Where the number came from. In October 1994, financial planner William Bengen published a study in the Journal of Financial Planning asking a simple question: what constant, inflation-adjusted withdrawal rate would have survived even the worst 30-year stretch in U.S. market history? Running every rolling 30-year period from 1926 on a 50/50 mix of large-cap stocks and intermediate Treasuries, he found the answer was about 4.15% — which the industry rounded to 4% and dubbed the "SAFEMAX," the maximum safe rate. The binding worst case was a retiree who started around 1968 and got hammered by a bear market followed by the high inflation of the 1970s. Four years later, three Trinity University professors (Cooley, Hubbard, and Walz) extended the analysis and found a 4% inflation-adjusted withdrawal had a 95%+ success rate over 30 years for stock-heavy portfolios. That's the "Trinity Study," and together the two cemented "4%" in the culture.
What "the rule" actually instructs. The mechanics matter and are widely misunderstood. You take 4% once, in year one. After that you ignore your balance and simply give the prior year's dollar withdrawal a raise equal to inflation. On a $1 million portfolio, that's $40,000 in year one; if inflation is 3%, you withdraw $41,200 the next year regardless of what the market did. The rule is a spending rule, not a percentage you recompute annually.
The number has moved — upward, from its own author. Bengen never intended 4% to be permanent. In his 2025 book, A Richer Retirement: Supercharging the 4% Rule to Spend More and Enjoy More, he rebuilt the study with a more diversified, seven-asset-class portfolio (55% stocks, 40% bonds, 5% cash) and a two-factor model that considers stock valuations (the Shiller CAPE ratio) and expected inflation. The new "Universal SAFEMAX" — the worst-case-proof rate — rose to 4.7%. On $1 million that's $47,000 instead of $40,000, a 17.5% raise. And Bengen goes further: because the truly catastrophic 1968-style setup is rare, the average safe rate across all his historical retirees was about 7.1%, and he argues most retirees today can reasonably start around 5.25%–5.5% even with elevated valuations.
...and downward, from the forecasters. Not everyone agrees, and the disagreement is instructive. Morningstar's late-2025 research put the 2026 safe starting rate near 3.9% for a moderate stock allocation with a 90% success target — lower than the classic 4%. Why the gap? Bengen's method is backward-looking (what survived history); Morningstar's is forward-looking (Monte Carlo simulations off today's capital-market assumptions, which bake in high U.S. equity valuations and lower expected bond returns than the 20th century delivered). Vanguard lands in between at roughly 3.5%–4.5%. Neither approach is "right." The historical camp says the future will rhyme with a very tough past; the forecasting camp says starting conditions matter and today's are expensive. Knowing which assumption you're leaning on is more useful than memorizing a single percentage.
The catch the rule hides: sequence risk. Two retirees can earn the exact same average return over 30 years and get wildly different outcomes if the order differs. Bad returns early, while you're also withdrawing, permanently shrink the base that later good years compound on. This is why a fixed 4% "rule" can still fail a 1968 retiree and why flexibility — trimming spending in down years — buys so much safety. Chapter 8 is devoted to it.
How to actually execute. Treat your withdrawal rate as a starting dial, not a vow. Pick a rate consistent with your horizon (longer retirements need lower rates — Bengen's own work shows a 60-year horizon drops the safe rate to about 4.1%), your allocation, and your tolerance for adjusting. Build in a plan to cut modestly if the first few years go badly, and to spend more if they go well. And remember Bengen's own verdict: inflation, not market crashes, is "the greatest enemy of retirees" — which is why the next chapter is about getting your inflation assumption right.
Same portfolio, three "safe" rates
Priya retires with $1,250,000. Her year-one withdrawal depends entirely on which camp's starting rate she trusts:
| Starting rate | Source | Year-1 income |
|---|---|---|
| 3.9% | Morningstar (forward-looking) | $48,750 |
| 4.7% | Bengen 2025 SAFEMAX | $58,750 |
| 5.4% | Bengen "reasonable start" | $67,500 |
Same money, an $18,750/year spread — before the market moves at all. Say she picks 4.7% → $58,750. The rule now fixes the dollar amount, not the percentage: if inflation is 3%, next year she withdraws $60,513 regardless of her balance.
The execution habit that matters most: if the first few years bring a bad market, freezing that raise — taking $58,750 again instead of $60,513 — protects the portfolio more than any fine-tuning of the starting rate. Treat the rate as a starting dial you can turn down in rough years, not a vow.