The 4% Rule: Where It Comes From and Where It Breaks

The 4% rule is probably the most repeated number in retirement planning, and also one of the most commonly repeated without its conditions attached. Stated on its own — withdraw 4% of a portfolio in the first year of retirement, then adjust that dollar amount for inflation every year after — it sounds like a universal formula. It was never meant to be one. It's the output of a specific historical test, run against a specific set of assumptions, and understanding what those assumptions were is the difference between using the rule as a starting point and using it as a promise it never made.

The logic behind the number

The reasoning behind the 4% figure works backward from a question: looking across historical periods, what withdrawal rate — as a percentage of an initial portfolio balance, adjusted for inflation each subsequent year — would a retiree have been able to sustain for a set retirement length, roughly three decades, without running out of money, even in the worst-starting periods on record? Testing a range of withdrawal rates against historical sequences of market returns, a rate around 4% held up across most of the historical starting points examined, including the least favorable ones, periods where a retirement began just before a prolonged downturn.

The key mechanism is what's often called sequence-of-returns risk: it's not the average return over thirty years that determines whether a withdrawal rate survives, it's the order those returns arrive in. A portfolio that experiences poor returns in its first several years of withdrawals, while money is also being pulled out, can be permanently damaged in a way that the same average return spread differently across the decades would not have caused. The 4% figure is essentially a rate calibrated to survive the worst historical orderings, not the average ones, which is also why it looks conservative in hindsight for the many historical periods that weren't worst-case.

What the rule assumes, explicitly

The reasoning rests on several conditions that don't always make it into the shorthand version. It assumes a specific time horizon, commonly framed around a thirty-year retirement, which fits a retirement starting in the mid-sixties but describes a much larger fraction of the drawdown for someone retiring earlier. It assumes a particular portfolio composition — a mix of stocks and bonds roughly balanced, not a portfolio concentrated heavily in one asset class or held entirely in cash. And it assumes annual withdrawals that rise with inflation regardless of what the portfolio is doing that year, a fixed, non-adaptive spending pattern that spends the same real amount whether markets are up sharply or down sharply.

Where each assumption breaks

Change any of those conditions and the rate that historically "worked" changes with it. A retirement horizon meaningfully longer than three decades, an early retirement, for instance, needs a lower starting rate to have similar odds of lasting the distance, simply because there are more years for a bad sequence to do damage. A portfolio mix skewed away from the balanced allocation the original testing assumed will have a different risk profile, for better or worse, than the rate was calibrated against. And a spending rule that never adapts — withdrawing the same inflation-adjusted dollar amount in a year the portfolio is down sharply as in a year it's up sharply — is more conservative than how most real retirees actually behave, because most people who watch a portfolio drop tend to pull back discretionary spending somewhat rather than mechanically continuing the same withdrawal.

Using it as a starting point, not a verdict

None of this means the 4% figure is wrong — it's a reasonable, historically grounded starting point for thinking about sustainable withdrawal rates, and it remains useful precisely because it's simple enough to reason about. The mistake is treating it as a fixed rule that applies identically to every retirement length, every portfolio, and every spending style, when the number was calibrated against one specific version of all three. The more useful takeaway isn't "withdraw exactly 4%" — it's the reasoning underneath it: sequence-of-returns risk matters more than average returns, time horizon changes the safe rate, and a spending plan that can flex in bad years carries less risk than one that can't. Anyone using the rule as a planning input should treat those conditions as inputs to revisit, not fine print to ignore.

Why flexible spending changes the picture more than any tweak to the rate

Of the assumptions behind the 4% figure, the fixed, non-adaptive spending rule may be the one that diverges furthest from how households actually behave, and it's worth dwelling on because the fix doesn't require a different percentage at all — it requires a different posture toward spending in bad years. A retiree willing to trim discretionary spending, skip a planned upgrade, or delay a large purchase during a period when the portfolio is down meaningfully reduces the odds of depleting the portfolio compared with a retiree who withdraws the identical inflation-adjusted amount regardless of what the market is doing. This flexibility doesn't show up in the historical testing behind the 4% figure, because that testing modeled a fixed, rule-following withdrawal pattern on purpose, to find the worst-case rate that would survive even without any such adjustment. In practice, a willingness to flex spending in the hardest years is often a bigger lever on a retirement's actual safety than the difference between withdrawing 3.5% and 4.5%.

A different lens: guardrails instead of a fixed percentage

Partly in response to the fixed-rate rule's rigidity, some approaches to retirement withdrawals use guardrails instead of a single locked percentage — a starting rate similar in spirit to 4%, paired with predetermined rules for adjusting spending up or down as the portfolio's actual performance diverges from the original plan. If the portfolio grows well ahead of plan, spending is allowed to rise; if it falls meaningfully behind, spending is trimmed before the shortfall compounds. The appeal of a guardrail approach is that it responds to what's actually happening rather than assuming, at the outset, that every year for three decades will draw down at the identical inflation-adjusted figure. It doesn't eliminate the need for a starting assumption — the guardrails still have to be set somewhere — but it converts a single static rate into an ongoing process, which more closely matches how most retirees would actually want to respond to a portfolio that's running either better or worse than expected.