You live in France or you spend long stretches there, you walked into your French bank to open a savings account, and you walked out with an assurance-vie. In France this is the most ordinary savings product in the country: a life-insurance contract that works, in practice, as a tax-advantaged investment envelope. Your banker probably told you that after eight years the tax treatment becomes very gentle, that you can name anyone you like as beneficiary, and that the money passes outside your estate. All of that is true under French law. None of it protects you from the United States tax system. If you are an American citizen or a United States tax resident, that same contract is very likely a passive foreign investment company, usually shortened to PFIC, in the eyes of the Internal Revenue Service. That single classification changes everything: the French tax advantage is invisible to the IRS, the gains can be taxed as ordinary income with a backward-looking interest charge going back over every year you held the contract, and each year of silence adds another unfiled information return, Form 8621, to the pile. This article explains both sides of the trap and the way out. On the American side, you will learn how the PFIC test works, why a French assurance-vie and the French stock-savings plan called the plan d’épargne en actions almost always fail it, what the default excess-distribution regime costs, which elections can soften the blow, and which forms document the fix. On the French side, you will learn why unwinding the contract in a hurry can cost real French tax, how the eight-year clock, the single flat-rate levy, the one-PEA-per-person rule, and the estate rules on beneficiary designations shape every exit, and what three recent court decisions teach about second PEAs, contested beneficiaries, and premiums that heirs claw back into the estate. Read the whole piece before touching the contract, because the right sequence is to measure the American exposure first, price the French exit cost second, and only then redeem, transfer, or reorganize.
I. Why Your French Assurance-Vie Looks Like Tax Evasion to the IRS
A. The PFIC Test Your French Insurer Has Never Heard Of
Start with the American definition, because it is the one that bites. Section 1297 of the Internal Revenue Code treats any foreign corporation as a passive foreign investment company when either of two thresholds is met. The income test is met where, in the words of the statute, at least seventy-five percent of the corporation’s gross income for the year is passive income, or the asset test, satisfied when at least half of the average assets held during the year produce, or are held to produce, passive income. The full text is published at 26 U.S. Code section 1297. A French insurance company that manages your assurance-vie, and the fund structure inside many unit-linked contracts, earns overwhelmingly dividends, interest, and capital gains, which are textbook passive income, and holds overwhelmingly portfolios of securities, which are textbook passive assets. So the foreign corporation behind your contract will in most cases satisfy one or both tests every single year. Your contract does not need to be exotic, leveraged, or aggressive. An ordinary euro-denominated fund, called fonds en euros, invested mostly in bonds, and an ordinary unit-linked offering, called support en unités de compte, invested in collective funds, both sit behind foreign corporations whose income and assets are passive. That is the entire trap in one sentence: France sees a favored household savings envelope, while the United States sees shares in a foreign passive investment vehicle.
Three consequences follow from that classification, and each one surprises Americans. First, the French tax wrapper is transparent to the IRS in the worst way: as long as you do not withdraw, France taxes nothing, so you feel compliant, while the United States treats the inside buildup as accumulating inside a PFIC and waits to tax it punitively at exit. Second, the PFIC regime applies per shareholder, which means it applies to you personally even if the insurer pools thousands of savers and even if you never thought of yourself as owning corporate stock. If you hold units whose value tracks foreign collective investments, the look-through logic of American law can treat you as holding the underlying funds. Third, nothing about French compliance satisfies American compliance. Your French insurer sends the French tax administration an annual printout, called imprimé fiscal unique, and withholds nothing while the money stays invested. That document never reaches the IRS, it never substitutes for an American return, and your banker’s assurance that “you pay nothing until you withdraw” describes French law only. The practical result is a file that looks clean in France and silently deteriorates in America with every passing year, because every year of PFIC ownership is a year that should have carried its own Form 8621 disclosure, and every eventual withdrawal is measured against your entire holding period rather than against the year you cash out.
The second product to check in the same drawer is the plan d’épargne en actions, universally called the PEA. This is France’s stock-savings plan: you fund it with cash, you buy European shares and qualifying funds inside it, and after a holding period the French tax on the gains drops sharply. Americans love it because it feels like a Roth-style account with a French flag on it. The IRS has no equivalent drawer for it. The PEA is not a qualified American retirement plan, it is a taxable brokerage account with French training wheels, and where it holds French collective funds, called OPCVM, which include the SICAV and FCP fund formats, each of those funds is tested separately as a potential PFIC. An American who diligently buys European index funds inside a PEA can therefore own a dozen PFICs without ever receiving a single statement that uses the word. The same goes for a plain French securities account, called compte-titres ordinaire, holding French or Luxembourg funds: the account itself is not the PFIC, the funds inside it are, and each one needs its own analysis and potentially its own Form 8621. Before reading further, list every French envelope and every fund line you own: assurance-vie contracts with their euro fund and unit-linked lines, PEA with its fund lines, employee-savings plans if you work in France, and any direct fund holdings. That inventory is the indispensable first step, because the PFIC analysis is done holding by holding, not account by account, and the cleanup sequence depends on counting exactly how many PFICs you own and since when.
B. What the Default American Bill Looks Like, and the Elections That Can Shrink It
When a United States person receives a distribution from a PFIC without having made any special election, section 1291 of the Internal Revenue Code imposes the default regime, and its mechanics explain why advisers call it punitive. The statute spreads the excess amount day by day across the taxpayer’s entire holding period for the position. The full passage is readable at 26 U.S. Code section 1291. In plain terms, the gain is not taxed as this year’s income. It is sliced day by day across every year you owned the position, the slice assigned to each prior year is taxed at the highest ordinary rate in effect for that year, and an interest charge for late payment runs from each of those years to today, as if you had underpaid tax every year since purchase. A gain that French law would tax once, gently, after eight years can therefore produce an American bill composed of ordinary rates plus years of interest, and the longer you held the contract, the heavier the interest tail. This is why discovering the problem early matters so much: every additional year inside an unremediated PFIC lengthens the allocation period and fattens the interest charge on the eventual exit. It is also why the worst response is a panicked full redemption ordered before understanding the regime, because a redemption is precisely the distribution event that crystallizes the calculation.
Two elections can change the picture, and both must be evaluated before any money moves. The first is the qualified electing fund election, universally called the QEF election, under which you agree to include each year your share of the PFIC’s ordinary earnings and capital gains as they arise, in exchange for escaping the excess-distribution interest machine going forward. The catch is brutally practical: a QEF election requires the foreign fund to supply an annual information statement with the figures computed under American tax principles, and French insurers and French fund managers do not issue that document. Without the statement, the election is unavailable in practice, however attractive it looks on paper. The second is the mark-to-market election, available for PFIC stock that is marketable, typically funds with published daily net asset values on a qualifying exchange. Under mark-to-market you include each year’s unrealized gain as ordinary income and you take ordinary loss, within limits, on unrealized declines, which converts the backward-looking interest charge into a pay-as-you-go annual inclusion. Many French listed funds qualify in principle, while euro funds and insurer general assets do not, and the election only applies prospectively from the year made unless combined with a purging election that closes the pre-election taint. The purging step itself triggers a deemed sale under the excess-distribution rules, so it must be priced, not assumed. Neither election is a form to file blindly: each demands eligibility analysis holding by holding, and each interacts with the foreign tax credit so that French tax paid on the same income, if any was paid that year, offsets the American inclusion within the correct limitation basket.
Paperwork is the third front, and it runs on its own track regardless of the tax number. Each PFIC you own generally calls for its own Form 8621 attached to your annual return, with the election status, inclusions, and distributions computed separately. The bank and brokerage accounts that hold or fund these positions carry their own reporting: the FBAR, which is FinCEN Form 114 for aggregate foreign accounts over ten thousand dollars at any point in the year, and IRS Form 8938 for specified foreign financial assets above its higher thresholds, plus the French side, form 3916, which French residents file to declare foreign accounts, a form your French bank will not file for you. Our companion walkthrough of FBAR, Form 8938, and French Form 3916, with missed-year costs and fixes, covers that disclosure stack in detail, and you should read it alongside this article because PFIC cleanup years are exactly the years when the disclosure stack must be rebuilt completely. Missed Forms 8621 do not fix themselves when you close the account; depending on the program used, you may need amended returns, a quiet attach-and-disclose going forward with careful penalty analysis, or one of the IRS voluntary correction paths, and the choice turns on facts no article can prejudge: willfulness, amounts, number of years, and whether the IRS has already made contact. What can be stated as a firm rule is the sequencing: do not redeem first and ask questions later. Measure the PFIC population, model the section 1291 cost of each exit, test QEF and mark-to-market eligibility line by line, align the disclosure repair, and only then give the redemption or transfer order, because once the distribution lands, the calculation year is fixed and the interest clock on the planning stops mattering.
II. What French Law Does to the Same Contract When You Try to Leave
A. The French Exit Toll: Eight-Year Clock, Flat-Rate Levy, and the One-PEA Rule
Turn now to the French side, because it sets the price of every American remedy. Under article 125-0 A of the French General Tax Code, and I quote, “Les produits attachés aux bons ou contrats de capitalisation ainsi qu’aux placements de même nature souscrits auprès d’entreprises d’assurance établies en France sont, lors du dénouement ou d’un rachat du bon, contrat ou placement et quelle que soit sa date de souscription, soumis à l’impôt sur le revenu.” The current text is readable on article 125-0 A of the Code général des impôts. A rachat is a partial or total withdrawal from the contract, and the taxable produit is the difference between what you receive and the premiums you paid. The statute then builds the famous maturity calendar: contracts subscribed since January 1990 face the eight-year threshold the provision describes, with an annual allowance of 4,600 euros for a single taxpayer and 9,200 euros for a jointly taxed married couple applying to the taxable products, and the levy mechanics now run through the single flat-rate levy, called prélèvement forfaitaire unique, whose income-tax leg sits in article 200 A of the Code général des impôts. In practice this means an American who redeems a young contract pays French tax at the full flat rate plus social charges, while an American who redeems a mature contract keeps the allowance and the reduced rate on the rest, before social charges that Americans resident in France generally still owe. The lesson for sequencing is direct: redeeming alles sofort to stop the American bleeding can manufacture a French tax bill that dwarfs the American saving, especially where the contract is young, the gain is large, and the allowance is unavailable. Price the French levy year by year, allowance included, before choosing the American exit year.
The PEA has its own French clock, and it is stricter than the assurance-vie calendar in one decisive way: leaving early destroys the wrapper. The public service portal explains that money taken out of a PEA before its fifth anniversary loses the wrapper’s income-tax advantage: the net gain since the plan opened is taxed at 12.8 percent. The page is the official service-public.fr PEA guidance, and it adds that social charges apply on top and that any withdrawal before five years closes the plan. After five years the income-tax leg falls away and only social charges remain on compliant operation. The statute behind the envelope is article 163 quinquies D of the Code général des impôts, which ties the plan to the operating rules of the Monetary and Financial Code, whose eligible-investment list is set by article L. 221-31 of the Code monétaire et financier, and ordinary securities gains outside any wrapper fall under article 150-0 A of the Code général des impôts, which provides, and I quote in substance, that “les gains nets retirés des cessions à titre onéreux, effectuées directement, par personne interposée ou par l’intermédiaire d’une fiducie, de valeurs mobilières, de droits sociaux” and related rights “sont soumis à l’impôt sur le revenu”. For the American holder this creates a genuine conflict of clocks: the IRS rewards early recognition through a mark-to-market election or an early purging distribution, while France rewards patience through the five-year and eight-year thresholds. There is no universal winner. A PEA stuffed with French funds that are also PFICs may be worth closing early despite the French cost where the American interest tail is enormous, while a mature assurance-vie with a large euro-fund gain may be worth unwinding gradually, allowance by allowance, across French tax years even though the American inclusions dribble on. Model both clocks on the same spreadsheet, line by line, before deciding.
One French trap deserves its own warning because Americans stumble into it while trying to be diligent: holding two PEAs. French law allows only one stock-savings plan per taxpayer, and the Cour de cassation enforces the consequence without mercy. In a judgment of its commercial chamber dated 9 February 2022, appeal number 20-16.471, readable at Cass. com., 9 February 2022, no. 20-16.471, the facts were that on 15 December 2011 the tax administration notified the taxpayer of a reassessment, and I quote, “lui contestant le bénéfice de l’exonération de l’imposition sur les plus-values réalisées sur la cession, le 30 mai 2008, de titres figurant sur un plan d’épargne en actions (PEA) ouvert le 14 décembre 2001”, for the reason that, again quoting, “il était titulaire d’un second PEA, ouvert le 1er mars 2002”. The taxpayer then sued the bank for failing to warn him, and lost on the dozens of branches of argument the court methodically rejected. For Americans the moral is concrete: expatriates who open a PEA at one bank, forget it, open another at the bank near the new apartment, or accumulate a PEA plus a so-called PEA-PME without understanding the articulation, can lose the French exemption on the whole position and still keep the full American PFIC problem on the funds inside. Before any cleanup, confirm with each institution in writing how many PEAs stand in your name, because a second PEA poisons the French advantage while changing nothing on the American side, and closing the duplicate has its own French tax and dating consequences that must be sequenced with the PFIC elections rather than improvised at the counter.
B. Beneficiaries, Forced Shares, and the Premiums Your Children Can Claw Back
The estate dimension is where French assurance-vie looks most generous and where American families make the costliest assumptions. The starting rule is article L. 132-12 of the Code des assurances, which provides, and I quote: “Le capital ou la rente stipulés payables lors du décès de l’assuré à un bénéficiaire déterminé ou à ses héritiers ne font pas partie de la succession de l’assuré.” A designated beneficiary is therefore, and the article continues, deemed to have owned the right from the day of the contract even if acceptance came after death. Americans hear “passes outside probate” and stop reading, which is a mistake on both sides of the Atlantic. On the French side, the exclusion from the succession governs the civil transmission, not the tax: premiums paid before age seventy fall under the special levy of article 990 I of the Code général des impôts, which taxes the payable sums per beneficiary on the redeemable fraction and post-seventy premiums on the non-redeemable fraction, while premiums paid after the seventieth birthday fall back into ordinary death duties by kinship degree under article 757 B of the Code général des impôts, which covers, and I quote, “la fraction des primes versées après l’âge de soixante-dix ans”. An American parent who funds a French contract at seventy-five thinking the beneficiary clause settles everything has therefore built two different tax computations into one gesture, and the American estate-tax side adds a third layer, since the United States taxes its citizens’ worldwide assets at death with its own exemption and treaty-credit mechanics. Map all three before designating: French civil exclusion, French tax inclusion under 990 I or 757 B by premium vintage, and American estate inclusion with available credits.
The beneficiary clause itself is litigation territory, and the Cour de cassation gave the controlling reading in a published judgment of its First Civil Chamber dated 30 September 2020, appeal number 19-11.187, readable at Cass. 1re civ., 30 September 2020, no. 19-11.187. The court held, and I quote: “Selon l’article L. 132-8 du code des assurances, le capital ou la rente garantis peuvent être payables lors du décès de l’assuré à un ou plusieurs bénéficiaires déterminés.” Designating “my heirs” counts as designating determined beneficiaries, and the court continued, again quoting: “Les héritiers, ainsi désignés, ont droit au bénéfice de l’assurance en proportion de leurs parts héréditaires et conservent ce droit en cas de renonciation à la succession.” Identifying who “heir” means in a given clause belongs to the trial judges, who interpret the subscriber’s will, taking the testament into account where appropriate. For an American family this means a clause drafted in English-language logic, such as “my heirs at law” or “my children in equal shares,” will be read through French interpretive rules, against the backdrop of the French will if one exists, and the renouncing heir keeps the insurance benefit even while walking away from the debts of the estate. Review every clause after every family event, birth, marriage, divorce, death, or move across the Atlantic, and align the French beneficiary designation with the French will and with the American revocable trust if one exists, because three documents telling three stories about the same money is the standard fact pattern of the lawsuits that follow.
The final French limit is the one disinherited children invoke: premiums that were manifestly excessive for the subscriber’s means fall back into the estate and can infringe the forced share, called réserve héréditaire. The Court of Appeal of Douai restated the test on 7 May 2026, docket number 24/00958, ordering each of two brothers to pay 20,000 euros back into their mother’s succession as manifestly excessive premiums, with the mirror order against his brother and the reminder that reintegrated capital benefits all heirs in proportion to their rights. The court’s legal basis was article L. 132-13 of the Code des assurances, which provides, and I quote: “Le capital ou la rente payables au décès du contractant à un bénéficiaire déterminé ne sont soumis ni aux règles du rapport à succession, ni à celles de la réduction pour atteinte à la réserve des héritiers du contractant. Ces règles ne s’appliquent pas non plus aux sommes versées par le contractant à titre de primes, à moins que celles-ci n’aient été manifestement exagérées eu égard à ses facultés.” Excessiveness is judged concretely against the subscriber’s wealth, income, age, health, family situation, and the usefulness of the transaction at the time, following the mixed-chamber precedent of 23 November 2004 that the Douai court expressly cites. An American parent who empties a lifetime of savings into a French contract for one child, or who subscribes late in life while dependent on the capital for living expenses, builds exactly the file that a bypassed child will attack after death, and the American dimension adds irony: the premiums most vulnerable to clawback are often the same late, large subscriptions whose PFIC interest tail is heaviest. Keep subscriptions proportionate, document means and motive at signing, revisit designations after every major asset move, and never use a French contract as a quiet substitute for the forced share, because French courts look through it and American tax ignores it.
Conclusion
Your French assurance-vie and your PEA are not mistakes; they are excellent French products owned by the wrong taxpayer profile for their default settings. The American fix is not a single form but a sequence: inventory every envelope and every fund line, test each holding under the income and asset thresholds of section 1297, price the section 1291 excess-distribution tail before triggering any distribution, test QEF feasibility against the missing-statement reality and mark-to-market eligibility holding by holding, rebuild the Form 8621 stack together with the FBAR, Form 8938, and French form 3916 filings, and only then execute the French side in the order French law rewards, respecting the eight-year and five-year clocks, the single-PEA rule, the 990 I and 757 B premium vintages, and beneficiary clauses drafted to survive French interpretation and forced-share review. Done in that order, most files end with a finite, fundable American cost, a minimized French levy, and designations that hold. Done backwards, starting with a panicked redemption at the bank counter, the same file ends with the maximum of both taxes and a clause the heirs will litigate. Measure first, elect second, redeem last.