TL;DR: The “4% rule” is a rough rule of thumb from a 1994 study by financial planner William Bengen. It says a retiree can withdraw 4% of their portfolio in year one, increase that dollar amount by inflation each year after that, and — based on U.S. stock and bond history — the money should last at least 30 years. It is a starting point, not a guarantee, and it hides several big assumptions.
Where the 4% rule came from
In October 1994, financial planner William Bengen published “Determining Withdrawal Rates Using Historical Data” in the Journal of Financial Planning. Bengen ran historical simulations across many different 30-year retirement windows in U.S. market history and asked a simple question: what is the highest constant withdrawal rate that would have survived the worst 30-year period on record?1
His answer, using a roughly 50/50 mix of U.S. stocks and intermediate government bonds, was about 4% of the starting portfolio, adjusted upward by inflation every year. Bengen has repeatedly said the 4% number was intended as a worst case, calibrated around a hypothetical retiree who stopped working in 1968 — right before the stagflation of the 1970s crushed both stock and bond real returns.1
Four years later, three finance professors at Trinity University in Texas — Philip Cooley, Carl Hubbard, and Daniel Walz — published a related study (now known as the “Trinity Study”) that reached similar conclusions. Using U.S. return data from 1925 to 1995, they showed that withdrawal rates of 3% and 4% were “extremely unlikely to exhaust any portfolio of stocks and bonds” over a 30-year horizon.2 Together, these two papers gave retirement planning its first widely-cited “safe” starting number.
How the math actually works
The 4% rule is not “withdraw 4% of the balance each year.” It is a specific, mechanical procedure:
- Year 1: Multiply your starting portfolio by 4%. That dollar amount is your first-year withdrawal.
- Every year after: Increase last year’s withdrawal by the prior-year inflation rate (the standard reference is the U.S. Consumer Price Index published by the BLS3).
- Ignore the balance. You do not cut spending in bad years or splurge in good ones. The rule tests whether a fixed, inflation-linked income stream is sustainable.
A simple illustrative example: a retiree starts with a $1,000,000 portfolio and applies a 4% initial rate. Year one, they spend $40,000. If inflation runs 3% that year, year-two withdrawal is $41,200. If inflation the following year is 2.5%, year-three withdrawal is $42,230. The number ratchets up regardless of what the portfolio does.
| Assumption | Bengen (1994) baseline |
|---|---|
| Retirement horizon | 30 years |
| Portfolio | ~50% U.S. large-cap stocks / ~50% intermediate U.S. government bonds |
| Initial withdrawal | 4% of starting balance (Bengen later revised upward to 4.5% in his 2006 book, and more recently suggested up to 4.7% in favorable conditions) |
| Adjustment | Prior-year U.S. CPI, applied to the dollar withdrawal (not the balance) |
| Return data | U.S. historical stock and bond returns; rebalanced annually |
| Fees and taxes | Not modeled |
| Success criterion | The portfolio was not fully depleted before year 30 (Trinity phrased it as “unspent assets at the end of the period”) |
A visual: what the withdrawal schedule looks like
The chart below shows the dollar withdrawal every year for a $1,000,000 starting portfolio, at a 4% initial rate, with a hypothetical steady 3% inflation. The line curves upward because each year’s withdrawal is last year’s number times 1.03 — small at first, then meaningfully larger by year 20. This is why the 4% rule is not a fixed-income strategy; the whole point is that the income has to keep up with prices.
Why the “worst case” matters
Bengen did not pick 4% because it always worked. He picked it because it was the highest rate that survived the worst 30-year window his data included. That worst window turned out to be someone retiring around 1968: they got a nasty bear market in 1973–74, negative real bond returns through the high-inflation late 1970s, and a portfolio that was seriously dented before U.S. markets recovered in the 1980s. Anyone who was withdrawing more than about 4% in that window ran out.1
What the 4% rule quietly assumes
- U.S. history, not global. Both studies used U.S. return data during a century in which the U.S. stock market delivered exceptional real returns. Studies on non-U.S. developed markets find lower “safe” rates.
- A 30-year clock. If retirement lasts 40 or 45 years — plausible for someone retiring in their early 50s — the safe rate drops materially, because the tail risk of a long drawdown grows with the horizon.
- You keep rebalancing. The rule assumes you sell some of whatever went up to buy whatever went down, every year. It is not a “hold and hope” strategy.
- You ignore your account balance. The rule does not tell you to spend less in a downturn or more in a bull market. In real life most retirees adjust — which usually makes higher starting rates viable.
- No fees, no taxes. A 1% annual fee and unfavorable tax treatment can easily push a “safe” 4% rate below 3.5% in practice.
Where the rule breaks down
Three failure modes come up in practice:
1. Sequence-of-returns risk. A poor first decade of returns is far more damaging to a portfolio being withdrawn from than a poor second decade with the same average return, because you sell shares when they are cheap and never own them for the recovery. This is the mechanism behind Bengen’s 1968 worst case, and we cover it in detail in our companion piece on sequence-of-returns risk.
2. Starting valuations. Both studies used all historical starting points equally, but retirees who start with rich equity valuations and low bond yields have historically faced lower safe rates. Multiple researchers, including Wade Pfau and Michael Kitces, have argued that starting-valuation-adjusted safe rates can be closer to 3.0–3.5% when the CAPE ratio is elevated and Treasury yields are low.
3. Longevity. If a 62-year-old couple has a meaningful chance of one of them living to 95+, the applicable horizon is closer to 35 years, not 30. The safe rate falls.
A second visual: where the 4% rule sits
The bar chart below plots the widely cited “safe” starting rates across a few horizons, illustrating why a longer retirement forces you toward a lower initial percentage. These are illustrative planning-industry rules of thumb, not guarantees.
Common mistakes people make with the 4% rule
- Treating it as an ongoing 4% of the balance. The rule fixes the year-one dollar amount and escalates by inflation. Withdrawing 4% of the balance every year is a completely different strategy (and, mechanically, cannot fail — it can only shrink).
- Applying it to any horizon. The 4% figure is calibrated to 30 years. Early retirees using the “FIRE” playbook usually target a lower initial rate.
- Forgetting rebalancing. The historical simulations rebalance annually. A pure buy-and-hold portfolio behaves differently.
- Confusing “safe” with “optimal.” A rate that never depleted the portfolio in any historical window will also, on average, leave a large ending balance. The 4% rule is engineered for the bad path, not the median one.
What modern research says
Bengen himself has revisited his own number several times. In his 2006 book Conserving Client Portfolios During Retirement, he raised the safe rate to 4.5% for tax-advantaged accounts (with 4.1% for taxable) after adding small-cap stocks to the mix, and he has more recently pointed to 4.7% under favorable starting valuations.1 Other researchers have pushed in the opposite direction, arguing that today’s combination of higher equity valuations, longer lifespans, and lower long-run real bond returns justifies something closer to 3.0–3.5% at the start, with dynamic adjustments over time.
The upshot: 4% is a good starting mental model, not a precise personal number. Real planners increasingly use dynamic withdrawal strategies — spend a bit less in bad markets, a bit more in good ones — which materially improves sustainability.
Related concepts to learn next
- Sequence-of-returns risk: why the order of returns matters more than their average when you are withdrawing.
- Bond duration and convexity: what actually happens to the “bond half” of a 60/40 portfolio when yields move.
- The yield curve: the market’s own read on where rates and growth are going, and why it matters for retirement bond allocations.
Sources
- William Bengen — biography and summary of the 1994 paper “Determining Withdrawal Rates Using Historical Data”
- Trinity study (Cooley, Hubbard, Walz, 1998) — U.S. 1925–1995 withdrawal-rate simulation summary
- U.S. BLS Consumer Price Index (CPIAUCSL) — inflation reference series hosted on FRED
- SEC Office of Investor Education (investor.gov) — general retirement planning resources
Disclosure: This article is for informational purposes only and is not investment advice.