The 4% Rule: How Much Can You Safely Withdraw in Retirement?
Where the 4% rule comes from, how it held up in every retirement since 1928, what Monte Carlo simulations say, and safer ways to plan your withdrawals.
Published
Free tool: Monte Carlo SimulatorThe 4% rule says you can withdraw 4% of your savings in the first year of retirement, raise that amount with inflation each year, and expect the money to last 30 years. It's the most quoted rule in retirement planning. This guide tests it against nearly a century of history and thousands of simulations, using the free Monte Carlo Simulator. For how simulations work, see Monte Carlo Simulation for Retirement, Explained.
Where the rule comes from
In 1994, financial planner William Bengen tested withdrawal rates against historical US returns. He found that a first-year withdrawal of 4%, raised with inflation, lasted at least 30 years in every period he tested back to 1926. A 1998 study by three professors at Trinity University, known as the Trinity study, reached similar conclusions.
How it works
With $1,000,000, you withdraw $40,000 in the first year. If prices rise 3%, you take $41,200 the next year, and so on, regardless of how the market does. The rule is about the first-year rate: later withdrawals follow inflation, not the balance.
Every retirement since 1928
We tested every 30-year retirement that started from 1928 to 1996, 69 in all, using the S&P 500 and 10-year Treasury returns and inflation from those years, with withdrawals at the start of each year:
| First-year withdrawal | 60/40 | 40/60 | 80/20 | 100% stocks |
|---|---|---|---|---|
| 3% | 69 of 69 | 69 of 69 | 69 of 69 | 69 of 69 |
| 3.5% | 69 of 69 | 69 of 69 | 69 of 69 | 68 of 69 |
| 4% | 65 of 69 | 62 of 69 | 66 of 69 | 64 of 69 |
| 4.5% | 58 of 69 | 49 of 69 | 60 of 69 | 62 of 69 |
| 5% | 50 of 69 | 41 of 69 | 54 of 69 | 54 of 69 |
| 6% | 36 of 69 | 24 of 69 | 43 of 69 | 47 of 69 |
With a 60/40 mix, 4% lasted 30 years in 65 of 69 periods. The four that failed started in 1965, 1966, 1968 and 1969, and ran out after 25 to 27 years, as high inflation and weak markets in the 1970s took their toll. Bengen's study, using different bond data and withdrawal timing, found 4% survived even those periods. Small changes in data and timing move the edge cases, so treat 4% as a borderline rate in the worst historical cases rather than a guaranteed one.
In the typical case, though, the money did far more than last: the median 60/40 retiree at 4% ended 30 years with 1.39 times the starting balance after inflation, and 43 of the 69 ended with more than they started with.
Longer retirements
Retiring early means planning for 40 years or more. Over the 59 forty-year periods starting from 1928 to 1986, with 60/40:
| First-year withdrawal | Lasted 40 years |
|---|---|
| 3% | 59 of 59 |
| 3.5% | 58 of 59 |
| 4% | 50 of 59 |
| 4.5% | 40 of 59 |
| 5% | 32 of 59 |
That's why people planning to retire early often use 3% to 3.5%. See How to Calculate Your FIRE Number.
What simulations say
History offers only a few dozen independent 30-year periods. A Monte Carlo simulation tries thousands of possible sequences. Results from the Monte Carlo Simulator, with 10,000 simulations over 30 years, withdrawing from $1,000,000 with amounts raised for inflation:
| First-year withdrawal | Your assumptions: 7% average, 12% volatility, 2.5% inflation | US history, 60% stocks |
|---|---|---|
| 3% | 95.9% | 98.0% |
| 3.5% | 90.4% | 94.8% |
| 4% | 81.5% | 90.3% |
| 4.5% | 70.2% | 83.8% |
| 5% | 57.5% | 75.0% |
Over 40 years, the same assumptions give 65.8% at 4% and 78.3% at 3.5%. The answer depends heavily on the returns you assume: lower expected returns argue for a lower starting rate.
Criticisms of the 4% rule
- Fees and taxes aren't included; a 1% yearly advisory fee is roughly like withdrawing 5% instead of 4%.
- Future returns may be lower than the 20th century's.
- It's rigid. Real retirees spend less after bad years and more after good ones.
- It often leaves a lot behind, which means many retirees could have spent more.
More flexible approaches
- Guardrails: cut withdrawals by about 10% if the withdrawal rate drifts well above the starting rate, and raise them if it drifts well below. The best-known version is from Jonathan Guyton and William Klinger (2006).
- Fixed percentage: withdraw a set percentage of the current balance each year. The money never runs out, but income swings with the market.
- Floor and upside: cover essential spending with guaranteed income, and invest the rest for growth.
Try it with the Monte Carlo Simulator
- Open the Monte Carlo Simulator and enter your balance and first-year withdrawal.
- Choose your return assumptions, or US history with your stock and bond mix, and the number of years.
- Read the chance your money lasts, and try lower withdrawals or a different mix to see how the odds change.
Frequently asked questions
Is the 4% rule still safe?
It has held up in most historical periods, but it was borderline in the worst ones, and it doesn't include fees. Many planners now suggest 3.5% to 4% for a 30-year retirement, with flexibility to cut back.
Does the 4% rule include Social Security?
No. It applies to your savings. Subtract Social Security and any pension from your spending first, and apply the rule to the rest.
What if I retire in a bear market?
Starting after a fall isn't necessarily worse: prices are lower and future returns may be higher. The danger is a fall in the first years after you start withdrawing. See Sequence of Returns Risk.