The 4% rule: how much can you withdraw each year?
Short answer
The rule: take 4% of your savings in your first year of retirement, then the same dollars adjusted for inflation each year after. Replayed through every start year in US market history since 1871, that lasted 30 years in 122 of 126 and ran out in 4. It is a place to start, not a promise.
Run your numbers
The Retirement Calculator opens with $5,000 a month of spending and the $1,500,000 the rule asks for already filled in. Change them to yours.
Open the retirement calculator →Test how you take the money out
What the rule says
William Bengen replayed retirements from each past start year, with a mix of stocks and bonds, and looked for the highest first-year withdrawal that never ran out. The Trinity Study (Cooley, Hubbard and Walz) tested the same idea across more mixes and rates. 4% is the round figure that came out of that work for a 30-year retirement.
The rule turns spending into a target. Spending $5,000 a month is $60,000 a year, and $60,000 is 4% of $1,500,000. Read the other way, that is 25 times a year of spending, which is why it is also called the 25x rule. After the first year you do not take 4% of the new balance: you take the same dollars again, raised with prices.
What it assumes
- A retirement of 30 years. A longer one needs a lower rate.
- A portfolio of US large-company stocks and 10-year Treasuries, here 60% in stocks, rebalanced every year.
- The same inflation-adjusted dollars every year, whatever the market does. No cutting back after a bad year.
- The year’s money is taken out at the start of the year, before that year’s return.
- No tax on withdrawals, no other income and no fund fees, unless a row below says otherwise.
- The future looks like the part of US history in the data.
How it held up in US history
Each row starts with $1,000,000 and replays every start year with a full retirement inside Shiller’s 1871 to 2025 data, from 1871 to 1996 for 30 years. It counts past start years. It is not a forecast.
| Plan | Start years it lasted | Worst start year |
|---|---|---|
| A lower start$35,000 a year, 30 years, 60% stocks | 126 of 126 | 1966: fell to 20% of the start |
| The 4% rule$40,000 a year, 30 years, 60% stocks | 122 of 126 | 1966: ran out in year 26 |
| A higher start$50,000 a year, 30 years, 60% stocks | 96 of 126 | 1966: ran out in year 19 |
| A longer retirement$40,000 a year, 40 years, 60% stocks | 107 of 116 | 1966: ran out in year 26 |
| With fund fees$40,000 a year, 30 years, 60% stocks, 1% fund fee | 113 of 126 | 1966: ran out in year 22 |
| A bond-heavier mix$40,000 a year, 30 years, 40% stocks | 119 of 126 | 1966: ran out in year 25 |
The 4% plan ran out for the start years 1965, 1966, 1968 and 1969. The highest first-year rate that lasted 30 years from every start year in this data was 3.7%, held down by the 1966 start.
When it fails
Bad years early on
You sell shares while prices are low, so there is less left when markets recover. This is sequence-of-returns risk, and the Sequence of Returns calculator tests it on your own numbers.
A longer retirement
Stretched to 40 years, the same 4% lasted in 107 of 116 start years, against 122 of 126 for 30. Stopping work in your forties or fifties asks for a lower rate.
Costs and a higher rate
A 1% yearly fund fee took the 4% plan from 122 to 113 of 126. Starting at 5% instead lasted in 96 of 126. Taxes on the withdrawals come on top of both.
Spending that is not flat
The rule pays the same real amount every year. Real spending moves, and so does the part Social Security covers. A plan built on your own spending and benefit is a better test than a single percentage.
How to use it
- Use 4% as a first estimate of the savings you need, then test your own spending, retirement age and Social Security.
- Plan for a lower rate if you will retire early or expect a long retirement.
- Decide now what you would trim after a bad year, so a fall in the early years does not force a bad choice.
- Keep fees low and choose which accounts to draw from first, since both change what you keep.
Sources
- Bengen (1994), "Determining Withdrawal Rates Using Historical Data", Journal of Financial Planning; Cooley, Hubbard and Walz (1998), the Trinity Study
- Robert J. Shiller, "Irrational Exuberance" online data (shillerdata.com)
Updated November 4, 2026.
This guide is for educational purposes only and does not constitute financial advice. Rules and figures change. Check with a licensed professional before you act on it. Full disclaimer →