The 4% Rule: Where It Comes From and What It Assumes
Educational explainer, not financial advice. The 4% rule is a research finding — a heuristic from historical back-testing — and this page describes what the research measured, no more. Statutory ages and account rules mentioned here are current as of September 2026 and change over time.
Few numbers in retirement math travel as far from their source as this one. "The 4% rule" gets quoted as if it were a law of nature or a promise, when it is actually a compact summary of two academic back-tests from the 1990s. Knowing what those studies computed — and what they held fixed while computing it — is the difference between using the number as a starting point for arithmetic and mistaking it for a guarantee. This guide covers the origin, the exact mechanic with a worked example, the assumptions underneath it, and where the research's own boundaries lie.
Where the rule comes from
The rule originates with William Bengen, a financial planner who published "Determining Withdrawal Rates Using Historical Data" in the Journal of Financial Planning in 1994. Rather than assuming an average return, Bengen replayed history: for each starting year in his U.S. market data, he asked what initial withdrawal rate — with the dollar amount adjusted for inflation annually — a stock-and-bond portfolio could have sustained for 30 years. An initial rate of about 4% survived every rolling 30-year period in the data he studied, including retirements that began just before the worst market stretches in the sample.
In 1998, three professors at Trinity University — Cooley, Hubbard, and Walz — published what is now called the "Trinity study," testing a grid of withdrawal rates, stock/bond mixes, and payout periods against historical data. Its results broadly reinforced Bengen's: inflation-adjusted initial withdrawals near 4% showed high historical success over 30-year horizons. Together the two papers are the entire foundation of the rule. It is research about past U.S. market sequences — nothing in either paper predicts future returns.
The exact mechanic, worked through
The convention has two steps, and the second is the one people miss. First, the initial withdrawal is set at 4% of the portfolio's starting value: from a $1,000,000 portfolio, that is $40,000 in year one — about $3,333.33 per month. Second, in every later year, the dollar amount is adjusted for inflation. The portfolio's current balance never re-enters the calculation; the 4% figure is used exactly once.
The arithmetic of the adjustment, written symbolically — actual inflation varies year to year, so no example rate is assumed here:
| Year | Withdrawal | How it is computed |
|---|---|---|
| 1 | $40,000 | 4% of the $1,000,000 starting portfolio — the only time 4% is used |
| 2 | Year 1's amount, adjusted | $40,000 × (1 + that year's measured inflation) |
| 3 | Year 2's amount, adjusted | Year 2's amount × (1 + that year's measured inflation) — the balance is never consulted |
This design holds purchasing power constant: the retiree's spending stays level in real terms while the portfolio absorbs all the market risk. It also explains why the rule can fail — in a bad early stretch, the withdrawals keep growing with inflation while the balance shrinks. A strategy that instead withdraws 4% of each year's current balance behaves completely differently: it can never fully deplete, but income falls whenever the market does. The Inflation Calculator translates dollar amounts across years if the adjustment step itself is the unfamiliar part.
What the back-tests held fixed
The 4% result is inseparable from its assumptions, and both papers state them plainly. The portfolio is a specific mix of U.S. stocks and bonds, rebalanced and held throughout — not cash, not a single stock, not whatever allocation a given retiree happens to hold. The horizon is 30 years. The withdrawals follow the mechanical schedule above with no skipped or reduced years, no spending shocks, and no fees or taxes modeled. Change any of these — a more conservative allocation, a longer retirement, a habit of raising withdrawals in good years — and the historical success rates the studies reported no longer apply as stated. The exact allocations and period grids are in the original papers, which remain the authoritative source for what was tested.
The standard criticisms, as research boundaries
The criticisms of the 4% rule are less attacks than reminders of where the studies stop. Sequence-of-returns risk is the first: because withdrawals are fixed in real dollars, the order of returns matters as much as their average. A deep downturn in the first years of retirement forces the sale of assets at depressed prices to fund a withdrawal that inflation is simultaneously pushing up — the exact stress that determined the historical worst cases in Bengen's data. A back-test can only contain the bad sequences that actually happened; it says nothing about worse ones.
The 30-year horizon is the second. Someone who retires early, or simply lives long, may need a portfolio to fund inflation-adjusted withdrawals well beyond the studied 30-year window — more withdrawals through more market cycles than the studies measured. The research literature treats longer horizons as a separate, harder problem, not a case the 4% figure covers. Timing interacts with the other fixtures of a retirement plan too — Social Security claiming ages and required minimum distributions run on their own statutory clocks, laid out in the retirement age milestones guide.
Finally, the studies model a portfolio in a vacuum: no advisory or fund fees, and no taxes. Which account a withdrawal comes from changes what it nets after tax — traditional 401(k) and IRA withdrawals are generally taxed as ordinary income, while qualified Roth withdrawals are tax-free. The retirement accounts guide walks through those mechanics, and irs.gov carries the current rules.
Running the arithmetic on other inputs
Because the rule is just arithmetic, its sensitivity to inputs is easy to see directly. The Retirement Withdrawal Calculator computes the first-year income for any initial rate from 2% to 6% — showing immediately how much income a half-point of withdrawal rate is worth on a given portfolio — and its savings-runway mode computes how many years a balance lasts at a chosen annual spending and real return. On the accumulation side, the Retirement Calculator projects what a nest egg a savings pattern builds and what an inflation-adjusted drawdown from it looks like, and the Compound Interest Calculator isolates the growth math itself. None of these change what the 4% rule is — a description of what survived in historical data — but they make its moving parts visible one input at a time.
Frequently asked questions
Does the 4% rule mean withdrawing 4% of the balance every year?
No — that is the most common misreading. The convention sets the first year’s withdrawal at 4% of the starting balance ($40,000 from a $1,000,000 portfolio), then adjusts that dollar amount for inflation each year regardless of what the balance does afterward. Withdrawing 4% of each year’s current balance is a different strategy with different behavior: it can never fully empty the account, but the income swings with the market.
Is the 4% rule a guarantee that money lasts 30 years?
No. It is a summary of a historical back-test: in the U.S. market data William Bengen studied in 1994, and in the broader grid the 1998 Trinity study examined, an inflation-adjusted 4% initial withdrawal survived every rolling 30-year period in the data. That is a statement about the past sequences tested, under a specific stock-and-bond allocation. Future return sequences, different allocations, and longer retirements all sit outside what the studies measured.
What is sequence-of-returns risk?
The risk that poor returns arrive early in retirement, while withdrawals are being taken. Two portfolios can earn the same average return over 30 years, yet the one that hits a deep downturn in the first few years — selling assets at depressed prices to fund the inflation-adjusted withdrawal — can deplete while the other finishes with a surplus. It is the standard first criticism of any fixed-withdrawal convention, because an average return hides the order in which returns arrive.
Does the 4% rule account for taxes?
The studies model portfolio survival, not tax. What a withdrawal costs in tax depends on the account it comes from: withdrawals from traditional 401(k) and IRA balances are generally taxed as ordinary income, qualified Roth withdrawals are tax-free, and a taxable brokerage account owes tax on dividends and realized gains in the year they occur. The mechanics are laid out in the retirement accounts guide on this site; irs.gov is the authority on the current rules.
What about retirements longer than 30 years?
The 30-year horizon is baked into the headline result — Bengen’s paper and the Trinity study measured survival over roughly 30-year payout periods. A retirement meaningfully longer than 30 years asks the same portfolio to fund more inflation-adjusted withdrawals through more market cycles, which is a harder test than the one the studies ran. The research literature treats longer horizons as a separate question, not a case the 4% figure was shown to cover.
Not financial advice: this page describes the mechanics and boundaries of a published research heuristic — it does not recommend a withdrawal rate, an allocation, or any strategy. The historical findings belong to Bengen (1994) and the Trinity study (1998); account and tax rules are summarized as of September 2026 and change over time, with irs.gov and ssa.gov as the authoritative sources. See the methodology page for how this site handles figures.