How Much Money Do You Need to Retire?

The short answer

The most widely used starting point is the 4% rule. Estimate what you expect to spend in a year of retirement and multiply it by 25. That product is the nest egg supporting a first-year withdrawal of 4% of the portfolio: $40,000 of annual spending points to $1,000,000, $60,000 to $1,500,000, and $80,000 to $2,000,000. Dividing your spending by 0.04 gives the identical answer, because 1 ÷ 0.04 is 25.

That figure is a planning heuristic, not a promise and not a law. It rests on a portfolio that keeps growing, on a withdrawal that rises with inflation every year, and on a retirement of roughly 30 years. Change any of the three and the target changes with it — retiring two decades earlier than the rule assumes makes the same 4% a far more aggressive bet. Even so, having a number beats having a feeling, and this one you can then interrogate.

• Target nest egg = expected annual spending × 25.
• Equivalently: expected annual spending ÷ 0.04.
• The driver is spending, not salary.
• Guaranteed income reduces what the portfolio must cover, so it reduces the target.
• A longer retirement argues for a lower initial rate, which raises the target.

Where the 4% rule comes from

The rule traces to a 1998 study by Michael Bengen, who tested what happens when a retiree withdraws a fixed percentage from a portfolio of US stocks and bonds, then raises that withdrawal each year to keep pace with inflation, using actual historical US returns rather than a projection. He was looking for the initial rate at which the money did not run out across the retirement periods in that data. The 4% figure that entered popular use came out of that work.

Two details of the origin are the honest limits of the tool. First, it is a backtest: it tells you which rates survived the periods that already happened, and it cannot guarantee the next few decades repeat them. Second, it was built around a 30-year retirement. Someone stopping work at 65 is close to the tested case; someone stopping at 40 or 45 is asking that rate to cover a horizon perhaps two-thirds longer, which is a different question. That gap, not any flaw in the arithmetic, is why lower starting rates are commonly quoted for early retirees.

Start with spending, not with income

The most common mistake in this calculation is planning from your salary. Salary is what you earn; the multiplier applies to what you spend. Someone earning $120,000 who lives on $55,000 needs the same pot as someone earning $70,000 who lives on $55,000 — $1,375,000 in both cases, because that is what $55,000 × 25 demands. Replacing 100% of your pre-retirement income in your head almost always overshoots the number.

So the real work is producing a spending figure and being alert to what changes when you stop. Costs that usually fall: commuting, work clothes and lunches near the office, pension contributions and payroll taxes you no longer pay, the mortgage if it clears first. Costs that usually rise or become visible: all the time you used to spend in five weeks a year, healthcare, travel. Costs that lump rather than spread: a car, a roof, home repairs, major dental work. A sound method is last year's spending, minus what genuinely disappears, plus what replaces it, plus a margin for the lumpy items rather than pretending they average out.

• $30,000 a year of spending → $750,000.
• $50,000 → $1,250,000.
• $60,000 → $1,500,000.
• $80,000 → $2,000,000.

How much you need at each retirement age

Age does not change the multiplier directly — 4% of a pot is 4% of a pot at any age. What age changes is how many years the pot must last, and therefore how safe a 4% first withdrawal is, because the rule was tested on roughly 30-year retirements. Lower starting rates mean larger targets: at 3.5% the multiplier is about 29, at 3% about 33.

The examples below hold spending at $40,000 a year so the effect of the horizon is visible on its own, and each assumes no guaranteed income covering any of it. If a pension or state benefits will cover part of your spending, apply that income to the spending first and multiply only the remaining gap — the guaranteed-income section shows exactly how.

Retiring at 40

The 25× version gives $1,000,000 for $40,000 of spending, and that is what a calculator will show you first. But it means planning for a horizon that could run 50 years or more, nearly double the period the underlying study examined. A rate that survived every historical 30-year window says nothing directly about a 50-year one: the extra two decades give inflation more time to erode the same purchasing power and the portfolio more decades in which it must keep growing.

Hence the lower rate commonly advised here. At 3.5%, $40,000 of spending needs roughly $1.14 million rather than $1,000,000; at 3%, about $1.33 million — a third more than the headline. Neither is safe, exactly; both are simply more conservative than a test built for a shorter life. The alternatives to a bigger pot are a smaller gap to fund, some later part-time income, or the willingness to spend less in a bad year.

Retiring at 45

The arithmetic is unchanged — $1,000,000 at 4%, around $1.14 million at 3.5% — but a structural problem shows up: the bridge. If the state pension or Social Security benefits you will eventually get do not start until well after 45, every dollar spent between now and then comes out of the portfolio alone, with no guaranteed income underneath it, and it is the same pot that must still exist decades later.

Two consequences. The first is that sequence-of-returns risk peaks immediately after you stop, which is why many early retirees keep several years of spending in cash or short-dated assets rather than all growth assets. The second is that the years between stopping and eligibility are worth pricing explicitly, because benefit ages change and vary by country and work history — check what applies to you instead of assuming a figure you read online.

Retiring at 50

Fifty is an awkward middle: a horizon that plausibly reaches 40 years or more, so still past the 30-year assumption, but with less pure portfolio-only funding ahead than a 40-year-old faces. At 4%, $40,000 of spending is a $1,000,000 target; wanting a 3.5% start for a 40-plus-year horizon puts it nearer $1.14 million.

This is also the age where the gap between saved and needed can look unbridgeable and then turn out not to be, because ten to fifteen years of saving is still a real runway. That is precisely the question the savings inputs answer: given current age, target age, what you already have and an assumed return, what monthly amount closes the distance. Run several monthly figures before concluding the target is out of reach.

Retiring at 55

At 55, $40,000 of annual spending still points to $1,000,000 under the 4% rule. The horizon is long — plausibly 35 years or more — so a rate a little below 4% is a defensible conservative choice rather than excessive caution, and the distance between 4% and 3.5% on this spending level is roughly $140,000 of extra target.

What changes materially at 55 is access. In several systems, leaving work in the mid-fifties opens earlier-but-reduced benefit options or lets a workplace pension be drawn, and an earlier claim is usually permanently smaller, not temporarily. Whether taking income earlier and smaller beats funding the gap yourself is a genuinely personal calculation that depends on rules which change, so get it answered for your own situation before you fix the number.

Retiring at 60

Sixty brings the horizon to something like 25 to 30 years, which is close to the case the rule was built for. That does not make 4% safe — it makes 4% the number the heuristic was designed around. A $40,000 spend is a $1,000,000 target; an $80,000 spend is $2,000,000.

The more useful refinement at this age is guaranteed income. If part of your spending will be covered by benefits or a pension starting now or shortly after, the portfolio only has to fund what is left. Spending $60,000 with $20,000 of reliable guaranteed income leaves a $40,000 gap, and $40,000 × 25 is $1,000,000 — not the $1,500,000 the full spending figure would demand.

Retiring at 62

Sixty-two gets searched constantly because it tends to line up with an eligibility milestone: an early-retirement age in a workplace scheme, or the earliest claim age for a state benefit. The maths is the same as at 60 — $40,000 of spending is a $1,000,000 target at 4%, over a horizon of roughly two and a half decades or a little more.

The real question at 62 is the trade-off inside that milestone: claiming earlier generally means a smaller, permanently adjusted amount, so the portfolio must cover a larger share of spending for longer. Model it rather than eyeballing it — put your after-guarantee spending gap through the calculator and test the same gap against a few different assumed returns.

Retiring at 65

At 65 you are squarely inside the case the original study examined: about a 30-year retirement from US stocks and bonds with an inflation-adjusted withdrawal. $40,000 of spending is a $1,000,000 target, $60,000 is $1,500,000, $80,000 is $2,000,000. This is the age at which 25× gets quoted most often, and the age at which it is most fairly quoted — being close to the tested case is still not the same as being guaranteed.

Two refinements outrank the multiplier itself here. Longevity: a 30-year test does not promise the money lasts to 95, so people with a long family history, or the surviving half of a couple, often plan below 4%. And the guaranteed-income base: where benefits and any pension cover part of the spending from the start, the portfolio funds only the gap, which is frequently the biggest lever available at this age.

What retiring “comfortably” actually means

“How much do I need to retire comfortably” has no number answer, because comfortable describes a relationship between your spending and your income rather than a sum. Anyone quoting a single figure for it is guessing at somebody else's lifestyle. The workable reformulation is: take what you need to spend, subtract the income you can count on without selling assets, and multiply only the remaining gap by the multiplier. Guaranteed income usually means a workplace or personal pension, state benefits such as Social Security, an annuity, rent, or a part-time role you intend to keep.

So one lifestyle produces two very different numbers depending on what sits underneath it. Spending $80,000 a year with $30,000 of guaranteed income leaves a $50,000 gap, and $50,000 × 25 is $1,250,000. The same $80,000 with nothing guaranteed is $2,000,000. That is not rounding — it is two plans funding the identical life. The caveat is that income is only guaranteed to the extent it really is: benefit rules and ages change, and a pension promise is only as solid as the scheme standing behind it.

• Portfolio target = (annual spending − guaranteed income) × 25.
• $60,000 spending, $20,000 guaranteed → $40,000 gap → $1,000,000.
• $80,000 spending, $30,000 guaranteed → $50,000 gap → $1,250,000.
• $80,000 spending, nothing guaranteed → $2,000,000.
• Comfort is the margin: the pot that covers the plan plus the shocks the plan cannot forecast.

Inflation: the term that makes the rule work

The 4% is a first-year rate. You are not taking 4% of the balance each year; you take a dollar amount set in year one and raise it every year afterwards to match inflation, so purchasing power stays level instead of sliding. That is exactly why the rule needs the growth assumption to hold: the portfolio must return more than inflation plus withdrawals, on average, across the whole retirement. If the pot only keeps pace with inflation, a constant real withdrawal from a pile of 25× spending depletes it in roughly 25 years with nothing going wrong at all.

The compounding of inflation over a long horizon is easy to underestimate because it is quiet. At a sustained 2%, prices double over about three and a half decades — a whole retirement for someone stopping at 60, and only two-thirds of one for someone stopping at 40. The same $40,000 of purchasing power at 40 is a materially larger nominal withdrawal by 70, and the plan has to survive both halves of that. One practical consequence: be consistent about whether your assumed return is nominal (inflation inside it) or real (net of inflation), and match it to the spending figure you entered. Mixing the two is the most common way a confident projection turns out to be wrong.

Sequence of returns: why the same average can fail

Two portfolios can earn an identical average return over 30 years and end in very different places, and the difference is the order. A retiree withdrawing while markets fall must sell more units to raise the same dollars, permanently shrinking the holdings that later recoveries would have grown; a retiree whose bad years arrive in decade two keeps more units through the rebound. Withdrawals invert the usual comfort about long horizons, because they convert a temporary price decline into a permanent reduction in what you own.

This is why horizon and withdrawal rate interact so strongly, and it is the real reason early retirement is harder than 25× arithmetic implies. A 30-year window contains relatively few severe early-decade sequences; a 50-year window contains many more chances for one, with more years to compound downward from it. The responses planning practice generally offers are not exotic: several years of spending in cash or short-dated assets so a downturn does not force sales, genuinely flexible spending, keeping part-time income available, and being willing to move the retirement date if the opening years go badly. None of those make a number guaranteed. They make it more likely to survive.

Running the numbers: the three calculators

Hand arithmetic is where people quit, and where it quietly goes wrong. The FIRE Retirement Calculator takes the whole framing at once: current age, retirement age, monthly savings, initial savings, an expected return and your expenses. From those it produces a nest-egg target using the 4% rule, the monthly savings you would need, and a year-by-year balance projection — which is the part that actually answers the age-qualified questions, since you can watch what a given horizon does to the required rate instead of reasoning about it in the abstract.

The other two answer narrower but necessary questions. The Compound Interest Calculator takes principal, rate and years and shows monthly, quarterly, yearly and continuous compounding side by side with an APY comparison, so you can see how much of your projection is assumption rather than arithmetic. The Savings Calculator works backwards from a fixed goal: enter the goal, current savings, the number of months and an expected annual rate, and it returns the exact monthly deposit required. Its stated assumptions are worth reading — a constant annual rate compounded monthly, with no fees or taxes — and the tool itself labels its output a planning estimate rather than financial advice, which is the right framing for everything on this page.

• Step one: get a spending figure from last year's real numbers, adjusted for what stops and what starts.
• Step two: subtract income you can count on without selling assets, to get the gap the portfolio must fund.
• Step three: feed that gap and your horizon into the FIRE Retirement Calculator and read the required monthly savings and the year-by-year projection.
• Step four: re-run it at a lower assumed return and a lower withdrawal rate. If the conservative version is not wildly out of reach, you have a plan.

Mistakes that make the number wrong

Most bad retirement numbers come from an input nobody examined, not from the 4% rule being imperfect. The frequent ones: planning from income instead of spending, which inflates the target; ignoring guaranteed income, which inflates it again; or assuming guaranteed income in full when the eligibility ages and rules do not match what you read online — benefit ages and reduction factors change and differ by country and work history, so verify yours rather than carrying a forum figure into a 30-year plan.

The remainder are omissions with predictable names: forgetting that the withdrawal rises with inflation every year after the first; treating the pot as covering routine spending while also paying for cars, repairs, healthcare and travel out of it; counting a home you live in as though it were liquid; assuming a multiplier tested on 30 years stretches to 50 unchanged; and never re-running the calculation once circumstances shift. Each of those is a ten-second change in a calculator, which is the cheapest discipline available in this whole exercise.

• Planning from salary instead of spending.
• Leaving out guaranteed income, or overstating how much of it is actually guaranteed.
• Assuming 4% is safe at any horizon.
• Forgetting the inflation adjustment, or the lumpy one-off costs hiding inside the spending figure.
• Counting a home you live in as though it funded withdrawals.
• Building the number once and never re-running it as spending, savings or the target date move.

Frequently asked questions