4% Rule vs 3.5% Rule: Which Is Right for You?

The difference between a 4% safe withdrawal rate and a 3.5% safe withdrawal rate is not 0.5 percentage points. On a $120,000 annual spending target, it is the difference between needing a $3.0 million portfolio and needing a $3.43 million portfolio — a $430,000 gap, or roughly three extra years of work for a household saving $150,000 a year. That spread is the entire debate, and most coverage treats it as a rounding error.

William Bengen’s 1994 paper in the Journal of Financial Planning never claimed 4% was universal. It tested a 30-year retirement. The FIRE math that $150k+ households actually face — leaving work at 45 or 50 — runs 40 to 50 years, and at that horizon the research points somewhere lower. This is a question of how long your money has to survive, not which number sounds safer.

The two numbers and where they come from

Start with the source, because the rule is almost always cited without it. Bengen examined historical market returns to find the highest withdrawal rate that never exhausted a portfolio over a 30-year retirement. how to calculate your FIRE number begins here: his worst-case sustainable rate came out to 4.15%, rounded down to 4% in publication, and the round number stuck. The portfolio he modeled was 50% large-cap US stocks and 50% intermediate-term US Treasuries. Those assumptions matter — change the horizon and the number moves.

Bengen himself has moved it. In his 2025 book, he raised his default safe withdrawal rate for a 30-year retirement to 4.7%, arguing the original figure was always the conservative floor for the worst historical start date, not a target. CNBC reported in December 2025 that he considers retirees sticking rigidly to 4% to be “cheating themselves.” That revision applies to a 30-year window. It does not rescue the early retiree.

Why does the horizon change everything? A longer retirement means more inflation-adjusted withdrawals drawn against the same starting balance, and more chances for an early market decline to permanently shrink the base. Research extending Bengen’s method to longer horizons — Wade Pfau, Michael Kitces, and David Blanchett’s 2007 work — converges on roughly 3.5% as the safe withdrawal rate for a 40-year retirement, dropping to around 3.25% at 50 years. Kitces’s analysis notes the rate does not fall much further beyond 40–45 years; 3.5% effectively forms a floor in the US data. Pfau’s 2012 study in the Journal of Financial Planning found a 3.3% rate at a 95% confidence level for 40 years, slightly stricter still.

So the choice is not arbitrary. The 4% rule is a 30-year tool. The 3.5% rule is its 40-year-plus equivalent, derived from the same historical method by researchers who extended Bengen’s own framework.

What the gap costs in portfolio terms

Sources: Bengen (1994), Journal of Financial Planning 7(4):171–180; Pfau (2012), Journal of Financial Planning; Kitces analysis of horizon-adjusted withdrawal rates. FIRE number = annual expenses ÷ safe withdrawal rate.

The mechanics are simple division, and that is precisely why the gap is underappreciated. A FIRE number is annual expenses divided by the safe withdrawal rate. At 4%, $120,000 ÷ 0.04 = $3.00 million. At 3.5%, $120,000 ÷ 0.035 = $3.43 million. Lowering the rate by half a point raises the required portfolio by 14.3%, because you are dividing by a smaller denominator.

The dollar gap scales with spending. A lean FIRE versus fat FIRE cost gap shows up sharply here: at lean FIRE spending the half-point costs less in absolute dollars, while at the fat FIRE income requirements relevant to this audience, it compounds into the high six figures.

Calculation by Finluxy using FIRE number = annual expenses ÷ safe withdrawal rate. Tier spend thresholds per cluster entity definitions; lean FIRE sub-$40k, fat FIRE $100k+ annual spend.

Run the difference at $200,000 of spending and the half-point gap reaches $710,000. That is not a rounding error. It is a down payment on a second house, expressed as the cost of one assumption.

Why the half-point is really a question about sequence risk

The withdrawal rate is a proxy for something less visible: how much damage a bad first decade can do. how sequence of returns risk destroys plans is the actual reason the safe rate falls as the horizon lengthens. A retiree who hits a market decline in year one is selling shares at depressed prices to fund living expenses, permanently reducing the base that needs to compound for the next four decades.

Early Retirement Now’s modeling makes the asymmetry concrete: sequence risk is far more punishing for the retiree drawing down than for the saver still accumulating. The saver who sees a crash buys cheap; the retiree who sees the same crash is forced to sell cheap. A 5% decline in the first year of retirement measurably lowers the probability that a plan survives its full horizon — and the longer the horizon, the more of those vulnerable early years exist. Monte Carlo retirement failure rates quantify this: the same 4% withdrawal that succeeds in the vast majority of 30-year simulations shows meaningfully higher failure rates stretched to 50 years.

This reframes the 4% versus 3.5% choice. You are not picking a spending number. You are pricing insurance against the specific scenario where you retire at 47, the market drops 25% at 48, and you still need the portfolio to last until 90. The half-point reduction is what buys margin against that path.

Finluxy FIRE Timeline Estimate

The withdrawal rate you choose changes not only the target but how long it takes to reach it. The Finluxy FIRE Timeline Estimate measures years from current financial position to FIRE, building current net investable assets plus annual savings at a 7% real return until the portfolio equals annual expenses divided by a 3.5% withdrawal rate. Net investable assets means liquid and investment accounts, excluding primary home equity. The table below models a household with $500,000 in current net investable assets saving $150,000 a year, across three spending scenarios.

Finluxy FIRE Timeline Estimate, calculated per cluster methodology: net investable assets growing at 7% real return plus $150k annual contributions until portfolio = annual expenses ÷ 3.5%. Withdrawal-rate basis: Pfau/Kitces 40-year horizon research.

The fat FIRE timeline of roughly 13 years matches the cluster’s worked example and exposes the real lever. Moving from a 3.5% to a 4% withdrawal basis would cut the fat FIRE target to $3.0 million and shave the timeline by roughly a year — tempting, until you remember that the shorter timeline is purchased by accepting a withdrawal rate built for a 30-year retirement you do not intend to have.

The insight most coverage misses

Nearly every article frames this as “is 4% still safe?” — a yes-or-no verdict on a single number. The data says the question is malformed. The safe withdrawal rate is not a constant to be defended or debunked; it is a function of horizon, and the two numbers are answers to two different questions. Bengen raising his own figure to 4.7% for 30 years and Pfau anchoring 3.3–3.5% for 40 years are not contradictory findings. They are the same method applied to different lifespans.

What this means for a $150k+ household is counterintuitive: the higher your income and the earlier you retire, the more the lower number applies to you — not despite your wealth but because of your time horizon. The fat FIRE retiree leaving at 48 needs the 3.5% framework precisely because they have 42 years of sequence risk ahead, while the conventional retiree at 65 with a 25-year horizon can defensibly use something closer to Bengen’s revised 4.7%. The withdrawal rate is inversely related to how early you stop working. Most coverage flattens that into one rule for everyone.

Methodology

What this means for a $150k+ household

For a household earning $150k+ and targeting fat FIRE, the practical decision is not which rule is “correct” but how much extra capital and time you are willing to commit to insure against a bad early decade. The $430,000 gap at $120,000 of spending is the price of that insurance. A household saving aggressively can close it in roughly one to two additional working years — a real cost, but a finite one, against a 40-year tail risk that is effectively uninsurable any other way.

There is a middle path the binary framing obscures. Rather than committing to a fixed 3.5% for life, some early retirees use a lower initial rate as a buffer and revisit it after the first decade, when the most dangerous sequence-risk window has passed — a dynamic approach Pfau and Kitces both support. The companion levers matter too: healthcare costs before Medicare can add five figures annually to a fat FIRE budget that the withdrawal-rate math alone ignores, and early retirement bracket management can change how much of your withdrawal you actually keep. A coast FIRE posture — building enough that existing assets compound to the target without further saving — interacts with the same horizon logic, since a longer coast means a longer sequence-risk exposure. For the household choosing between retiring earlier on a thinner cushion or working two more years for the 3.5% margin, the data does not pick for you. It only makes the trade-off explicit: a half-point on the withdrawal rate is the single largest lever in the entire plan, larger than almost any spending cut, and for anyone leaving work before 55 it bends toward the lower number.