How FireMe’s Calculators Work

FIRE Plan

It projects one row per year from your current age to 70 and reports the age at which your withdrawable income first covers your inflated spending requirement. If your income already covers that requirement at your current age, or still does not by 70, the results say so and name the surplus or the shortfall. Open the FIRE Plan calculator.

Net income grows each year at your wage-growth rate. Spending and your monthly financial independence requirement both grow at your inflation rate. Whatever income is left after spending, rent or mortgage payments is added to the invested pot, which compounds at your investment return.

You can also add one-off lump sums, such as an inheritance or a house sale, landing at a chosen age before you reach FIRE. Each is added to the invested pot in the age it lands and compounds like the rest of your savings from then on, so a lump sum can only bring your FIRE age forward, never push it back.

If you own your home, the property grows at the housing-growth rate and the mortgage is amortised as a standard annuity, so equity builds as the loan is paid down. Realizable capital is the invested pot plus that equity. If you rent, rent grows with inflation and no housing capital accrues.

Each year the tool compares the monthly income that capital could support, meaning realizable capital times your withdrawal rate, reduced by the tax rate you set on withdrawals, against the inflated monthly requirement. The FIRE age is where those two lines cross, interpolated between the two years either side of the crossing.

The headline FIRE number is a closed form: twelve times your monthly requirement, divided by your withdrawal rate after the withdrawal tax.

The projection runs in future money, then divides every money figure it reports by inflation compounded to that year, so results read in today’s purchasing power and inflation is counted exactly once. The growth rates you enter are before inflation.

Starting values: Age 34, net income 4,000 a month, 100,000 invested, spending 1,800 a month, a 3,000 a month independence requirement, 4% withdrawal rate, 8% tax on withdrawals, 6% investment return, 3% wage growth, 2.5% inflation, and an owned home worth 400,000 against a 300,000 mortgage at 3% with 25 years left.

Tax: Two tax settings. A flat rate on FIRE-phase withdrawals, and an optional wealth tax with its own rate, threshold and property-valuation multiplier, switched off by default because most jurisdictions do not levy one.

Biggest thing it leaves out: Income tax during the accumulation years is not modelled at all. The income you enter is already net, so a change in your marginal rate is something you apply yourself before typing the number in.

FIRE Simulation

It replays your plan against real market history, once for every start year the data covers, and reports how many of those retirements survived. Open the FIRE Simulation calculator.

This is a rolling-window backtest, not a random simulation. For each start year with a full window of data, the plan is run through that year’s actual returns and the years that followed. Retiring in 1929 and retiring in 1982 are two of the runs, with the real sequences attached.

Within a year, the withdrawal comes out first, then the remaining balance earns the blended return of your target allocation. Applying the target-weighted return to the whole balance is the same as rebalancing once a year, which is what the engine assumes you do.

Everything is computed in nominal money and then deflated by the actual inflation of those same years, so results are shown in today’s purchasing power and inflation is counted exactly once. A 1966 retirement and a 1982 one are directly comparable as a result.

A start year is only used if every asset you hold, plus any series the chosen strategy needs, has data covering the entire window. A longer retirement is therefore tested against fewer historical windows, holding any gold shortens the tested history to 1971 onward, and a valuation-linked strategy needs the valuation series to reach back that far too.

The withdrawal strategies that ship run from the fixed inflation-adjusted amount that most studies benchmark against through to valuation-linked and guardrail rules that change spending as the portfolio moves.

Each one has a page of its own, covering how it sets a withdrawal and where it struggles:

• Constant Amount
• Percent of Portfolio
• 1/N
• Variable Percentage Withdrawal (VPW)
• Dynamic SWR
• Endowment Strategy
• Guyton-Klinger
• 95% Rule
• CAPE-based
• Sensible Withdrawals
• Hebeler Autopilot II
• Vanguard Dynamic Spending

Starting values: A 30-year retirement on a 1,000,000 portfolio held 80% stocks, 15% bonds and 5% cash, withdrawing 40,000 in the first year and raising it with inflation thereafter. That first withdrawal is 4% of the portfolio.

Tax: None. Withdrawals are gross, and no account type or holding period is modelled.

Biggest thing it leaves out: The history is US market history. The currency selector changes the symbol in front of the numbers and nothing else, because no exchange-rate series is applied. A high historical success rate says a plan survived every sequence that has happened, which is not the same as saying it will survive the next one.

Lifestyle Change

It is the drawdown loop with the things a career break actually involves bolted on: side income over a date range, one-off lump sums, and two separate tax rates. Open the Lifestyle Change calculator.

The yearly loop matches the drawdown calculator, with three additions. Side income applies between a start and end year you choose and can optionally be inflated over that span. Lump sums land in a single named year. Both are taxed at your income rate before they reach the balance.

Investment gains carry their own separate rate. Only the after-tax gain compounds, and the monthly passive income the tool reports is after-tax too, so the summary, the table and the chart all read the same numbers.

The loop inflates your expenses as it runs, then divides every reported figure back by the same factor, so the table and chart are in today’s purchasing power and inflation is counted exactly once. The return you enter is before inflation.

Starting values: 5% return, 2.5% inflation, 1,000,000 starting balance, 2,000 a month of spending, over 25 years, with 25% tax on earned income and 15% tax on investment gains.

Tax: The only calculator here that taxes investment gains. Earned income and lump sums use one rate; gains use another. Both are flat rates you set, not a bracket table, because brackets differ by country and go stale every year.

Biggest thing it leaves out: Flat rates mean no allowances, no thresholds and no capital-gains treatment that differs from income. Returns are still a fixed rate every year, as in the drawdown calculator.

Die with Zero

It solves the earliest quit age whose chance of going broke before a chosen plan-to age stays inside a chosen tolerance, using the same exact probability engine as the retirement probability calculator. Open the Die with Zero calculator.

Each candidate quit age defines a full cashflow schedule: a monthly saving amount lands until that age, a monthly spending amount is drawn after it, and any dated one-time amounts or income periods sit at their own years, unmoved by the solve. The engine propagates the exact distribution of the balance through that schedule and reads the cumulative chance it has hit zero by the plan-to age.

A later quit swaps a drawdown year for a saving year, so risk never rises with the quit age; the earliest age whose risk fits under the tolerance is therefore a well-defined boundary, and the calculator searches straight to it. The tolerance presets are plain odds (1 in 50 down to a coin flip), and every computed probability is shown as a whole percent. The plan-to age is the reader’s own input by default, and the period life tables this site ships are read only when the reader opts in.

That opt-in adds a second way to measure the same risk. Given a country and sex, the death distribution for someone of the reader’s age is multiplied through each candidate’s cached ruin curve, giving the chance of running out at some point while still alive: the sum over years of the chance of dying in that year times the chance of being broke by the end of that same year. Pairing them that way charges the year of death as a year the money still had to cover, which is what the model spent it on, and it never reads the risk lower than the alternative. It assumes the two are independent. Because every candidate is already propagated to the fixed horizon, this is a re-reading of curves already computed and costs no further engine runs. Deaths from age 105 onward, past the last year the horizon carries, are counted at the odds for 105, which understates the result slightly.

Each candidate also keeps a ladder of balance quantiles per year, from which the distribution of the balance at one age is rebuilt by differencing. That histogram is the one figure on the page reconstructed rather than read off directly: against a reference ladder forty times finer, every bar lands within about a fifth of a percentage point, and the roughest is always the bar nearest zero, where the distribution is steepest. Outcomes that have run out are reported as their own bar rather than as a small balance, and that bar carries the same ruin figure the rest of the page shows.

Starting values: 500,000 saved at age 40, 2,000 added per month while working, 3,000 spent per month after quitting, planning to age 90 at a 1 in 20 tolerance, 100% stocks, and the normal return fit.

Tax: None. Returns are index returns with no fees or taxes, so the model is optimistic by roughly your total cost drag, and the solved quit age is correspondingly early.

Biggest thing it leaves out: The quit age moves only the monthly saving; every dated entry stays put. Spending is fixed in real terms after quitting, the optional life-table weighting treats lifespan and market returns as independent when they are not quite, and the return distribution inherits every caveat of the probability engine it shares.

Simple FIRE

It divides a year of spending by your withdrawal rate to get a target, then compounds your net worth forward until it clears that target. Open the Simple FIRE calculator.

The target is twelve times your monthly spending divided by your withdrawal rate. At the default 4%, that is the familiar 25 times annual spending.

Net worth then rolls forward one year at a time, until it clears the target or reaches age 70: the balance grows by your expected return, and twelve months of the gap between income and spending is added. The FIRE age is the first year the balance exceeds the target, and the tool says which of the two things stopped it when no year in that span does.

Every figure is real, in today’s money. The return entered is a real return, above inflation, so no separate inflation rate appears anywhere in this calculator. Income and spending are held flat in real terms too, so a raise that only keeps pace with prices does nothing here.

Starting values: Age 30, income 5,000 a month, spending 3,000 a month, 100,000 net worth, 6% return, 4% withdrawal rate. Those inputs give a target of 900,000.

Tax: None. Monthly income goes in after tax and spending is what you actually spend, so no rate is applied anywhere in the calculation. The return you enter is likewise the return you keep, with nothing taken off gains or withdrawals.

Biggest thing it leaves out: The target is a fixed line in today’s money: spending is not inflated as the years pass, and neither is the target. That makes it a back-of-envelope figure rather than a plan, which is the trade the tool is making on purpose.

Retirement Probability

It computes the exact probability distribution of your future balance, year by year, and reports the share of outcomes in which the plan survives the horizon. Open the Retirement Probability calculator.

Each simulated year has two steps: the entire distribution of possible balances is multiplied by the year’s uncertain return (a convolution in logarithmic space), then the year’s real spending is subtracted, and any probability mass reaching zero moves into a permanent depletion state. After the final year, that state’s weight is the exact chance of failure. Nothing is sampled, so there is no run-to-run noise: the same inputs always produce the same number, unlike a Monte Carlo simulation of the same model.

The return distribution is estimated from the same Shiller-derived annual real returns the simulation replays, in the reader’s choice of two forms: a normal fit to them (the default) or the historical distribution with its fat left tail intact. A stock/bond mix is blended within each historical year first, so the assets’ real joint behaviour is preserved rather than assumed away. The measurements behind the model’s independence assumption are in the evidence section below.

Starting values: 1,000,000 invested, 40,000 a year of real spending, a 30-year horizon, 100% stocks, and the normal return fit.

Tax: None. Returns are index returns with no fees or taxes, so the model is optimistic by roughly your total cost drag.

Biggest thing it leaves out: Each year is an independent draw from a distribution fitted to one country’s unusually successful history, and real spending stays fixed no matter what markets do. Both are named on the page itself, next to the number they bend.

Sequence of Returns Risk

It separates the effect of the ORDER of returns from their average, using two different methods, because the probability engine alone cannot show an order effect at all. Open the Sequence of Returns Risk calculator.

The order demonstration involves no model. One real window of annual real returns is replayed twice from the same starting balance, once in the order it happened and once reversed, subtracting the same real spending each year and stopping at zero. With spending switched off the two endpoints are identical, because the ending balance is the starting amount times a product of returns and multiplication commutes. With spending on they separate, and that separation is the whole subject of the page.

The danger-zone curve is a conditional probability, computed exactly. For a chosen opening return, the first few years are walked deterministically, and the remaining years of the horizon are then propagated on the full return distribution, giving the chance of running out given that opening. The engine draws each year independently, so it averages over orderings and cannot generate a sequence effect by itself. Conditioning on the opening years is what makes position matter, and it is exact rather than approximate.

The headline figure alongside it conditions on the worst opening run in the record using that run’s actual year-by-year returns, in the order they occurred, rather than a flat run averaging the same amount. The two differ, which on a page about order is the reason to use the real one.

Starting values: 1,000,000 invested, 40,000 a year of real spending, a 30-year horizon, 100% stocks, normal distribution, conditioning on the first 3 years. The replayed window starts in 1969.

Tax: None. Returns are index returns with no fees or taxes, so the model is optimistic by roughly your total cost drag.

Biggest thing it leaves out: The curve conditions on the OPENING years only, so it says nothing about a poor stretch arriving later in a retirement, which would need a model that varies the return distribution year by year. The replayed window is a single stretch of one country’s history, and real spending stays fixed no matter what markets do.

Coast FIRE

It discounts the retirement target back over the years remaining, reporting the balance that grows into that target on its own with nothing further added. Open the Coast FIRE calculator.

The retirement target is annual spending divided by the withdrawal rate, the same definition the simple FIRE calculator uses. The coast number is that target divided by growth over the years to retirement, which is the same compounding run in reverse.

Every figure is real, in today’s money. The return entered is a real return, above inflation, so no separate inflation rate appears anywhere in this calculator. The requirement rises each year at exactly the return, because each year of waiting is one less year of compounding.

Contributions arrive monthly and compound at the geometric monthly rate, so twelve of them add up to exactly the annual figure entered rather than slightly more. The age you reach coast is found by stepping the balance forward year by year and taking the first year it covers that year’s requirement.

The barista setting adds the present value of the spending part-time income does not cover, because the portfolio funds that gap before retirement. It therefore reports a HIGHER number than plain coasting, not a lower one: the pot is being drawn on years before it has finished growing. Once past coast, the earliest age already funded is solved in closed form, accounting for that same drawdown.

The historical section is a different calculation with a different engine. It runs the same plan through the rolling-window backtest behind the FIRE simulation, expressing the coast years as years whose withdrawal is cancelled by an equal income so the portfolio compounds untouched. It uses real market returns and a fixed 80/20 stock and bond mix, so it ignores the expected return entered above. Windows overlap heavily and a sixty-year window cannot start recently, so the count and the last start year are both reported.

Starting values: 35 years old aiming at 65, with 150,000 invested, 40,000 of annual spending in retirement, a 4% withdrawal rate, a 5% real return and 500 a month still going in. The barista toggle starts off, with 20,000 of part-time income ready behind it.

Tax: None. Every account is treated as one untaxed pot, so money locked in a retirement account until a set age counts the same as money you can reach today. That distinction matters more here than in most of these tools, because the whole premise is a long wait.

Biggest thing it leaves out: A single smooth real return applied for decades with no further contributions, which is the assumption carrying almost all the weight and the reason the historical section exists. No state or workplace pension is counted either, so any later income would reduce what the portfolio has to cover.

Quit My Job

It steps a balance forward one month at a time from the day you quit and reports how many complete months it survives, alongside the month a return to work stops being affordable. Open the Quit My Job calculator.

Each month the balance earns one month of your expected return, that month’s spending comes out, and any side income goes in. Spending grows every month at the inflation rate you set. Side income climbs by a flat amount each month until it reaches a ceiling and then stays there, which is a straight line rather than a forecast.

Annual rates are converted to monthly by compounding rather than by dividing by twelve, so twelve months add up to exactly the annual figure entered. The projection stops after fifty years, and a plan still standing then is reported as outlasting the projection rather than as lasting forever.

Three figures are read off that one projection. The break-even month is the first month side income covers spending. The point of no return is the last month the closing balance still holds the reserve you set aside for a job search, fixed at the first month that reserve is breached so a balance that recovers later does not move it. In sabbatical mode the question is reversed and solved in closed form: the savings that would land on zero in the final month of a break of the length you asked for.

Spending is inflated month by month inside the loop and every reported balance is deflated by the same monthly factor, so the chart is in today’s purchasing power and inflation is counted exactly once. The return entered is before inflation.

Starting values: 45,000 of savings against 3,000 a month of spending, 2.5% inflation on that spending, a 3% return on what is left, and three months of spending held back as a reserve. Business mode starts a side income at zero and climbs it 100 a month to a ceiling of 2,000; sabbatical mode starts on a twelve-month break.

Tax: None. Savings are treated as spendable and side income as money you keep, so the figures you enter should already be net of whatever you would owe.

Biggest thing it leaves out: Side income is a straight line to a ceiling, and real freelance or product income arrives in bursts. A quiet stretch landing while the balance is low can end a plan that the same average carried comfortably, and a straight line cannot show that.

Savings Rate

It solves for the number of years at which a stream of annual savings, compounded at your expected return, first equals the retirement target your spending implies. Open the Savings Rate calculator.

The savings rate is derived rather than entered: take-home pay minus spending, over take-home pay. That keeps the rate on screen consistent with the two figures it came from, and the slider writes spending rather than storing a rate of its own.

The arithmetic is done in units of one year’s take-home pay rather than in currency. With nothing invested at the start, income cancels out entirely, which is why two people on very different salaries who save the same share reach independence in the same number of years. Money already invested enters as a multiple of take-home pay and is the only input that breaks that independence.

Contributions arrive at the end of each year and compound annually, so the balance after n years is the standard ordinary-annuity future value. Setting that equal to the target and solving for n gives the answer in closed form, as a fractional number of years, rather than by stepping a loop until it crosses.

The year-by-year projection behind the chart is a separate forward loop rather than a rendering of that formula, so the two are independent calculations of the same quantity. The reference table on the page, and the rates it lists for each horizon, are generated by the same function the calculator runs, so neither can drift from the tool.

A rate of zero with nothing invested, or spending above take-home pay that outruns the growth, reports no timeline rather than a very large number. The projection itself stops at 80 years for display, which never changes the figure reported.

Starting values: 60,000 take-home pay, 36,000 spending, nothing invested, 5% real return, 4% withdrawal rate. That is a 40% savings rate against a target of 900,000, and it carries no starting balance on purpose so the default scenario is exactly one row of the reference table.

Tax: None. The inputs are net rather than gross: take-home pay and spending both go in after tax, so the rate is measured on money that has already been taxed, and no rate is applied anywhere in the calculation. A savings rate quoted on gross income elsewhere will look lower than the same plan does here.

Biggest thing it leaves out: Income and the savings rate are held flat in real terms for the whole period. Careers rarely work that way, and a rising income only raises the rate if spending stays put, which is the assumption that usually breaks first.

Simple Bag O'Money

It runs a drawdown year by year and shows the balance falling, so you can see how long a pot lasts against rising costs. Open the Simple Bag O'Money calculator.

Each year the balance earns your expected return, then a full year of spending is taken out. Inside the projection spending is inflated by compounding your inflation rate from year zero, so the withdrawal grows every year while the return rate stays fixed.

The balance is floored at zero rather than going negative, so the chart shows the pot running out rather than an implied debt. The projection runs for as many years as you ask for.

Spending is inflated year by year inside the projection and every reported figure is then divided by that same factor, so the balance and spending columns are in today’s purchasing power and inflation is counted exactly once. The return you enter is before inflation.

Starting values: 5% return, 2.5% inflation, 1,000,000 starting balance, 2,000 a month of spending, over 40 years.

Tax: None. Returns compound gross and withdrawals are untaxed.

Biggest thing it leaves out: Returns are a flat rate every single year. A real portfolio delivering the same average through a bad first decade would run out sooner, which is the gap the simulation exists to close.