centrality makes Circle a target. The more systemically important a stablecoin becomes, the stronger the case for treating its issuer like a bank, with bank-like supervision. The endgame of stablecoin regulation, if you follow the logic far enough, is that successful issuers get regulated into the banking perimeter they were trying to avoid. Circle's current lobbying is, in a sense, a bid to define the terms of its own eventual containment before someone else defines them. That's a rational strategy. It's also a sign that the company sees the wall coming.
Contrarian: The Safety Narrative Is Backwards
Here's my contrarian read, and I want to state it cleanly because it runs against the instinct of almost everyone in this debate.
The framing that bank deposits are the "safe" reserve and market instruments are the "risky" reserve is historically backwards.
We are two decades into an era in which sovereign paper has repeatedly proven more reliable than bank credit in moments of acute stress. 2008. 2011. 2023. Each episode taught the same lesson: when the system seizes, the thing that holds value is the thing backed by a sovereign's taxing power, not the thing backed by a private institution's balance sheet. Deposit insurance is a promise — a good promise, a promise that has mostly held — but it is a promise made by a system that has itself required rescue.
So when a regulator says "put your reserves in banks to be safe," they are outsourcing the stablecoin's safety to the banking system's safety. That's not eliminating risk. It's relocating it — concentrating it inside a sector that is explicitly backstopped by the state and therefore implicitly Too Big To Fail, which creates its own moral hazard. The stablecoin becomes safe only insofar as the banks are safe, and the banks are safe only insofar as the state will rescue them, and the state's willingness to rescue them is a political variable that no prospectus can capture.
Circle's liquidity argument, stripped of self-interest, is actually the more intellectually honest one. It says: let us hold assets whose value is transparent, whose market is deep, whose risk is measurable and diversifiable. Don't force us to concentrate our safety inside a single industry's solvency.
The catch — and this is why I can't fully endorse Circle's position either — is that liquidity risk and credit risk are not interchangeable, and the framework Circle proposes quietly assumes that duration risk can be managed through discipline. That assumption is exactly the one that failed in 2022, when a reserve that looked liquid on paper met a redemption wave that turned paper liquidity into realized losses. Discipline is a promise too. Promises fail under stress.
So the genuinely contrarian conclusion is not "Circle is right" or "the EU is right." It's that both frameworks are optimizing for a different catastrophe, and the choice between them is a choice about which disaster you'd rather survive. The EU fears a stablecoin failure becoming a banking failure. Circle fears a banking failure becoming a stablecoin failure. They are both right to fear what they fear. And there is no reserve rule that eliminates both risks simultaneously, because the risks are structurally opposed.
Anyone selling you a clean answer here is selling you a story, not an analysis. Which brings me to the second blind spot: the assumption that Circle's success would be the industry's success. It wouldn't. Circle's competitive advantage is its compliance maturity — its license, its disclosure, its listed status. A rule change that helps Circle helps Circle. Whether it helps the broader stablecoin ecosystem depends on whether the change lowers barriers for everyone or simply rewards the incumbent who lobbied hardest. Lobbying power is itself a form of centralization, and it is rarely distributed evenly.
Takeaway: The Rule Is a Proxy
Watch the MiCA review not for the reserve percentage it lands on, but for the signal it sends about the EU's true priority. If the final framework relaxes reserve requirements toward liquidity, the EU has decided it trusts stablecoin issuers to manage market risk and would rather they hold high-quality liquid assets. If it tightens toward deposits, the EU has decided that the banking perimeter is the only perimeter it trusts — and that stablecoins are, ultimately, a banking product in disguise.
That decision will outlast Circle. It will outlast USDC. It will set the template that every other jurisdiction — Washington, Hong Kong, London — reads and reacts to. The reserve rule is not really about reserves. It's about who gets to define what safety means in the digital monetary system, and whether the incumbents of the old system or the builders of the new one get to write that definition.
Circle is trying to be in the room when the definition is written. Whether it succeeds matters less than the fact that it's trying — because the next five years of stablecoin regulation will be decided by who shows up to these arguments first, and who shows up armed with the better narrative.
The technical details will be negotiated in committee rooms nobody watches. The story will be told everywhere else. And in this industry, as I learned in Prague long before any of this, the story is usually the thing that moves first.
What the EU decides about reserves is a detail. What it decides about safety is the whole game.