Safe withdrawal rate research: what Bengen, Trinity and friends really found

By Tom, CPA · Updated October 25, 2026

TL;DR: The 4% rule comes from US historical data over 30-year retirements. Bengen (1994) found a 4% inflation-adjusted first-year withdrawal never exhausted a portfolio in under 33 years from 1926 on. Longer early-retirement horizons, expensive markets, and non-US data all point to lower or more flexible withdrawal rates. Treat 4% as a starting point for thinking, not a guarantee.

Educational only, not tax, legal, or investment advice. Examples are illustrations, not predictions. Rules and limits change; verify current figures and talk to a qualified professional about your situation.

Everyone in FIRE has heard of the 4% rule. Fewer people have read where it came from, which is how a careful 1994 finding about 30-year retirements became a bumper sticker for 50-year retirements. Lavish Larry quotes it like scripture. Frugal Fran has read the footnotes. Let's be Fran.

This page summarizes the major studies, what each one actually tested, and the practical lessons for early retirees. For the basics of the rule itself, see the 4% rule explained.

CPA Warning Label: Every safe withdrawal rate is a statement about the past. Your retirement will be one new path, and it will not show up in any backtest until it's over.

Trinity and Bengen

Bengen (1994)

Financial planner William Bengen's article "Determining Withdrawal Rates Using Historical Data" (Journal of Financial Planning, October 1994) used actual US stock, bond and inflation history from 1926 on, drawn from Ibbotson Associates' data. He tested a fixed first-year withdrawal, as a percentage of the starting portfolio, then raised that dollar amount with inflation every year.

His headline finding: assuming a minimum need of 30 years, a 4% first-year withdrawal with inflation adjustments "should be safe." In his data it never exhausted a portfolio in less than 33 years, and in most cases lasted 50 years or longer. A 4.25% rate could have run out in as little as 28 years. He also found that about 3% to 3.5% had never lasted fewer than 50 years. He favored stock allocations roughly in the 50% to 75% range.

The Trinity study (1998)

Three professors at Trinity University, Philip Cooley, Carl Hubbard and Daniel Walz, published "Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable" in the AAII Journal in February 1998. They used S&P 500 stocks, long-term high-grade corporate bonds and inflation data from 1926 to 1995, tested withdrawal rates from 3% to 12% over 15- to 30-year payout periods, and reported a "portfolio success rate," the share of historical periods in which money remained at the end.

Their conclusions include that 3% and 4% rates from stock-dominated portfolios were very conservative, that retirees who insist on inflation-adjusted withdrawals must accept lower rates, and that higher rates looked sustainable only over short payout periods. The study ignored taxes and transaction costs.

Notice what both studies are: 30-year (or shorter) horizons, US data only, and a rigid rule that never cuts spending. Each of those assumptions matters for early retirees.

Longer horizons

A 35-year-old retiree may need money for 50 or 60 years. Two things happen when you stretch the horizon:

1. There's simply more time for bad sequences to compound.
2. Under a 30-year definition of success, a plan that ends with $1 left in year 30 "succeeds." A 60-year retiree can't accept that.

Bengen's own data hinted at this: lower rates of around 3% to 3.5% were the ones that never lasted fewer than 50 years. Independent researcher Karsten Jeske's long-running Safe Withdrawal Rate Series at Early Retirement Now extends the historical approach to 60-year horizons and different end-of-retirement targets. It generally finds that the historically safe fixed withdrawal rate falls as the horizon lengthens and as the required ending balance rises, and that the worst outcomes cluster in retirements that started when markets were expensive.

The gap between a 4% and a 3.5% withdrawal rate sounds small. On $60,000 of spending it's the difference between needing $1,500,000 and about $1,714,000. That's why many early retirees pair a moderate rate with flexibility rather than chasing an ultra-low number. See guardrails and sequence of returns risk.

Valuation-based rates

Not every starting year is equal. Several researchers have noted that the worst historical retirements tended to begin when stock valuations were high, commonly measured by the cyclically adjusted price-to-earnings ratio (CAPE). The logic is intuitive: expensive starting prices tend to be followed by lower long-run returns, and low early returns are what wreck a withdrawal plan.

Valuation-based approaches use the starting CAPE, or a related earnings yield, to suggest a higher or lower initial withdrawal rate. Supporters argue it tailors the plan to conditions you can actually see on day one. Critics point out that valuations can stay high for long stretches, that measurement choices change the answer, and that the relationship is noisy for any single retiree.

Practical read: if you're retiring into a market that looks expensive by historical standards, that's an argument for some combination of a lower starting rate, a bigger cash or bond buffer (see bond tents), or rules that cut spending if early returns disappoint. It's not a reason to try to time your exit to the month.

International data

US stocks had an exceptional 20th century. Building a rule from the luckiest major market is a little like setting your lottery budget based on past winners.

Wade Pfau's "An International Perspective on Safe Withdrawal Rates: The Demise of the 4 Percent Rule?" (Journal of Financial Planning, December 2010) applied the historical approach to 109 years of data for 17 developed-market countries, using 30-year retirements. His findings:

• Even assuming "perfect foresight" about the best fixed asset allocation, the maximum safe withdrawal rate exceeded 4% in only 4 of 17 countries: Canada, Sweden, Denmark and the United States.
• A fixed 50/50 stock/bond portfolio would have failed at some point in all 17 countries at 4%.
• In the US data, 4% had failed in 0% of 30-year periods, while 5% failed in 22.5% of them.

Pfau also noted that some of the worst outcomes coincided with the world wars, and that flexible spending rules and global diversification can improve results. Global diversification is one reason many FIRE portfolios hold international stocks.

What to take away

• 4% is a 30-year, US-history, never-adjust number. For longer horizons, many early retirees plan around a somewhat lower starting rate, more flexibility, or both.
• Flexibility is worth a lot. Guyton and Klinger's 2006 decision-rule research and other studies found that willingness to trim spending after bad years supports higher starting withdrawals. See variable percentage withdrawal.
• Costs come straight off the top. The classic studies ignored fees. A 1% annual advisory or fund fee on a 4% plan is a quarter of your spending. See expense ratios and advisor fees.
• Taxes aren't in the rate either. A 4% gross withdrawal from a traditional 401(k) isn't 4% spendable.
• Other income changes everything. Social Security, part-time work or rental income later in retirement reduce what the portfolio must cover.

Illustration: Fran wants $50,000 a year from her portfolio. At 4%, that implies $1,250,000. At 3.5%, about $1,428,600. At 3.25%, about $1,538,500. If she expects $20,000 a year of Social Security from age 67, the portfolio only needs to cover $30,000 after that, which can justify a higher early withdrawal rate if she plans the bridge carefully. Try the scenarios with our withdrawal rate calculator and the guardrails calculator.