The 4% number¶

Archived — September 5, 2026

Snapshot of Trending on X that day. Current · All snapshots

Updated September 5, 2026. Trending in Wealth on X.

X spent the week arguing over a single multiplier. One camp posts charts: annual spend times 25, withdraw 4%, you are done. The other camp — including men who actually left work in their 40s — says that number is a 30-year classroom result, a million dollars is $40k a year, and chasing 50× is how you donate a decade you do not get back. The default here is simple: 25× for a ~30-year plan, 33× if you want out in your 40s or early 50s, a cash sleeve so you are not a forced seller, and a savings rate that closes the gap. Do not treat 4% as a law. Do not treat $1M as retirement.

Not financial advice

This is planning arithmetic from public research and this week's argument on X, not a recommendation to buy, sell, withdraw, or quit work. If you need a plan, use someone licensed.

What X is actually arguing¶

The loud post this week was a spend-to-portfolio chart. $35,000 a year at 4% needs $875,000 invested. $100,000 a year needs $2.5 million. The mechanic is: take 4% of the pile, let returns refill it, live on that. Thousands of bookmarks later, the replies split.[1]

Camp A treats 4% as the number. White Coat Investor’s thread went at the opposite panic: someone online will tell you that you need 50× spending. 50× is a 2% withdrawal rate. The Trinity-style tests — Depression, WWII, stagflation, 2008 — had a 4% inflation-adjusted withdrawal surviving a 30-year retirement 96% of the time. That is 25× spending, not 50×. When it survived, the leftover pile was often larger than day one, not a near-miss. The cost of chasing 50× is years of extra work the data does not require.[2]

Camp B is the early-retiree bench. Dad is FIRE, retired at 42: if he had run a rigid 4% plan, he is back at work around 72. His wife may live to 100. A 30-year model is not a 50-year life. Another long-time early retiree said the same thing this week: he has yet to meet a single person who left work early and followed 4% exactly. They cut travel in a crash and take the trip after a few good years. 4% is a starting point. It is not an autopilot.[3]

Camp C is the “millionaire with a middle-class paycheck” post. A $1M portfolio at 4% is about $40,000 a year. That is less than a lot of full-time jobs, before tax, before health insurance you no longer get from an employer, before a kid still in the house. The number did not get smaller. The lifestyle the word “millionaire” used to buy did.[4]

Underneath the charts is the original paper. In 1994 Bill Bengen asked how much you could take without the money dying first. Historical answer: about 4.15%. The internet dropped the .15. Two details almost nobody in the replies keeps: it was a 50/50-style 30-year test, and it is 4% of the starting portfolio in year one, then that dollar amount adjusted for inflation. It does not promise to preserve principal. If you start at $1M, spend $40k, and the market goes nowhere, you now have $960k. That is the design.[5]

A fourth camp wants out of the model entirely — higher withdrawal on a concentrated bet, or “live off dividends and never touch principal.” Sequence-of-returns risk does not care. A 70% drawdown in year two of retirement is not a chart with a purple box. It is groceries. Diversified 25× is boring on purpose.

Why this lands after 40¶

After 40 the 4% number stops being trivia and becomes a remaining-years problem.

If you retire at 65, you are in the study’s neighborhood: ~30 years, Social Security as a floor later, Medicare in range. 4% / 25× is a planning default, not a guarantee.

If you want out at 45–52, you are not in that neighborhood. You may need the pile for 40–50 years. Social Security is 15–22 years away. Employer health insurance ends the month you quit. Sequence risk is worse because a bad first decade cannot be worked off as easily. Stretch the horizon and the ceiling drops; one post this week put a 40-year, 3% real-return case near 4.2% with almost no margin. Practical translation: 3% / 33× if you are leaving in your 40s, unless you will flex spending hard.[6]

You also still have the accumulation problem. The 4% rule only matters after there is something to withdraw from. After 40 the lever is still savings rate versus lifestyle, with fewer compounding years than a 28-year-old. Peak earnings, if you have them, are the window. Kids, a mortgage, and aging parents are the leak. A 4% debate while you save 8% of take-home is entertainment.

Taxes and healthcare are not in the original 4%. A $72k withdrawal from a taxable brokerage is not $72k of groceries. An ACA premium in your 50s is not a line in the 1994 paper. After 40, S has to be all-in spend, not “rent plus food.”

The body is a parallel clock. Working five extra years to hit 50× is a health bet as much as a money bet. Under-saving and quitting on 8% of a thin pile is the other bet. Both are common in this age band.

The default, in enough detail to actually do it¶

Do this on paper. Not in your head.

1. Get S, actual spend.
Pull 12 months of checking, credit, and cash. Housing, food, insurance, cars, kids, travel, gifts, the “one-offs” that happen every year. Annualize. That number is S. If you do not have 12 months, use 90 days × 4 and mark it as a draft. Recalc in January.

2. Split S into two columns.
Column A: years 1–10 after you stop a W-2 (no Medicare yet, maybe no Social Security). Column B: years 11+. Subtract only income that will actually show up in that column — a pension in pay, net rent you already collect, Social Security you will claim later. Do not put age-67 Social Security into the age-48 budget. S_net for the first decade is the number that kills most “I hit 25×” stories.

3. Pick the multiplier by horizon, not by Twitter mood.

Horizon	Multiplier	Withdrawal	Use
~30 years (stop ~65)	25 × S_net	4%	Default planning number
40–50 years (stop in 40s / early 50s)	33 × S_net	3%	Default if you want out early
You will not cut spending in a crash / want a fat leftover	40 × S_net	2.5%	Ceiling before 50× theater

50× is 2%. The research camp this week called that a way to die richer than you need to be, after extra years on the job. Skip it unless the point is a dynasty, not your life.[7]

Worked example, planning only: S_net = $72,000.

25× = $1,800,000 → year-1 withdrawal $72,000
33× = $2,376,000 → year-1 withdrawal $71,280
Cash sleeve (below) = $72,000 to $144,000, held outside the risk portfolio

If you have $900,000 today, quitting is an 8% withdrawal. That is not a 4% plan. The gap is $900k–$1.48M depending on horizon. Close it with S and savings rate, not a hotter story.

4. How 4% actually runs, if you ever take it.
Year 1: 4% (or 3%) of the starting portfolio.
Every later year: last year’s dollar amount × (1 + inflation).
You do not recast 4% of a crashed balance. Recasting 4% of a smaller pile is a different rule. It cuts groceries in the year you can least stand it. Bengen’s version holds real spending and lets principal flex. Principal is allowed to fall. That is the part people skip.[8]

5. Cash sleeve.
Hold 12–24 months of S in cash or cash-equivalents, separate from the invested pile. Year the market is −30%, you spend the sleeve. You do not sell the 25× at a bad print. Refill the sleeve after recovery. This is the sequence-of-returns patch. A 4% classroom portfolio survived crashes; your grocery store still wants dollars that year.[9]

6. Guardrails, not a personality.
Portfolio −20% from the retirement-day high: cut 10% of spend. Cut travel, restaurants, the boat, the new truck. Not insulin, not the mortgage.
Portfolio +20%: you may raise spend.
Almost no early retiree this week claimed they ran a rigid 4%. Flexibility is the real rule. The 4% number is the budget you start from.[10]

7. Payday script until you hit the number.
The 4% rule does nothing during accumulation. After 40, if you are behind:

Paycheck hits.
Same day: housing, minimums on any high-interest debt.
Same day: automatic transfer of the gap. If you want to stop in 10–15 years, 25–40% of take-home is the range that actually moves 25×. 8% is a raise the market has to save for you. It will not.
Remainder: living.

Recalculate S every January. If S rose $6,000, 25× rose $150,000. Lifestyle creep is how the number runs away while you “invest.”

8. Tax and health add-ons, still not a plan.
Model withdrawals as taxable, not headline 4%. A $72k IRA withdrawal is not $72k spend. Price health insurance as a line in S for every year before Medicare. If that line is $12k, 25× on it is another $300k. Ignore it and your 4% is a 5% in disguise.

9. What “done” looks like on a Saturday.
You can point at: S, S_net, the multiplier, the target, current invested assets, current implied withdrawal if you quit tomorrow, cash sleeve in months, and the automatic transfer. If any of those is a vibe, you do not have a 4% number. You have a tweet.

What people get wrong¶

Treating 4% as “never touch principal.” It was a depletion schedule that usually left money. It was not an endowment. 2% is the perpetual-ish number people reach for when they want the pile to grow forever. Different job.
Using a 30-year rule for a 45-year retirement. Leaving at 42 and spending like you are 65 is how you meet year 72 without a plan. Drop to 3% / 33× or commit to cutting spend in bad years.
Chasing 50×. 2% withdrawal. No historical 2% failure in the usual tests. The failure mode is you. Extra years of work for a leftover you will not spend.
Calling $1M “retired.” At 4%, that is ~$40k before tax. Run S first. If S is $90k, you are not close.
One concentrated bet as a substitute for 25×. A 70–80% drawdown in year two is sequence risk with no sleeve. Retirement is a time-horizon problem, not a prediction problem.
Recasting 4% of the new balance every January. That is a variable-percentage rule. It feels safer on a spreadsheet and wrecks a down year. Bengen inflates the original dollar amount.
Skipping the cash sleeve. 25× in a fund is not Friday’s cash. Groceries do not wait for the recovery.
Letting S drift. A 4% target on last year’s lifestyle plus a new truck is a 4.5% target you did not admit to.

This week¶

Write S from the last 12 months. One number. No rounding toward comfort.
Write S_net for years 1–10 after a W-2 stop. Strip out income that is not actually there yet.
Compute three targets: 25×, 33×, and 40×. Circle one based on when you actually want to stop, not when it would be nice.
Divide current invested assets by S_net. That is today’s implied withdrawal if you quit. If it is 6–8%, you are not at the 4% number.
Count cash that is not the market. If it is under 12 months of S, that is the first hole, before allocation arguments.
Set or raise the payday auto-transfer by a fixed dollar amount this coming payday. Recalc the gap to your circled target. Do not open a new account for this. Move the money.