Does the 4% rule still work?

The full picture

First, the attribution, because almost everyone gets it wrong. The 4% figure is William Bengen’s, published in “Determining Withdrawal Rates Using Historical Data” in the Journal of Financial Planning in October 1994. He was not calculating a probability. He was looking for the worst starting year in the historical record and asking what withdrawal rate survived it — a figure he called SAFEMAX. On that basis 4% never exhausted a portfolio in under 33 years.

The Trinity Study (Cooley, Hubbard and Walz, AAII Journal, February 1998) came four years later and did something different: it reported success as a probability across overlapping historical periods. That framing is Trinity’s contribution. The number was already Bengen’s.

Trinity never tested a 60/40 portfolio. It tested five mixes, and the result depends enormously on which one you pick. For a 4% inflation-adjusted withdrawal over 30 years:

Bengen revised upward, not downward. Widening the asset mix beyond large-cap stocks and intermediate Treasuries — adding small-cap, mid-cap, micro-cap and international — raised his worst-case figure from 4% to 4.5% and then to 4.7%, which is where he put it in his 2025 book. He stresses that this remains a worst-case number; he places the historical average sustainable rate closer to 7%. If you have been told the 4% rule is dead because its author disowned it, the opposite happened.

Morningstar’s number moves every year, so always ask which edition. Its annual State of Retirement Income analysis has given 3.3% (2021), 3.8% (2022), 4.0% (2023), 3.7% (2024) and 3.9% in the edition published December 2025. Those assume a 90% success probability over 30 years with roughly 40% in equities. A figure quoted without its edition date is close to meaningless here.

The flexibility claim is usually garbled. The research behind it is Guyton and Klinger, Journal of Financial Planning, March 2006. Their capital-preservation rule cuts the withdrawal by 10% — not 10 to 15% — and it triggers when the current withdrawal rate has risen more than 20% above the initial rate, which is not the same as “a down market.” And their 99% is a confidence standard used to solve for a higher starting rate, not a success rate you achieve by bolting flexibility onto a 4% withdrawal. Willingness to cut spending buys you a higher initial rate; it does not buy near-certainty at the same one.

What the rule still cannot tell you. Sequence-of-returns risk means two retirements with identical average returns end very differently if one takes its losses early. Healthcare costs do not track CPI. And none of this research models a spending path that falls in later retirement, which is what most people’s actually does. Treat any single percentage as a starting point for a plan you revisit, not a setting you choose once.

What the withdrawal rate does not decide

A withdrawal rate governs how much you take. It does not govern how much you are required to take, and it does not govern what the withdrawal is worth after tax. Required minimum distributions generally begin at age 73, and they apply to the account balance regardless of what any planning rate says (IRS). A retiree following a 3.5 percent plan can still be obliged to withdraw more than that from a traditional account once the requirement starts.

Tax is the other gap. Benefits are not automatically tax free; depending on other income, up to 85 percent of a Social Security benefit can be taxable (SSA). Two retirees drawing the same gross amount from differently taxed accounts do not have the same spendable income, which is why a rate expressed as a percentage of a portfolio is only ever half of a plan.

Common questions

Does the 4 percent rule still work?

There is no single rate, and the rule is routinely misattributed. The 4 percent figure is William Bengen's, from a 1994 paper in the Journal of Financial Planning; the 1998 Trinity Study contributed the success-rate framing rather than the number itself, and Bengen has since revised his own worst-case figure upward. Morningstar's recent annual analysis puts a safe starting rate below 4 percent for a 30-year retirement at a high success threshold. What actually moves the number is your horizon, your allocation, and whether you will cut spending when markets fall.

What should I use instead of a fixed 4 percent?

A range, tested against your own horizon, rather than a constant. The published figures vary because the assumptions behind them vary: the length of retirement, the equity allocation, the success threshold and whether spending is allowed to flex. Treat a withdrawal rate as one input to be stress-tested rather than a rule to be followed, and re-check it against conditions rather than setting it once at retirement.

Related questions

When should I claim Social Security? →

How much money do I need to retire? →