The 4% rule in the UK: does it still work?

As a first estimate, yes: 4% of your pot in year one, rising with inflation, is a sensible starting point. As a plan, no: it was built on American market history, ignores tax, fees and the State Pension, and assumes you never adjust. Research using UK market history lands nearer 3% to 3.5% for the same confidence. The fix is not a different percentage. It is to test your own plan, with your own pots, spending and State Pension, and read off the probability.

Figures for the 2026/27 tax year · Published 4 October 2026

The 4% rule is the most quoted number in retirement planning and one of the least understood. Here is where it came from, what it does and does not include, and how to use it without being caught out.

Where the rule came from

In 1994 an American financial planner, William Bengen, asked a simple question: looking at every 30-year retirement in US history since 1926, what is the most you could have withdrawn in year one, raised each year with inflation, without ever running out? With a portfolio of half shares and half bonds, the answer was a little over 4%. The Trinity study in 1998 reached a similar conclusion. Four per cent became the rule.

Read that carefully, because the fine print is the whole point. The rule is 4% of the starting pot, not 4% of whatever is left each year. A £500,000 pot gives £20,000 in year one; in year two you take £20,000 plus inflation, whatever the market did. It assumed a 30-year retirement, a portfolio at least half in shares, no fees, no tax, and a retiree who never changed their spending whatever happened.

Four reasons it travels badly to the UK

• Different history. American shares and bonds had an unusually good twentieth century. Studies that repeat Bengen’s method on UK or global market data generally find the safe figure nearer 3% to 3.5%, because the worst UK sequences (the 1970s above all) were harsher than the worst US ones.
• Fees. Bengen’s portfolio cost nothing to run. A typical UK pension in a workplace default fund costs 0.3% to 0.75% a year; an old personal pension or an adviser-managed portfolio can cost 1.5% or more. Every 0.5% of fees takes roughly 0.3 percentage points off the safe withdrawal rate over a long retirement.
• Tax. The 4% is gross. Once the State Pension has used your £12,570 personal allowance, every pound you draw from a pension beyond the tax-free quarter is taxed at 20% or more. £20,000 gross from a pension is closer to £17,000 in your hand, depending on how you take it.
• Longer than 30 years. Retire at 60 and a 30-year horizon takes you to 90; a healthy 60-year-old has a reasonable chance of living past it. Extend the horizon to 40 years and the historical safe rate falls further, towards 3%.

And three reasons it is too cautious for many people

The rule is often attacked for being too generous. For a lot of UK retirees it is the opposite.

• The State Pension. The full new State Pension is £12,548 a year in 2026/27, inflation-linked and guaranteed. A couple on two full pensions has £25,000 a year that is not subject to any withdrawal rate at all. The 4% only has to cover the gap, not the whole budget.
• Nobody spends in a straight line. Real spending tends to fall through the seventies and eighties, as travel and big purchases tail off, before care costs may push it back up. A rule that assumes flat real spending for 35 years overstates what most people actually draw.
• You will adjust. The rule’s worst cases come from a retiree who keeps raising withdrawals with inflation through a crash. Almost no one does that. Skipping the inflation rise for a year or two after a bad year, or trimming spending by 10%, transforms the odds. Guardrail rules that do this systematically support starting rates nearer 5%.

A worked example

Jo retires at 65 with £500,000 in pensions and ISAs, a full State Pension from 67, and wants £30,000 a year after tax.

• The 4% rule says £20,000 a year from the pot. With the State Pension from 67 that is £32,500 gross, or about £29,500 after tax if a quarter of each pension withdrawal is taken tax-free. On the rule, Jo is just about there.
• For the two years before 67, Jo needs the whole £30,000 from the pot: more than 6% a year, not 4%. If those happen to be bad market years, the pot takes a lasting hit. Our guide to bridging the gap covers how to fund those years from cash and ISAs instead.
• If Jo’s pension charges 1% a year and the horizon is 35 years, the historically safe rate is nearer 3.3%, or £16,500. The rule overstates the budget by about £290 a month.
• If Jo is willing to skip the inflation rise after any year the pot falls, 4% looks comfortable again on the same history.

Four different answers from one pot, and none of them wrong. That is the problem with a rule of thumb: it cannot see your State Pension date, your fees, your tax position or how you would behave in a crash.

How to use the rule well

• Use it to size the pot you need, early on. Take the annual income you want from investments, after the State Pension and any final salary pension, and multiply by 25 (for 4%) or 30 (for 3.3%). That is the range to aim for. It is a good target to carry through your forties and fifties.
• Do not use it to set your spending once retired. At that point you know your real pots, dates and tax. Model them.
• Count the State Pension separately. It is the best inflation-linked annuity you will ever hold. Fill any gaps in your National Insurance record before you worry about the withdrawal rate.
• Decide your adjustment rule in advance. “If the pot ends the year lower than it started, I freeze spending next year” is simple, and it is most of the difference between 3.5% and 4.5%.
• Replace the percentage with a probability. A stress test runs your plan through a thousand futures and every retirement year on record, and tells you the chance it lasts. That number already includes everything the rule leaves out.

The short version

The 4% rule is a good way to work out roughly how much you need. It is a poor way to decide how much to spend. Multiply by 25 while you are saving, and when you get close, swap the rule for a test of your own plan.