The 4 percent rule is the single most-cited piece of retirement advice on the internet, and it is also one of the most misapplied. It was calibrated for a specific scenario: a 30-year retirement horizon, a portfolio held in the mid-20th-century US stock and bond markets, and an annual inflation-adjusted withdrawal. If your retirement plan does not match those parameters, the rule is not lying to you exactly, but it is doing a different math problem than the one you are trying to solve.
For anyone planning to retire in their forties or early fifties, three specific assumptions in the original derivation quietly break, and the direction they break in is not favorable. This piece walks through where the mismatch is, why it matters, and what to plug in instead.
Where the rule came from
The 4 percent rule traces back to the Trinity Study, published in 1998 by three Trinity University finance professors. They looked at rolling 30-year historical periods and asked: given a starting portfolio and an annual inflation-adjusted withdrawal, what withdrawal rate survived essentially every historical window.
The answer, for a 50/50 stock/bond portfolio over 30 years, was around 4 percent. Even the worst 30-year sequences (starting in 1929, 1966, and 1969) survived at that rate. That is the entire foundation of the rule. It is empirical, calibrated to US historical data, and specific to a 30-year time frame.
Almost no one who uses the rule uses it in the specific way it was derived.
Break number one: the 30-year assumption
If you retire at 65 and live to 95, the 30-year assumption holds. If you retire at 45, your horizon is more like 45 or 50 years. The failure rate at any given withdrawal rate rises with horizon length in a way that is easy to underappreciate.
The original 4 percent rate had close to a 95 percent success rate for 30 years. The same rate over 45 years has a success rate in the low 80s, and over 50 years drops into the mid 70s. That difference is not decorative. It means one out of every four people who retire at 45 and follow the rule literally will run out of money before they die.
The ongoing analysis at Early Retirement Now has spent years working through this specific mismatch, and the answer they converge on is that the "safe" rate for a 45-year horizon is closer to 3.25 to 3.5 percent, not 4.
Break number two: the historical base is optimistic
The Trinity Study calibrated against US markets in a period that was, by global historical standards, extraordinarily good. The US did not have a currency collapse, did not have a period where equity markets closed for years, did not have hyperinflation, did not have property confiscation.
Most other major economies did have at least one of these in the same window. Japanese equity markets are still below their 1989 peak, four decades later. Any Japanese retiree who used the US 4 percent rule in 1989 ran out of money.
The point is not that the US is due for a Japan-style outcome. The point is that a rule calibrated on the best historical run in developed-market history is not a floor. It is the outcome you got from an unusually good base case. If the next 50 years are merely average by global historical standards rather than exceptional the way the US 20th century was, the 4 percent rule's headroom is smaller than the number suggests.
Break number three: bonds do not act the way they used to
The 4 percent rule assumes bonds behave as a stable ballast when stocks fall. Between 1980 and 2020 that was mostly true, because bond yields spent 40 years falling from double digits to near zero. Falling bond yields mean rising bond prices, which meant bonds delivered strong real returns and offset stock drawdowns.
That regime is over. Starting bond yields today are higher than they were five years ago, which is good for bond expected returns going forward, but the sustained tailwind of falling yields is not repeating. And 2022 demonstrated that in a specific kind of inflation shock, bonds fall alongside stocks rather than offsetting them. The Bogleheads community has extensive discussion on this exact regime shift, and the practical answer they converge on is to hold more diversified real assets alongside the traditional stock/bond mix.
For an early retiree specifically, the implication is that the 50/50 portfolio the Trinity Study modeled is not the same 50/50 portfolio you can build today. Same allocation, different expected behavior.
The compounding effect
Take these three breaks together. A rule calibrated for a 30-year horizon (you have 45), against unusually strong historical returns (future returns are likely more average), assuming bonds hedge stocks (which they do less reliably now). Each individual mismatch is a couple of percentage points of tail risk. Together they can turn a "95 percent safe" rate into a "70 percent safe" rate.
This is why the mainstream FIRE community has been quietly migrating its default assumption from 4 percent to somewhere between 3 percent and 3.5 percent over the last decade. It is not that the 4 percent rule was wrong. It is that using it outside its calibration range is wrong. See the extensive threads at Mr Money Mustache for the practical drift in community defaults over time.
The specific numbers to plug in instead
For a defensible early-retiree plan, three concrete substitutions.
Start with a 3.5 percent withdrawal rate, not 4 percent, for horizons of 40 years or more. If your horizon is 50 years or more, start at 3.25 percent. These numbers give you the same tail-scenario protection the original 4 percent rule offered for 30 years, adjusted for the longer horizon.
Use a return assumption tied to observable inputs, not a fixed 7 percent. A reasonable starting point is the current ten-year Treasury yield plus 3 to 4 percent for equity exposure, weighted by your allocation. You can pull current yield data from public sources like the St. Louis Fed FRED database any time you want to recalibrate.
Add a spending guardrail rule that reduces discretionary spending by 10 to 15 percent when the portfolio falls below 80 percent of its starting value. This one adjustment often adds more tail-scenario safety than lowering the withdrawal rate by another half percent, and it lets you keep a slightly higher baseline.
What actually works for early retirees
Three practical adjustments that consistently improve outcomes.
Lower the starting rate. If you were planning to withdraw 4 percent, plan on 3.25 or 3.5 percent for the first decade of retirement. Once you are past the first 10 years without a bad sequence, the flexibility to raise the rate opens up.
Use guardrails. Instead of a fixed inflation-adjusted withdrawal, tie your withdrawal to portfolio performance. Reduce spending 10 percent when the portfolio falls below a threshold, restore it when it recovers. This alone can add 10 to 20 percentage points of success rate over long horizons.
Model your specific plan. The 4 percent rule is a general answer to a general question. Your plan is specific. Plug your actual portfolio, actual expected spending, actual horizon, and actual asset allocation into a proper Monte Carlo tool. The free FIRE calculator by EvvyTools will run the simulation on your actual inputs and show you where the pressure points are.
For a full walkthrough of how to stress-test a plan against the specific sequences that historically broke retirements, see the detailed guide at EvvyTools.
The one line summary
The 4 percent rule is not wrong. It is right about the specific problem it was solving. If you are 45 and planning a 50-year retirement, that is not the problem you are solving. Use the rule as a starting benchmark, then adjust down for the horizon, adjust down for likely lower future returns, and adjust down for reduced bond hedging power. What you end up with is not the 4 percent rule anymore, but it is a plan that survives.
Top comments (0)