Coast FIRE Lab

How the Coast FIRE Lab calculator works

By Shuping Xu · Last updated 2026-10-28

Every number this calculator returns comes from four inputs and three assumptions. This page documents all seven: what each one is set to, where the value came from, and the condition under which it stops being true. If you find a figure on this site that cannot be traced back to this page, treat it as an error and tell me.

The four inputs that decide your number

Current invested assets. Retirement accounts plus taxable brokerage, at market value. Excluded: home equity, cash you need for emergencies, and anything you could not sell this year.

Monthly spending in retirement, in today's dollars. This is the single most sensitive input in the model. A 10% error here moves the answer more than a 1% error in the return assumption.

Your age now, and the age you stop contributing. Coast FIRE asks a specific question: at what balance can you stop adding money and still arrive at your target by retirement age, with growth doing the rest? The calculator defaults the retirement age to 67, which is the full retirement age for anyone born in 1960 or later under current SSA rules.

Expected return and inflation. Defaults are 6% nominal and 3% inflation, which give roughly 2.9% real. More on both below.

The calculator deliberately does not ask for your income or your savings rate. Those determine how fast you get to your coast number, not what the number is.

The 25× rule, and the 4% rule behind it

The target is annual spending divided by the withdrawal rate. At 4%, that is 25 times spending: $5,400 a month is $64,800 a year, and $64,800 × 25 = $1,620,000. That is the target before Social Security, which the calculator subtracts first — see the Social Security section below.

The 4% rate comes from two papers, not from a rule of thumb someone invented:

What those papers actually tested matters more than the number everyone quotes from them. They ran rolling 30-year windows of US market history, with portfolios between roughly 50% and 75% equities, and defined success as not running out of money before the window closed. They did not test 40- or 50-year retirements, non-US markets, or the effect of fund fees.

Where this assumption fails. Beyond a 30-year horizon, 4% has historically been too high. On a long horizon, use 3.5%. And the failure mode that matters most is not the average return — it is the order in which returns arrive. A 4% rate that survives the average path can still fail if the first decade is bad, because withdrawals during a drawdown sell shares permanently. That risk is real, and this calculator does not model it.

Growth, inflation, and why the default is 2.9% real

The nominal return is 6%. Inflation is 3%. Because the cost of living compounds too, what matters for a coast calculation is the gap between them, not either number alone: (1.06 ÷ 1.03) − 1 ≈ 2.9% real. Working in real terms is what lets the target be expressed in today's dollars, which is the only version of the number a person can actually interpret.

Where this assumption fails. 6% nominal is a historical US average for a mostly-equity portfolio, net of nothing. Fund fees come straight out of it. Bonds have not returned anywhere near that over the last decade. And 3% inflation is a recent-run average that looked generous in 2022, when CPI peaked above 9%.

Default spending: where $5,400 a month comes from

$5,400 a month is $64,800 a year. The published benchmark is the BLS Consumer Expenditure Survey: in 2024, households with a reference person aged 65 or older spent $61,432 a year on average, which is $5,119 a month. The narrower 65–74 bracket — closer to a household that has just retired — averages $65,354 a year, or $5,446 a month.

Our default sits about 5.5% above the 65-and-over figure, and within 1% of the 65–74 one, deliberately, for two reasons. First, the 65-and-over bracket pools households in their sixties with households past eighty, whose spending is materially lower, so the published average understates what a newly retired 67-year-old spends. Second, a coast number that is slightly too high fails safe — you end up saving a little more than you needed, rather than discovering the gap at 67 with no time left.

This is a default, not a recommendation. Replace it with your own actual spending. If you have never tracked it, take last year's bank and card statements rather than an estimate.

Where this assumption fails. If you own your home outright, $5,400 may be far too high, because housing is the largest line in the survey. If you carry a mortgage past 67, it is too low. If you plan to retire outside the United States, the entire benchmark is irrelevant.

Source: U.S. Bureau of Labor Statistics, Consumer Expenditure Survey 2024, Table 1300 — bls.gov/cex

Social Security: what we subtract, and when subtracting is wrong

The default is $2,084 a month. That is the Social Security Administration's own published figure for retired workers: the OASDI fact sheet dated 30 June 2026, Table A, reports an average monthly amount of $2,084 for retired workers in current-payment status. The figure moves as new claims join the rolls and cost-of-living adjustments land, so it is dated to the month it was published rather than rounded to a year.

The calculator subtracts annual Social Security from annual spending, then builds the target on the shortfall. It does not convert your benefit into a lump sum and add it to your assets. The reason is that Social Security is inflation-indexed and lasts for life, which is closer to a bond-like income stream than to a pile of money you could spend down — and converting it to a lump sum quietly embeds a discount rate you never chose.

When subtracting is wrong, or at least wrong at this value:

Source: Social Security Administration, Fact Sheet on the Old-Age, Survivors, and Disability Insurance (OASDI) Program, 30 June 2026, Table A — ssa.gov/OACT/FACTS

Pensions and one-time lump sums

A defined-benefit pension is entered as a fixed monthly amount and subtracted the same way Social Security is.

One difference matters: most private pensions are not inflation-indexed, so a $2,000 monthly pension is worth about $1,480 in today's purchasing power after fifteen years at 3% inflation. The calculator flags this rather than modelling it — if your pension has no COLA, the safe move is to enter a number lower than the one on the statement, or to leave Social Security in and the pension partially out as a buffer.

A lump sum is entered as invested assets on the date you expect to receive it, not on the date it is announced. If you are modelling a pension buyout offer, do not enter the buyout figure as current assets unless you could actually invest it today.

What the model does not do

This list is the honest half of the methodology. Everything here can change your answer by more than the assumptions above.

None of these gaps is hidden: they are listed here so you can price them yourself before trusting the number the calculator returns.

When our own defaults would be the wrong choice

  1. Horizon over 35 years. Drop the withdrawal rate to 3.5%. The 4% rate was tested on 30-year windows.
  2. You want high confidence, not a coin flip. Use 3% real growth instead of 2.9% nominal-minus-inflation, or hold a 10–20% buffer on the resulting number.
  3. Your spending is not average. Add a mortgage that runs past 67; subtract housing if the house is paid off; adjust for a move to a lower-cost country.
  4. A large share of your assets sits in one employer's stock or one sector. The model assumes a diversified portfolio. Concentration changes the distribution of outcomes, not just the expected value.
  5. You are within ten years of retirement. Historical sequence matters more than the average, and a bond glide path belongs in the plan at that point.
  6. You are not in the United States. Every source on this page is a US source: US market returns, US inflation, US Social Security, US tax treatment.

How the public by-age table differs from your calculator

The Coast FIRE number by age table on the homepage is a fixed reference, not a personalised result. It uses one assumption set for everyone: a 2.9% real return (6% nominal minus 3% inflation), a 4% withdrawal rate, retirement at 67, $2,084 a month of Social Security claimed at 67, three fixed spending levels — $3,000, $4,000 and $5,000 a month — and no pension or lump sum.

Your own calculation uses your inputs and the same 2.9% real default. Where your inputs match the table's, the two answers agree: a 45-year-old spending $4,000 a month with default assumptions reads the same figure in both places. They part company the moment one of the six fixed assumptions changes — a pension, a lump sum, a different claiming age, or a return you set yourself. The table exists so that a reader with no inputs yet can see the shape of the curve: how steeply the required balance falls as you get older.

Sources

Every figure above is re-checked against the publisher's own release: the BLS publishes the Consumer Expenditure Survey once a year, and the SSA updates its OASDI fact sheet twice a year. The date at the top of this page changes when either number does.