Every financial independence calculation on this site, and on every other site, rests on one number: the rate at which you can draw down a portfolio without running out. Change it from 4% to 3.5% and a $50,000 budget goes from needing $1.25 million to needing $1.43 million. That is a difference of nearly two hundred thousand dollars, produced by moving a single decimal, and it is worth knowing where the decimal came from.
Where it came from
The rule descends from two pieces of work in the mid-1990s. William Bengen, a financial planner, ran historical US market data through a series of simulated retirements in 1994 and found that a portfolio of stocks and bonds had survived every 30-year window in the record at an initial withdrawal of about 4%, adjusted for inflation each year afterward. A few years later three professors at Trinity University published a broader version of the same exercise, testing various stock/bond mixes across historical periods and reporting the share that survived. That paper is where the phrase “the 4% rule” attached itself.
Two details of the method matter more than the headline.
The withdrawal is fixed at the start, then inflation-adjusted. You take 4% of the balance in year one and then that same real amount every year afterward, regardless of what the portfolio does. You do not take 4% of the current balance each year. This assumption does most of the work, and it is also the one least like how people actually behave.
Success meant not hitting zero. A retirement that ended with one dollar counted as a success, identically to one that ended with five million. The rule was never a claim about comfort or about leaving an estate. It only addressed not running out inside the window.
The window is the problem
The window was thirty years. That is a sound planning horizon for someone retiring at 65. It is an unsound one for someone retiring at 40, who may need the portfolio to last fifty years or more.
This is the largest single reason to treat 4% carefully in a FIRE context. The research doesn’t say a portfolio survives indefinitely at 4%. It says the portfolio survived thirty years in the historical US record. Extending the horizon lowers the rate that clears the same bar. Analyses of longer periods generally land somewhere between 3% and 3.5% for a portfolio expected to last indefinitely, though the answer is sensitive to how you ask the question.
There is a countervailing point, and it is a real one. In the large majority of the historical windows tested, the portfolio did more than survive. It finished with substantially more than it started with, often several times more. The rule is calibrated to the worst case in the record, so it is too conservative for most outcomes and, if the future is worse than the past, not conservative enough for the one that matters.
Sequence of returns
The average return over a retirement matters far less than the order the returns arrive in. Two retirements with identical thirty-year average returns can end very differently depending on when the bad years land.
Withdrawals and losses compound against each other. Selling into a fallen market to fund a year of spending removes shares that are not there to recover when the market does. A 30% drawdown in year two of retirement is a different event from the same drawdown in year twenty-five, because the first one permanently shrinks the base that everything afterward grows from.
Every calculator on this site, including all four in the FIRE section, assumes a smooth constant return. That is what makes them arithmetic and not simulation, and it is the honest limit of what they can tell you. They will give you the right order of magnitude for a target. They cannot tell you the probability that a particular retirement survives, because they do not model sequence at all.
What real retirees do instead
The fixed-withdrawal assumption is the rule’s weakest link and also the easiest to fix. Nobody with an intact survival instinct spends the same inflation-adjusted amount through a 40% market decline. Real flexibility improves outcomes substantially, and it does so precisely in the scenarios where the fixed rule fails. That flexibility includes cutting discretionary spending in bad years, deferring a large purchase, and taking on some work.
Several formalised versions of this exist. Guardrail strategies set an upper and lower band around the withdrawal and adjust when the portfolio crosses one. Others tie the withdrawal to the current balance instead of the starting one, which cannot run out by construction but produces an income that moves with the market. A plan able to bend is worth more than a higher starting number.
How to use the number
Treat 4% as a reference point and run the range:
- 4%: the conventional figure, calibrated to 30-year historical windows.
- 3.5%: a common adjustment for a retirement expected to run much longer.
- 3%: cautious, and roughly what a very long horizon with no flexibility implies.
The FIRE calculator takes the withdrawal rate as an input for this reason. Run your spending at all three and look at the spread instead of any single figure. If the gap between them changes your plans, that gap is the most useful thing the calculation has told you, and it is more informative than a precise-looking target that quietly depends on a decimal nobody examined.
The withdrawal rate decides the multiple, but you choose the thing being multiplied. At 4% the target is 25× spending, and at 3.5% it is roughly 28.6×. Both multiply the same input, and reducing your spending pulls down the target under every assumption at once. It is the only lever on this page that doesn’t depend on being right about the future.