Safe Retirement Withdrawal Rate: What the 4% Rule Actually Said

The 4% rule was never a rule. It was the answer to a stress test, and the number everyone repeats is the worst result that test produced — the floor beneath a bad-luck retiree, not a target for a normal one. Somewhere between the research and the retirement seminar, a survival threshold got remembered as a promise.

That mix-up matters, because what determines whether a portfolio lasts isn’t the percentage you pick in year one. It’s the order your returns arrive in, and what you’re willing to do when that order goes badly.

What the Study Actually Measured

In October 1994, financial planner William Bengen published “Determining Withdrawal Rates Using Historical Data” in the Journal of Financial Planning. He built a portfolio of S&P 500 stocks and intermediate-term Treasury notes, ran it through real U.S. market returns and real CPI inflation, and asked one narrow question: starting in any given historical year, how large could the first withdrawal be — raised annually for inflation, never adjusted for how markets behaved — before the money ran out?

Roughly 4% held up across every starting year he tested, lasting at least three decades in all of them. Not “produced a comfortable retirement.” Lasted. He also argued for holding between 50% and 75% in equities, on the grounds that owning too few stocks did more damage than owning too many.

The boundary conditions are the part that gets dropped:

• A roughly 30-year horizon — nothing useful for someone retiring at 55 who may need 40 years of income.
• A specific two-asset portfolio. Large-cap U.S. stocks and intermediate Treasuries, not your target-date fund.
• U.S. market history only — one country’s unusually good century, replayed.
• Rigid spending. The withdrawal never flexes, and that rigidity is what makes 4% look conservative.
• No taxes, no fees. Every dollar of that withdrawal is gross.

Bengen has revised his own figure upward in later work. The point isn’t that 4% is wrong; it’s that it was a fire drill, and you don’t plan a life around the fire drill’s clock.

Why Two Identical Averages End in Different Places

Here is the risk almost nobody has had explained to them. Two retirees can earn the identical average return over 25 years and end up in wildly different positions, purely because of when the bad years landed.

Averages don’t care about order. Withdrawals do. When you sell to fund living expenses during a downturn, you liquidate more shares to raise the same dollar — and those shares are permanently gone. They aren’t there when the market recovers, so the rebound lifts a smaller base. A 30% loss in year two compounds a wound; the same loss in year twenty-two hits a portfolio that has already done most of its growing.

This is sequence-of-returns risk, and it concentrates the danger into the first decade of retirement. Same person, same savings, same long-run average — a comfortable retirement or a frightening one, depending on nothing but timing.

Bengen’s own worst cases make the point. The most punishing starting years weren’t 1929. They were retirees who walked into the 1973–74 bear market, because falling prices arrived alongside severe inflation — the portfolio shrank while the withdrawal it funded kept climbing.

Sequence risk is largely a function of the volatility you carry into those first years, so it’s worth knowing what your mix implies. Our Investment Risk Calculator shows how allocations behave in a drawdown.

The Three Levers That Actually Help

You can’t control the sequence. You can control your response to it.

Flex the withdrawal instead of fixing it. In March 2006, Jonathan Guyton and William Klinger published “Decision Rules and Maximum Initial Withdrawal Rates” in the Journal of Financial Planning, formalizing what advisors now call guardrails. Their capital-preservation rule cuts the withdrawal 10% when the current rate drifts more than 20% above where it started; a prosperity rule raises it 10% when the rate falls 20% below. The higher sustainable starting rates in their work came with those rules attached — flexibility buys the higher start.

Hold a buffer you can spend from. A cash or short-duration bond reserve covering a year or more of expenses funds a bad market year without selling equities into it. It doesn’t raise returns. It removes the forced sale, which is the mechanism doing the damage.

Be honest about how much of your budget can move. This is the most powerful variable and the one nobody wants to audit. A retiree who can genuinely cut in a bad year can safely start higher than one who can’t — but “flexible” means discretionary travel and a delayed car replacement, not property tax, Medicare premiums, or insulin. Add up your true fixed floor before assuming you can flex. If that floor already crowds out everything else and the monthly bills are the problem rather than the plan, that’s a more urgent situation — the Can’t Pay My Bills guide triages which obligations to protect first.

The Tax Order Is a Default, Not a Law

The conventional sequence — taxable accounts first, then tax-deferred, then Roth last — is a reasonable starting point and a bad fit for some retirees. Followed blindly, it wastes the low-income years right after you stop working, leaving the tax-deferred balance to keep growing until it lands on you all at once.

Because it will land. The SECURE 2.0 Act of 2022 moved the required-minimum-distribution age to 73 for people born from 1951 through 1958, and to 75 for people born in 1960 or later. The statute left 1959 genuinely ambiguous — it reads as both ages — and the IRS reserved that provision in its July 2024 final regulations, addressing it in a separate proposed rule. Roth IRAs have never required distributions during the owner’s lifetime, and SECURE 2.0 removed lifetime RMDs from designated Roth accounts in 401(k) and 403(b) plans starting in 2024.

The squeeze comes when RMDs, Social Security, and any pension switch on within a few years and stack. Retirees who coasted in a low bracket for a decade get pushed up several brackets involuntarily, with more of their Social Security becoming taxable as a side effect.

Which makes the gap years before those streams turn on the most valuable tax real estate most people will ever own. Deliberately realizing income to fill a lower bracket — through tax-deferred withdrawals or partial Roth conversions — can cut lifetime tax even though it raises this year’s bill. Our Tax Bracket Calculator shows where your next dollar falls and how much headroom your bracket has left. State treatment matters too, since some states tax retirement income and others don’t; the state guides cover the local rules.

Claiming and Withdrawing Are One Decision

Social Security timing isn’t separate from withdrawal strategy. Delaying benefits means drawing harder on the portfolio meanwhile, which raises early sequence risk — and it opens exactly the low-bracket window that makes conversions cheap. Model both together. Our Social Security Estimate Calculator gives the benefit at each claiming age, and When to Claim Social Security covers those trade-offs in full.

None of this is a recommendation about your portfolio; the right numbers depend on your balances, health, and tax picture. But the framing is general: your starting rate matters less than your flexibility, your first decade matters more than your average, and account order is a lever most people never pull. If you’re still building the balance, the Retirement Savings Calculator works the accumulation side of the same problem.

Run your own numbers with the Retirement Withdrawal Calculator — enter balances, expected return, horizon, and inflation, and it projects how long the portfolio lasts at a given withdrawal, so you can see what a rate change does to the finish line.