Life Assurance or a Pension Plan: the Lock-Up Decides, Not the Tax Break

The deduction is a loan of the tax, not a gift of it

Everything sold about a deductible retirement plan leads with the same figure: pay in 10,000 euros, save 4,100 euros of tax at the top of the 41 % band. That figure is true and it is also, on its own, meaningless, because the deducted contribution is taxed on the way out. What you have actually done is postpone the tax from this year to the year you draw the money, at whatever marginal rate applies then. Framed as a gift, the deduction invites the wrong comparison — deduction against no deduction — and the wrong conclusion, which is that the plan is worth it only if your tax rate falls in retirement.

The right comparison is between two ways of investing the same gross income. Route one: 10,000 euros go into the plan untaxed and grow; the tax is paid on the way out, decades later, on a base that has itself grown. Route two: the 10,000 euros are taxed today at 41 %, leaving 5,900 euros, and those 5,900 euros grow inside a life-assurance contract whose only later tax is on the gain. Set out like that, the source of the advantage becomes visible: in route one, 4,100 euros of the state's money is working for you the whole time. At 3.5 % over twenty years that borrowed 4,100 euros nearly doubles, and although the tax eventually claims its share, what it claims is the original 4,100 euros' worth at the future rate, not the growth on it. The deduction is a loan, and the interest on the loan is yours.

What the arithmetic says at each combination of rates

The assumptions, so you can disagree with them precisely: 10,000 euros of gross income, a net return of 3.5 % a year in both wrappers, no difference in fees, the pension plan drawn as capital with the contribution component taxed at the income-tax scale and the gain at the flat rate, and the life-assurance contract redeemed after at least eight years with the annual allowance set aside. At a 41 % marginal rate today and 30 % at exit, the plan ends at 13,929 euros against 10,297 for life assurance after twenty years, and 9,874 against 7,724 after ten. At 30 % and 30 %, with no rate differential at all, the plan still ends ahead: 13,929 against 12,217 at twenty years, a margin of 14 %, and 9,874 against 9,164 at ten years, a margin of 7.7 %. At 30 % today and 11 % at exit the margin is about 30 % at both horizons.

Two results in that set deserve to be pulled out because they contradict the usual advice. The first: because the advantage comes mostly from deferral rather than from a rate differential, the pension plan keeps winning at twenty years even if your marginal rate rises — at 30 % in and 41 % out it is still 5 % ahead. Solve for the exit rate at which the two draw level and you get 47.1 % for someone deducting at 30 %, and 66.3 % for someone deducting at 41 %. The fear that tax rates will be higher in thirty years is a reasonable fear, but it has to be a very large rise before it changes this answer. The second: at an 11 % marginal rate the whole thing collapses. Deducting at 11 % and paying 11 % back leaves a twenty-year advantage of 1.9 %, or about 300 euros on 10,000, in exchange for locking the money until retirement. That is a bad trade, and it is the trade most often recommended to the people for whom it is worst.

Where the horizon matters, and where it does not

The usual framing of this choice promises a crossover point: a number of years before which one wrapper wins and after which the other does. For a French saver at 30 % or 41 % there is no such point, and saying so plainly is more useful than inventing one. The pension plan is ahead from about five years onwards in every scenario except one, and the gap widens with time because it is a compounding advantage rather than a fixed bonus — 7.7 % at ten years and 14 % at twenty in the flat-rate case. Time is not the variable that switches the answer; it is the variable that multiplies whichever answer the rates already gave.

The one exception is worth the paragraph. If your marginal rate rises between the contribution and the withdrawal — 30 % now, 41 % later — the plan is behind at ten years by 4.3 % and only pulls ahead by about 5 % at twenty. That is the single case in this comparison where the horizon flips the answer, and it is a realistic case for someone early in a career whose income will keep climbing. The practical response is not to avoid the plan but to time the contributions: deduct in the years when your marginal rate is at its highest, not in the years when your income is still rising. The same reasoning applies in reverse at the exit, where taking the whole pot as capital in a single year piles the entire deducted contribution onto that year's income and can push you into a band you never expected — the reason the staged withdrawal, or an annuity, is a tax instrument and not just a lifestyle preference.

The lock-up, in the words of the statute

Article L224-4 of the code monétaire et financier, in the version in force on 14 June 2026, sets a closed list of situations in which a French retirement savings plan can be released before retirement, and it is worth reading rather than paraphrasing: the death of the holder's spouse or civil partner; disability of the holder, of their children, or of their spouse or partner meeting the social-security criteria; a serious illness, disability or particularly grave accident affecting a dependent child; over-indebtedness under the consumer code; the exhaustion of unemployment rights, or two years without employment or a corporate mandate for former company officers; a judicial liquidation or a conciliation procedure; the acquisition of a principal residence; and the case of a holder still under eighteen at the moment of the request. Nothing else works. Not a change of mind, not a business opportunity, not a family emergency outside that list.

The other side is looser but not unconditional, and the two qualifications are worth naming. A life-assurance contract can be redeemed at any time, in whole or in part, which is the feature you are paying for. But its favourable tax treatment depends on an eight-year clock that runs from the opening of the contract, after which withdrawals fall under an annual allowance of 4,600 euros for one person or 9,200 for a couple and the rate on the rest drops to 7.5 % on premiums up to 150,000 euros; and Article L631-2-1 of the same code lets the Haut Conseil de stabilité financière temporarily limit the payment of surrender values for up to three months, renewable, with a hard stop at six consecutive months. That power has never been exercised, and it remains the reason a life-assurance contract is a liquid savings vehicle rather than a bank account.

The same choice in Spain and Italy

Spain runs the same trade with a much smaller deduction and a softer lock. The individual contribution eligible for reduction in the general tax base has been capped at 1,500 euros a year since 2022, with up to 8,500 euros more from employer contributions or matched employee contributions to an occupational plan, and the reduction is additionally limited to 30 % of net earned and business income. The lock-up has been loosened in a way the French plan has not: since 1 January 2025, contributions at least ten years old can be redeemed without any qualifying event, which in 2026 means everything paid in during 2016 or earlier. A Spanish saver is therefore trading much less tax relief for a much less absolute commitment, and the arithmetic above tilts correspondingly further towards keeping the money accessible.

Italy sits between the two, and its numbers changed this year. The annual deductibility limit for contributions to a complementary pension scheme, unchanged at 5,164.57 euros for almost two decades, rose to 5,300 euros with the 2026 budget law, legge n. 199 of 2025, applying to the whole 2026 tax period, with the additional quota for workers first employed after 2007 rising in step to 2,650 euros. What makes the Italian version distinctive is the exit: benefits are taxed by a substitute tax of 15 %, which falls by 0.30 points for each year of membership beyond the fifteenth, down to a floor of 9 % after thirty-five years. That is the one European system in this comparison where the horizon really does change the tax rate rather than merely multiplying an advantage, and it rewards starting early in a way the French and Spanish plans do not.

The verdict, and what would change it

If you are taxed at 30 % or 41 %, you are confident the money will stay untouched until retirement, and you can control the shape of the exit, the pension plan wins and it is not close: 14 % to 35 % more after tax over twenty years depending on the rate you deduct at and the rate you draw at. If you are taxed at 11 %, life assurance wins, because the plan's arithmetic advantage falls to about 2 % over twenty years and you would be surrendering complete access for it. If there is any real chance you will need the money — a business you might start, a move you might make, an illness in the family that does not meet the statutory definitions — life assurance wins regardless of your tax band, because the alternative is not a worse return, it is not having the money at all.

Three things would change that verdict, and it is worth knowing which are under your control. The first is a change to the deduction ceilings or the exit taxation, both of which move by statute: the French ceiling for 2026 contributions is 10 % of 2025 professional income capped at 37,680 euros, or 4,710 euros if that is higher, and the same rules withdrew the deduction entirely from the age of seventy. The second is the fee difference between the two contracts you are actually offered, which this comparison assumed away and which in practice can be worth more than the tax advantage — a plan charging one point a year more than a life-assurance contract erases the flat-rate case's entire twenty-year margin. The third, and the only one that is fully yours, is honesty about the lock-up. Read the eight cases in Article L224-4 and ask whether the reasons you might realistically need this money are on that list. If they are not, you are not choosing between two returns; you are choosing between a return and an option, and options are worth paying for when the future is uncertain.

Frequently asked questions

Is a deductible pension plan worth it if I am in the lowest tax band?

On these assumptions, almost never. Deducting at 11 % and being taxed at 11 % on the way out leaves a twenty-year advantage of about 1.9 %, roughly 300 euros on a 10,000-euro contribution, and you have surrendered all access to the money in exchange. Worse, if your income rises during your career the deduction you took at 11 % may come back at 30 %, turning a small advantage into a loss. The deduction is worth its marginal rate, and at 11 % that is eleven cents on the euro — too little to buy a lock-up that lasts decades.

Can taking the whole plan as a lump sum push me into a higher tax band?

Yes, and this is the most under-discussed risk in the whole comparison. The deducted contributions come back into your taxable income in the year you draw them, so a pot representing twenty years of contributions can add a very large sum to a single year's income and be taxed at rates well above the one you deducted at. The arithmetic in this article assumed a marginal rate at exit, and if a lump sum pushes that rate above about 47 % for someone who deducted at 30 %, the advantage disappears entirely. The management is straightforward: draw over several years, or take part of it as an annuity, both of which spread the contribution component across tax years instead of stacking it into one.

Can I get money out of a pension plan early to buy a home?

Yes for a principal residence, which is one of the cases listed in Article L224-4 of the code monétaire et financier, and no for anything else in property — not a second home, not a rental investment. The release is not free of consequences: the deducted contributions come back into that year's taxable income, which is precisely the bracket risk described above, and the gains are taxed at the flat rate. Someone unlocking a large plan in the same year they buy a home can face a substantially higher marginal rate than the one they deducted at. Check the current tax treatment before assuming this route is cheap, and read the case as written, because the statutory wording is narrower than the way it is usually summarised.

Does the eight-year rule mean life assurance is locked for eight years?

No, and this is the most common confusion about the product. A life-assurance contract can be redeemed on day one; the eight-year mark is a tax threshold, not a lock. Before it, the gain in a withdrawal is taxed at 12.8 % of income tax plus social contributions; after it, the income tax falls to 7.5 % on premiums up to 150,000 euros and an annual allowance of 4,600 euros for one person or 9,200 for a couple applies to the income-tax part only. Social contributions are due either way. So an early withdrawal costs you a better rate you would have had by waiting, which is a real cost but a completely different thing from being unable to get the money.

Should I simply hold both?

For most people with a 30 % or 41 % marginal rate, yes, and the split follows from the criteria rather than from a rule of thumb. Money you are certain not to need before retirement goes into the plan, because that is where deferral compounds and where the deduction is worth its marginal rate. Money that has any chance of being needed goes into life assurance, because the option to withdraw is what you are buying and it is worth between one and two per cent a year of foregone advantage. The mistake is putting the emergency fund in the plan for the deduction, and the opposite mistake is refusing the plan entirely on the grounds that the tax rate might rise — which, as the break-even rates above show, would have to rise a great deal.