The 4% rule is dead — here’s what replaced it in 2026

1. The retirement rule everyone learned was wrong

For thirty years, almost every introduction to retirement income planning used the same number: 4%. The rule, developed by financial planner Bill Bengen in a 1994 article in the Journal of Financial Planning, was simple. Withdraw 4% of your retirement portfolio in year one. Each subsequent year, increase the dollar amount by the prior year’s inflation rate. Over a 30-year retirement, even in the worst historical market scenarios, your money would last.

The rule was elegant. It was easy to remember. It became the foundation of nearly every retirement planning conversation in the industry. And in 2025, the man who invented it published a book saying the number should be higher.

Bengen’s “A Richer Retirement: Supercharging the 4% Rule to Spend More and Enjoy More” — released in 2025 and the basis of his 2026 media tour — revised the safe withdrawal rate from 4% to 4.7%. The change isn’t subtle. On a $1 million portfolio, the difference between 4% and 4.7% is the difference between $40,000 and $47,000 in your first year of retirement — and that 17.5% increase compounds across every subsequent year.

But the 4.7% revision isn’t where the story ends. Just months earlier, in February 2026, Morningstar’s “State of Retirement Income” research arrived at a different conclusion: 3.9% is the baseline safe withdrawal rate for a balanced portfolio over a 30-year retirement. Morningstar’s number is LOWER than the original 4% rule, not higher.

How can the man who invented the rule say 4.7% while one of the largest retirement-research institutions says 3.9%? Both numbers can be right because they’re answering different questions. Bengen’s 4.7% is the worst-case historical floor. Morningstar’s 3.9% is a forward-looking estimate that incorporates current market valuations and bond yields. They’re not actually contradicting each other — they’re measuring different things.

What this means for anyone planning retirement income in 2026: the single 4% number that dominated retirement conversations for 30 years is no longer the right answer. The actual safe withdrawal rate for your specific retirement depends on factors that the original rule didn’t capture. This article walks through what changed, why both Bengen and Morningstar are correct, and how to think about your own withdrawal rate without relying on a single number.

2. What Bengen actually changed in 2025

Bengen’s 2025 revision wasn’t a dramatic reversal of his earlier work. It was an upgrade based on better data and a more diversified portfolio. Three specific changes drove the move from 4% to 4.7%:

1. More asset classes in the portfolio. The original 1994 research used a simple 50/50 split between US large-cap stocks and intermediate-term US Treasury bonds. The 2025 revision uses a diversified equity portfolio split across five categories: large-cap, mid-cap, small-cap, micro-cap, and international equities. The additional diversification — particularly the small-cap and international exposure — improved the historical worst-case outcome enough to support the higher withdrawal rate.

2. More recent data. The original analysis used market data through the early 1990s. The 2025 analysis includes another 30+ years of market history, including the dot-com bust, the 2008 financial crisis, the 2020 pandemic crash, and the subsequent recoveries. The expanded data set provided more stress-test scenarios — and the diversified portfolio survived all of them at 4.7%.

3. Refined definition of “safe.” Bengen now calls 4.7% the “Universal SAFEMAX” — the worst-case historical maximum that would have worked for every starting date in the analyzed period. The number applies specifically to a retiree who began retirement in the absolute worst market environment in US history. For more typical starting environments, Bengen’s research shows higher rates are safe:

Bengen’s own framing is telling. In a 2026 interview, he said that retirees who rigidly stick to 4% (or even 4.7%) when conditions don’t warrant the worst-case assumption are “cheating themselves a little bit.” The point isn’t that 4.7% is the new universal answer — it’s that 4.7% is the universal SAFE floor, and most retirees can safely withdraw more.

The math impact, for a $1 million portfolio:

• 4% (original rule): $40,000 in year one
• 4.7% (Bengen 2025): $47,000 in year one — +$7,000/year, +$175,000 over 25 years (before COLA)
• 5.25% (typical case): $52,500 in year one — +$12,500/year, +$312,500 over 25 years
• 6% (favorable case): $60,000 in year one — +$20,000/year, +$500,000 over 25 years

The dollar differences are substantial. Choosing 4% when 5.25% would have been safe means accepting $312,500 less in lifetime spending — voluntarily, in exchange for a level of caution the math doesn’t require.

3. Why Morningstar disagrees — and how both can be right

Morningstar’s 2026 “State of Retirement Income” report, published in February 2026 by researchers Amy Arnott, Christine Benz, and Jason Kephart, reached a different conclusion: 3.9% is the baseline safe withdrawal rate for retirees using a balanced portfolio with a fixed spending strategy over 30 years.

That’s a slight increase from Morningstar’s 2024 number of 3.7% (improved capital-markets assumptions produced the small upward revision). But it’s lower than both the original 4% rule AND Bengen’s revised 4.7%. The apparent contradiction dissolves when you understand the methodology difference:

Bengen’s approach: historical worst case. Bengen analyzes what actually happened in past retirement scenarios — using historical market returns from 1926 to present — and identifies the worst single starting year (October 1968). His 4.7% is the rate that would have worked even for that worst-case retiree.

Morningstar’s approach: forward-looking expected case. Morningstar uses forward-looking return forecasts for current market conditions. With US equity valuations historically elevated and bond yields at moderate levels, Morningstar’s models project lower expected returns going forward than historical averages. The 3.9% reflects what the math says is safe given those forward assumptions.

Both approaches are legitimate. Both are answering useful questions. They just answer them differently:

The Morningstar 5.7% rate — for retirees willing to adjust spending based on market conditions — is the most actionable single number in recent retirement-income research. It tells us that retirees who can flex their withdrawals (spending less in bad markets, more in good ones) can safely START at significantly higher rates than the 4% or 3.9% conservative baselines. This is the real shift in 2026 retirement planning: the conversation is moving from “find the right fixed rate” to “use a flexible approach that adjusts to actual conditions.”

Bengen’s 4.7% is the worst-case historical floor. Morningstar’s 3.9% is a forward-looking conservative baseline. Both are correct — they answer different questions. The single 4% rule that dominated retirement planning for 30 years was an oversimplification of what was always a more nuanced calculation.

4. The 10 variables that actually determine your safe rate

Bengen’s revised research identifies 10 variables that determine the right withdrawal rate for any specific retiree. Eight are within the retiree’s control. Two are not. Understanding all ten reveals why a single “rule” was always going to be inadequate.

Variables within your control:

1. Withdrawal scheme. Fixed-dollar (4% rule style), fixed-percentage (constant % of current balance), guardrails (adjust based on conditions), or income-only (live on dividends and interest). Different schemes produce different sustainable rates.
2. Asset allocation. Stock-bond split, diversification across asset classes, international exposure. More diversification typically supports higher withdrawal rates.
3. Retirement time horizon. A 30-year retirement is the standard assumption. Early retirees planning 40-50 years need lower rates; late retirees planning 20-year horizons can take higher rates.
4. Legacy goals. Withdrawing to deplete the portfolio at end of life supports higher rates than withdrawing while preserving capital for heirs.
5. Tax efficiency. Optimal withdrawal sequencing (taxable accounts first, Roth last) extends portfolio longevity by 5-10 years compared to suboptimal sequencing.
6. Spending flexibility. Retirees willing to cut spending during bad markets can sustain higher initial withdrawal rates than retirees with rigid spending needs.
7. Investment costs. Every 1% in annual fees reduces the safe withdrawal rate by roughly 0.5-1%. Index funds and low-cost approaches preserve significantly more income than actively-managed alternatives.
8. Other income sources. Social Security, pensions, annuities, and rental income create a guaranteed-income floor that lets the portfolio absorb higher withdrawal rates without risk of depletion.

Variables NOT in your control:

9. Stock market valuation at retirement. Higher starting valuations correlate with lower expected returns going forward. Retirees who start in high-valuation environments (like 2024-2026) face lower expected returns than retirees who started in low-valuation environments (like 2009-2010).
10. Inflation rate during retirement. Bengen identifies inflation as “the greatest threat to all retirees” — high inflation erodes purchasing power faster than nominal returns can replenish it. The retirees who triggered the original 4% worst-case (October 1968) faced a decade-plus of high inflation following retirement.

The combined effect: a retiree with optimal control of the 8 controllable variables, retiring into average-to-favorable market conditions, can safely withdraw substantially more than 4.7%. A retiree with suboptimal control of the variables, retiring into high-valuation/high-inflation conditions, faces a real risk that even 4% may be too aggressive. The single number was always going to be wrong for most people in one direction or the other.

5. The “guardrails” approach that’s replacing the rule entirely

The most significant development in 2026 retirement-income research isn’t a new number — it’s a new framework. The guardrails approach (formally the Guyton-Klinger rules, developed by financial planners Jonathan Guyton and William Klinger) replaces the rigid “fixed withdrawal rate” with dynamic adjustments based on portfolio performance.

The basic structure:

1. Start at a higher initial rate than the conservative fixed-spending baseline. For Morningstar’s 2026 analysis, this can be as high as 5.7% (versus 3.9% for fixed-spending).
2. Set upper and lower guardrails. Typically, the upper guardrail is 20% above the initial rate as a percentage of CURRENT portfolio value, and the lower guardrail is 20% below.
3. Adjust spending when guardrails are hit. If portfolio growth pushes the current withdrawal rate below the lower guardrail (you’re being too conservative), increase spending. If market losses push it above the upper guardrail (you’re depleting too fast), cut spending.
4. Other refinement rules. Skip inflation adjustments in bad years. Take inflation-plus increases in great years. Adjust for excessive longevity.

The math implication: a retiree using guardrails can safely start at substantially higher rates than the fixed-spending baseline — because they have a built-in mechanism to course-correct if conditions deteriorate. The flexibility itself produces the higher safe initial rate. For most retirees, this is the most actionable change from the 4% rule discussion. Instead of asking “what’s my safe fixed rate?”, the right question is “what’s my safe initial rate given that I can adjust if conditions change?” The answer is consistently higher than the fixed-rate baseline — typically 5-6% rather than 3.9-4.7%.

The guardrails approach has practical limitations. It requires retirees to actually monitor portfolio performance against the guardrails. It requires willingness to cut spending in bad market years (which retirees often resist emotionally). And it requires having flexibility in the household budget — retirees with fixed costs they can’t easily reduce can’t fully implement the approach. For retirees with substantial flexibility in their spending (discretionary travel, dining, gift-giving), guardrails produce dramatically better outcomes than fixed withdrawals.

6. Worked examples at $500K, $1M, and $1.46M portfolios

The practical impact of the different withdrawal rates depends on your portfolio size. Here’s the math at three common portfolio levels:

Reading the table: the choice between withdrawal-rate frameworks produces substantial dollar-difference outcomes. For a $1 million portfolio, choosing 4% versus 5.25% means $12,500/year less spending — every year, for 25-30 years. That’s a $312,500-$375,000 lifetime spending difference for a household whose only “mistake” was being overly conservative.

The flip side. The higher rates also assume the underlying conditions support them. A retiree using 5.7% (guardrails) in a falling market environment may need to cut to 4.5% temporarily, then resume higher rates when markets recover. The flexibility produces the higher safe rate, but it also requires the flexibility to be executed.

One important reality check. Most American retirees have nowhere near the $1.46 million Northwestern Mutual benchmark. As covered in the $955 median savings article, the median 65-74 household has $409,900 in total net worth (including home equity). The typical retirement portfolio is closer to $186,000. For most retirees, the question isn’t whether 4% vs 5.7% is the right rate — it’s how to maximize the meaningful retirement income from a portfolio that’s smaller than the planning frameworks assume.

For a $186,000 portfolio at various rates: 4.0% produces $7,440/year; 4.7% produces $8,742/year; 5.25% produces $9,765/year; 6.0% produces $11,160/year. At these portfolio sizes, Social Security is the primary retirement income, and the portfolio’s role is supplementation rather than foundation. The withdrawal-rate framework still matters, but its dollar impact is smaller than for retirees with larger balances.

7. The federal employee advantage that changes the math

For federal employees specifically, the safe-withdrawal-rate calculation is fundamentally different from the math facing most American retirees — because the FERS pension and Social Security together create a guaranteed-income floor that the TSP portfolio doesn’t need to cover.

A worked example. A FERS retiree with 30 years of service and a high-3 of $130,000:

For this household, the TSP portfolio’s role is to fund spending ABOVE the $63,000 floor. If the household wants to spend $90,000/year in retirement, the TSP needs to produce only $27,000/year — meaning a $540,000 TSP balance at a 5% withdrawal rate produces the target spending.

Why this changes the safe-rate math. The portfolio is essentially a “supplemental” rather than “foundational” income source. Three specific implications:

1. Higher safe initial rate. Because the household has a substantial guaranteed-income floor, the portfolio can absorb higher withdrawal rates without risking complete depletion of essential income. Even if the TSP runs out, the household still has $63,000/year in pension and Social Security.

2. More room for the guardrails approach. A retiree depending entirely on the portfolio for income may feel unable to cut spending during bad market years. A federal retiree with a $63,000 guaranteed floor has more room to flex TSP withdrawals up or down without affecting basic spending.

3. Less sensitivity to sequence-of-returns risk. The “worst case” that drove Bengen’s 4% rule was a retiree facing bad markets and inflation while depleting their portfolio for essential needs. A federal retiree facing the same conditions still has the pension and Social Security covering essentials — making the portfolio-depletion risk substantially less severe.

The practical result: federal employees can typically support higher TSP withdrawal rates than the general framework suggests. A retiree using guardrails on a TSP balance that’s only one component of total retirement income can often safely sustain initial rates of 6-7% rather than 4-5%, because the pension and Social Security backstop produces a much lower risk of catastrophic depletion.

For the detailed federal retirement income strategy, including how the FERS pension calculation works and how Social Security claiming-age decisions interact with TSP withdrawals, see the FERS Annuity Supplement guide and the federal retirement application process.

8. Five questions retirees ask about safe withdrawal rates in 2026