What Is the 4% Rule for Retirement?

Withdraw 4% of your savings in year one, then raise that dollar amount with inflation each year after. In the historical data it never ran out over 30 years. That is the whole rule — and every word of it matters.

The short answer

On a $1,000,000 portfolio the 4% rule says you take $40,000 in the first year — $3,333 a month. In year two you do not take 4% again. You take last year's $40,000 plus inflation, so at 2.5% inflation you take $41,000, regardless of what the portfolio did. The percentage is applied once, at the start. Everything after that is an inflation adjustment.

That distinction is the single most misunderstood thing about the rule, and it is what makes it a spending plan rather than a formula you re-run each year.

Where the number came from

In 1994, a financial planner named William Bengen published Determining Withdrawal Rates Using Historical Data in the Journal of Financial Planning. He took every 30-year retirement window in the U.S. record — someone retiring in 1926, 1927, 1928 and so on — and asked which starting withdrawal rate would have survived the worst of them with a portfolio of roughly half stocks and half bonds.

The answer was a little over 4%. Bengen called that worst-case number SAFEMAX. It is not the rate that usually works; it is the rate that worked even for the unluckiest retiree in the sample, someone who stopped working straight into the 1929 crash or the 1966–1982 stagnation. The 1998 Trinity Study, by three Trinity University professors, tested a similar range of portfolios and reported success rates that broadly agreed.

So the 4% rule is a floor derived from a bad case, not an average. That is precisely why it feels too cautious most of the time — it is designed to be.

What the rule actually promises

• Thirty years, not forever. The study horizon was 30 years. Retire at 65 and that takes you to 95, which is a reasonable plan. Retire at 55 and the rule was never tested on your situation.
• Your starting balance, not today's. The 4% is computed once. Later withdrawals ignore the portfolio's value entirely.
• Before tax. Withdrawals from a traditional 401(k) or IRA are ordinary income. A $40,000 withdrawal at a 20% blended rate is $32,000 to spend.
• Before fees. Bengen's data had no fund expenses or advisory fee subtracted. A 1% annual fee comes directly out of the safe rate.
• U.S. returns. The record it draws on is the best-performing large market of the twentieth century.

The four things that break it

A longer retirement. The safe rate is a function of horizon. Morningstar's 2026 research puts the 30-year figure at 3.9% and the 35-year figure at 3.5% — that is a 10% pay cut for five more years of life. Early retirees planning 40 or more years should be thinking closer to 3%.

A different portfolio. The rule assumed a substantial equity allocation. Hold mostly bonds or cash and you lose the growth the rule depends on; hold 100% stocks and you gain volatility without much extra safe spending, because the worst case gets worse.

Bad luck early. A poor decade at the start does far more damage than the same decade at the end, because you are selling into it. This is called sequence-of-returns risk, and it is the mechanism the 4% rule was built to survive. It is also the reason two retirees with identical average returns can land in completely different places.

Fees. A 1% advisory fee plus 0.5% in fund costs is 1.5% a year off the top, and it comes out of the same return the safe rate is calculated from.

What the safe rate is now

Nobody serious treats 4% as fixed. Morningstar recalculates it annually against current bond yields and equity valuations; their 2026 base case is 3.9% for a 30-year retirement, a 30–50% equity allocation and a 90% probability of success. It has moved between 3.3% and 4.0% over the past six years, entirely because the starting conditions moved.

Bengen himself has revised upward over the years — adding asset classes such as small-cap and international to the mix pushed his own figure above 4.5% in later work. Two credible researchers landing on different numbers is not a contradiction. It is a reminder that the answer depends on assumptions you get to choose, which is the argument for running your own numbers rather than adopting anyone's rule of thumb.

Try it against your own balance

The calculator below starts at the 4% rule on $1,000,000. Change the balance to yours and watch the "to last 30 years" figure — that number is the 4% rule recomputed for your actual return, inflation and tax assumptions, which is more useful than the rule itself.

When 4% is the wrong number for you

Too conservative if you have a pension or claim Social Security late, so guaranteed income covers your essentials and savings only fund discretionary spending. Also if you are willing to cut spending in a bad year — Morningstar found that flexible strategies, which adjust withdrawals to market conditions, supported starting rates as high as 5.7%. Flexibility buys more than allocation ever will.

Too aggressive if you retired before 60, if you pay 1%+ in fees, if your portfolio is mostly bonds, or if you may face significant long-term-care costs. Morningstar's figure drops to 3.5% once long-term care is modelled for someone retiring at 67.

The honest use of the 4% rule is as a sanity check: a fast way to know whether you are in the right postcode. If your plan needs 7%, no amount of asset allocation will rescue it, and you have learned something important in ten seconds. If it needs 3%, you can probably afford more life than you are allowing yourself.