Why does the 4% rule face challenges in current markets?

The short answer

The 4% rule comes from a single 1994 paper by William Bengen in the Journal of Financial Planning. The exact finding was that 4% (originally 4.15%, rounded down) of initial portfolio value, with subsequent inflation adjustments, would have sustained a 50/50 stock-bond portfolio through any 30-year period from 1926 to 1976. The rule was a worst-case backstop, not an average.

Two challenges face the rule today. First, current US equity valuations (Shiller CAPE near 40 in 2026 versus historical median 16) suggest below-average forward returns for stocks. Second, real bond yields, while higher than 2020-2021, remain volatile, and the assumption of a stable real bond return embedded in Bengen’s analysis no longer holds reliably.

Modern retirement researchers have produced wide ranges of revised guidance, from Morningstar’s 3.3% (under low return assumptions) to Bengen’s own current 5.25-5.5% baseline (using broader asset diversification). The point is not the precise number — it is that 4% is neither a floor nor a ceiling, but a single point in a distribution.

→ New to retirement withdrawal frameworks? Financial education hub

What the data shows

The reference numbers come from Bengen’s original 1994 paper and his subsequent updates, the Trinity Study (1998), Morningstar’s 2021 “State of Retirement Income” report, and Pfau-Finke-Blanchett work on low-return scenarios.

The reference numbers (Bengen 1994 to Morningstar 2021):

• Bengen 1994 — 4% sustained any 30-year retirement starting 1926-1976 with 50/50 portfolio (success rate above 95%)
• Bengen revised SAFEMAX (2006-2020) — 4.5% as worst case using diversified portfolios with mid- and small-cap exposure
• Bengen current view (2024-2025) — historical average safe rate roughly 7%, with 5.25-5.5% as a comfortable modern baseline
• Morningstar 2021 — 3.3% recommended starting rate given low yields and elevated valuations at the time
• Pfau, Finke, Blanchett — under -1.4% real bond return and 4.6% real stock return assumptions, a 4% withdrawal failed in 57% of Monte Carlo simulations

The exception that complicates the picture: across 100 years of US data, the average safe withdrawal rate has been close to 7%, not 4%. The 4% rule was the worst-case scenario for a 1968 retiree facing 1970s stagflation. Most retirees historically could have withdrawn far more without exhausting their portfolio.

→ Dataset: Real interest rates vs CAPE ratio dataset

Why it happens — the macro mechanism

The 4% rule is a function of three inputs: starting valuation, expected real returns by asset class, and the duration of retirement. Changes in any of these shift the sustainable rate.

The first channel is starting valuation. Bengen’s worst case (1968) coincided with high US equity valuations on the eve of stagflation. Kitces has documented that Shiller CAPE at retirement start has an R² of approximately 0.77 with subsequent 30-year SWR for portfolios with significant equity exposure. With CAPE near 40 in 2026 versus historical median 16, the implied SWR sits well below the historical average.

The second channel is the expected real bond return. Bengen’s 1994 model used historical intermediate Treasury returns averaging meaningful real yield. The assumption that real bond returns will average 2-3% over a 30-year retirement may not hold in some forward scenarios. Pfau-Finke-Blanchett showed that even modestly negative real bond returns can produce failure rates above 50% for the standard 4% rule. Real yields matter more than nominal yields here.

A brief transition: many practitioners have shifted toward dynamic withdrawal frameworks (Guyton-Klinger guardrails, valuation-aware floor-and-ceiling rules) rather than abandoning percentage-based withdrawal entirely.

The third channel is retirement duration. Bengen modeled a 30-year retirement. Modern early retirees aiming for 40-50 years (FIRE community) face structurally different math. Conversely, retirees starting at 70 or later may sustain higher rates simply because the horizon is shorter.

Synthesis by regime: in regimes of low CAPE plus high real bond yields (1982 retirees), 4% was extremely conservative — actual sustainable rates exceeded 8%. In regimes of high CAPE plus low real bond yields (2000 retirees, potentially 2021-2022 retirees), 4% has been pressured and dynamic frameworks become more attractive. In regimes of high CAPE plus high real bond yields (uncertain about 2026-onward), the verdict is open. The transition parameter is the joint state of valuations and real bond yields at retirement entry.

The 4% rule was never an average — it was the worst case from 1968. Treating it as a default is mathematically defensible but historically pessimistic.

→ Framework: Asset allocation and resilient portfolios

What it means for different economic actors

Pre-retirees facing high-valuation environments may want to anchor planning at the lower end of the SWR distribution rather than the historical average, accepting some tradeoff in initial spending for a wider safety margin.

Retirees already several years into withdrawals can use realized returns and current portfolio level to recalibrate, since the worst sequence-risk window is partly observable rather than purely forward-looking.

Financial planners increasingly use dynamic frameworks (variable percentage, floor-and-ceiling, guardrails) that adjust withdrawals based on portfolio performance, which respects the sequence-risk reality without requiring a single fixed-rate decision upfront.

A common error is to treat the 4% rule as either a floor (under-withdrawing in benign environments) or a ceiling (over-withdrawing without checking valuations). Both interpretations reflect misreading of the original research as a universal rule rather than a worst-case backstop.

Practical observation

Go deeper

Related questions

• What is sequence of returns risk?
• How do safe withdrawal rates vary with starting valuations?
• What is longevity risk and how can it be managed?
• Is the CAPE ratio reliable?

Frequently asked questions

Did Bengen really say 4% should be the default rate?

No. Bengen has clarified repeatedly that 4% (originally 4.15%) was the worst-case rate from his 1926-1976 dataset, anchored on a 1968 retiree. He has stated publicly that the average safe rate across his roughly 400 historical scenarios is closer to 7%, and that current retirees can typically take 5.25-5.5% comfortably. The 4% number stuck in popular finance because of the rounding and its memorability, not because it was Bengen’s central recommendation.

How does CAPE valuation affect the SWR debate?

Kitces and others have shown a strong negative correlation (R² near 0.77 for equity-heavy portfolios) between starting Shiller CAPE and subsequent 30-year SWR. The mechanism is that high starting valuations imply lower subsequent equity returns, which compounds with sequence-risk dynamics in early retirement. With CAPE near 40 versus historical median around 16 in 2026, the historical relationship suggests SWRs in the lower part of the distribution.

Why does Morningstar suggest a different number than Bengen?

The two are answering slightly different questions. Bengen’s framework uses historical sequences as the worst case. Morningstar uses Monte Carlo simulation with forward-looking return assumptions. When forward equity and bond return assumptions are below historical averages, Monte Carlo produces lower SWRs than historical worst case. The 3.3% Morningstar 2021 figure assumed below-average returns for both asset classes. Updated 2024-2025 figures from Morningstar have since revised upward as bond yields rose.

Last updated — 4 June 2026