Test Your FIRE Number With a Pre-Retirement Spending Pilot

You have rebuilt your FIRE spreadsheet a dozen times. You have toggled the safe withdrawal rate between 4% and 3.5%, replayed the historical sequences, and watched the output barely move relative to the one input you have never actually observed: the annual spending your entire FIRE number exists to fund.

A FIRE number is one multiplication, a chosen multiplier times an assumed annual spend. The multiplier has been litigated for three decades. The spending figure is usually a guess, drafted over coffee and never tested, about a life you have not lived yet. That asymmetry is the real problem, and it is fixable, because your spending is observable while you still have a paycheck.

Which is why the fix is mechanical rather than philosophical: to test your FIRE number before quitting, you live on your retirement budget for a defined window while the salary still runs, log every failure in both directions, and recompute. This article gives the protocol, then prices the two ways the test can come out wrong, extra working years if the target is padded, sequence-of-returns risk if it is lean.

Your FIRE Number Is a Spending Estimate Times 25

Under the 4% rule, your financial independence number is 25 times the annual spending you assume. The 25 is not a mystery; it is just 1 divided by 0.04. Assume a $40,000 early retirement budget and your target is $1,000,000. Assume $48,000 and the same rule hands you $1,200,000.

Notice what that structure does to error. People defend their multiplier to two decimal places and estimate the input it multiplies with a shrug. A 10% mistake in assumed spending moves the target 10%, which is $100,000 on a million-dollar plan, and every annual dollar you misjudge moves the portfolio requirement by roughly $25. The calibration problem in FIRE is mostly a spending-estimate problem wearing a math costume.

That is also why the debate you keep having, 4% versus 3.5%, is the smaller lever. Both multipliers are defensible for the right horizon. No multiplier rescues a spending number nobody has ever lived on.

The Cost of a Mis-Set FIRE Number, Both Directions

Being wrong is not one risk, it is two, and they bill you in different currencies.

Aim too high and you buy working years

Suppose your true comfortable retired spend is $52,000, but you padded the budget to $60,000 out of fear. Eight thousand dollars of padding sounds like prudence. At 25 times, it inflates your FIRE number by about $200,000, and now your savings rate sets the price. Saving $60,000 a year, the gap takes 3.3 years to close with zero real returns, and still roughly 3.2 years if the portfolio compounds at 3% real while you save (the arithmetic: $5,000 a month against a $200,000 gap at 0.25% monthly real growth is about 38 months). The padded target costs roughly three to four additional working years, and if the padding was never needed, those years bought nothing.

The same savings-rate logic that tells you how many years separate you from financial independence also prices the detour, the point behind the classic shockingly simple math of early retirement.

Assumptions: $60,000 saved per year and a 3% real portfolio return. Your numbers will differ; the structure of the cost will not.

Aim too low and you buy failure odds

The other direction forces the question nobody enjoys: what if your FIRE number is too low? The bill arrives decades later, as sequence-of-returns risk. The 4% rule's evidence base, the Trinity study, tested 30-year retirements, as did Bengen's 1994 analysis that first worked out the 4% initial withdrawal guidance. You are probably not retiring for 30 years. A 45-year-old quitter is planning a 40-to-50-year horizon, and follow-up work on longer retirements, including Wade Pfau's research, commonly lands on initial withdrawal rates closer to 3% to 3.5% at those durations.

Now watch a lean budget quietly break that. Build a $1,000,000 number on an assumed $40,000 spend, then discover your real retired spend is $46,000. Your actual initial withdrawal rate is 4.6%, sustained over 45 years, well above what long-horizon historical work supports. The mechanism that kills such portfolios is timing: as Kitces explains in his breakdown of sequence of returns risk, returns in the first decade dominate the outcome, because withdrawals sold into an early bear market permanently impair a portfolio even if the later decades average out fine.

The cruel part is that this failure is invisible in your spreadsheet today. The projection shows the average. The risk lives in the sequence.

The Pre-FIRE Spending Pilot Protocol

Call it a trial retirement. The design goal is falsification: run your early retirement budget as a live experiment while failure is still cheap, because a breach discovered at 42 is a data point, and the same breach at 47 with no paycheck is a withdrawal-rate problem.

Set the window

Minimum six months; twelve is materially better, because a full year captures the seasonal lumps, insurance premiums, holiday spending, annual travel, the vet bill that arrives like clockwork every October. Run the pilot during ordinary working life, not during a sabbatical or a renovation. You are testing the budget you will live on, embedded in the life you actually have.

Set up the money mechanics

On payday, transfer one-twelfth of your projected annual retirement budget into a dedicated checking account. Everything else sweeps to investments automatically, exactly as it does today. You keep saving normally; the pilot changes only the spendable pool.

One rule does the heavy lifting:

No top-ups. If the account runs dry on the 19th, you log a breach and finish the month on what remains, or draw on a pre-declared emergency line, but you record it either way. A bailout is a data point, not a fire alarm.

Log both failure directions

Your spreadsheet projects. It cannot tell you which months hurt. Keep two ledgers:

• The breach ledger. Date, amount, category, and a one-line diagnosis: one-off (the water heater) or structural (groceries, and it will recur monthly). Flag whether the driver would even exist in retirement.
• The under-spend ledger. Months you finished under budget with zero felt deprivation, and which categories you simply did not miss.

Months of painless under-spending are the padding signal. Breaches are the lean signal. Both are answers, and most people who model only the average never learn anything about either tail.

Converting Pilot Results Into a Recalibrated FIRE Number

Raw pilot months are not yet a target. Three conversions get you there.

Annualize honestly

Do not average blindly. Take the median month as your base, then adjudicate breaches one by one: a structural breach that would exist in retirement raises the budget; a one-off routes to the capital reserve instead. The output is an adjusted annual spend, say $44,000, that you have actually lived inside. The question "how much can I spend in early retirement" turns out to mean "how much do I spend when nothing stops me," and now you have measured data instead of a hunch.

Adjust for retired-life structure

The pilot ran while you were employed, so subtract what work costs and add what freedom costs. Typical subtractions: commuting, professional wardrobe and dues, convenience spending bought because you were time-starved. Typical additions: health insurance before Medicare at 65, now a real premium quote rather than a guess, plus weekday leisure you currently postpone. For a sanity check, compare your adjusted budget against what households 65 and older actually report in the Consumer Expenditure Survey tables, or against how current retirees describe their habits in EBRI's 2024 spending survey. Use both as checks, not targets. Their lives are not yours, but if your budget is half the typical retired household's, you should be able to say why in one sentence.

Pick the multiplier for your horizon

Your horizon is longer than you think. Check your age in the SSA actuarial life table: at 45 the period table's median already implies roughly three and a half more decades of life, and half of everyone outlives the median. Planning a 45-to-50-year retirement is prudence, not paranoia, and a safe withdrawal rate for a 50-year retirement sits near the bottom of the research band, 3% to 3.5%, rather than the 30-year-tested 4%.

Your recalibrated FIRE number is adjusted spend times the horizon-appropriate multiplier. It may come out higher than your old one even though measured spending came in lower, which is exactly the kind of honest surprise this exercise exists to surface.

Where the Pilot Misleads You

An instrument that cannot be wrong cannot tell you anything. Here is where this one bends.

• Earning-state bias. Spending measured while employed differs systematically from retired spending: commute costs, stress purchases, and social anchoring to colleagues who spend like you. Even after the adjustment step, remember that the pilot scores budget livability more than exact retired outlays, and the adjustment itself is an estimate.
• Lumpy multi-year costs. Twelve smooth months can hide the cycles that recur across a 40-to-60-year retirement: vehicles, roofs, major dental, health events. Smooth is not the same as safe. Convert each into a sinking-fund line inside the budget; for illustration, a $30,000 car every 12 years is $2,500 a year, and a $20,000 roof every 25 years is $800 a year. Fund the reserve deliberately instead of treating each arrival as an emergency.
• Novelty frugality. Month one runs on enthusiasm; month eight runs on habits. Weight the back half of the window more heavily when you annualize.
• Partner drift. If one partner runs the pilot and the other keeps spending as usual, the data is corrupted at the source. Both of you live inside the budget, or the pilot measures one person's discipline and nothing about the household.
• The un-simulatable tail. No twelve-month window prices a health shock or long-term care. That residual is what the conservative multiplier and the reserve are for. A pilot shrinks your unknowns; it does not abolish them.

Decision Rules for Quitting, Delaying, or Resetting

The pilot should hand you one of three verdicts. Set the thresholds before you look at results, or you will negotiate with yourself afterward.

Two cautions on the delay verdict. First, price the cost of working one more year before FIRE honestly. After you reach a validated number, an extra year adds savings, growth, and cushion, but a body of retiree wealth research keeps finding that many retirees die holding most of their savings, so aiming too high carries a real and documented cost too. Second, ask what the extra year is for. When ESI Money profiled a couple who sold their house and businesses to move aboard a sailboat, the most common regret they heard from older cruisers was not the risk they had taken. It was the decade they waited.

Enough turns out to be measurable: a budget you have lived in for twelve months, times a multiplier you can defend for your horizon, plus a reserve for the things that arrive on their own schedule. You can measure all three while you still have a salary, and that is the entire point of running the test now instead of answering the question in hindsight.