France's 2027 Budget Bill Ends the Apport-Cession Deferral at Death

I. What Article 5 of the 2027 Finance Bill actually changes

A. One word replaced, one paragraph repealed

The apport-cession regime, codified in Article 150-0 B ter CGI, places under deferral (report d'imposition) the capital gain arising when an individual contributes shares to a company subject to corporate income tax that he or she controls. In the version in force since February 21, 2026, paragraph I lists the events that end the deferral: the transfer for consideration, redemption, repayment or cancellation of the shares received in exchange for the contribution (1°); the sale of the contributed shares by the holding within three years of the contribution without reinvestment of at least 70% of the proceeds (2°); the sale of interests in interposed entities (3°); and the transfer of the taxpayer's tax residence outside France (4°). Paragraph IV extends the mechanism to successive contributions, and paragraph V provides that the deferral ends "in proportion to the shares transferred for consideration, redeemed, repaid or cancelled" (CGI, art. 150-0 B ter).

The reform is built on two operations. Point a) of 1° of Article 5(I) of the Bill replaces, in 1° and 3° of paragraph I and in 1° of paragraph IV, the words "cession à titre onéreux" with the word "transmission". Point b) repeals paragraph II of Article 150-0 B ter, which governed the deferred gain where the holding shares were gifted to a donee taking control of the holding. Nothing else in Article 150-0 B ter is amended: not 2° of paragraph I, which keeps the reference to a transfer for consideration of the contributed shares, not paragraph V, not the conditions of paragraph III. The text applies to transfers made on or after October 1, 2026, that is, before parliamentary debate has even opened.

B. The end of the step-up at death and of the transfer of the deferral to the donee

The reach of the substitution is considerable. The word "transmission" covers the transfers for consideration already caught, but also gratuitous transfers (transmissions à titre gratuit): gifts, manual gifts and devolution on death. Until now, the contributor's death was not an event ending the deferral; the contribution gain died with the contributor, and the heirs received the holding shares at their value on the date of death. This is what practitioners call the purge, the French equivalent of a step-up in basis at death. The Bill's impact assessment acknowledges it plainly: the gain "is also exempt at the taxpayer's death or on a manual gift, after a holding period of six or eleven years by the donee".

A gift becomes a taxable event in the hands of the donor. The repealed paragraph II provided that, where the holding shares were gifted to a person taking control of the holding, the deferral was transferred to the donee, who became liable to tax if the shares were sold within six years, or eleven years where the holding had reinvested in funds. After that period, the gain was permanently exempt. That mechanism disappears. From now on, any gift of the shares received in exchange for the contribution ends the deferral in the hands of the donor, who becomes immediately liable to income tax and social contributions on the contribution gain. On death, the deferral lapses and the tax becomes, in the words of the impact assessment, "a tax debt of the estate". The tax is computed at the rate frozen in the year of the contribution, under paragraph 2 ter of Article 200 A CGI, to which the last sub-paragraph of Article 150-0 B ter(I) refers.

To soften the cash-flow impact, the Bill creates a new Article 1681 G CGI allowing the taxpayer to request a five-year instalment plan (plan de règlement échelonné) for the income tax attributable to a deferred gain that crystallises on a gratuitous transfer. The plan is conditional on compliance with current tax obligations and on the provision of security to the public accountant, and it is revoked on default. As we shall see, its scope, limited to income tax alone, significantly reduces its value.

C. The exit tax component: a symmetry that is not one

Article 5 also amends paragraph VII of Article 167 bis CGI, which lists the events ending the payment deferral (sursis de paiement) under the French exit tax. Three changes are made. Point f) of VII(1), which referred to the transfer for consideration of shares carrying an apport-cession deferral created before departure, is rewritten as a direct cross-reference to the events in 1° to 3° of paragraph I and 1° of paragraph IV of Article 150-0 B ter: a gift by, or the death of, the expatriated taxpayer now ends the payment deferral and makes the tax due, whatever the State of residence. Correspondingly, the second sub-paragraph of VII(3), which granted a discharge where such shares were transferred gratuitously from abroad, no longer refers to Article 150-0 B ter (CGI, art. 167 bis, VII).

A new point f bis for contributions made after departure. The third change goes further. A new point f bis provides that the payment deferral ends on a gratuitous transfer of shares received in exchange for a contribution made, under Article 150-0 B ter, after the transfer of tax residence out of France, where the contributed shares are those on which an unrealised gain was assessed on departure. The ordinary discharge of the second sub-paragraph of VII(2), which extinguishes exit tax on unrealised gains in the event of death or of a gift made from an EU Member State or a treaty State, is expressly excluded in that case. In other words, a taxpayer who left France and then placed his shares in a holding company loses an extinction of exit tax that the law grants to a taxpayer who did nothing. We return to this below, because in our view it is the most objectionable provision in the text.

II. Why the "refocus on investment" is a contradiction

A. What the 2012 legislature actually intended

The explanatory memorandum invokes the "original spirit" of the regime. That spirit deserves an accurate account. Article 150-0 B ter was introduced by the third amending Finance Act for 2012 to defeat apport-cession arrangements that the Conseil d'État had been striking down case by case as an abuse of law (abus de droit): contribute shares to a holding under rollover relief (sursis d'imposition), have the holding sell the shares free of tax, then enjoy the proceeds through the structure. The legislature codified the dividing line drawn by the courts: the operation is legitimate if the holding reinvests a significant part of the proceeds in an economic activity within a reasonable time. The reinvestment threshold, set at 50%, then raised to 60% and now 70%, is the condition for maintaining the deferral where the holding sells within three years. It was never presented as the price of a final exemption.

The step-up results from the silence of the text, not from a choice. The 2012 text never provided that the deferred gain would be exempt at death. That outcome flows from the general principle that a gratuitous transfer is not a taxable event for capital gains purposes, combined with the silence of Article 150-0 B ter on the contributor's death. Paragraph II, by contrast, expressly regulated gifts by transferring the deferral to the donee with a holding period, precisely so that a gift could not serve as an immediate purge. To say the step-up was the legislature's objective would be as inaccurate as to say it was foreign to it: it was a known and accepted effect and, in the impact assessment's own words, part of "the attractiveness of the regime". The real debate is therefore not about fidelity to an original intent. It is about the effect of the reform on investment behaviour, which the Bill claims in its title and ignores in its impact assessment.

B. The taxpayer's decision after the reform

Consider the position of an entrepreneur about to sell her company, choosing between a direct sale and a prior contribution to a holding. Before the reform, the contribution offered a deferral whose price, in the event of a quick sale, was the obligation to reinvest 70% of the proceeds in eligible assets held for at least five years, with the prospect of a permanent extinction of the tax at death or after a transfer to the children. The constrained, illiquid and risky reinvestment was justified by that prospect. After the reform, the same reinvestment constraint remains, but the contribution gain will be taxed, at the latest, in the entrepreneur's estate, in full and at the rate of the year of the contribution.

The deferral becomes a postponement locked inside the holding. The postponement retains a financial value: the tax is not paid and the gross sale proceeds can compound inside the holding. But that value is eroded by the cost of extracting the cash. To recover the proceeds personally, the entrepreneur must distribute, and therefore bear a second layer of tax at the 30% flat rate. As long as the step-up existed, that second layer was never paid by someone who kept the holding until death: the holding was a transfer vehicle at zero tax cost on the contribution gain. Without the step-up, a taxpayer with no genuine investment project compares a deferral coupled with constrained reinvestment and trapped cash against a direct sale with 30% paid immediately and complete freedom over the balance. For that profile, the direct sale often wins. And that profile was precisely the one feeding the funds eligible under point d) of 2° of paragraph I and the minority stakes in SMEs eligible under points b) and c).

The impact assessment contradicts itself on this point. Its section 4.1.1 states that the measure has "no" micro- or macroeconomic impact. Yet the title of the article claims a "refocus on its objective of supporting investment", and the explanatory memorandum an effect on "optimisation strategies". If the measure changes behaviour, it has an economic impact; if it has none, it refocuses nothing. One of the two statements is false. In our view, it is the first: the reform will reduce the flow of capital directed to the productive economy through the apport-cession channel, with no substitute mechanism on offer.

C. The lock-in effect on lifetime transfers

The most serious issue is not budgetary but patrimonial. A donor who gifts the shares of her holding becomes liable to tax on a gain attached to assets she has just given away. Her estate is reduced by the value transferred and no cash is generated by the transaction. The instalment plan of Article 1681 G covers income tax only, requires security and can be revoked. The rational response of a head of family is to stop gifting holding shares and wait for death, when the tax will fall on an estate that does hold the value. The text thus creates a lock-in effect that encourages the retention of capital in the hands of the oldest generation.

An inconsistency with the policy of early business succession. The same Tax Code elsewhere encourages the early transfer of businesses, through renewable gift allowances, the Dutreil regime for business succession and transfers of bare ownership. For animating holding companies held under a Dutreil pact, combining a partial exemption from gift tax with immediate taxation of the contribution gain in the donor's hands is economically incoherent and practically dissuasive. The Bill offers no articulation between these regimes.

Finally, the revenue estimate, EUR 500 million in 2027, of which EUR 200 million in income tax and EUR 300 million in social contributions, rests on a static method: the average deferred gains of taxpayers over sixty across three years, spread over five years. That method assumes no behavioural change whatsoever. It is therefore inconsistent with the stated aim of curbing optimisation strategies: if the measure achieves its purpose, contributions stop and the base melts away. If it does not, the refocus is an empty word.

III. A poorly coordinated and legally fragile text

A. Coordination gaps that hit the very case targeted

The technique chosen, replacing one word in three places and repealing one paragraph, leaves the rest of Article 150-0 B ter untouched. Paragraph V, which determines the proportion in which the deferral ends, still refers only to shares "transferred for consideration, redeemed, repaid or cancelled". A gift or a devolution on death covering a fraction of the holding shares falls under 1° of paragraph I but under none of the cases listed in paragraph V. Does the deferral end for the entire gain on the first partial transfer, or does the proportionality rule extend by analogy? The text does not say, and the gap hits exactly the case the reform intends to catch.

The fate of deferrals already transferred to donees is not addressed. Paragraph II is repealed by Article 5(I), which applies "to transfers made on or after October 1, 2026". For earlier gifts, the donee in principle remains bound by the six- or eleven-year holding period and the gain is taxable in the donee's hands on an early sale. No transitional provision confirms this. A literal reading of the repeal would remove the legal basis for taxing the donee for the entire existing stock. The interpretation attaching the regime to the original taxable event is the more likely one, but it needs to be secured in the text; otherwise litigation is guaranteed.

Other defects affect the exit tax component and the new Article 1681 G. The new point f) of Article 167 bis(VII) cross-refers to 1° to 3° of paragraph I and 1° of paragraph IV of Article 150-0 B ter, but neither to 2° nor to 3° of paragraph IV: a sale, by a company receiving a successive contribution, of the contributed shares within three years without reinvestment (3° of IV) would end the deferral under domestic law without ending the exit tax payment deferral, the reverse of the intended result. Paragraph VII(1 bis), which defines the shares referred to in points a) and b) and already covers shares received in a post-departure contribution, is not coordinated with point f bis, whose scope does not match it. Above all, Article 1681 G covers income tax only, whereas EUR 300 million of the EUR 500 million expected are social contributions: the instalment plan covers less than half the burden.

B. The unequal treatment of expatriates under point f bis

Point f bis calls for a substantive objection. Take two taxpayers who have moved their tax residence to an EU Member State with a portfolio subject to exit tax under payment deferral. The first keeps his shares directly; on his death, the exit tax on unrealised gains is discharged under the second sub-paragraph of Article 167 bis(VII)(2). The second contributes his shares to a holding after departure, a transaction that point a) of VII(1) expressly declares neutral for the payment deferral; on his death or on a gift of the holding shares, point f bis ends the deferral and the tax becomes due, with no discharge and no possibility of showing the absence of a principally tax-driven motive.

An unrealised gain taxed solely because of a neutral restructuring. The gain caught by point f bis is not a contribution gain within the meaning of Article 150-0 B ter; it is the unrealised gain assessed on the day of departure, which by definition has never been realised. The post-departure contribution, which generated no cash, becomes the indirect trigger for a tax that the law extinguishes for everyone else. There is no connection with the anti-abuse purpose of the apport-cession regime. The precedent is well known: in 2012, the Conseil constitutionnel struck down the so-called "gift-and-sale" provision, which imposed on the donee a tax "unrelated to his situation" on the basis of a time criterion that was "in itself insufficient to presume irrebuttably" an intention to evade tax (Cons. const., Dec. 29, 2012, No. 2012-661 DC, para. 24). Point f bis rests on a presumption of the same kind, with no possibility of rebuttal, and triggers the tax on an event, death, that is never chosen. In our view, this difference in treatment, based solely on the interposition of a company, is hard to justify under the principle of equality before public burdens, and raises a serious question under the freedoms of movement guaranteed by EU law, since it hits taxpayers established in other Member States on account of a transaction that domestic law treats as neutral.

C. Retroactivity in two stages: the filing date and the existing stock

Article 5(II) makes the regime applicable "to transfers made on or after October 1, 2026", the date the Bill was filed, for a law that will be enacted, at best, at the end of December. The technique is not new. The 2012 text that created Article 150-0 B ter already applied to contributions made on or after November 14, 2012, the filing date of the amending Finance Bill, and the Conseil constitutionnel accepted it: by choosing that date, the legislature "intended to prevent the filing of the bill with the National Assembly from producing, before the law entered into force, effects contrary to the objective pursued", so that the retroactive effect was "justified by a sufficient reason of general interest" (Cons. const., Dec. 29, 2012, No. 2012-661 DC, para. 19).

The 2012 justification does not cover death. The reason accepted in 2012 is the prevention of forestalling: stopping the announcement of a text from triggering a wave of transactions designed to escape it. That reasoning holds for gifts, which a taxpayer can indeed bring forward between filing and enactment. It does not hold for death. Nobody arranges to die in order to escape a bill. By applying the text to estates opened between October 1, 2026 and enactment, the legislature retroactively taxes a gain in the hands of a taxpayer who died under a text that did not make it taxable, without being able to invoke the only reason of general interest accepted so far for this technique. The heirs of a contributor who dies in the autumn of 2026 will spend the parliamentary debate uncertain as to the very existence of a tax debt whose triggering event will have occurred before the law creating it.

The second stage of retroactivity runs deeper. Article 5 applies to all deferrals created since 2012, a stock the impact assessment valued at EUR 127 billion in 2023. Those deferrals were created under a text that expressly provided for the transfer of the deferral to the donee and allowed the step-up at death to operate, and, for many, at the price of irreversible five-year reinvestment commitments. The measure is not retroactive in the strict sense, since the taxable event is a future transfer. But it upsets the effects that could legitimately be expected from legally acquired situations, and the Conseil constitutionnel has held, precisely with respect to deferred gains, that a taxpayer whose gain is deferred by operation of law, unlike one who elected for a deferral, cannot be deemed to have accepted the consequences of rules set later: "only a sufficient reason of general interest can justify subjecting the gain retroactively to liquidation rules that were not determined at the date of its realisation" (Cons. const., Apr. 22, 2016, No. 2016-538 QPC, paras. 14 and 15). The Article 150-0 B ter deferral applies by operation of law. The 2016 case concerned rules of base and rate; the 2026 text concerns the event that makes the tax due, which is another liquidation rule, and the legislature has clearly understood as much, since it freezes the rate at that of the year of the contribution. To justify application to the existing stock, the impact assessment offers no reason other than budget revenue and the concentration of the regime among the wealthiest taxpayers. Its section 3.3, on compatibility with EU law, does not devote a line to the protection of acquired rights.

D. A question of method: how French tax law is made

This text does not come alone. Since the summer of 2025, the French tax authorities have been reassessing presidents of single-shareholder simplified companies (SASU) that elected for income tax treatment, subjecting their profits to social contributions on investment income without any express statutory basis, and the Government confirmed that position in a ministerial answer of June 2, 2026 (Rép. min. Bergantz, No. 12673, JOAN June 2, 2026), whose legal weaknesses we have analysed elsewhere (SASU under income tax: the Government's answer is legally wrong). In one case, the Finance Ministry taxes by interpretation where the law is silent; in the other, it legislates without rereading the article it amends, presenting as a refocus a revenue measure whose own impact assessment denies any economic effect.

The common symptom is the erosion of review. A finance bill article that amends three sub-paragraphs of a six-paragraph regime without adjusting the proportionality rule, that creates an instalment plan for income tax while forgetting the social contributions that make up 60% of the expected yield, and that applies to estates an effective date designed to counter gifts, is not an ill-intentioned text. It is a text that has not been reread by anyone who practises the subject. In our view, the question is no longer the political direction of the measure, which is for Parliament to judge, but the capacity of the administration to produce tax rules that are coherent, predictable and applied without de facto retroactivity. That capacity, more than the rate of any particular levy, determines whether entrepreneurs continue to invest and to transfer businesses in France.

IV. What affected taxpayers should do now

A. Holders of deferred gains: revisit the arbitrage

The chosen effective date, October 1, 2026, closes the door on gifts of holding shares made to pre-empt the text. Holders of deferred gains must therefore reason under the new law, subject to the outcome of the parliamentary debate. Three questions arise. The first concerns the value of the postponement: when will the contribution gain be taxed, at what frozen rate, and does the return on the holding's assets justify keeping the structure rather than crystallising the gain through an early sale of the holding shares or a capital reduction? The second concerns transfers: a gift of holding shares, formerly neutral, becomes a taxable event in the donor's hands; pending gift plans must be reviewed, particularly where they are combined with a Dutreil pact. The third concerns succession: the latent tax liability must be built into estate projections and into the heirs' liquidity needs, since the Article 1681 G instalment plan does not cover social contributions.

Document the tax base now. A deferral created in 2014 or 2016 will be taxed on the contributor's death, sometimes thirty years later, at the rate of the year of the contribution and for an amount that will have to be substantiated. Taxpayers should assemble a complete file today: contribution deed, contribution auditor's report, declaration of the deferred gain, annual follow-up returns, evidence of reinvestments and of their holding period. The burden of proof will fall on the estate, at a time when the direct witnesses of the transaction will be gone.

B. Taxpayers established outside France: post-departure contributions become risky

Taxpayers subject to the exit tax who contemplate restructuring their holdings after leaving France must factor point f bis into their analysis. Contributing shares carrying an unrealised gain to a French or foreign holding company, until now neutral for the payment deferral, now deprives the taxpayer of the discharge of exit tax at death and on gifts. Where the restructuring meets a genuine need, its deferred tax cost must be measured and, where appropriate, alternative holding arrangements preferred. Taxpayers who have already made such a contribution should inventory the shares concerned and assess the exit tax that would become due, without delay.

Expatriates who held, before departure, holding shares carrying an apport-cession deferral are also affected by the new point f) and by the removal of the discharge under VII(3): a gift or death will make the tax due whatever their State of residence, and the Article 1681 G instalment plan will be available to them on the same terms as to residents, security included. The interaction with inheritance and gift taxation in the State of residence, and with applicable tax treaties, will need to be examined case by case.

C. What a genuine refocus would have looked like

We do not defend the step-up as such. That a gain of several tens of millions of euros can permanently escape income tax through the mere passage of time is a loophole the legislature may legitimately wish to close. But the option chosen is the only one of the two examined in the impact assessment that is disconnected from the stated objective. Option 2, which would have subjected heirs to the same conditions for maintaining the deferral as those applicable to donees, was rejected on the sole ground that it "would preserve a possibility of purging all taxation". That admission reveals that the real objective is revenue, not the direction of savings.

A refocus would condition the step-up on effective reinvestment. A regime true to its title would have made the maintenance or extinction of the deferral, on a gratuitous transfer, conditional on demonstrated effective and durable reinvestment in the productive economy: a minimum proportion of eligible assets on the holding's balance sheet, assessed over a significant period, with the deferral passing to the donee or the heirs for as long as that proportion is met. Such drafting would have preserved the incentive to invest, closed the step-up to pure cash-management holdings and answered the parliamentary criticism about the concentration of the regime. It is on that line, rather than on a frontal defence of the exemption, that we believe an amendment stands a chance.