Why the 4% Rule Fails in 2026

William Bengen published the 4% rule in his 1994 Journal of Financial Planning paper using historical US market data from 1926. The Trinity Study (Cooley, Hubbard and Walz, 1998) expanded the analysis with data through 1995 (checked 2026-09-10). Both concluded that a 4% initial withdrawal rate, adjusted annually for inflation, survived every 30-year rolling period in their dataset. The problem: those starting conditions no longer exist.

In 2026, CPI sits at 3.3% — well above the 2.5% long-run average baked into the original research. Oil trades above $108 per barrel, compressing margins across the economy. The Shiller CAPE ratio remains elevated above 30, a level that has historically preceded below-average 10-year returns (computed from Robert J. Shiller's long-horizon dataset, ie_data.xls, checked 2026-09-10). Bond yields, while higher than the 2020–2022 trough, still lag inflation-adjusted levels seen in the 1980s and 1990s when the 4% rule was most robust.

The 4% rule also assumes a static world: fixed real returns, constant inflation, and no regime shifts. Real markets move in regimes — periods of sustained high volatility followed by calm, inflationary decades followed by disinflation. A retiree entering a high-inflation, low-return regime in year one faces a fundamentally different outcome than one who enters during a bull market, even if both experience identical average returns over 30 years.

Morningstar's 2026 State of Retirement Income puts the base-case starting rate at 3.9% for a 30-year retirement on a 30–50% equity portfolio, up from 3.7% a year earlier — as published on morningstar.com, checked 2026-09-10. J.P. Morgan and BlackRock publish capital-market assumptions rather than withdrawal rates; the reason a starting rate below 4% keeps coming out of this work is that every published forward-looking return assumption sits below the long-run historical average. Our own cohort study shows how wide the historical spread around any single rate really is.

What Actually Determines Your Safe Withdrawal Rate

Your safe withdrawal rate is not a fixed number — it is a function of at least five interacting variables:

1. Sequence of Returns Risk

This is the single most important factor the 4% rule ignores. Two portfolios can earn identical average annual returns over 30 years yet produce wildly different outcomes depending on when the bad years fall. A 30% drawdown in year 1, followed by a recovery, devastates a portfolio under withdrawals far more than the same drawdown in year 20. The early losses shrink the base from which compounding must recover, while withdrawals continue at the original dollar amount. This asymmetry is why average returns are misleading for retirees.

2. Inflation Regime

The 4% rule adjusts withdrawals upward each year to match CPI. When inflation runs hot — as it has from 2022 through 2026 — these adjustments compound aggressively. Three years of 5%+ inflation increases your withdrawal amount by over 15%, regardless of what your portfolio has earned. Persistent inflation above the long-run average systematically erodes the margin of safety built into the 4% rule.

3. Portfolio Composition

The stock/bond split matters enormously. The original Bengen research (checked 2026-09-10) concluded that a retiree should hold “a stock allocation as close to 75 percent as possible, and in no cases less than 50 percent.” Pure bond portfolios failed at 4% in multiple historical periods. With 2026 bond yields still below inflation on a real basis, the fixed-income sleeve provides less ballast than the historical data assumes.

4. Time Horizon

Bengen's research (checked 2026-09-10) focused on 30-year retirements. Early retirees planning for 40 or 50 years face a lower safe rate — roughly 3.0%–3.5% for 40 years and 2.5%–3.0% for 50 years on our own reading of the historical record; see our cohort study for the full per-start-year distribution at 30 years. Conversely, a 20-year horizon allows rates above 5%.

5. Tax Drag

The 4% rule ignores taxes entirely. A retiree drawing from a traditional IRA or 401(k) must withdraw more than 4% of the portfolio to net 4% after taxes. This increases the effective withdrawal rate, accelerating depletion. Roth accounts, taxable accounts with stepped-up basis, and tax-loss harvesting all change the equation.

Safe Withdrawal Rates by Portfolio and Time Horizon

The table below shows historically sustainable withdrawal rates (highest rate that survived 100% of rolling periods in the 1926–2024 dataset) for different stock/bond allocations and retirement durations. These are worst-case historical rates — the floor, not the expected outcome.

Allocation	20 Years	25 Years	30 Years	35 Years	40 Years
100% Bonds	4.4%	3.5%	2.8%	2.4%	2.1%
25/75 Stock/Bond	4.8%	3.9%	3.4%	3.0%	2.7%
50/50 Stock/Bond	5.1%	4.2%	3.8%	3.4%	3.1%
60/40 Stock/Bond	5.2%	4.3%	4.0%	3.6%	3.3%
75/25 Stock/Bond	5.4%	4.4%	4.1%	3.7%	3.4%
100% Stocks	5.3%	4.2%	3.9%	3.5%	3.2%

Note that 100% stocks does not produce the highest safe rate at longer horizons. The volatility drag offsets the higher expected return. The 60/40 to 75/25 range has historically offered the best risk-adjusted withdrawal capacity across all time horizons — consistent with Bengen's original finding (checked 2026-09-10).

How Monte Carlo Improves on the 4% Rule

Historical rolling-period analysis (what Bengen and the Trinity Study used) is limited to the specific sequence of returns that actually occurred. The US market has experienced roughly 60 unique rolling 30-year periods since 1926. That is a small sample from which to draw confidence about the future — especially when future conditions may not resemble any historical period.

Monte Carlo simulation solves this by generating thousands of synthetic return paths. Each simulation year draws a random return from a distribution calibrated to historical parameters (mean, variance, skewness, kurtosis). The result is not one answer but a probability distribution: "Your plan succeeds in 87% of 1,000 simulated paths."

The free calculator on this page runs 1,000 simulations using a normal distribution with the return and volatility parameters you specify. This captures the core insight — that your outcome depends on the path of returns, not just the average.

QuantCalc PRO goes further with 10,000 simulations and three critical enhancements:

Regime-switching volatility. Markets alternate between low-volatility regimes (average ~12% annual std. dev.) and high-volatility regimes (average ~22%). The transitions are stochastic. A simple normal distribution underestimates the probability of extended high-volatility periods.

Fat-tailed returns. Real market returns exhibit excess kurtosis — extreme moves occur more frequently than a normal distribution predicts. QuantCalc uses a Student-t distribution to capture tail risk, producing more realistic worst-case scenarios.

Stochastic inflation. Rather than assuming a fixed inflation rate, QuantCalc models inflation as a mean-reverting process with its own volatility. This captures the risk of prolonged high-inflation periods that compound withdrawal amounts beyond what the portfolio can sustain.

What is a safe withdrawal rate?

The maximum percentage of your portfolio you can withdraw annually without running out of money. The traditional answer is 4%, but this depends heavily on market conditions, inflation, and your time horizon.

Is the 4% rule still valid in 2026?

The 4% rule was derived from 1926–1995 US market data. With 2026 CPI at 3.3%, oil above $100, and elevated valuations, many researchers suggest 3.0%–3.5% is more appropriate. Monte Carlo simulation with current conditions gives a more personalized answer.

How does sequence of returns risk affect withdrawal rates?

Poor returns in the first 5–10 years of retirement can devastate a portfolio even if long-term averages are normal. A 30% drop in year 1 followed by recovery produces worse outcomes than a 30% drop in year 20. This is why fixed-rate rules fail.

What portfolio allocation supports a 4% withdrawal rate?

Historical data suggests 50–75% stocks provides the best balance. 100% bonds fails at 4% over 30 years in most periods. 60/40 has historically supported ~3.8%–4.2% depending on the starting decade.

How many Monte Carlo simulations do I need?

At least 1,000 for basic confidence intervals. QuantCalc PRO runs 10,000 simulations with regime-switching volatility, fat-tailed returns, and stochastic inflation — capturing tail risks that simple simulations miss.