What is a sustainable withdrawal rate for your retirement portfolio?

01 What the 4% rule actually says

William Bengen's 1994 paper asked a narrow question: what is the largest first-year withdrawal, increased each year by inflation, that would have survived every rolling thirty-year period in US market history? His answer was slightly above 4%, and it became shorthand for a rule he did not state.

The Trinity study extended the work in 1998, publishing success rates for different withdrawal rates, horizons and asset mixes. Both were historical survival analyses, not forecasts, and both reported the worst case rather than the typical one — a 4% rate succeeded in every window, and in most windows the retiree died with more money than they started with.

Four things are baked in: US returns, a thirty-year horizon, a withdrawal that never varies in real terms, and no charges. Change any one of them and the number moves.

Source: Determining Withdrawal Rates Using Historical Data

02 Why the number is different in the UK

UK equity and gilt returns over the twentieth century were not the same as the US series Bengen used, and the US record is the most favourable major market of the period. Applying a rate calibrated on the best available history to a different market is a survivorship problem, not a translation.

Charges are the second adjustment and the one most often ignored. Bengen's analysis was gross. A platform fee plus a fund charge of 0.7% in total comes directly off the sustainable rate — which is the whole reason fees matter more in drawdown than in accumulation.

The third is the horizon. A 30-year window from age 65 runs out at 95, which is no longer a safe assumption for a couple; the chance of at least one of them reaching 95 is not small. A 40-year horizon lowers the safe rate materially.

There is a fourth adjustment nobody makes and everybody should: the withdrawal in the studies is inflation-linked with no upper limit, so a decade of high inflation raises the real cost of the plan exactly when the portfolio is least able to bear it. UK retirees who lived through 2022 saw their required withdrawal rise by around a tenth in a single year while the portfolio fell. A rule that increases spending automatically in that situation is the opposite of a safety mechanism, and it is embedded in every quoted safe withdrawal rate.

Source: National life tables, UK

03 What makes it higher: the State Pension

A UK retiree is not asking a portfolio to fund their whole retirement. The State Pension arrives at 66 or 67, is inflation-linked, and is guaranteed for life — which means the portfolio's job is to cover the gap years in full and only the shortfall thereafter.

That structure raises the sustainable rate on the portfolio considerably, because the hardest period for a withdrawal strategy is the early years and the portfolio is not carrying the load indefinitely. A retiree at 60 with a full State Pension record has seven years of heavy withdrawals followed by a permanent step down.

This is why applying a single flat percentage across the whole of retirement is a poor model of a UK plan. The withdrawal pattern is naturally front-loaded, and a fixed-percentage rule cannot express that.

Source: The new State Pension

04 The rate depends on the rule, not the other way round

A retiree who will never change their withdrawal needs a low rate, because the plan has to survive the worst sequence without adaptation. A retiree who will cut spending by 10% after a bad year can start considerably higher, because the strategy self-corrects before the damage compounds.

The size of that difference surprises people: the gap between a rigid rule and a modestly flexible one is more than a percentage point of starting income, which on a £500,000 portfolio is over £5,000 a year. No fund selection decision available to a retail investor is worth that much.

So the first question is behavioural rather than financial. What would you actually do in a year the portfolio fell a quarter? An honest answer to that determines which rate you are entitled to use.

The other reason the rule matters more than the rate is that it is testable in advance. A rate is a claim about the future; a rule is a description of your own behaviour, and you can check it against what you did in 2008, 2020 and 2022. Someone who sold at the bottom in each of those years should not be planning on a strategy that requires them to hold through a fourth.

On the defaults above, the worked example shows £16,500 a year. Charges take £3,500 of that before you spend anything. A 10% cut after a bad year would take the income to £14,850.

Source: Retirement income market data

05 Sequence risk is the real enemy

Two retirees with identical average returns can end up in completely different places if the bad years fall at different times. Losses in the first five years of drawdown are permanent in a way that later losses are not, because the withdrawal sells units at depressed prices and those units never recover.

That is sequence risk, and it is the reason a safe withdrawal rate is so much lower than an average return. The average is irrelevant to someone taking money out; the order is everything.

It also explains why the fixes that work are all about the early years: a cash buffer to avoid selling in a downturn, a lower initial rate that rises later, or an annuity floor covering essentials so that portfolio withdrawals become discretionary.

Source: Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable

06 A practical UK starting point

Cover essential spending with guaranteed income first — State Pension, any defined benefit pension, and an annuity if there is a gap. Then treat the portfolio as funding discretionary spending, where a variable withdrawal is tolerable because the consequence of a cut is a cheaper holiday rather than a cold house.

On the portfolio itself, start between 3.5% and 5% depending on how flexible you can genuinely be, hold one to two years of withdrawals in cash so a bad year does not force sales, and review annually against the actual balance rather than against the original plan.

And revisit at State Pension age, because the portfolio's job changes on that date. The withdrawal that was funding everything becomes a top-up, and the rate can usually fall.

Two numbers make this concrete. On a £500,000 portfolio, moving from a rigid 3.5% to a flexible 4.5% is £5,000 a year of extra income, permanently. Holding two years of withdrawals in cash costs perhaps £700 a year of forgone return. The flexibility is worth roughly seven times what the buffer costs, which is the ratio that should decide where the effort goes.

Source: Plan your retirement income

07 What to ignore

Ignore any single number presented without its assumptions, including the ones in this post. A withdrawal rate is a function of horizon, asset mix, charges, flexibility and guaranteed income, and quoting it alone is like quoting a speed without a distance.

Ignore backtests that stop at 2021, because a decade of falling yields flatters bond returns in a way that will not repeat from here. And ignore the instinct to solve this once — a withdrawal strategy is a rule you apply annually, not a decision you make at 65 and file.

Source: MoneyHelper: pensions and retirement