The plan that fails with money in the bank
Picture the moment it goes wrong. You stopped working at 53 because the total looked sufficient, and only afterwards did you discover that two thirds of your money sits in a pension you cannot legally touch for another four years. The arithmetic was never wrong. It answered a question about totals when the binding constraint was access, and that is precisely why this failure catches careful savers: they checked the sum, not the timing.
The years between stopping work and unlocking a pension are the bridge, and they must be funded entirely from money you can spend at any age without penalty: ISAs, general investment accounts, savings, premium bonds. Not pensions, and not money already committed elsewhere. If the accessible pot runs dry before the unlock date, the plan fails, however large the pension sitting beside it. (And if you plan to retire after your access age, you have no bridge at all and this page has nothing to tell you; the FIRE calculator is the one you want.)
Two pots, two phases
Enter your accessible and pension pots separately, with what you add to each, the age you intend to stop, and the age your pension unlocks. While you are still working, both pots grow and receive contributions. Once you stop, only the accessible pot is spent; the pension keeps compounding untouched, and the page shows what it reaches by the unlock age, because being unable to spend money is not the same as it sitting idle.
Then comes the question that matters: does the accessible pot survive to the unlock age, and if not, at what age does it run out? Alongside that sits the requirement, the amount you need in accessible savings on the day you stop. It is calculated as the present value of the withdrawals rather than spending multiplied by years, because money left in the pot keeps earning while you draw it down; the naive multiplication overstates the target, sometimes badly.
The model runs one steady, real after-inflation rate, and a bridge is exactly where that assumption is most dangerous. You are selling assets in a fixed window with no flexibility to wait out a downturn, so a bad first two years of a four-year bridge is a materially worse outcome than the same average return delivered smoothly. This is sequence risk in its most concentrated form, and the retirement crash test is the right place to stress it. Tax is not modelled either, and it is significant here: ISA withdrawals are tax-free, drawing from a general investment account may realise capital gains, and at 57 the pension itself is typically 25% tax-free with the remainder taxed as income. A bridge that works before tax may not work after it, so treat the output as gross and take advice before acting on a marginal case.
Spending is flat in the model too, though in practice bridge spending is often lower than later retirement spending, and one-off costs land where they land. Other income (rental, a partner still working, a defined benefit pension starting mid-bridge, the state pension arriving later) is ignored, and any of those changes the picture. Even the two pots are a simplification, since real portfolios hold several wrappers with different rules and some have partial access; model your genuinely accessible money as one figure and everything locked as the other. And the tool assumes you stop completely, which most early retirees do not. Even modest earnings shorten the bridge considerably.
Fifty-five, fifty-seven, and April 1973
Private pensions in the UK currently unlock at 55, rising to 57 in April 2028; if you were born after roughly April 1973, plan on 57. The government's published rationale for the rise sets out the change. Some older schemes carry a protected lower age, and public sector schemes have their own rules, so check your own scheme's terms rather than assuming: the difference between 55 and 57 is two full years of bridge funding, which is a large number. The state pension is separate again and arrives much later, currently 66, rising to 67 and then 68; GOV.UK's checker gives your own date. Other countries differ substantially, which is why the access age here is a slider rather than a fixed assumption.
When the pot falls short
A gap leaves you pulling on one of only four levers, and they trade against each other. Working longer is the most direct, and its effect is larger than it first appears because it shortens the bridge from both ends at once: more accumulation, fewer years to fund. Spending less during the bridge works too, and only the bridge years need to be lean; once the pension unlocks, spending can rise again, which is why many people find a temporary, bounded squeeze more palatable than delaying.
The third lever is shifting contributions from the pension toward accessible accounts. That fixes the access problem directly and costs you the tax relief, which for a higher-rate taxpayer is substantial. Rarely does it make sense to redirect all of it, because you would also give up any employer match, usually the best return available anywhere; a common approach keeps the match and directs the surplus toward an ISA. The bridge guide works that split through with numbers attached, alongside how the bridge pot is sized. This is the genuine trade-off, with no universally correct answer. It depends on your marginal rate now versus in retirement and on how long the bridge is, and the tax treatment of drawdown is complex enough that a qualified adviser is genuinely worth their fee here.
Last, earning something during the bridge. Part-time or seasonal work does not need to cover everything, only the gap, and that approach has a tool of its own: the Barista FIRE calculator.
If the tool shows a large surplus instead, run the same logic in reverse: money beyond what the bridge requires could go into the pension and collect relief you are currently declining.