The 4% Rule and What It Actually Tells You

Where it came from

In 1994, a financial planner named William Bengen published a study in the Journal of Financial Planning. He was trying to answer a question his own clients kept asking: how much can I take out without running out?

Rather than assume an average return, he ran a portfolio through actual historical sequences, starting in every year for which data existed, and tracked which starting withdrawal rates survived a 30-year retirement and which did not. A few years later a group of professors at Trinity University ran a related exercise across different portfolio mixes and horizons, and the shorthand that emerged from both was 4 percent.

That work was genuinely important, and for a reason that often gets lost. Its contribution was not the number. It was the method: testing a plan against sequences of returns rather than against an average, which is the insight underneath everything we cover in how investment management changes when you retire.

The question it answered

Being precise about the question matters, because almost every misuse of the rule comes from answering a different one.

The study asked: taking a percentage of the starting balance in year one, and then raising that dollar amount each year by inflation regardless of what the portfolio did, what is the highest starting percentage that would not have exhausted the portfolio within 30 years, in the worst starting year in the record?

Two things follow from that framing. The rate is a starting rate: after year one you are not taking 4 percent of anything, you are taking last year's dollar amount plus inflation. And because the answer is anchored to the worst historical case, in most historical periods a retiree following it would have died with substantially more money than they started with. That is not a failure of the rule. It is what optimizing against the worst case produces.

The assumptions inside it

Every one of these is defensible as a research choice and consequential as a planning assumption.

Historical analysis and simulation describe how a plan behaved under past or modeled conditions. Neither predicts future results.

The taxes point deserves its own sentence, because it is the one that most often changes the conclusion. A withdrawal from a tax-deferred account funds your spending and your tax bill together. Two households withdrawing the identical percentage can be living on very different amounts depending on which accounts the money comes from, which is the subject of the Social Security tax torpedo and of withdrawal sequencing generally.

How it gets misread

Four misreadings come up constantly.

• Treating it as a promise. It is a description of what would have survived historically. The future is not required to resemble the worst case in the record, in either direction.
• Taking 4 percent of the current balance each year. That is a different strategy entirely. It never runs out, because you are always taking a share of what remains, but your income falls with the market.
• Reading it as a target. A starting rate that survived the worst case is a ceiling in that study, not a goal to spend up to.
• Ignoring that it was designed to fail rarely. Planning to the worst historical case means accepting lower income in most scenarios in exchange for durability in the bad one. That trade-off should be a decision, not a default.

Bengen himself has revisited the figure repeatedly over the years as data and methods changed. Treating any single number from that literature as fixed misrepresents the research it came from.

What is used instead

Most current practice keeps the method and drops the fixed number. The common approaches all share a feature: the withdrawal responds to what the portfolio actually does, within limits stated in advance.

• Guardrails. The withdrawal adjusts when the portfolio crosses defined thresholds, with a stated cut in a bad stretch and a stated raise in a good one. The bounds keep income predictable enough to budget against.
• A floor plus a variable layer. Essential spending is covered by income that does not depend on markets, such as Social Security and any pension, and the portfolio funds the discretionary layer, which can flex.
• Reserve-based approaches. A stated number of years of withdrawals held outside the growth portfolio, so an early decline does not have to be met by selling into it.
• Annual recalculation. Rather than a rate set once, the sustainable amount is recomputed each year against the current balance, current age, and remaining horizon.

None of these removes the underlying risk, and each gives something up. A guardrail gives up certainty of income. A floor gives up growth. A large reserve gives up growth as well. The right combination depends on how much of your essential spending is already covered by income that does not move with markets, which is a household-specific fact rather than a general one.

Where it is still useful

The rule earns its place as a first sanity check, and it is a good one.

If a plan requires drawing substantially more than the historical worst-case figure from year one, with no flexibility and no other income, that is worth knowing early, while there are still choices available: working longer, saving more, spending less, delaying Social Security, or some combination. The useful function is diagnostic.

What the rule cannot do is serve as the plan. A plan states what happens when the assumptions are wrong: what the reserve is, what governs the withdrawal in a down year, and how often the whole thing is revisited against actual results. A number on its own states none of that.

Richard Casolari, Financial Advisor, Investment Advisor Representative, CFP®, has spent roughly fifty years in financial services. The practice is founder-led and based in Palos Heights, and portfolios are built with a documented buy and sell discipline rather than preset risk buckets. The work coordinates retirement income planning, investment management, tax-aware withdrawal sequencing in coordination with your CPA, and Medicare enrollment timing, into one written plan that is revisited on a schedule.