The 4% Rule Is Wrong: What Monte Carlo Shows
The 4% rule — withdraw 4% of your portfolio in year one, adjust for inflation thereafter — has a roughly 15-20% chance of leaving an early retiree broke. William Bengen designed it in 1994 for 30-year retirements. The 1998 Trinity Study confirmed it for the same timeframe. Neither study tested the 40-60 year horizons that FIRE retirees face, and Monte Carlo simulation shows the gap is not small.
The actual safe withdrawal rate for a 40-year retirement is closer to 3.25-3.5%. But even that framing is flawed. A fixed withdrawal rate is the wrong tool for a multi-decade retirement. Variable strategies that adjust spending based on portfolio performance dominate fixed rates in every Monte Carlo test. The real question isn't "what percentage is safe?" — it's "what withdrawal strategy maximizes lifetime spending while protecting against ruin?"
The 4% Rule Was Built for 30-Year Retirements
William Bengen's 1994 paper tested withdrawal rates against every rolling 30-year period in US market history from 1926 to 1992. He found that a 4% initial withdrawal rate, adjusted annually for inflation, survived every single 30-year period — even those starting in 1929 and 1966. The worst-case scenario (retiring in October 1968) came close to failure but held.
The Trinity Study, published by Cooley, Hubbard, and Walz in 1998, expanded on Bengen's work. It tested portfolios ranging from 100% stocks to 100% bonds across 15, 20, 25, and 30-year periods. For a 75/25 stock/bond portfolio over 30 years, the success rate at a 4% withdrawal was 98%. For 50/50, it was 95%.
What the 4% rule actually says: Withdraw 4% of your portfolio value in year one of retirement. In each subsequent year, increase that dollar amount by inflation — not 4% of the current portfolio. On a $1,000,000 portfolio, you withdraw $40,000 in year one. If inflation is 3%, you withdraw $41,200 in year two regardless of what the portfolio did. This distinction matters: you're not recalculating 4% each year.
Three assumptions are baked into these results. The studies used US-only historical data, which includes the best-performing equity market of the 20th century. They assumed no taxes on withdrawals. And they capped the time horizon at 30 years. All three assumptions break for a 35-year-old FIRE retiree pulling from a traditional IRA.
Why 4% Fails Over 40-60 Year Horizons
Each additional decade of retirement compounds the probability of encountering a catastrophic return sequence. The ERN Safe Withdrawal Rate Series — the most rigorous public analysis of long-horizon withdrawal rates, running 55+ posts of analysis — finds that extending from 30 to 60 years roughly triples the failure rate at a 4% withdrawal. A 60/40 portfolio at 4% over 30 years fails about 5% of the time. Over 50 years, that jumps to 15-20%.
The reason is mathematical, not speculative. A fixed-rate withdrawal strategy is a bet that your portfolio will recover from any drawdown within the remaining time horizon. Over 30 years, you have 20+ years of recovery time after a bad first decade. Over 50 years, a bad decade at year 15 or year 20 can still be fatal because cumulative withdrawals have already depleted the base.
| Time Horizon | 4% SWR Failure Rate | 3.5% SWR Failure Rate | 3.25% SWR Failure Rate |
|---|---|---|---|
| 30 years | ~5% | ~2% | ~1% |
| 40 years | ~10-12% | ~5% | ~3% |
| 50 years | ~15-20% | ~8-10% | ~5% |
| 60 years | ~20-25% | ~12-15% | ~7-8% |
Approximate failure rates based on Monte Carlo simulation of a 60/40 US stock/bond portfolio with historical return distributions. Sources: ERN SWR Series, Bengen 1994 extended analysis.
Sequence-of-Returns Risk: The First Decade Decides Everything
Two retirees can earn the same average annual return over 30 years and have radically different outcomes. If you earn -15%, -10%, +25%, +20% in your first four years, your portfolio is in a fundamentally different position than if those returns arrived in reverse order — because you're withdrawing a fixed dollar amount from a shrinking base.
The sequence-of-returns trap: A retiree withdrawing $40,000/year from a $1,000,000 portfolio who earns -20% in year one starts year two with $760,000. That $40,000 withdrawal is now 5.3% of the portfolio, not 4%. A second bad year can push the effective withdrawal rate above 7%, creating a death spiral that strong later returns cannot reverse. Monte Carlo simulation captures this risk by testing thousands of different return orderings.
This is why average-return calculators are dangerous. They assume you earn 7% every year, which never happens. Monte Carlo simulation draws returns from a distribution — including the possibility of two or three bad years in a row early in retirement — and tells you how often your plan survives across all those scenarios.
Safe Withdrawal Rate by Age and Time Horizon
The safe withdrawal rate is not a universal constant. It depends on how long your money needs to last. The table below shows approximate initial withdrawal rates at a 95% success probability for a 60/40 stock/bond portfolio, based on Monte Carlo simulation using historical US return distributions.
| Retirement Age | Time Horizon | Safe Withdrawal Rate (95% success) | Portfolio Needed for $50,000/yr |
|---|---|---|---|
| 35 | 55 years | ~3.15% | $1,587,000 |
| 40 | 50 years | ~3.25% | $1,538,000 |
| 45 | 45 years | ~3.35% | $1,493,000 |
| 50 | 40 years | ~3.50% | $1,429,000 |
| 55 | 35 years | ~3.75% | $1,333,000 |
| 60 | 30 years | ~4.00% | $1,250,000 |
| 65 | 25 years | ~4.30% | $1,163,000 |
| 70 | 20 years | ~4.70% | $1,064,000 |
Assumes 60/40 US stock/bond portfolio, inflation-adjusted withdrawals, no taxes, no Social Security. These are pre-tax gross withdrawal rates — your after-tax spending will be lower. Sources: ERN SWR Series, Bengen 1994 methodology extended.
A 40-year-old retiring with $1,500,000 can safely withdraw about $48,750 per year (3.25% × $1,500,000) in pre-tax dollars. That same person at 60 could pull $60,000. The 15-year difference in age buys a 23% increase in spending — which is why the Coast FIRE concept of letting invested assets grow untouched for a decade or more is so powerful.
Note these are pre-tax withdrawal rates. If you're pulling from a traditional IRA or 401(k), federal and state taxes take a meaningful bite. A $48,750 gross withdrawal does not buy $48,750 of groceries. We'll quantify the tax drag below.
How Monte Carlo Simulation Actually Works
A Monte Carlo retirement simulator doesn't predict the future. It stress-tests your plan against thousands of plausible futures. The mechanics: draw a random annual stock return from a distribution matching historical data (mean ~10.3%, standard deviation ~19.7% for US large-cap stocks, per Damodaran/NYU Stern), draw a correlated bond return, apply them to your portfolio, subtract your withdrawal, and repeat for each year of retirement. Do this 1,000+ times. Count how many trials end with money remaining.
The correlation between stock and bond returns matters. Stocks and bonds don't move independently — their correlation has ranged from -0.3 to +0.4 historically, and the sign matters for diversification benefits. Generating properly correlated return pairs requires Cholesky decomposition of the covariance matrix, which ensures the simulated return sequences reflect realistic joint behavior rather than treating stocks and bonds as independent coin flips.
What Monte Carlo tells you that backtesting can't: Historical backtesting is limited to the ~100 years of US market data available. That gives you roughly 70 overlapping 30-year periods. Monte Carlo generates thousands of unique return sequences drawn from the same distributions but in orderings that never actually happened. This reveals tail risks that the historical record — by luck — may not contain. The US has never experienced a Japan-style 30-year flat market, but Monte Carlo shows how your plan would fare in one.
CoastIQ's Monte Carlo Projection runs 1,000+ parallel trials with correlated stock and bond returns, outputting P10 (bad luck), P50 (median), and P90 (good luck) portfolio trajectories. The Historical Backtester complements this by testing against every actual rolling historical period. Using both together gives you the full picture: how does history judge your plan, and how does a broader stress test judge it?
Fixed vs. Variable Withdrawal Strategies
A fixed 4% withdrawal rate is the financial equivalent of driving with cruise control on a mountain road. It works on flat highway. Over 30 years of mostly-average markets, it's fine. But when the terrain changes — a 40% drawdown in year three, a decade of flat returns — you need to steer.
Variable withdrawal strategies adjust spending based on portfolio performance. The three most studied approaches:
Guyton-Klinger Guardrails (2006): Start with an initial withdrawal rate (can be higher than 4%). Set a ceiling and floor — if your current withdrawal rate rises above 20% of the initial rate (portfolio dropped), cut spending by 10%. If it falls 20% below (portfolio grew), increase spending by 10%. Guyton and Klinger's original paper found this supported initial rates of 5.2-5.6% over 40 years.
VPW (Variable Percentage Withdrawal): Each year, recalculate your withdrawal as a percentage of the current portfolio value using an actuarial table based on remaining life expectancy. This can never deplete the portfolio to zero by design — it's mathematically impossible. The tradeoff: spending volatility. A 30% market crash means a 30% spending cut.
Constant-percentage with floor: Withdraw a fixed percentage of the current portfolio each year (say, 4%), but set a dollar floor below which spending never drops (say, $35,000 from Social Security + part-time work). This blends the safety of VPW with a spending floor.
| Strategy | Initial Withdrawal Rate | Can Run Out? | Spending Volatility | Best For |
|---|---|---|---|---|
| Fixed 4% (Bengen) | 4.0% | Yes (~5% over 30yr) | None (fixed) | Traditional 30-year retirement |
| Fixed 3.25% (FIRE-adjusted) | 3.25% | Yes (~5% over 50yr) | None (fixed) | Conservative early retirees |
| Guyton-Klinger guardrails | 5.0-5.5% | Yes (very low) | Moderate (±10% steps) | FIRE retirees willing to flex spending |
| VPW | ~4-5% initial | No (by design) | High (tracks portfolio) | Retirees with guaranteed income floor |
| Constant % with floor | 4.0% of current value | No (if floor is covered) | Moderate-high | Retirees with Social Security/pension |
The case against a single number: Obsessing over whether your "safe withdrawal rate" is 3.5% or 3.75% misses the point. A retiree using Guyton-Klinger guardrails can start at 5%+ and cut spending by 10% during downturns — supporting higher average lifetime spending than someone rigidly withdrawing 3.25% who dies with $2M unspent. The right question is not "what fixed percentage is safe?" but "what strategy maximizes spending while keeping ruin probability below my threshold?"
Worked Example: How Taxes Reduce Your Effective Withdrawal Rate
Most SWR analysis ignores taxes entirely. The 4% rule doesn't account for the fact that withdrawals from traditional retirement accounts are taxed as ordinary income. This means your effective spending rate is lower than your gross withdrawal rate.
Consider a married couple filing jointly in 2025, retired at 50, withdrawing entirely from a traditional IRA. They need $60,000 in after-tax spending. Using 2025 federal tax brackets (Rev. Proc. 2024-40):
- Standard deduction (MFJ): $30,000
- 10% bracket: first $23,850 of taxable income
- 12% bracket: $23,850 to $96,950 of taxable income
To net $60,000, you solve for the gross withdrawal W where W minus taxes on W equals $60,000. Working backward through the brackets:
- Gross withdrawal needed: $63,549
- Taxable income: $63,549 - $30,000 = $33,549
- Tax on first $23,850 at 10%: $2,385
- Tax on next $9,699 at 12%: $1,164
- Total federal tax: $3,549
- Effective tax rate: 5.6%
- After-tax spending: $63,549 - $3,549 = $60,000
At a 3.5% SWR, they need a portfolio of $63,549 / 0.035 = $1,815,700 to spend $60,000 per year after federal tax. Without accounting for taxes, the naive calculation says $60,000 / 0.035 = $1,714,300. That's a $101,400 gap — 6% more savings required just from federal tax drag on withdrawals.
Add state income tax and the gap widens. In California (9.3% marginal rate at this income level), the gross withdrawal climbs to ~$69,800 and the required portfolio to $1,994,300. Texas or Florida residents pay zero state tax and need only $1,815,700 — a $178,600 difference driven entirely by geography. This is why tax-accurate retirement calculators matter.
Tax drag on your SWR: A 3.5% gross withdrawal rate at a 5.6% effective federal tax rate yields only 3.3% in after-tax spending. On a $1,500,000 portfolio, that's $52,500 gross but $49,560 after tax — $2,940/year lost to federal taxes alone. Roth conversions during the early retirement years before Social Security kicks in can eliminate this drag entirely on converted amounts, which is why the Roth conversion ladder is central to FIRE tax planning.
What This Means for Your FIRE Number
The 4% rule gives a tidy formula: multiply your annual spending by 25 and you have your FIRE number. For $50,000 in spending, that's $1,250,000. The problems compound when you adjust for reality:
- Time horizon correction: At age 40 with a 50-year horizon, 3.25% SWR means multiplying by 30.8 instead of 25. Target: $1,540,000.
- Tax correction: Withdrawing from traditional accounts at a 5.6% effective federal rate means you need ~$52,950 gross to spend $50,000. At a 3.25% SWR: $52,950 / 0.0325 = $1,629,200.
- Healthcare: Pre-Medicare retirees (under 65) must budget for ACA premiums — $500-$1,500/month for a family — which are highly sensitive to MAGI, the same income that determines your tax bracket.
These corrections push the real FIRE number 25-35% above the naive 25× calculation. A 40-year-old targeting $50,000 in after-tax spending from traditional accounts needs $1,630,000-$1,700,000 — not $1,250,000.
The antidote to this inflation isn't saving more forever. It's using variable withdrawal strategies that allow higher initial rates, Roth conversions that eliminate tax drag on future withdrawals, and Social Security income that reduces the portfolio's burden starting at 62-70. These levers interact in ways that a single SWR percentage cannot capture, which is why Monte Carlo simulation with tax-aware modeling produces fundamentally different answers than back-of-envelope math.
Frequently Asked Questions
CoastIQ Team
The team behind CoastIQ, building the most tax-accurate retirement calculator on iOS.
