The FIRE Number: The 25x Rule Neither Founding Paper Contains (2026)

Ask a FIRE forum how much you need before you can quit, and the answer is usually "25 times your annual expenses." The multiple gets cited as settled research, with a pedigree running back to William Bengen's 1994 withdrawal-rate paper and the 1998 Trinity study. It has no such pedigree. We read both papers in full for our guide to the 4% rule and reread them for this one, hunting for the multiple specifically. It is not there. Not in Bengen, not in Trinity, not as "25x", not as any savings target stated in years of spending.

The 25x rule is arithmetic: 1 ÷ 0.04 = 25. Invert a withdrawal-rate convention and you get a savings multiple. The inversion is useful, and this guide is not here to talk you out of it. But a derived number inherits every assumption of its parent study, which carries more fine print than the multiple ever discloses. This guide traces what the founding papers establish, what the shortcut silently imports, and who carried the multiple into folk canon. We will also look at how the two retirement systems closest to home, CPF and EPF, answer the same question with different arithmetic. This guide continues our pedigree series, which walks famous numbers back to their primary documents.

Before anything else: this guide is informational. It never tells you what your savings target or withdrawal rate should be. Personalised recommendations on investing or drawing down retirement savings are regulated financial advice in both Singapore and Malaysia; for your own numbers, talk to a licensed adviser.

The multiple neither paper contains

Bengen's arithmetic runs in one direction only: portfolio first, withdrawal second. He described the mechanics himself: "The withdrawal dollar amount for the first year (calculated as the withdrawal percentage times the starting value of the portfolio), will be adjusted up or down for inflation every succeeding year. After the first year, the withdrawal rate is no longer used for computing the amount withdrawn; that will be computed instead from last year's withdrawal, plus an inflation factor." At no point does the paper reverse that arrow and tell a saver how large the portfolio should be as a function of spending.

Every "25" in the 1994 paper is something else. Clients retiring at 60 or younger, he wrote, "should live at least 25–30 years" — a life expectancy. A "25-percent stocks" row sits in his allocation grid. A 1929-retiree example measures one portfolio's wealth as "25 percent greater" than another's. Years, allocations, a wealth comparison; never a savings multiple.

Trinity is the same. Cooley, Hubbard and Walz tested withdrawal rates from 3% to 12%, and every one of their success-rate tables is headed "Withdrawal Rate as a % of Initial Portfolio Value" — the rate is defined on the portfolio, full stop. Their 25s are payout periods, a 25% stocks/75% bonds allocation and a 25-year inflation illustration. Their success test is blunt: "If an investor's portfolio outlives the investor's planned payout period, then it is counted a success." Nowhere does the paper convert any of this into a savings target.

The 25x shortcut simply runs Bengen's year-one formula backwards. If the first-year withdrawal is 4% of the starting portfolio, then a portfolio worth 25 years of spending covers year one. Only year one — after that, per the mechanics above, the rate is abandoned and withdrawals become last year's dollars plus inflation. The multiple applies exactly at the starting line, but says nothing about withdrawals in subsequent years.

Expenses, not income

The base of the multiple is annual spending. Not salary, not take-home pay. The founding studies define their rates against the portfolio, and the portfolio's job is to replace what you spend, so an inverted version has to target spending. "25 times your income" appears in neither paper and produces a very different number: a saver banking half their take-home lives on the other half, so 25x income works out to 50x expenses — double the target for the same life. The higher your savings rate, the further the income version overshoots.

The FIRE community's focus on savings rates is coherent for this reason. Cutting spending does double duty, raising the amount saved each year while shrinking the target those savings must reach.

What 25x inherits from its parents

Multiply your expenses by 25 and you have implicitly signed the founding studies' terms, which include these main clauses.

• A 30-year window. Bengen's headline finding: "In no past case has it caused a portfolio to be exhausted before 33 years, and in most cases it will lead to portfolio lives of 50 years or longer." Trinity tested payout periods of 15 to 30 years. Neither tested the 50-year horizons early retirees face; for very long horizons Bengen's 1994 band was 3% to roughly 3.5%, which converts to about 33.3x down to 28.6x — not 25x.
• US market history. Bengen used the Ibbotson Associates 1992 Yearbook, retirements starting each year from 1926 to 1976, on 50% stocks and 50% intermediate-term Treasuries. Trinity used S&P 500 and long-term high-grade corporate bond returns, 1926–1995. One country, one unusually good century.
• Specific allocations, none of them 60/40. Bengen's baseline was 50/50, with counsel to hold stocks "as close to 75 percent as possible, and in no cases less than 50 percent". Trinity ran five mixes from 100% stocks to 100% bonds. The 60/40 portfolio that often gets glossed onto the rule appears in neither paper.
• Inflation-adjusted dollar withdrawals. The rate sets year one; every later year is last year's dollars plus inflation, whatever the market did.
• No taxes, no fees. Bengen assumed all assets tax-deferred; Trinity states plainly: "The study did not adjust for taxes or transaction costs."

The allocation clause is not advisory. Trinity's famous 4%, 30-year row, with withdrawals adjusted for inflation:

A bond-heavy saver who reaches 25x has hit a target whose own evidence base gives them 71% or 20% odds, not 95%-plus. The multiple is a dollar figure with the assumptions rubbed off.

Slide the rate and the multiple moves

Because 25x is 1 ÷ 4%, it is hostage to the rate, and the rate has been under revision since the day it was published. Here is the ladder at a concrete spending level:

The endpoints are live positions, not hypotheticals. Bengen revised his own figure to 4.5% in a 2020 Financial Advisor article — his "SAFEMAX", anchored on a worst-case retiree of 1 October 1968 — which converts to about 22.2x, and then to 4.7% in a December 2025 CNBC Make It interview, on a broader multi-asset portfolio: roughly 21.3x. Morningstar's The State of Retirement Income: 2025 lands on the other side, recommending a 3.9% starting rate for 2026 retirees from forward-looking Monte Carlo simulation at a 90% success target over 30 years, before fees and taxes — about 25.6x. Its base case has ranged from 3.3% to 4.0% across five annual reports, and adding 1% in annual fees drops the 3.9% to 3.4%.

A "correct" multiple that runs from about 21x to beyond 33x depending on method, portfolio, horizon and fees is clearly not a constant of nature. It is a unit conversion of an argument that is still in progress. The full history of that argument, including why Bengen and Morningstar can both be right, is in the sibling guide.

The convention's paper trail

The papers never said 25x, and nobody owns the number; there is no standard-setter for FIRE arithmetic. The most influential single carrier is Mr. Money Mustache's post of 13 January 2012, "The Shockingly Simple Math Behind Early Retirement", and the folk version has drifted a long way from what that post says.

The post's central claim: "It turns out that when it boils right down to it, your time to reach retirement depends on only one factor: Your savings rate, as a percentage of your take-home pay". His stated assumptions: "You can earn 5% investment returns after inflation during your saving years", retirement spending funded at the 4% withdrawal rate with flexibility in recessions, and a stash that is never drawn down at all — only the gains get touched. At a 50% savings rate, his own figure is "about 16" years to financial independence. The folk version of the rule drops three details from his post. The return assumption is 5% real, not 7%. The target is stricter than anything the research tested, since a never-depleted stash is a higher bar than 30-year survival. And the post never once uses the acronym "FIRE" or the figure 25x. His term is "financially independent".

The "17 years" figure that circulates online, sometimes stapled to a 7% assumption, is a mash-up of his answer with an assumption he did not use. Run the algebra at 7% real — save one year of expenses annually, compound, stop at 25 years of expenses — and you land near 15 years, not 17. Every years-to-FIRE figure is three assumptions in a trench coat; our FIRE number calculator will run the projection at whatever savings and return figures you actually hold.

The rest of the FIRE taxonomy is even less official. Lean FIRE and fat FIRE have no authoritative budget cutoffs; the dollar lines you see quoted are folk conventions from community forums, and they drift. Coast FIRE — already having enough invested that compounding alone will fund a conventional retirement — traces to no identifiable coiner; a March 2020 post by The Fioneers is a representative statement, not an origin. Barista FIRE, part-time work layered over a partial stash for income and health cover, sits in the same position, with Financial Samurai as its representative source. All four are conventions and we label them that way. Our Coast FIRE guide, published alongside this one, works through the coast variant's maths.

Singapore and Malaysia: institutions that answer differently

25x assumes a saver who accumulates a private portfolio and later spends it down alone. Neither national system closest to home is built that way, and each publishes its own answer to the question the multiple claims to settle.

Singapore's anchor numbers are the CPF retirement sums. A member turning 55 in 2026 faces a Full Retirement Sum of S$220,400 — the published schedule runs S$213,000 in 2025, S$220,400 in 2026, S$228,200 in 2027, with the Basic sum at half, S$110,200. The FRS was estimated in 2022 to deliver about S$1,780 a month from age 65 for a male member on the CPF LIFE Standard plan. Since 2025 the Enhanced Retirement Sum has been set at double the year's FRS — S$440,800 for 2026 — with a payout CPF puts at up to S$3,400 a month for life. That phrase carries the weight. CPF LIFE is, in CPF Board's own words, "a national longevity insurance annuity scheme that provides you with monthly payouts no matter how long you live." An annuity has no portfolio to deplete and no opening bear market to survive; longevity risk is pooled. A 25x-style calculation therefore applies only to savings held outside that floor. In Singapore the tax-advantaged home for such savings is SRS, with contributions capped at S$15,300 a year for citizens and PRs (S$35,700 for foreigners), contributions deductible against taxable income, and half of withdrawals taxable at or after statutory retirement age. The two national schemes are compared head to head in our EPF vs CPF guide.

Malaysia's EPF answered with a framework rather than a multiple. Its Retirement Income Adequacy tiers, announced in December 2024 and in force since January 2026, set three savings benchmarks at age 60: Basic RM390,000, Adequate RM650,000, Enhanced RM1.3 million. The middle tier's construction is printed in the announcement itself: the Belanjawanku budget for a single elderly person in the Klang Valley, RM2,690 a month, "multiplied by 240 months (reflecting the average life expectancy of Malaysians) and rounded to the next RM10,000, equivalent to RM650,000". Basic is 60% of Adequate; Enhanced is double it. The spending model is explicit too: "The framework encourages a monthly drawdown for 20 years, aligned with average life expectancy in Malaysia." EPF's illustration has the RM650,000 paying RM2,708 a month in year one and RM7,389 by year 20 — the account spends to zero.

EPF's formula is expenses × 240 months, a 20-year full drawdown. The 25x formula is expenses × 300 months at the year-one rate, on a portfolio built to survive the whole window and, in most historical cases, outlast it. A national provident fund looked at "how much is enough" and chose a structurally different model. This makes "multiply by 25" one convention among several, not a law of retirement.

Our calculators, and what we fixed today

Our FIRE number calculator does what this guide describes and nothing more. It divides your annual expenses by the withdrawal rate you choose, shows the resulting targets side by side at 3%, 3.5%, 4% and 4.5%, and projects years-to-target from your current portfolio and annual savings using a simple year-end contribution model; if the pace would take a century or more it reports "Never at this pace" rather than printing a fantasy date. The return input is a real return, labelled after inflation, and you should feed it one — mixing a nominal return with today's expenses is the fastest route to a fictitious FIRE year. The tool models no taxes, no fees, no CPF or EPF flows and no market variance. It is an estimator for exploring the arithmetic, not an advice engine.

Auditing it against the primary sources for this guide turned up prose that needed correcting, and as of today (22 July 2026) it is corrected: the tool no longer credits the 4% figure to the Trinity study (Bengen 1994 is the origin; Trinity corroborated it with success rates), no longer describes a 60/40 evidence base that neither paper used, and now pairs Mr. Money Mustache's savings-rate timeline with his actual 5%-real assumption. Our safe withdrawal rate calculator went through the same treatment when the sibling guide shipped. The Coast FIRE calculator discounts your FIRE number back to a coast target at your assumed real return, and our retirement calculator deliberately uses a flat 20-year drawdown — the same convention EPF's framework adopted.

FAQ

Is the 25x rule in the Bengen or Trinity study?

No. We checked both papers in full. Bengen 1994 defines withdrawals as a percentage of the starting portfolio; Trinity 1998's tables are headed "Withdrawal Rate as a % of Initial Portfolio Value". Neither states a savings target as a multiple of expenses. 25x is the community's inversion of the 4% convention.

Do I multiply my income or my expenses?

Expenses. The portfolio's job is to fund spending, so the multiple targets spending. An income-based version has no basis in either paper and overshoots badly for savers: at a 50% savings rate, 25x income equals 50x expenses.

How long does reaching 25x take at a 50% savings rate?

Mr. Money Mustache's 2012 answer was "about 16" years, on his stated assumption of 5% real returns. At 7% real, the same algebra gives about 15 years. The widely repeated "17 years" figure pairs his answer with an assumption he did not use. Our FIRE number calculator projects the timeline from your own savings and return inputs.

Does 25x apply to my CPF or EPF savings?

Not directly. CPF LIFE pays out for life, so there is no portfolio to deplete, and EPF's adequacy tiers are built on a 20-year drawdown — a different model again. The multiple is a tool for savings you manage yourself outside those schemes, and decisions there belong with a licensed adviser.