Bengen, 1994
Bengen asked a simple question with the data then available: for every year since 1926, if someone had retired with a portfolio of half US stocks and half intermediate-term government bonds, rebalanced yearly, what first-year withdrawal rate, raised each year by inflation, would have lasted 30 years? The answer was that 4% lasted at least 33 years for every start year, with the worst case, a 1966 retirement, just scraping through, while 4.25% could fail in 28. (His data ran to 1992, so the latest start years were partly projected.) In later work, in 2006, he named the worst-case rate SAFEMAX and put it at 4.15%, for a retirement starting in October 1968.
The Trinity study, 1998
Cooley, Hubbard and Walz at Trinity University reframed the question as a success rate rather than a worst case. They tested many withdrawal rates, several stock and bond mixes and periods of 15 to 30 years, between 1926 and 1995, using corporate rather than government bonds. For a 30-year retirement, 4% inflation-adjusted withdrawals from a 50/50 portfolio succeeded 95% of the time; a 75/25 portfolio did a little better. The study ignored taxes and fees, which matters: a 1% fund fee is a quarter of a 4% withdrawal.
What the rule assumed
Thirty years. US stocks and bonds, which had an unusually good century. Yearly rebalancing. No fees, no taxes. A withdrawal that never changes in real terms, even in a crash. Each assumption is a place the rule can break: a 50-year retirement has more chances to hit a bad run; a portfolio of other countries’ assets has a different history; fees come straight off the top; and a retiree who cannot cut spending in a bad year has no safety valve.
Why early retirees use less
Over 30 years the rule has margin; over 50 it has less, because the portfolio must survive the bad first decade and then another forty years. Studies that extend the method to longer periods tend to land between 3.25% and 3.5% for the same confidence. The safe withdrawal rate calculator shows how long a portfolio lasts at each rate under a steady return, which is kinder than history; treat its answer as a comparison between rates, not a promise about any one of them.
Using the rule in practice
Before retirement, multiply your planned spending by 25 (or 28.6 at 3.5%) and you have a target: the FIRE number. After, set the first withdrawal, write it down, and raise it with inflation. Many retirees add a guardrail, such as skipping the inflation rise after a year in which the portfolio fell, which most research finds extends the money considerably. The 4 percent rule calculator does the arithmetic both ways.
The rule is a starting point
It gives a target that is simple, historically tested and roughly right, which is more than most financial rules of thumb can claim. It is not a guarantee, and it was never meant to be followed through a crash without a glance at the balance. Use it to set the number, a lower rate to add margin, and flexibility to cover what the history books cannot. Revisit the figures every year with real balances: a plan that is checked often needs less margin than one that is set once and forgotten.