Safe Withdrawal Rate for Retirement 2026: Is 4% Still Safe?

The safe withdrawal rate for a 2026 retiree is 3.9% for a 30-year plan with fixed, inflation-adjusted spending and a 90% chance of money left at the end, according to Morningstar's 2026 research. Bill Bengen, who created the 4% rule, now puts his historical SAFEMAX at 4.7% (Firstlinks). Retirements longer than 30 years need a lower rate, and retirees who will cut spending after a bad year can start higher.

At 4.0%, the classic rule falls between the 3.9% and 4.7% figures for a 30-year retirement. Morningstar runs forward-looking simulations on its current return assumptions, with fixed real withdrawals and a portfolio of 30% to 50% stocks, and asks what works in 90% of futures, while Bengen replays U.S. market history and asks what survived the worst starting year on record, using a portfolio of 55% stocks, 40% bonds and 5% cash. On $1 million that is $39,000 a year versus $47,000, and Bengen himself told Money.com he would "probably recommend something around 5.25% to 5.5%" for today's retirees.

2026 safe withdrawal rates on a $1 million portfolio

Every row uses the same $1,000,000 portfolio and shows the year-one withdrawal before tax, which then rises with inflation.

The flexible rows come from the same Morningstar State of Retirement Income report as the 3.9% base case, and they assume your spending moves with the market. The UK figure is from Morningstar's UK edition, useful if you draw down in pounds.

Sarah retires at 65 with exactly $1,000,000. At Morningstar's 3.9% she draws $39,000 in year one, about $3,250 a month. At Bengen's 4.7% SAFEMAX she draws $47,000, or $3,917 a month. The difference is $8,000 a year, or $16,000 if she takes Bengen's 5.5% suggestion. If you want to see how long a given balance survives at a given draw, the free retirement money calculator plots the depletion path year by year, no sign-up.

Data: Morningstar, The State of Retirement Income (Dec 2025); Bengen via Money.com and Firstlinks (Aug 2025). Year-one withdrawal on $1M, before tax.

Where did the 4% rule come from?

Bengen's withdrawal study appeared in the Journal of Financial Planning in October 1994. He took every 30-year period in U.S. market history and asked how much a retiree could withdraw in year one, then raise with inflation every year, without running out even when retirement began in the worst year on record. The worst start year was 1968, and he called the result the SAFEMAX, the maximum historically safe starting rate (Wikipedia on Bengen). His original figure was 4.15%, which the industry rounded to 4%.

From 4% comes the 25x rule: divide 1 by 0.04 and you need 25 times your annual spending invested. Spending $40,000 a year means a $1,000,000 target. The full arithmetic, with other rates and examples, sits in how to calculate your FIRE number.

Bengen meant 4% as the worst-case number, the rate that survived even the 1968 start. Across all the start years he tested, the average safe rate was closer to 7%. His 2025 book raised the floor itself from 4.15% to 4.7% after adding small-, mid- and micro-cap and international stocks to the original large-cap and government-bond mix (Boldin summary).

Why do Morningstar and Bengen disagree?

Bengen's number is what the worst actual sequence in U.S. history allowed, and with a broader portfolio that worst case is now 4.7%. A retiree who does not start in a 1968-style year could, on his history, have spent more.

Morningstar simulates thousands of futures from its own capital-market assumptions for the coming decades and reports the highest fixed real withdrawal that still leaves money in 90% of them. Its base case has moved with those assumptions: 3.3% for 2021 retirees, 3.8% for 2022, 4.0% for 2023, 3.7% for 2025 and 3.9% for 2026 (Morningstar).

The Shiller CAPE ratio, the S&P 500's price against ten years of inflation-adjusted earnings, stood at 41.04 on 28 October 2026 against a long-run mean of 17.44, and high starting valuations like that are one reason a forward-looking model comes in under history. Bengen's own 2026 scenarios for longer horizons explicitly assume today's high valuations (WTOP / U.S. News).

Around 2020, in what he described as a low interest-rate environment, Wade Pfau told Morningstar's The Long View podcast that 3% was "a lot more realistic than 4%", and that under lower projected returns a rigid 4% would work "more like 60% to 70% of the time".

Safe withdrawal rate for a 30-year retirement in 2026

For a 30-year retirement starting in 2026, such as 65 to 95, the published rates run from 3.9% to 4.7%. Morningstar's forward-looking answer is 3.9%, and its highest rates come from portfolios holding 30% to 50% stocks. Bengen's historical worst case is 4.7% on a 55/40/5 mix. The 1998 Trinity study by three Trinity University professors (Cooley, Hubbard and Walz) found the plain 4% rule lasting 30 years in 95% of historical periods on a 50/50 stock-bond portfolio and 98% on 75/25. A UK retiree gets 4.1% from Morningstar's UK model.

A 30-year retiree who wants a fixed, inflation-adjusted income and accepts no pay cuts can treat 3.9% to 4% as the safe starting rate, which is $39,000 to $40,000 a year per $1 million. A retiree who trusts the historical record and holds a diversified portfolio can start at 4.7%. Starting above 4.7% depends on cutting spending in bad years, which the guardrails section below covers.

Morningstar published the 3.9% figure for 2026 retirees in December 2025 and usually publishes the next edition in early December, so check for a 2027 figure if you retire next year.

Longer retirements need a lower starting rate

The 4% rule was built for 30 years. Someone leaving work at 45 and planning to 90 is funding 45 years, and every extra decade lowers the safe starting rate and raises the multiple of spending you need.

In the Morningstar report, stretching the horizon from 30 to 35 years cuts the rate from 3.9% to 3.5%. The 40-year 3.2% comes from Morningstar's separate study, How Much Can You Safely Withdraw If You Retire Early?. The historical rows are collected by Michael Kitces, who notes that going from 30 to 45 years took Bengen's rate from about 4.1% to 3.5%, and that 3.5% "effectively forms a safe withdrawal rate floor" in the historical data.

For 50 years the research is thin, and Bengen's 2026 work says moving from a 30- to a 50-year horizon "requires a cut in safemax of about 13%" (WTOP / U.S. News). In practice a 50-year plan starts below every 30-year figure above, and the 3.2% to 3.5% range for 40 to 45 years is the nearest published guide.

Data: Morningstar, The State of Retirement Income (Dec 2025) and early-retirement study; historical rates from Bengen (1994, 1996) via Kitces.

Long horizons also leave more years for sequence-of-returns risk, the damage bad returns do early in retirement. Withdrawing $40,000 from a portfolio that has fallen from $1,000,000 to $600,000 is a 6.67% draw, and selling shares at those prices in the first few years breaks plans that look fine on average returns. The cash-bucket and bond-tent defences are in the sequence of returns risk playbook. If you are still saving and choosing between a 25x and a 31x target, the Coast FIRE calculator shows the balance you need today for compounding alone to reach that target by your retirement age.

Guardrails and flexible withdrawal strategies

The rigid 4% rule withdraws the same inflation-adjusted amount whatever the market does. Guardrails tie spending to the portfolio's value, and because you commit in advance to cutting after bad years, you can start higher.

The best-known version is Guyton and Klinger's 2006 decision rules, published in the Journal of Financial Planning, shown here on a 5% starting rate.

Dave starts at 5% on $1,000,000, a $50,000 withdrawal. Markets fall and $50,000 becomes more than 6% of what is left, so he cuts to $45,000. Years later a bull market lifts his balance until the withdrawal is under 4% of it, and he raises it 10%, to $49,500 (inflation raises left out to keep the arithmetic plain).

The paper's own maximum starting rates, for a 40-year retirement with at least 65% in stocks, were 5.2% to 5.6% at 99% confidence and 5.7% to 6.2% at 95%. A 50% stock portfolio dropped the rate toward 4.6%. In Morningstar's 2026 report, the guardrails strategy starts at 5.2% on a 40/60 portfolio and the most flexible methods it tested reach 5.7%.

Fitzpatrick and Tharp, writing on Kitces.com, found that Guyton-Klinger retirees who started in October 1965 paid for the higher start with real spending 54% below plan, and 45% for a 1936 start. Karsten Jeske's 1966 example bottomed at $16,400 a year on a $40,000 plan, a 59% cut. Pfau found that at a 5.3% start there was a 10% chance of a 48% cut within ten years, and January 2000 retirees starting at 4%, 5% or 6% all faced cumulative cuts above 50%.

Guardrails work best when Social Security, a pension or an annuity already covers essential costs, so a 40% or 50% cut lands on travel and gifts rather than rent and food. Holding one to two years of spending in cash or short bonds also means a cut year does not force stock sales at the bottom.

Is the 4% rule too cautious?

The 4% rule protects against running out, and in most historical periods it underspent, ending with a large surplus. Kitces found that with a 4% start on a rebalanced 60/40 portfolio, across rolling U.S. 30-year periods since the 1870s, "the median wealth at the end of 30 years is almost 2.8X principal". For retirements starting since the 1920s it was 3.7x.

Yet a MetLife survey from February 2026 found 51% of retirees who still have money in their workplace plan worry about running out, up from 30% less than a decade earlier.

David Blanchett's retirement spending smile shows real spending dropping about 1% a year early in retirement, closer to 2% a year in the slow-go years, then easing back toward 1%, so holding real spending flat for 30 years, as the 4% rule does, overstates what many households spend. The trade-off between that safety margin and money actually spent is the subject of the Die With Zero vs FIRE debate.

How to pick your own withdrawal rate

1. Retirement length. 30 years: start at 3.9% to 4% (25x). 35 to 45 years: 3.2% to 3.5% (about 29x to 31x). Beyond that, stay near the 3.5% historical floor or below.

2. Portfolio mix. Morningstar's best fixed-spending results came from 30% to 50% stocks; Bengen's 4.7% assumes 55% stocks, 40% bonds and 5% cash, with small-cap and international holdings. A portfolio of mostly cash, or one concentrated in a single stock, supports less than either table.

3. Guaranteed income. If Social Security, a pension or an annuity covers housing, food, insurance and healthcare, the portfolio only funds the discretionary gap, and that part can run at a higher rate.

4. Spending flexibility. If you could live through a cut of 40% or more in a bad decade, guardrails justify a start of 5% or so. If your budget has no slack, stay at 3.9% to 4%.

5. Taxes and fees. Every table above shows rates before tax and fees. As an example, if 20% of each withdrawal goes to tax, a 4% draw from a pre-tax account leaves 3.2% to spend, and a 1% annual advisory fee on a 4% plan leaves roughly 3% for you.

A 67-year-old with a pension covering essentials, a balanced portfolio and a flexible budget can reasonably run 5% with guardrails. A 45-year-old with no pension, a 45-year horizon and a tight budget belongs near 3.2% to 3.5%.

Turning the rate into a monthly paycheck

Year-one monthly income is the starting rate times the portfolio, divided by 12.

Jake has $1,200,000 and picks a 5% start with guardrails. Year one is $60,000, or $5,000 a month. With U.S. inflation at 3.7% over the 12 months to September 2026, a fixed-real plan raises year two to $62,220, about $5,185 a month, whatever the market did.

A guardrail plan recalculates against the actual balance every January. If Jake's portfolio has grown to $1,400,000, 5% of it is $70,000; if it has dropped to $1,000,000, 5% is $50,000, and the guardrail table above tells him whether a cut applies. Either way he needs an up-to-date balance once a year, and a rate review every two or three years as valuations, yields and inflation move.

Data: annual spending divided by withdrawal rate; 3.3% is Pfau's 2012 40-year rate at 95% confidence, 3.9% Morningstar's 2026 base case, 4.7% Bengen's SAFEMAX, 5.5% the top of Bengen's 2025 suggestion.

Starting from the spending you want, the same chart gives how much money you need to retire at each rate. To set your own number, put your spending and chosen rate into the free FIRE planner to see the portfolio you need and the year you reach it, then test that balance against your planned withdrawals in the how long will my money last calculator.