On a Tuesday in the French National Assembly, a budget amendment passed that will be reported, in English, as "France taxes stablecoin transactions." That sentence is wrong in almost every clause. It is wrong about the tense — the law is not in effect. It is wrong about the object — the text targets "certain stablecoin swaps," a phrase with no defined boundary. And it is wrong about the mechanism — what passed is not a new tax but a re-characterization of an existing one. In my 2017 audit of fifteen early Ethereum ICO contracts, I learned to distrust the summary and read the diff. Three of those projects had reentrancy holes their whitepapers described as "advanced security architecture." The gap between the document and the code was where the money died. The same gap now sits between a press release and a tax statute, and the market is reading the press release.
The Regime Nobody Bothered to Read
France does not have a crypto tax problem. It has a crypto tax definition problem, and the two are routinely confused by people who have never opened Article 150 VH bis.
Under the current regime, French tax residents are taxed on crypto gains at a flat 30 percent — the prélèvement forfaitaire unique, or PFU. That rate is not the interesting part. The interesting part is the trigger. Under existing law, a taxable disposal occurs only when a crypto asset is exchanged for fiat currency. Crypto-to-crypto swaps are treated as deferred — no taxable event, no gain recognized, no paperwork beyond the eventual fiat exit. This is the architecture that has governed French crypto taxation since the 2019 reforms, and it is the architecture the new amendment is quietly trying to demolish.
That matters because everything downstream — the enforceability, the economic distortion, the political contradiction — flows from a single unstated premise: if a crypto-to-stablecoin swap becomes a taxable event, then the French state has decided that a stablecoin is a fiat equivalent for tax purposes. This is not a parameter adjustment. It is a reclassification of an entire asset class, executed through a budget amendment, with no public impact assessment and no defined scope.
The timing is not accidental. The EU's DAC8 directive — the tax administrative cooperation framework that obliges crypto asset service providers to report client transactions — applies from January 1, 2026, with first reporting in 2027. The OECD's CARF framework begins first exchanges in 2027. France is legislating now because for the first time, the state will possess a data pipeline it can actually cross-check. You cannot audit what you cannot see, and until 2026, the French tax authority could not see very much at all. The amendment is domestic-law preparation for a surveillance capability that has not yet fully arrived.
The Asymmetry That Should Have Led the Story
The reporting on this amendment has bundled three proposals together: the stablecoin swap tax, an exit tax on departing residents, and a self-hosted wallet reporting obligation. Read together, they look like a coordinated tightening. They are not. They are three provisions at wildly different stages of maturity, and conflating them is the single most expensive analytical error available to a reader right now.
Consider enforceability as a spectrum rather than a binary.
A stablecoin swap executed on a centralized exchange or through a MiCA-licensed crypto asset service provider leaves a centralized, queryable record. The DGFiP — the French tax directorate — can request that record directly. The data already exists because the provider is already obliged to collect it under anti-money-laundering rules. Enforceability here is high, and it is high today, before DAC8 even lands.
A self-hosted wallet, by contrast, has no third-party intermediary. There is no exchange holding the ledger. Enforcement would rest entirely on the taxpayer's voluntary declaration, reinforced by the threat of penalty and supplemented — in theory — by chain-clustering analysis. That is not enforcement in any operational sense. It is paper compliance, a self-reporting obligation whose detection rate depends on the taxpayer's own honesty and the state's forensic budget, neither of which is reliable at scale.
These two provisions are not peers. One rides on infrastructure that already exists. The other is a wish. Yet they are reported as a pair, and the pairing invites a false conclusion: that self-custody is about to become taxable in France. It is not. What is close to landing is the swap tax, because the data is close to landing. What is far from landing is the self-custody regime, because the data does not exist and will not exist without a level of surveillance that would trigger its own constitutional fight.
The exit tax sits even further out. France's existing exit tax applies to taxpayers departing the country while holding more than €800,000 in equity. Extending it to crypto requires resolving three problems that have no clean answer: at what moment do you value a portfolio that can move in seconds, how do you reconcile the tax with constitutional protections on freedom of movement and property, and how do you prevent double taxation under bilateral treaties that were never drafted with self-custodied assets in mind. Each of those problems is a litigation vector. The probability that the exit tax survives intact is materially lower than the probability that the swap tax does.
The Real Technical Problem: A Stablecoin Is Not Always a Stablecoin
Here is where the amendment's drafters have walked into a trap of their own construction, and where my audit background makes me flinch.
If the taxable event is "crypto to stablecoin," then the natural question is: which stablecoin, and against what cost basis? The naive assumption is that the taxpayer converts volatile crypto into a dollar-pegged token and books a clean gain or loss. But the actual behavior of anyone who has used a decentralized exchange is messier. People swap USDT for USDC. They swap EURC for USDC to move between euro and dollar liquidity. They rotate between yield-bearing and non-yield-bearing stablecoin wrappers. These are same-asset conversions — the economic substance does not change. You started with a dollar and you ended with a dollar.
Taxing that is taxing a non-event. And the technical cost of complying with a non-event tax is not zero. To report a USDT-to-USDC swap correctly, a taxpayer must maintain a multi-currency cost-basis ledger, track acquisition history across wallets, apply a consistent accounting method (FIFO or LIFO — the law does not yet specify), and reconcile all of it against exchange statements that may not agree with each other. The compliance cost of taxing a no-gain swap can exceed the tax owed on it by an order of magnitude. That is not a theoretical concern. It is the operational reality of anyone who has tried to produce a clean crypto tax filing from a year of DeFi activity, and I have watched accountants give up on exactly this problem.
Worse, the reclassification creates a philosophical contradiction the state has not acknowledged. MiCA — the regulation France spent years negotiating and championing — classifies stablecoins as electronic money tokens or asset-referenced tokens. It treats them as payment instruments. The French tax code, if this amendment passes as drafted, would treat them as taxable exit points. The same asset, in the same jurisdiction, under the same government, would be simultaneously a means of payment and a trigger for capital gains. Regulatory philosophy and tax philosophy would point in opposite directions, and the taxpayer would be caught in the gap.
There is a hidden assumption buried in the amendment that deserves to be dragged into the light: the drafters almost certainly intend to capture crypto-to-stablecoin, not stablecoin-to-fiat. The latter is already taxable under current law — no new amendment is needed to tax a fiat exit. So the only reason to legislate is to close a path that was previously tax-free: the crypto-to-stablecoin leg. And the only reason that path matters is that stablecoins have become a de facto fiat substitute — a buffer layer people use to sidestep the disposal trigger. The state is not taxing stablecoins because it dislikes stablecoins. It is taxing stablecoins because people started using them to avoid a tax the state thought it had already closed. That is a reasonable diagnosis of a real leak. The remedy is where it goes wrong.
The Lock-In Effect Is the Point, Not the Revenue
Let me be blunt about what this tax will actually do, because the fiscal framing obscures it.
The core economic effect of taxing stablecoin swaps is not revenue generation. It is a lock-in effect. Every conversion between asset forms becomes a friction point, and rational actors respond to friction by not moving. They hold. They stop rebalancing. They leave collateral in place rather than rotating it. The tax does not extract money from the market so much as it freezes the market in place.
This is the Laffer curve arriving in crypto. Raise the cost of a taxable action and the action shrinks — and when the action shrinks, the tax base shrinks with it. A French resident who would have realized a gain and paid 30 percent on it now has a strong incentive to simply not realize it, either by holding indefinitely or by relocating their tax residency. The state collects less, not more. The people who remain compliant are the ones with the least ability to leave, which is to say the least mobile and often the least wealthy. The tax is regressive in practice even though it is flat in rate.
The distortion is sharpest in DeFi, and this is where the market has badly underestimated the damage. DeFi's core operations — collateral posting, liquidation, rebalancing, cross-protocol arbitrage — almost all route through stablecoins as the intermediate medium. If every hop constitutes a taxable event, then the marginal cost of participating in DeFi as a French resident rises structurally. This is not "pay a little more tax." It is a permanent increase in the cost of the basic unit of DeFi activity. A French user facing that cost has three options: stop using DeFi, use it through offshore structures, or leave. None of those options is what a government trying to build a crypto hub wants.
We didn't build a tax system for assets that move. We built one for assets that sit still — real estate, equities held for years, art in a vault. Crypto assets are defined by their mobility, and applying a static-asset tax logic to a mobile-asset class produces a predictable outcome: the assets move, and the tax base stays empty. France is about to rediscover this the way every jurisdiction that taxed friction before it did.
The Contradiction France Refuses to Name
Here is the part of the story that the French crypto press will underplay and the English press will miss entirely: France is taxing stablecoins while simultaneously trying to become Europe's stablecoin capital.
Paris has spent years positioning itself as the EU's crypto hub — first through the AMF's PSAN licensing regime, then through its influence in shaping MiCA. Société Générale's crypto arm, SG Forge, issues EURCV, a euro-denominated stablecoin. Circle's EURC operates in the same market. These are the flagship European stablecoin projects, and they are French-adjacent by design. Euro stablecoins already suffer from thin liquidity; their order books are shallow, their market-making is fragile, and any additional friction gets amplified rather than absorbed. If the amendment taxes stablecoin-to-stablecoin swaps, the market makers who provide euro-stablecoin liquidity will face a direct cost to the exact activity that keeps their books alive. They will not pay it. They will route around France.
A government cannot credibly promote a stablecoin industry with one hand and tax its use into illiquidity with the other. The signal this sends to SG Forge and to every euro-stablecoin issuer weighing a French footprint is that the tax authority and the industrial policy authority are not talking to each other. That is not a small coordination failure. It is the kind of contradiction that takes years to unwind because once a market maker leaves a jurisdiction, the liquidity does not come back just because the tax is later repealed.
And then there is Ledger.
I have spent enough time inside French crypto circles to know that Ledger is the country's most important crypto company, and its entire brand rests on a single promise: not your keys, not your coins. Self-custody is not a feature of Ledger's product. It is the product's reason for existing. If the French state enforces a self-hosted wallet reporting obligation, it does not merely inconvenience Ledger's users — it places the company in a legitimacy bind with its own home government. Every line of code writes a history of power, and a hardware wallet company whose founding narrative is distrust of centralized authority, being asked by that authority to help enumerate self-custodied holdings, is a narrative collision that no marketing budget can paper over. This is a reputational exposure, not a financial one, which is precisely why the market will not price it until it is too late to matter.
The Legislative Chain Is Longer Than the Headline
"Parliament passed" is the most misleading four words in this entire story, and it is the phrase that will anchor every retail read of the amendment.
The French budget process is not a single vote. It is a gauntlet. The National Assembly's passage of an amendment is the opening move, not the closing one. From here the text must survive the Senate — whose finance committee has a long tradition of independent scrutiny of crypto provisions and has rewritten them before. It must then survive the commission mixte paritaire, the joint committee that reconciles the two chambers' versions and where clauses routinely disappear. It must survive constitutional review by the Conseil constitutionnel, which has struck down tax provisions before on procedural grounds — most notably when they are attached as riders that have no genuine budgetary purpose. And it must survive the possibility that the government invokes Article 49.3 to force the budget through without a vote, in which case the clause's fate depends on political bargaining rather than technical merit.
The real trading signal is the final text, not the parliamentary passage. Everything between now and then is noise, and the market is currently pricing the noise. My estimate is that less than five percent of the implications are priced in — not because the market is irrational, but because the information chain is so long that most participants have correctly decided to wait. That is the right instinct. The wrong instinct is to treat the amendment's passage as a fait accompli and act on it.
What I would actually watch is narrower than the headline. First, whether the Senate finance committee defines "certain stablecoin swaps" or leaves it empty. Second, whether the swap tax and the self-custody reporting survive as a pair or get separated — because if they are separated, the self-custody provision is almost certainly dead and the swap tax is almost certainly alive. Third, whether the final text includes any transition period, because a transition period is the tell that the drafters understood the lock-in effect and chose to soften it. Fourth, and most telling, whether the DGFiP publishes implementation guidance before the law takes effect — because a tax authority that issues guidance is a tax authority that intends to enforce, and one that stays silent is a tax authority that is still building its data pipeline.
The Contrarian Read: This Is a Data Bill Wearing a Tax Bill's Clothes
Everyone is analyzing this amendment as a revenue measure. I think that is the wrong frame, and the wrong frame leads to the wrong forecast.
Consider what the self-hosted wallet reporting obligation can actually achieve. The state cannot effectively tax self-custodied assets — there is no intermediary, no withholding point, no clean valuation moment. So what does a reporting obligation produce, if not tax revenue? It produces data. It produces a registry of who holds what, at what addresses, in what quantities. In a jurisdiction that cannot yet see self-custodied wealth, the first move is not to tax it. The first move is to map it. Taxation comes later, once the map exists.
This is the classic progressive path of tax administration. You begin with a declaration requirement. You use the declarations to establish a baseline. You then cross-reference that baseline against the data that DAC8 and CARF will deliver from 2026 onward. You find discrepancies. You open audits. You tighten. At no point does the state need to tax self-custody directly on day one — it needs only to see it, and then let the data do the work.
The same logic explains why France is legislating ahead of the EU's data infrastructure rather than waiting for it. Legislating now buys political soundness and a head start on the compliance apparatus. It also creates a risk the drafters have not addressed: duplicate reporting and conflicting definitions once DAC8 and CARF come online. A French taxpayer could end up reporting the same transaction twice under two different standards, with two different valuation methods, to two different authorities. That is not efficiency. That is bureaucracy layered on bureaucracy, and it is exactly the kind of friction that pushes activity offshore — reducing the very regulatory visibility the amendment was designed to increase.
Which brings me to the deepest irony. If the goal is to make French crypto activity legible to the state, taxing stablecoin swaps and mandating self-custody reporting may achieve the opposite. Users who find onshore compliance too costly do not stop using crypto — they move to offshore platforms or peer-to-peer channels where the state sees nothing at all. The amendment would convert a visible, regulated, reportable population into an invisible one. That is a self-defeating design, and it is the outcome I would bet on if the text passes as drafted.
What Comes Next
Truth emerges from transparency, not from silence, and the French state has chosen silence on the three questions that determine this amendment's actual weight: which swaps, at what rate, from when. Until those are answered, every forecast is a guess dressed as analysis.
My forward judgment is narrow and testable. The stablecoin swap tax will likely survive in some form, because the data to enforce it exists and the political cost of taxing crypto is low. The self-custody reporting obligation will likely be delayed, diluted, or deleted, because it cannot be enforced without a surveillance apparatus France has not built and may not be allowed to build. The exit tax will likely be the first casualty, because it carries the heaviest constitutional baggage. And Ledger will face a legitimacy question it has not yet been asked, in a home market that built its brand on the promise the amendment quietly threatens. Watch the Senate. Watch the definition. Watch whether the data pipeline arrives before the tax does — because in France, right now, the order of those two events is the whole story.