Almost every retirement conversation lands on one number: 4%. Save 25 times your annual spending, withdraw 4% in year one, raise that dollar amount with inflation, and history says you probably won't run out of money over a 30-year retirement.
That's a useful starting point and a terrible stopping point. The 4% rule was never a law or a guarantee — it answered one narrowly defined historical question. This guide covers what the research found, what its author says now, today's figures, and how to turn a static rule into a workable spending plan.
Where the number came from
The rule traces to financial planner William Bengen's paper "Determining Withdrawal Rates Using Historical Data," published in the Journal of Financial Planning in October 1994. Bengen asked a narrow question: across every 30-year window in US market history, what is the highest first-year withdrawal rate that never would have exhausted a portfolio? He called the answer SAFEMAX. For a 50/50 stock/bond portfolio it came to roughly 4.15%, which the industry rounded to 4%.
A 1998 study by Cooley, Hubbard and Walz at Trinity University — the Trinity study, published in the American Association of Individual Investors' journal — reframed the same data as portfolio success rates across rolling periods from 1926 to 1995. For a 4% inflation-adjusted withdrawal over 30 years from a balanced stock/bond portfolio, secondary summaries most often cite success around 95%.
Two things matter about that lineage. 4% was a worst-case figure, not a typical one; in most historical windows a retiree could have spent considerably more. And both studies excluded taxes, fees, and Social Security — a gross withdrawal from a frictionless portfolio, which is not a thing anyone owns.
How the rule works, and the mistake everyone makes
The most common misreading is treating 4% as an annual percentage of the current balance. It isn't:
- In year one, withdraw 4% of your starting portfolio. On $1,000,000, that's $40,000.
- Every year after, raise that dollar amount by inflation — don't recalculate it from the new balance.
If inflation runs 3%, year two is $41,200 whether the portfolio rose, fell, or went sideways. That rigidity is the point of the original test, and it's also the rule's biggest practical weakness. A true percentage-of-balance withdrawal can never run out of money, but it hands you a pay cut every time the market falls.
The inverse of the rule is the savings target: a 4% starting rate implies 25 times your first year of portfolio-funded spending. Portfolio-funded is the operative word. If Social Security will cover $30,000 and you need $70,000, the portfolio only has to produce $40,000 — a $1,000,000 target, not $1,750,000. That math needs an honest picture of which assets are actually drawable, which is what the Net Worth Calculator is for; home equity sits on the balance sheet but doesn't buy groceries. To work the accumulation side — whether what you've saved already grows into that target on its own — the Coast FIRE Calculator runs the same multiple backwards.
What the research says in 2026
The figure has moved in both directions since 1994, and the two most prominent current estimates disagree for instructive reasons.
Bengen revised his own number up. In his 2025 book A Richer Retirement, he raised SAFEMAX from about 4.1% to 4.7%, crediting diversification across more asset classes than his original two-asset portfolio. He now also weighs market valuation (the Shiller PE) and expected inflation, and frames 4.7% as a worst case rather than an expectation. He has called inflation, not a crash, the retiree's "greatest enemy" — though as CNBC noted, a figure derived from past performance is no promise about future retirees.
Morningstar's forward-looking estimate is lower. Its annual State of Retirement Income research puts the safe starting rate for 2026 retirees at 3.9%, up from 3.7% a year earlier. That assumes a 30-year retirement, inflation-adjusted withdrawals, roughly 30–50% equities, and a 90% probability of not running out of money. Horizon moves it sharply: at 50% equities, the safe rate falls to about 3.5% over 35 years and 3.2% over 40 years.
The gap isn't a contradiction. Bengen asks what would have survived the worst of the past; Morningstar asks what survives 90% of modeled futures given today's valuations and yields. The honest read is that a defensible starting rate for a 30-year retirement sits somewhere in the high 3s to mid 4s — and lower if you're retiring at 50 rather than 65.
Sequence-of-returns risk, what the rule is really about
Why not just withdraw your portfolio's average expected return? Because averages arrive in an order, and order matters once you're selling rather than buying.
While you're working, a bear market is mostly a paper loss and your contributions buy cheap shares. Retirement inverts that. Every withdrawal during a downturn converts a temporary decline into a permanent one — the shares you sold at the bottom aren't there for the recovery.
The danger concentrates in the first decade, for an arithmetic reason: that's when the portfolio is largest relative to withdrawals. A 20% drop on $1,000,000 costs $200,000; the same drop later, on $400,000, costs $80,000. Research from Michael Kitces narrows it further: what drives the sustainable rate is the real return over roughly the first ten years.
The standard defenses all reduce forced selling into weakness: hold one to three years of withdrawals in cash or short bonds, keep guaranteed income high (delaying Social Security raises the inflation-adjusted floor), and stay willing to trim spending in bad years. Modeling the drawdown with your real horizon and allocation is what the Retirement Planning Calculator is for.
Flexible strategies beat a fixed rate
Because rigid spending is what forces the starting rate down, relaxing it buys back real room. Morningstar's own work finds flexible withdrawal methods can lift a safe starting rate to as much as 5.7% — a far bigger improvement than any plausible change to asset allocation.
The best-known framework is the guardrails approach from Jonathan Guyton and William Klinger (2006). Set a starting rate, then react to drift: if strong markets push your current withdrawal rate 20% below where it started, give yourself a 10% raise; if weak markets push it 20% above, take a 10% cut. A companion rule skips the inflation adjustment after a year of negative portfolio returns. Kitces' critique is worth weighing — the cuts can be sharper than some retirees could absorb — so the cushion only counts if it's genuinely discretionary.
There's also evidence that constant inflation-adjusted spending overstates what retirees need. David Blanchett's research on the "retirement spending smile" found real spending tends to drift down through retirement: in Kitces' summary of the data, roughly 1% a year in the first decade, about 2% a year in the second, and about 1% a year in the third, as discretionary travel and dining fall faster than health costs rise. Planning for flat real spending across 30 years is conservative, and conservatism has a price — you work longer than you needed to.
Frequently asked questions
Is the 4% rule still safe?
For a 30-year retirement it remains a reasonable planning anchor, but it isn't a guarantee and never was. Morningstar's forward-looking estimate for 2026 retirees is 3.9%; Bengen's updated historical worst case is 4.7%. Treat the high 3s to mid 4s as a starting range, and the first decade's real returns as the thing that will actually decide it.
Does the 4% rule account for taxes?
No — the original research excluded taxes and fees entirely. A $40,000 withdrawal from a traditional 401(k) or IRA is taxable as ordinary income, so your spendable amount is less; Roth withdrawals are generally tax-free. Fees matter too: one analysis in the Journal of Financial Planning estimated that trailing the index by one percentage point lowers the safe rate by roughly half a point.
Can I withdraw 4% a year forever?
Not reliably. The rule was tested against 30-year retirements; at 50% equities, Morningstar's safe rate drops to about 3.5% for 35 years and 3.2% for 40. Early retirees facing 40-plus years should plan on a lower starting rate, more flexibility, or both.
How much do I need to retire using this rule?
Divide the spending your portfolio must cover — total spending minus Social Security, pensions, and other guaranteed income — by your rate. At 4% that's 25 times; at 3.5%, about 29. Run it on income you must self-fund, or you'll overshoot.
Disclaimer
This guide is for general educational purposes only and does not constitute personalized financial, tax, or legal advice. Every reader's situation is different — consult a licensed financial advisor, accountant, or attorney before making decisions based on this content. Figures and rules cited here reflect the most recent information available at time of publication and may change; verify current limits and regulations before acting.