Explainer
The 4% rule is a research finding, not a guarantee. It says that a retiree who withdrew 4% of a portfolio in the first year, then raised that dollar amount with inflation every year, would not have run out of money in any 30-year period in the U.S. market history that was studied. Past performance is no guarantee of future results.
The rule comes from a 1994 paper by William Bengen, a financial planner, in the Journal of Financial Planning. He tested 30-year retirements beginning in each year from 1926 onward, using a portfolio of roughly half to three quarters U.S. stocks with the rest in intermediate-term government bonds. A first-year withdrawal of about 4%, raised with inflation each year after, lasted at least 30 years in every period he tested. A 1998 paper by three professors at Trinity University, often called the Trinity study, reached similar conclusions. Bengen later raised his own figure as he added more asset classes to the research.
The rule describes the worst starting year in one country's market history. It is a description of the past. It was never a promise about the future, and its author did not present it as one.
Hypothetical illustration
A hypothetical retiree starts with $1,000,000. Under the rule, the first year's withdrawal is 4% of that, or $40,000. From then on the dollar amount rises with inflation, whatever the portfolio does. With a hypothetical inflation rate of 3% a year, the withdrawal is $41,200 in year 2, about $52,200 in year 10, and about $94,300 in year 30. The inflation rate is invented to show the arithmetic. It is not a forecast.
This is the most commonly misread part of the rule. After the first year, the withdrawal follows prices, not the portfolio. If the portfolio falls, the same dollar withdrawal becomes a larger share of what remains. If the portfolio rises, it becomes a smaller share. The rule never looks at the balance again.
Hypothetical illustration
Two hypothetical retirees each start with $1,000,000 and receive the same fifteen invented yearly returns used in the sequence of returns risk explainer, worst years first: −15%, −10%, −5%, 4%, 6%, 8%, 8%, 10%, 10%, 12%, 12%, 14%, 14%, 16%, 18%. The first follows the rigid rule: $40,000 at the start of year 1, raised 3% a year for hypothetical inflation. The second follows a fully flexible rule: 4% of whatever the balance is at the start of each year. These returns are invented to show the arithmetic. They are hypothetical, they do not represent any investment, and they are not a projection of future results.
| Rigid: $40,000 plus inflation | Flexible: 4% of the balance | |
|---|---|---|
| Withdrawal in year 1 | $40,000 | $40,000 |
| Smallest yearly withdrawal | $40,000 | about $25,700 (year 5) |
| Total withdrawn over 15 years | about $744,000 | about $496,000 |
| Balance after year 15 | about $753,000 | about $1,368,000 |
By the start of year 4, the rigid retiree's withdrawal of about $43,700 is 7.0% of a balance of about $622,000. It began as 4%.
Neither rule is better in the abstract. The rigid rule keeps the paycheck steady and lets the balance absorb the poor years. The flexible rule protects the balance and makes the paycheck absorb them, with a cut of more than a third by year 5 in this example.
One more piece of arithmetic is worth seeing. Under the flexible rule, the balance after year 15 is about $1,368,000 whether the same returns arrive worst first or best first. The order shows up in the income instead: about $496,000 withdrawn in total when the worst years come first, and about $1,017,000 when the best years come first. Sequence risk does not disappear under a flexible rule. It moves from the balance to the paycheck.
Guardrail approaches, described in research by Jonathan Guyton and William Klinger in 2006, start from a fixed dollar plan like the rigid rule and adjust it only when the withdrawal drifts outside preset bounds as a share of the portfolio. If the share climbs past an upper bound, spending is cut by a modest, stated amount. If it falls below a lower bound, spending is raised. The idea is to accept small, occasional adjustments in exchange for less strain on the balance than a fully rigid rule.
How much flexibility a household actually has depends on how much of its spending is fixed. That is a personal question, and a general page cannot answer it.
Nobody knows what future returns will be or in what order they will arrive, so nobody knows the safe withdrawal rate in advance. That is the honest starting point. It still leaves useful questions. How does a particular plan behave across many hypothetical orderings of returns? How much does the result change if spending is a little higher or lower? How much of the spending is covered by income that does not depend on the portfolio? A Monte Carlo analysis is one way to look at those questions. Its results are hypothetical in nature, they do not reflect actual investment results, and they are not guarantees of future results.
Multi-Cycle Planning is a free planning platform. With a free account, a person enters their own numbers and sees a Monte Carlo retirement analysis, tax estimates, and cash-flow projections built from them, including an estimate of sustainable spending at a stated success target. Every number is computed by deterministic software: the same inputs produce the same answer every time. The platform's AI explains those computed figures in plain English and is labeled as AI. It is never asked to produce a number itself. The platform is an educational planning tool, and no advisor reviews a self-serve account unless the person chooses to work with one.
FAQ
This page is educational only. It is not investment advice and makes no recommendations. Examples are hypothetical. More explainers: What is sequence of returns risk? · What does diversification actually do? · What makes inflation go up or down? · All free tools
The full planning platform is free. Create an account, enter your own numbers, and see a Monte Carlo retirement analysis built from them. Results are hypothetical and are not guarantees of future results.