The 4% Rule Answered a Narrower Question
The 4% rule is the most quoted number in retirement planning and one of the most misread. I've used it myself as a back-of-the-envelope check, and it's genuinely useful for that. But the more time I spent building a real plan, the clearer it got that the rule was the answer to a much narrower question than the way people quote it suggests.
What it actually was
In 1994 a financial planner named William Bengen published a study in the Journal of Financial Planning. He took US market history and asked a specific question: if you retire, withdraw some percentage of your portfolio in year one, and then raise that dollar amount with inflation every year after, what's the highest starting percentage that would have survived 30 years in the worst case history had produced?
His answer was 4%, and it was not a rounded-down version of something more precise. Working from a portfolio of 50% common stocks and 50% intermediate-term Treasuries, rebalanced annually, he found that a 4% first-year withdrawal had never exhausted a portfolio in under 33 years, while 4.25% could have run one dry in 28. Three years of headroom over his own 30-year requirement was the whole of the margin. The retirements that set that floor began in the mid-1960s and early 1970s and ran into the 1973–1974 collapse and high inflation in their first years, which is about the most hostile start a retirement can have.
The 50/50 split deserves a sentence of its own, because it gets quoted as a recommendation and it was not one. Bengen called it "an arbitrary asset allocation chosen for purposes of illustration", and by the end of the same paper he argues for holding somewhere between 50% and 75% in stocks — as close to 75% as the client can stand. The 60/40 portfolio that often gets attached to his name appears in the article exactly once, in the opening sketch of a planner who multiplies out an 8.2% average return and lands his client in trouble. It is the mistake the paper was written to correct, not the finding.
Read that back and notice what it is. It's not a forecast. It's not an optimum. It's a floor derived from a single unlucky cohort in one country's history. Bengen was answering "what's the worst that happened?" and people have been hearing "what should I do?"
The misreading that matters most
If you take one thing from this: the 4% rule is not "withdraw 4% of your balance each year."
It's 4% of your balance in year one, and after that you forget the percentage entirely and just raise last year's dollar figure by inflation. On a million-dollar portfolio that's $40,000 in year one, then roughly $41,200, then $42,400, whether the market went up 20% or down 30%.
Withdrawing 4% of the current balance every year is a completely different strategy. It can never fully run dry, because you're always taking a slice of what's left, but your income drops exactly when markets do. Neither approach is wrong. They just behave in opposite ways in the scenario you care most about, and the phrase "the 4% rule" gets used for both.
So is it still viable? Two credible answers disagree
This is where 2026 gets interesting, because the public disagreement is wider than usual and it's coming from serious people.
Morningstar says 3.9%. Their State of Retirement Income research puts the highest safe starting rate for steady, inflation-adjusted income over 30 years at 3.9%, aiming for a 90% chance of not running out. Their number moves year to year because it's forward-looking: it reflects stock valuations, bond yields and inflation expectations as they stand now, not as they averaged across history.
Bengen now says 4.7%. The man who produced the original figure revised it upward, first to 4.5% in 2006 and then, in his 2025 book A Richer Retirement, to a SAFEMAX of 4.7%. The change isn't optimism about the future. It's that he widened the portfolio from two asset classes to seven, including mid-, small- and micro-cap US stocks and international equities, and better diversification raised the worst historical case. He's been explicit that 4.7% is a conservative floor rather than a target.
Nearly a full percentage point separates them, which on a million dollars is $8,000 a year. They're not contradicting each other so much as answering different questions: one is "what do today's valuations imply going forward," the other is "what's the worst that history did to a well-diversified portfolio." If two people this careful land that far apart, the honest conclusion isn't that one of them is right. It's that the precision implied by any single number is false.
Why your own plan matters more than the number
Every version of this research shares a set of assumptions, and the assumptions are where real plans diverge from the model.
It assumes 30 years. If you're retiring at 55 and planning to 95, that's 40, and a longer horizon has historically supported a lower starting rate. Early retirement also adds years of funding your own health insurance before Medicare, which is usually the largest single line item in the gap.
It assumes one undifferentiated pot of money. Most households have three kinds: taxable, pre-tax and Roth. A dollar from each costs a different amount in tax, and some of them you legally cannot touch yet. Retire before 59½ and a chunk of your balance isn't spendable on the timeline the rule assumes.
It assumes no other income ever arrives. This is the big one. Social Security turns up partway through, and if you're married it turns up twice, often years apart. In my own plan the portfolio has to carry the full load for a handful of years, and then my wife's benefit starts and the two together cover the budget outright. The shape of that is nothing like a flat withdrawal for thirty years. A single percentage can't describe it, because the hard part isn't the whole retirement, it's a specific window in the middle.
It assumes your spending is flat in real terms. Most people's isn't. Spending tends to be higher in the early, more active years, drift down through the middle, and tick back up later with healthcare. Modelling it as a flat line is tidy but it isn't anyone's life.
What I'd actually do with it
None of this makes the rule useless. It makes it a sizing tool rather than a withdrawal plan, and that's still worth something.
Use it to size the problem. Multiplying your annual spending by 25 gets you a target in about four seconds. That's a reasonable first look at whether you're in the neighbourhood.
Then stop using it to make decisions. The questions that actually move the outcome are when you claim Social Security, which account you draw from in which year, what happens in a bad first decade, and whether you can flex spending. The rule is silent on all four.
Pay attention to the first decade. The entire 4% figure exists because of one cohort that got a bad start. That's sequence-of-returns risk, and it's the mechanism behind the number, not a footnote to it.
Value flexibility over precision. Morningstar's own work finds that retirees willing to adjust spending can support meaningfully higher rates than the fixed, inflation-adjusted approach. Being able to skip an inflation raise after a bad year is worth more than arguing about whether the right starting number is 3.9 or 4.7.
These are things to weigh, not instructions. Which apply depends on your own situation, and a fee-only planner can help you work through them for your circumstances.
See it on your own numbers
If you want to compare the rule of thumb against your actual plan, the free retirement withdrawal calculator will show you what a given rate does to your balance over time, and the Monte Carlo retirement calculator runs the same plan through a thousand different market histories so you can see how much of the answer depends on luck. For the whole picture, including when Social Security starts and which accounts you draw from first, planbend models it as one plan rather than one percentage. If you'd like a professional's read on your situation, there are directories on our Connect with a professional page.