How the 4% Rule Works — and Where It Breaks
It is the most famous number in retirement planning: spend 4% of your pot in year one, adjust for inflation, and your money should last 30 years. Here is where it came from, the vivid way it can fail, and how to use it without being ruled by it.
The rule in one paragraph
In your first year of retirement, withdraw 4% of your portfolio. In every year after that, withdraw the same dollar amount, adjusted for inflation — up years, down years, whatever the market does. If history is any guide, your money lasts at least 30 years. That's the whole rule.
Which makes the arithmetic simple enough to do on a napkin. Multiply the income you want by 25 and you get the pot you need:
| Portfolio | Year-one withdrawal (4%) | Or: spending goal | Pot needed (×25) |
|---|---|---|---|
| $500,000 | $20,000 a year | $40,000 a year | $1,000,000 |
| $750,000 | $30,000 a year | $60,000 a year | $1,500,000 |
| $1,000,000 | $40,000 a year | $80,000 a year | $2,000,000 |
| $1,250,000 | $50,000 a year | ||
| $2,000,000 | $80,000 a year |
Note what the rule is not: it is not "withdraw 4% of whatever the portfolio is worth this year." The dollar amount is set once, from the starting balance, and then only inflation moves it. On $1,000,000 with 3% inflation, your withdrawal is $40,000 in year one, about $52,191 in year ten, $70,140 in year twenty and $94,263 in year thirty — the pot has to fund a paycheck that more than doubles in nominal terms.
Where it came from: the worst year to retire in history
In 1994, a financial advisor named William Bengen asked a question nobody had rigorously answered: how much can a retiree safely spend? He took US stock and bond returns back to 1926 and simulated retiring at the start of every year, withdrawing an inflation-adjusted income for 30 years. Every start date survived 4% — and the binding case, the retiree who came closest to running dry, was the one who retired in 1966, directly into a decade of rising inflation and flat markets. That worst-case rate — about 4.15% in Bengen's original data — became famous rounded down: the 4% rule.
Four years later, three professors at Trinity University (Cooley, Hubbard and Walz) ran a broader version of the same test across different stock/bond mixes and withdrawal rates, and the "Trinity study" confirmed the headline: a 4% inflation-adjusted withdrawal over 30 years had succeeded in roughly 95–100% of historical periods, depending on the portfolio. Two important details hide inside that sentence. The failures cluster in the same neighbourhoods — the mid-1960s start dates. And "history" here means one country: the United States, the best-performing major market of the twentieth century.
See it with your own numbers
Everything above, live. Change the numbers to match your own situation:
Prefer the full-screen version? Open the Retirement & 4% Rule Calculator →
Where it breaks, part 1: the order of returns
Here is the rule's deepest weakness, and it isn't visible in any average. While you're saving, the order of returns doesn't matter — a −20% year followed by +25% leaves you exactly where +25% followed by −20% does. The moment you start withdrawing, order becomes destiny, because you are selling assets at whatever price the market is offering that year.
Take two retirees with $1,000,000 each, withdrawing a flat $40,000 a year (we'll hold inflation aside to isolate the effect). They get the same thirty annual returns — three bad years of −9%, −12% and −22% (the 2000–02 bear market's shape) and twenty-seven years of +7% — in different orders:
| Bad years first | Bad years last | |
|---|---|---|
| Balance after year 1 | $870,000 | $1,030,000 |
| Balance after year 3 | $525,968 | $1,096,447* |
| Balance after 30 years | $288,943 | $1,911,924 |
*After three good years. Same thirty returns in both columns; only the order differs.
The unlucky retiree's pot falls to about $526,000 by year three, and then a quiet tragedy unfolds: $40,000 a year is now more than 7% of the remaining balance, so even in good years the withdrawals roughly cancel the growth. The balance bleeds sideways for decades — $522,786 in year four, $519,381 in year five — too damaged to recover, even though the market spends 27 years being perfectly decent. The lucky retiree, with the identical returns, finishes with nearly $1.9 million. This is sequence-of-returns risk, and it is the main reason a single fixed withdrawal rate can never be truly "safe."
Where it breaks, parts 2–4
Longer retirements. The rule was built and tested for 30 years. Retire at 55 and you may need 40 — and every extra year is another draw on the pot and another chance to meet a bad sequence. Research consistently lands on roughly 3.3–3.5% as the more defensible rate for 40-year horizons.
It is an American result. When researcher Wade Pfau replayed the test across 20 developed countries, the US was close to the best case on Earth. In most countries, a 4% inflation-adjusted withdrawal would have failed outright in some period, and in several the worst-case safe rate was under 2%. World-war losses, inflation episodes and weaker markets are the norm of financial history, not the exception. If your retirement depends on one country's continued exceptionalism, know that it does.
Fees, taxes and real life. The original studies assume no investment fees and no taxes, and a spending pattern that never changes for 30 years. A 0.5% annual fee drag, a tax bill on withdrawals, one roof replacement in a down year — each nudges the truly safe rate down. And nobody actually spends like the model: real retirees cut back after crashes, which is precisely the flexibility that rescues plans the rigid rule would sink.
How to use the rule without being used by it
The 4% rule is best treated as a planning compass, not an autopilot. Use ×25 to size the goal while you save — it is simple, roughly right, and far better than guessing. In retirement, take the first-year 4% as a starting income, then behave like a human: trim withdrawals (or skip the inflation raise) after bad market years, spend a little more freely after good ones, and re-run the numbers every few years. Guardrail-style adjustments of that kind are the main finding of the research since Bengen — a flexible 4%-ish beats a rigid 4% by a wide margin.
And if the maths says your pot supports $38,000 a year and your life needs $45,000, the honest fixes are the unglamorous ones: save a little longer, spend a little less, or add an income stream — not assuming the market will be kind because you need it to be.
Origins: Bengen, "Determining Withdrawal Rates Using Historical Data," Journal of Financial Planning (1994); Cooley, Hubbard & Walz, "Retirement Savings: Choosing a Withdrawal Rate" (the Trinity study, 1998); Pfau, "An International Perspective on Safe Withdrawal Rates" (2010). Sequence example: author's calculations, flat (non-inflation-adjusted) withdrawals to isolate the ordering effect.
Frequently asked questions
Where did the 4% rule come from?
From a 1994 paper by financial advisor William Bengen, who tested every historical retirement start date using US stock and bond data and found the worst case — retiring in 1966, straight into stagflation — supported about 4.15% a year for 30 years. The Trinity study (1998) confirmed similar results and popularised the rule.
Does the 4% rule mean I withdraw 4% of whatever my portfolio is worth each year?
No — that is the most common misunderstanding. You withdraw 4% of the starting balance in year one, then adjust that dollar amount for inflation each year, whatever the market does. Withdrawing a fixed percentage of the current balance is a different strategy: it can never fully run out, but your income swings with the market.
Does the 4% rule work outside the United States?
Much less reliably. Researcher Wade Pfau applied the same test to 20 developed countries and found the US was close to the best case — in most countries, a 4% inflation-adjusted withdrawal would have failed at some point, and in several the worst-case safe rate was under 2%. The rule is built on the strongest market history in the world.
What withdrawal rate is safer for a 40-year retirement?
Most research points to roughly 3.3–3.5% for retirements of 40+ years, because a longer horizon has more time to hit a bad sequence of returns. Alternatively, keep 4% but stay flexible: skipping inflation increases after down years dramatically improves survival odds.
Is the 4% rule still valid today?
It is a reasonable starting point for a 30-year US retirement, not a guarantee. It assumes no fees, no taxes, a fixed spending pattern and US-style returns. Treat it as the centre of a range — and re-test your plan every few years rather than once at retirement.