Take what you spend in a year and divide it by a withdrawal rate. That is your FIRE number: the size of the pot that lets investment returns cover your spending, so working becomes optional rather than compulsory.
Spend £30,000 a year, use a 4% withdrawal rate, and the number is £750,000. One division. Everything else (the arguments, the blog posts, the spreadsheets) is about whether those two inputs are right, so that is where the rest of this piece lives.
The spending figure has to survive an audit
The most common mistake is building the number from salary. It does not matter what you earn once you stop earning it. What matters is what leaves your account each year, because that is what the portfolio has to replace.
So the first job is an honest spending figure, and honest means annual, not a good month multiplied by twelve. A year contains the car insurance renewal, the boiler service, Christmas, a dentist and at least one thing that breaks. Most people who track properly find their real annual spending is ten to twenty per cent above their estimate, which moves the FIRE number by the same proportion. If you have never measured, an expense diary kept for three months gets you a far better figure than any estimate. Guessing your spending and then computing a number to the nearest pound is precision built on a foundation of sand.
The figure also is not your current life, frozen. Take out the costs that stop: commuting, work clothes, and the mortgage if it will be repaid, since a mortgage that ends at 62 changes the shape of the whole plan. Then add the costs that start: health cover, more travel, more time at home with the heating on.
And expect the answer to move. Every change in spending moves the number by 25 or 28 times the annual difference, which is a brutal multiplier: a £200-a-month lifestyle creep is £60,000 on the target at 4%. Recalculate once a year, not once.
A judgement dressed up as a constant
The 4% figure comes from the Trinity Study, which tested US portfolios over rolling thirty-year periods and found that withdrawing 4% of the starting balance, rising with inflation, survived almost every one. That result is usually quoted without its conditions: thirty years, a specific US stock and bond mix, no fees, and a century of unusually strong American returns. Applied to a forty-five-year early retirement with a 0.5% platform charge, the same research does not support the same number.
So the rate is a judgement, and each choice implies a multiple of your spending. At 3.0% you need 33 times your annual spending, the territory of very long retirements and low tolerance for risk. At 3.5% the multiple is 28.6×, which suits an early retirement of 40 or more years. The classic 30-year horizon is where 4.0% and its 25× live. And 5.0%, a multiple of just 20×, only really belongs to people with flexible spending or other income arriving later.
The gap between those choices is not academic. On £30,000 of spending, 3% needs £1,000,000 and 5% needs £600,000: £400,000 of difference produced entirely by a number you chose, not by anything you did. The safe withdrawal rate explorer walks through where each figure comes from.
One assumption hides inside every one of those multiples: a constant real return. Real markets deliver the same average in wildly different orders, and the order matters enormously when you are drawing down at the same time. A bad first decade does far more damage than the same average spread evenly. That is sequence risk, and the retirement crash test is where to stress it.
Three people, three numbers
A single renter spending £24,000 has no mortgage to repay, so spending stays roughly flat into retirement. At 3.5% the number is £686,000. Rent is the risk here: it rises with inflation forever, where a repaid mortgage does not, which is the argument for the more cautious rate.
A couple spending £46,000, of which £14,000 is the mortgage, look like they need a fortune until you notice the mortgage ends in eleven years. Retirement spending is nearer £32,000, so at 4% the number is £800,000, not the £1,150,000 the pre-payoff figure implies. Failing to model the mortgage ending is one of the biggest single errors in amateur FIRE planning; it can overstate the target by a third.
A late starter aged 52, spending £38,000, has a state pension arriving at 67 that covers perhaps £11,500. The portfolio only needs to fund £38,000 for fifteen years and £26,500 after that, so treating the whole £38,000 as a permanent portfolio job overstates the number badly. This is why the FIRE calculator takes future income streams rather than a single flat figure.
The total can be right and the plan still fail
A pot of £750,000 that is 70% inside a pension you cannot touch until 57 will not fund retirement at 50. The number says nothing about access, so the total is correct and the plan fails anyway. That gap has its own arithmetic in the pension bridge calculator.
Tax plays the same trick. The number is what you must draw, not what you receive, and where the money sits (pension, ISA, general account) changes the after-tax income substantially. Two people with identical pots can end up with different spendable incomes.
The number you can actually pull on
Here is the uncomfortable part. Most people fixate on the FIRE number, and it is the least actionable figure in the whole exercise. You cannot do anything to £750,000 tomorrow.
What you can change is your savings rate, and it is a far better predictor of the date. Saving 10% of income takes around 51 years to financial independence on typical assumptions; 30% takes 28 years; 50% takes about 17. Not because the returns differ, but because saving more simultaneously builds the pot faster and lowers the pot you need, since the target is a multiple of what you spend. The savings rate calculator shows the curve, and it bends hard at the low end, which means the earliest percentage points are worth far more than the later ones.
And if the withdrawal rate is the part you are least sure of, the 4% rule explained covers where that number came from and when it stops holding.
So the working method is short. Measure your real annual spending over three months and annualise it honestly, subtract what stops in retirement, add what starts. Pick a withdrawal rate matched to how long the money must last, not to what sounds good, then divide, treating the result as an order of magnitude rather than a target to the pound. Check what fraction of it will be accessible before the pension unlocks. Then come back once a year and do it all again.
The number is a compass bearing, not a finish line. It tells you which direction counts as progress. What it cannot tell you is that you are safe: that depends on sequence, tax, access and how flexible you are willing to be, none of which survive being compressed into one division.