Safe Withdrawal Rate in 2026: Fixed Rules, Guardrails, and Which to Choose

The question sounds deceptively simple: what is a safe withdrawal rate? The frustrating honest answer is that three credible researchers will give you three different numbers — and all of them are correct.

Morningstar's 2026 State of Retirement Income report puts the baseline at 3.9% for a fixed spending strategy. William Bengen, who built the 4% rule in 1994, has since revised his own SAFEMAX estimate up to 4.7% using a broader asset class mix. And retirees using a flexible "guardrails" approach can legitimately start at 5.7%. That is not a disagreement about facts. It is a consequence of answering different questions.

What a withdrawal rate actually is

A withdrawal rate is the percentage of your starting portfolio you spend in year one, with future years stepped up for inflation. On a $1 million portfolio, a 3.9% rate means $39,000 in year one — about $3,250 per month before taxes. In year two, at 3% inflation, you'd spend roughly $40,170.

The research question underneath every withdrawal rate study: given this percentage, this time horizon, and this stock/bond allocation, what fraction of all historical 30-year market sequences would have left you with money remaining? "Safe" typically means surviving 90% or more of them — including 1929, 1966, and every other brutal starting decade on record.

The time horizon is where many plans go wrong. Bengen's 4% rule was stress-tested against 30-year retirements. Retire at 55 with a 40-year horizon and you are implicitly using a study calibrated for a shorter problem. The 4% number survives because it was tested on 30 years; whether it survives 40 depends on data that simply doesn't exist in sufficient quantity.

The 2026 baseline: 3.9% fixed

Morningstar's 2026 research used a 30-year horizon, roughly 50% equities and 50% bonds, and a 90% success threshold. Their result: 3.9%. That is up from 3.7% in 2023, primarily because higher bond yields over the past two years increased the expected return on the fixed-income portion of the portfolio. When bonds pay more, portfolios don't need equities to carry as much of the load, and the math gets modestly friendlier.

On $1 million, 3.9% produces $39,000 per year. Add a typical Social Security benefit — the average retired-worker benefit runs around $2,000 per month in 2026 — and household income climbs to roughly $63,000 before taxes. A paid-off house in most of the country is workable on that number.

The practical gap between 3.9% and the classic 4% is $1,000 per year on a $1 million portfolio. Whether that difference ever materializes into a real shortfall depends less on the rounding and more on which years the market decides to be bad. Bengen's revised 4.7% reflects what is achievable with a small-cap value tilt in the equity allocation — a legitimate adjustment, though one that adds complexity and tracking error most retirees won't bother with.

The problem fixed rates cannot solve

Fixed withdrawal rules treat retirement like an engineering problem: solve for the rate, plug it in, and the plan runs itself. Real retirements do not cooperate because the sequence of returns — the specific order in which good and bad years arrive — matters as much as the average.

Consider two hypothetical retirees, both starting with $1 million and a 4% initial rate. Both experience the same 7% average annual return over 30 years. The one who gets the bad years early runs out of money; the one who gets them late ends with a substantial estate. Same average. Opposite outcomes. Morningstar estimates that poor returns concentrated in the first five years of retirement account for 70% of retirement plan failures.

A fixed withdrawal rate does not respond to this. It prescribes the same $40,000 withdrawal whether the portfolio is sitting at $800,000 or $1.2 million in year two. The rule is simple; the sequence risk it ignores is real.

Guardrails: the systematic response to sequence risk

The Guyton-Klinger guardrails strategy — developed in a 2006 FPA Journal paper and still the most extensively studied flexible approach — makes the withdrawal rate dynamic instead of fixed. The mechanics:

1. Set a higher initial withdrawal rate. The research supports starting at 5.2%–5.6% with at least 65% equities — materially more than fixed alternatives.
2. Define the guardrails. If your current effective withdrawal rate rises more than 20% above the initial rate (bad markets shrink the portfolio), cut spending by 10%. If it falls more than 20% below (good markets grow the portfolio), take a 10% raise.
3. Between guardrail triggers, adjust only for inflation. No changes unless a guardrail is crossed.

On $1 million with a 5.2% initial rate, year one income is $52,000. If markets fall hard and the portfolio drops to $700,000 by year two, your effective rate has risen to roughly 7.4% — well beyond the 20% upper trigger — so you'd reduce to $46,800. That cut is the price of the higher starting rate.

Morningstar's 2026 research estimates that a well-implemented flexible strategy supports an initial rate of 5.7% at the same 90% success threshold as 3.9% fixed. The higher rate is not free: it comes with the obligation to actually reduce spending when the guardrail fires. Retirees who say they will cut but don't are not running the guardrails strategy; they're running a 5.7% fixed strategy, which is not safe by any research standard.

Which framework fits which situation

Situation	Better fit
Fixed obligations (mortgage, recurring medical costs)	Fixed rate — 3.9% to 4%
Genuine spending flexibility (can cut 10–15% in a bad year)	Guardrails — 5% to 5.7%
Retirement horizon under 25 years	Fixed closer to 4.5% is defensible
Retirement horizon 35–40 years (retiring at 55 or earlier)	Conservative fixed (3.5%–3.9%) or guardrails
Social Security covers 60%+ of annual spending	Higher portfolio rate is fine — SS is sequence-immune
Portfolio is mostly Roth, no RMDs to manage	Flexible guardrails work cleanly; no forced distributions distort the math

The Social Security point is underappreciated. A hypothetical couple receiving $80,000 per year in combined Social Security income and spending $90,000 per year needs their portfolio to produce only $10,000. Running a 5% withdrawal rate on a $200,000 portfolio for that gap is not a plan at risk — the analysis that matters for them is Social Security timing, not the withdrawal-rate debate. Guardrails vs. fixed is most consequential when your portfolio is doing most of the heavy lifting.

What the calculator can tell you — and what it can't

The framework above is a starting point, not a complete plan. Your actual sustainability depends on specifics a table cannot capture: the tax treatment of your withdrawals (a 4% rate on a $1 million Roth balance versus a $1 million traditional IRA produces very different after-tax income), the RMD surge that begins at 73 and forces taxable income upward whether you need the money or not, healthcare costs before Medicare at 65, and whether your Social Security timeline has been optimized.

The when can I retire calculator lets you enter your actual account types, planned annual spending, and retirement date and runs the full projection with taxes — including the RMD compulsion at 73 that consistently surprises people who planned only for voluntary withdrawals. For a direct read on portfolio longevity at different spending levels, the how long will $1 million last calculator plots depletion curves at 3%, 4%, and 5% rates side by side, with Social Security layered in.

The withdrawal rate is the right starting question. But the number worth optimizing is the probability that your specific plan — with your account mix, your tax situation, your Social Security timing — survives the full run. A 3.9% rate sounds conservative; 3.9% with 100% equities and a retirement date of January 2000 was still painful. The rate and the strategy are both only as good as the plan they sit inside.

For the full picture — accounts, taxes, Social Security timing, ACA premiums, RMD projections, and Monte Carlo simulation against your actual spending — Granary puts all of it in one place.

This article is educational content about retirement planning concepts, not personalized tax or investment advice.