Retirement · 7 min read · 2026-04-19
The 9-variable safe withdrawal rate.
There is no single safe withdrawal rate. There is a withdrawal rate that is safe for your portfolio, your horizon, and the regime you are retiring into.
A single number for spending in retirement is a fiction.
The 4% rule (Bengen 1994, Trinity Study 1998) was calibrated against U.S. equity history and 30-year horizons. Applied today — with longer life expectancy, different valuations, lower bond yields, and varied fee structures — the appropriate withdrawal rate ranges from 2.8 to 5.5 percent depending on inputs. Treating it as a constant produces predictable failures.
The nine indicators
The nine variables behind your safe rate.
Each pulls the appropriate withdrawal rate up or down. The composite is what gets implemented; the canonical 4% is just one possible answer.
Planning horizon (years)
Range: 25–50 years
Longer horizons require lower rates. A 50-year horizon (FIRE retirees) supports approximately 3.0–3.3%; a 25-year horizon supports 4.5%+.
Equity allocation (target)
Range: 40–80%
Higher equity allocations support higher long-run withdrawals but produce more drawdown variance. The trade-off is sequence-vs-longevity.
Starting CAPE valuation percentile
Threshold: top quintile
Retirements starting in top-quintile CAPE regimes have historically supported approximately 80 basis points lower safe rates than median-CAPE regimes.
Total fee drag (advisory + fund)
Threshold: < 50 bps total
Each 50 bps of fee drag reduces the safe rate by approximately 30 bps. Fee structure is durable input, not a one-time decision.
Variable spending rule in place
Pattern: guardrails enabled
Variable rules add 60–90 bps of safe rate by adjusting spending in adverse sequences.
Guaranteed income share of essential spending
Threshold: > 70%
Higher guaranteed-income share allows higher portfolio withdrawal rates because portfolio is funding discretionary, not essentials.
Tax-location sequencing
Order: taxable → tax-deferred → Roth
Optimal withdrawal order can extend portfolio life by 2–5 years equivalent at the same nominal withdrawal rate.
Healthcare cost reserve
Threshold: separately funded
Healthcare shocks are the most common forced-withdrawal trigger. A separately funded reserve allows the spending portfolio to operate at a higher rate.
Legacy / bequest priority
Threshold: high vs none
High legacy targets require lower rates to preserve principal. Spend-it-down retirees can run higher rates without violating goals.
What the 4% rule actually said
Bill Bengen's 1994 work and the subsequent Trinity Study (1998) examined historical U.S. equity and bond data and asked what initial withdrawal rate, adjusted for inflation, would have survived every 30-year period in the historical sample. The answer was approximately 4 percent, applied to a 50/50 to 75/25 equity-bond portfolio. The result was contingent on three explicit assumptions: a 30-year horizon, U.S. historical returns, and no fees.
The 4% rule was always a baseline, not a universal constant. The original authors emphasized that real-world implementation required adjustments. Over the subsequent thirty years, the rule has been treated as gospel by investors and dismissed as too aggressive (and too conservative) by various critics. Both reactions miss the point. The rule is a starting input; the nine-variable framework is the calibration.
Horizon as the dominant variable
The 30-year horizon assumed by the original Trinity Study was reasonable for someone retiring at 65. It is conservative for a healthy 65-year-old retiring today (where life expectancy at 65 is approximately 21 years for a man and 24 for a woman, with significant tail risk). It is dramatically inadequate for a FIRE retiree at 40 looking at a 50-year horizon.
Each additional decade of horizon reduces the appropriate withdrawal rate. A 40-year horizon supports approximately 3.5%; a 50-year horizon supports 3.0–3.3%; a 60-year horizon (rare but possible for early retirees with healthy genetics) supports 2.7–2.9%. The mechanism is purely longevity risk: longer horizons mean more potential adverse sequences within the holding period.
Valuation regime — the sequence-risk amplifier
Wade Pfau and others have shown that retirement starting CAPE valuation explains much of the historical variation in safe withdrawal rates. Retirements starting in top-quintile CAPE regimes faced safe rates of 3.2–3.5%; those starting in bottom-quintile regimes supported 5.0–5.5%. The mechanism is sequence-of-returns risk: high starting valuations imply low forward returns, and low forward returns plus inflation-adjusted withdrawals produce poor sequences.
The valuation adjustment is meaningful but not deterministic. The discipline is to start more conservatively in high-CAPE regimes and ratchet the rate up as the portfolio behaves well. Static rates set in high-CAPE regimes have historically produced the worst outcomes.
Variable rules outperform static rules
Static withdrawal rules — including the 4% rule — fail because they are insensitive to portfolio performance. Variable rules adjust the withdrawal rate based on portfolio outcomes. The Guyton-Klinger guardrails approach reduces withdrawals by 10 percent when the portfolio underperforms a target glide and increases withdrawals modestly when the portfolio overperforms.
Stochastic backtests consistently show that variable rules support 60 to 90 basis points more in starting withdrawal rate than equivalent static rules with the same failure tolerance. The cost is that the retiree must accept a 10 percent spending reduction in adverse years. The benefit is that the portfolio survives at higher starting rates than a static policy.
Fee drag — the silent reducer
Each 50 basis points of fee drag reduces the safe withdrawal rate by approximately 30 basis points in the historical sample. The mechanism is direct: fees come out of the same compounding base that produces the income. Advisory fees, fund expense ratios, and platform fees all aggregate. A retiree paying 100 bps in advisory plus 50 bps in funds plus 25 bps in trading effectively shaves 100 bps from the safe rate compared to a low-cost implementation.
The discipline is to construct the lowest-cost implementation consistent with the holder's actual needs. Index funds at 5–10 bps, advisory at 0–50 bps, and zero-commission brokerages produce all-in costs of 25–50 bps for a self-directed implementation, which preserves nearly the full safe-rate budget for spending.
Get the nine, every Monday.
Free weekly digest. The only U.S. equity letter that publishes a name only when nine independent signals align.
Common questions
Questions.
Is the 4% rule still safe?
Approximately, in median-CAPE regimes for 30-year horizons. In current high-CAPE conditions or longer horizons, 3.5–3.8% is more defensible without variable adjustments.
What about the 'first five years' approach?
Variants like Pfau's bond tent and Kitces' rising glide path concentrate defensive moves in the first five years of retirement, when sequence risk is highest. They support higher long-run rates.
How do I model my own safe rate?
Tools include FireCalc, Engaging Data's Rich Broke Dead, and Pralana Gold. Each runs Monte Carlo and historical simulation against the user's actual asset mix and spending.
Does the rule include taxes?
The original rule was pre-tax. Real implementation requires gross-up — the 4% rate becomes 5–6% pre-tax for most U.S. retirees in mid brackets, depending on tax-location.
What if the portfolio is bond-heavy?
Bond-heavy portfolios (40% equity or below) support lower safe rates — historically 3.2–3.5% — because long-run growth is constrained even in benign sequences.
Can I withdraw a percentage of current balance instead of inflation-adjusted?
Endowment-style withdrawals (e.g., 4% of current balance each year) eliminate sequence risk but produce highly variable income year-to-year. Most retirees prefer the smoother inflation-adjusted approach with variable adjustments.
One name. Sometimes weeks of silence. Always with conviction.
Free weekly digest. No credit card. Unsubscribe anytime.