Educational information, not individual financial advice.
Key Takeaways
The 4% rule is the single most-cited number in retirement planning. Its history, assumptions, and limitations are worth understanding precisely because it gets used as a universal answer when it's actually a narrow one.
William Bengen was a financial advisor who, in 1994, published "Determining Withdrawal Rates Using Historical Data" in the Journal of Financial Planning. He asked a simple question: if a retiree had withdrawn a fixed real amount from a 50/50 stock-bond portfolio, what's the highest starting withdrawal rate that would have survived every rolling 30-year period in U.S. history?
His data: U.S. large-cap stock returns and intermediate-term U.S. government bond returns, 1926–1992. The answer: about 4% survived every 30-year period, including periods starting in 1929 (Great Depression), 1937 (recession), and 1966 (stagflation onset).
Bengen's paper called this the "SAFEMAX" — the safe maximum withdrawal rate. The term "4% rule" came later.
In 1998, three professors at Trinity University (Cooley, Hubbard, Walz) extended Bengen's analysis. They tested multiple portfolio mixes, multiple withdrawal rates, and multiple horizons. Their findings:
Higher equity allocations, despite more volatility, support the 4% rule more reliably than bond-heavy portfolios over long horizons.
The Trinity Study is updated periodically. Results have been consistent with the original findings, with some softening as post-2000 data has tempered future return assumptions.
The retirement that ended in 1995 is the reason the 4% rule is 4% and not 5%. A retiree starting in 1966 experienced:
Starting at 4% survived this period. Starting at 5% would have run out. Starting at 4.5% came close to failure.
The 1929 retiree actually did better than 1966 in some analyses because the Great Depression was followed by decades of strong real returns. High inflation is often worse for retirees than deflation, counterintuitively.
The 4% rule is not "withdraw 4% of your current portfolio each year." That's a different rule (and it never fails; you just may have to live on very little after a crash).
The 4% rule is:
So the spending amount is driven by inflation, not portfolio value. A bad year doesn't make you cut back; a good year doesn't let you spend more.
This is the source of much criticism — real retirees don't behave this way. Most naturally reduce spending in bad years and increase it in good years, which makes higher starting rates sustainable. Dynamic withdrawal strategies formalize this.
Constant dollar (classic): described above.
Constant percentage: always withdraw 4% of current portfolio. Never runs out, but income varies wildly.
Guyton-Klinger guardrails: start at 5%, cut 10% if the rate exceeds 1.2× the original or add 10% if it falls below 0.8×.
Floor/ceiling: variants that adjust within preset bands.
Bucket strategy: cash bucket for current spending (2–5 years), intermediate bucket (5–10 years), growth bucket for remainder. Refill cash from growth during up markets.
Each has tradeoffs. Research generally supports higher starting withdrawal rates (5–5.5%) when combined with spending discipline during bad markets.
Wade Pfau and others have applied the Bengen methodology to non-U.S. historical data. Results are sobering:
The 4% rule is a U.S.-specific result. If you believe U.S. returns will continue to look like the 20th century, the rule is reasonable. If you're preparing for possibility of Japan-style outcomes, use 3% or plan with lower assumed returns.
Horizons Monte Carlo simulations use forward-looking expected returns (typically slightly below 20th-century averages) rather than pure historical bootstrapping. This produces withdrawal-rate guidance that's somewhat more conservative than straight Trinity Study results — typically 3.5% for 30-year horizons as the "comfortable" rate, with 4% being workable but aggressive.
The 4% rule says 'withdraw 4% year one, then raise that dollar amount with inflation.' A retiree starts in a strong market year — their portfolio grows 25%. Should they raise spending?
Try this next
Safe Withdrawal Rates
More related reading