How the calculator works
The maths behind your retirement age, and what to do if you want to change it.
How retirement age is calculated: the survival test
The calculator finds your earliest retirement age by running a survival test: if you stopped working at age R, would your pots last until age 100 without hitting zero?
It simulates year by year from R to 100. Before your pension access age, spending comes out of your ISA/GIA only: the bridge. From pension age onwards, the calculator draws from your pension first (grossing up for your tax rate), then tops up from your savings if there's still a shortfall. Once your State Pension starts, that income reduces how much needs to come from your own pots.
Your retirement age is the earliest year where this test passes on the expected (average-return) path. On the Monte Carlo run, that same test runs across thousands of randomised return sequences to find the realistic spread of outcomes.
Tax note: SIPP withdrawals are taxed as income: so if you need £30k in retirement and your tax rate is 20%, the calculator assumes you draw £37.5k gross. ISA and GIA withdrawals are always tax-free. The State Pension is treated as taxable income at the same rate.
Tax simplification: State pension income is taxed at the flat marginal rate you set, as the model does not account for your personal allowance (≈£12,570). This is a conservative simplification: your effective tax on state pension income may be lower.
This calculator does not use the 4% rule or a fixed safe withdrawal rate. Instead, it tests whether your pots survive to age 100 under a full drawdown simulation.
The pension bridge: why your ISA balance matters as much as your SIPP
In the UK, your pension (SIPP or workplace scheme) is locked until age 57, rising from 55 in 2028. If you want to retire at 50, you need 7 years of expenses sitting in accessible savings (ISA or GIA) before you can touch the pension.
The calculator checks two conditions separately: (a) is your total pot large enough to fund spending to age 100? and (b) is your accessible pot large enough to cover those bridge years before the pension unlocks? Both must pass for the survival test to succeed.
Tip: If the bridge is what's holding you back, where your total wealth is fine but your ISA runs dry before pension age, shifting more of your annual contributions from pension to ISA can bring your retirement date forward.
Monte Carlo simulation: why you get a range, not a single number
Markets don't return exactly 7% every year. The calculator runs 1,000 simulations, each one playing out a different random sequence of annual returns drawn from your expected CAGR and volatility. Returns are drawn from a lognormal distribution calibrated so the median outcome matches your stated expected return: half the simulations do better, half worse, and the typical path is exactly what you entered.
For each simulation the calculator runs the same survival test and finds the earliest viable retirement age. The headline figure is the median, the age by which half the simulations reach independence. The "likely range" on the chart and results panel spans the 25th to 75th percentile of the simulations, which represents the middle half of all outcomes. The dashed line shows the smoother expected-return path for comparison; it sits a little earlier than the median because it ignores the sequence-of-returns risk the simulations capture.
The "chance of retiring by age X" stat shows what fraction of simulations passed the survival test at or before your target age.
Some gap between the expected path (dashed green) and the MC median (solid blue) is normal: it reflects sequence-of-returns risk, where bad early years in retirement deplete your pot before good years can recover it. The higher your volatility, the wider this gap. It's an honest reflection of real-world uncertainty, not a pessimistic assumption.
Because the simulation is random, re-running it with the same inputs can shift the median by a year. That's statistical noise, not a bug.
Median vs mean: The central line shows the median path (50th percentile). Because returns are lognormally distributed, the mean outcome drifts higher, so the average investor does slightly better than the central line suggests.
Stress Tests: sequence of returns risk
While the Monte Carlo simulation shows a statistical range, Stress Tests allow you to model a specific "nightmare scenario" on your expected path. They help answer: "What if the market crashes on the exact day I retire?"
A Retirement Crash simulates a -30% drop in the first year of retirement, while a Lost Decade models 0% real growth for 10 years. These tests only apply to the green dashed "Expected Path" line: they show how your baseline plan would buckle under targeted pressure, potentially requiring you to work a few years longer to maintain the same spending.
Real terms vs nominal: why the chart figures don't "grow" with inflation
Everything runs in today's money (real terms). If you enter 7% expected return and 2.5% inflation, the model uses a real return of ≈4.4%: the purchasing power gain after inflation. Your spending target stays constant on the chart because both the pots and the withdrawals are expressed in today's pounds.
A nominal model would show numbers growing to impressive-looking sums by age 70, but that growth is mostly just inflation. Real terms strips it out so the chart is honest about what your money will actually buy.
Contributions are assumed to keep pace with inflation: you're putting in the same real amount each year, just with the pounds adjusted upward.