Every FIRE thread eventually arrives at the same argument. Is it 4%, or is it really 3.5%, or has the whole thing been dead since 2022? It is a good argument to have in the United States, where the rule was built. Carried into Europe unchanged, it answers a question most people are not actually asking, and it stays quiet about the thing that moves their date the most.
So we ran it. One European household, one portfolio, one set of assumptions, changing only the capital gains regime. Every number below came out of the free forecast calculator on this site, which runs the same engine as the app, and every scenario is a link you can open and edit.
The short version
- The 4% rule is stated before tax. It tells you what to withdraw, not what you get to spend. In Europe the gap between those two numbers is a policy decision made by your country.
- The same portfolio, same spending, same growth assumptions reached the earliest month it could stop in May 2039 under Belgian rules, August 2041 under German rules, and March 2042 under French rules. That is a spread of nearly three years, produced by nothing but the tax code.
- The rate everyone argues about moved that date by zero months. Setting the withdrawal rate to 3%, 3.5%, 4% or 5% changed the target pot, which is what the rule computes, and left the earliest stop date exactly where it was.
- If you want one number to be careful about, it is not the 4. It is the tax treatment of the account you plan to sell from.
Why the rule says nothing about tax
The 4% rule comes out of US research on US market history, and its withdrawals are gross. That was a reasonable place to stop: the tax you owe on a sale depends on your country, your account type, your other income and the year, none of which belong in a study about market survival.
The trouble is what happens when the number crosses the Atlantic. A rule stated before tax gets repeated as though it were a rule about spending money. In a country that taxes gains when you realise them, the two are different, and the difference compounds for every year of the drawdown.
The household
Deliberately modest numbers, and the same ones in every run.
| Age | 38 (born 1988) |
| Net salary | €4,000 a month, growing 2% a year |
| Spending now | €2,600 a month, indexed to 2% inflation |
| Retirement budget | €2,800 a month |
| Stocks | €180,000, plus €1,200 a month |
| Pension pot | €90,000, plus €200 a month |
| Cash | €20,000 |
| Assumed stock growth | 7% nominal, so 5% real |
At a 4% withdrawal rate, the rule's own arithmetic asks for €33,600 a year divided by 4%, which is a pot of €840,000. At 3% it asks for €1,120,000. That is the number the argument is about.
The same plan under three tax regimes
The calculator has a realistic tax mode: gains are taxed when they are realised, with an annual exemption and a separate rate on interest-type yield. The three settings below are the ones the app ships as country presets, so they are the rules the product itself applies.
| Capital gains treatment | Earliest month it could stop | Age |
|---|---|---|
| No tax on withdrawals | July 2038 | 50 |
| Belgium: 10%, €10,000 exempt | May 2039 | 51 |
| Germany: 26.375%, €1,000 exempt | August 2041 | 53 |
| France: 30%, no exemption | March 2042 | 54 |
Three years and eight months separate the first row from the last. The person is identical in all four. They save the same amount, hold the same assets, assume the same returns and want the same retirement budget.
Belgium's €10,000 annual exemption is doing a lot of the work in row two. A drawdown of roughly €33,600 a year that realises gains inside that exemption meets very little tax, which is why the Belgian answer sits so close to the untaxed one. Germany's allowance is €1,000, and France's headline rate applies from the first euro, so both of those plans hand over a real slice of every sale and need a bigger pot to survive the same spending. If you want the Belgian rules in detail, including the cost basis reset that came with them, we wrote them up here.
The part we did not expect
We ran the same Belgian scenario at a 3% withdrawal rate, at 3.5%, at 4% and at 5%.
The earliest stop date was May 2039 in all four.
That is not a bug, and once you see why, it is the most useful thing on this page. A withdrawal rate is an input to one specific calculation: how big a pot do I need. Divide your spending by the rate and you get a target. Change the rate and the target moves a lot, from €840,000 to €1,120,000 in this household's case.
But "when could I stop" is a different question, and a rate cannot answer it. The calculator answers it by simulating the months: selling what has to be sold, paying tax on the gains those sales realise, waiting for locked money to unlock, and checking whether the money outlives the plan. Nothing in that simulation asks what withdrawal rate you believe in.
So the forum argument is real, but it is an argument about a target number. Meanwhile the thing nobody in that thread mentioned, which country's rules apply to the sales, was worth thirty-four months.
The pension that will not open yet
The other European specific is that a large part of most people's wealth is legally locked until an access age. We set the pension pot in the scenario above to unlock at 65, which for this household is January 2053, and ran it again.
Nothing moved. Same stop date, and stop-now progress read 34% with the lock and 34% without it. The accessible side was carrying the bridge years on its own, so the lock never bound on anything.
It takes a different balance for it to show up at all. We tried a more pension-heavy version of the same €290,000, with €80,000 in stocks and €190,000 in the pension. There the same lock cut stop-now progress from 29% to 20%, and the earliest stop date still did not move.
So the conclusion from these runs is narrower than the story we set out to tell. A pension lock does not by itself push your date back. It changes how much of your pile counts toward stopping today, and the date only follows once the accessible side is too thin to cover the years before the pension opens. Which side of that line you are on is worth knowing, and no withdrawal rate will ever tell you.
What this leaves out
Plenty, and it matters.
- These are headline rates. Real tax codes have envelopes that change the answer: a French PEA, Belgian pension-savings products, German fund rules. The presets model the general regime, not your specific wrapper.
- Social contributions and healthcare in retirement are not in the model. In several countries they are not a rounding error.
- One expected path is not a guarantee. The same calculator has a Monte Carlo mode that runs 500 simulated markets and reports how often the plan survives. A single line through the middle always looks calmer than the thing it averages.
- Growth assumptions dominate everything. We used 7% nominal on stocks against 2% inflation. If you think that is generous, change it in the link and watch every date above move together.
The point is not that Europe makes early retirement harder. It is that the rule most people are using was never designed to answer the question they are asking it, and the input it leaves out is the one their government sets. If you would rather run this on your own numbers than on ours, the calculator is free, needs no account and stores nothing, and the whole scenario lives in the link. If you would rather it read your real balances and your real cost basis instead of typed-in figures, that is what the app does.
As always: these are estimates from your own assumptions, for planning. Not financial advice.