You retire on January 1 with $1,200,000 in your IRA and 401(k). How much can you spend this year without running out of money over the next thirty? The "4% rule" gives a clean answer: $48,000 in year one, then bump that dollar amount up with inflation every year afterward. It is the most widely repeated guideline in personal finance — and for the past two years, the researchers who actually study it have been telling retirees the safe number is closer to 3.9%, while the rule's original author argues it can safely be as high as 4.7%. Both can be right, depending on what you do with the rule. This guide walks through the math, the 2026 numbers, and the small set of decisions that determine which version applies to you.
Where the 4% rule came from
The rule was published in October 1994 by William Bengen, a financial planner in California, in the Journal of Financial Planning. Bengen took every rolling 30-year period from 1926 onward — the Great Depression, the 1970s stagflation, multiple recessions, and several long bull runs — and asked a single question: if a retiree had withdrawn X% of a portfolio (split 50/50 between U.S. large-cap stocks and intermediate Treasury bonds) in year one, then increased that dollar amount each year by the rate of inflation, what is the highest X that never ran the portfolio out of money over 30 years?
The answer in his original paper rounded to 4%. (The precise figure was 4.15%; the round number stuck.) Bengen called this the SAFEMAX — the highest safe starting withdrawal rate in the worst historical period he could find. Critically, the worst case was not 1929, as most people guess. It was a retirement that started in 1966, just before a brutal combination of bear-market returns and double-digit inflation that ate into both portfolio values and purchasing power simultaneously.
How the 4% rule actually works
Three pieces sit underneath the slogan, and skipping any of them changes the math:
- You compute the dollar amount once, in year one. Multiply your starting portfolio by 4%. That is your year-one withdrawal.
- You raise that dollar amount each year by inflation. Not by 4% of the new portfolio balance — by last year's CPI. The withdrawal is set on day one and only adjusts to keep purchasing power flat.
- The horizon is 30 years. If you retire at 65, the rule plans you to age 95. Earlier retirements need a lower rate; later retirements can support more.
Here is what the first five years look like for a $1,000,000 retiree if inflation runs at the long-term U.S. average of about 3%:
| Year | Withdrawal | How it was calculated |
|---|---|---|
| 1 | $40,000 | $1,000,000 × 4% |
| 2 | $41,200 | $40,000 × 1.03 |
| 3 | $42,436 | $41,200 × 1.03 |
| 4 | $43,709 | $42,436 × 1.03 |
| 5 | $45,020 | $43,709 × 1.03 |
The portfolio balance moves around in the background — markets give and take — but the withdrawal schedule above is fixed in advance. That is the entire mechanic. To see how a different starting balance and rate of return play out for you, plug numbers into our retirement calculator and watch the schedule.
The Trinity Study and the meaning of "safe"
In 1998, three professors at Trinity University (Cooley, Hubbard, and Walz) replicated and extended Bengen's analysis using a different bond mix and a longer time series. Their work, generally called the Trinity Study, popularized the language of "success rates": across all rolling 30-year periods in U.S. market history, what percentage of retirees following a given starting rate finished with a non-zero balance? A 4% rule with a 50/50 portfolio came in around 95% historical success. Bump the starting rate up to 5% and success drops; pull it down to 3% and it climbs to roughly 100%.
Two things to keep in mind here. First, "success" in these studies just means "did not hit zero." It does not mean "lived comfortably." Some 4% paths finished 30 years with $4 million. Others finished with $40,000 and white knuckles. Second, the success rates are based on U.S. data only. Studies that include international markets — where countries have lost decades to wars and inflation — tend to produce lower SAFEMAX numbers, because the U.S. data set is unusually generous.
Why 2026 looks different
Two camps now publish updated SAFEMAX numbers each year, and they disagree.
The lower number: 3.9%
Morningstar's annual State of Retirement Income report uses forward-looking capital-market assumptions rather than historical returns. For 2026, Morningstar's base case sets the safe starting withdrawal rate at 3.9% for a 30-year horizon and a 90% success threshold, with an equity allocation between 30% and 50%. That figure is up from 3.7% in their 2025 report and down from 4.0% in 2023, with the moves driven mainly by changes in expected bond yields and equity valuations. (Source: Morningstar, "What's a Safe Retirement Withdrawal Rate for 2026?", published December 2025.)
The higher number: 4.7%
Bengen, the rule's original author, has spent the last decade rerunning his model with broader portfolios — adding small-cap stocks and other asset classes that didn't make his original 1994 cut. His revised Universal SAFEMAX across that fuller diversification comes in at roughly 4.7%. Bengen's argument is that the original 4% was conservative because his 50/50 mix was conservative; a more diversified portfolio survives the 1966 retiree path with room to spare.
The two numbers reflect a real disagreement about the right baseline portfolio and the right way to estimate future returns. They are not contradictions of the underlying logic.
Worked example: a $1.2M retiree, three rules
Suppose you retire at age 65 with $1,200,000, and you want to translate the headline rates into year-one spending. Here is the side-by-side:
| Rule | Starting rate | Year-one withdrawal | Implicit assumption |
|---|---|---|---|
| Morningstar 2026 base case | 3.9% | $46,800 | Forward-looking returns; 30-year horizon; 90% success |
| Original Bengen (1994) | 4.0% | $48,000 | U.S. historical 50/50 portfolio |
| Bengen Universal SAFEMAX | 4.7% | $56,400 | Broader, more diversified portfolio |
The spread between the most cautious and the most optimistic figure on a $1.2M portfolio is $9,600 per year — and over a 30-year retirement that gap, growing with inflation, is roughly the cost of a small second car every two years. Which version you pick is not a rounding error.
Sequence-of-returns risk and why year one matters
The same average return can produce two very different retirements depending on when the bad years arrive. A new retiree who sees a 30% market drop in year one is forced to sell a larger share of the portfolio to fund that fixed inflation-adjusted withdrawal. The portfolio is permanently smaller when the recovery comes, and the math compounds against the retiree for decades. This is called sequence-of-returns risk, and it is the single biggest reason Morningstar's forward-looking SAFEMAX moves with bond yields and equity valuations: starting conditions in years 1–5 dominate the outcome.
Two practical implications:
- An emergency cash buffer of one to three years of spending lets you skip selling stocks during a bad year. Some planners call this a "bond tent" or "cash reserve" strategy.
- Working part-time for the first two or three years of retirement, even in a low-stress role, can reduce your effective withdrawal rate enough to cut sequence risk dramatically.
Flexible withdrawal strategies that beat the strict 4% rule
The strict 4% rule is mechanically simple, which is its virtue and its weakness. It does not let the retiree adjust to good or bad markets. Most modern planners use a flexible variant.
Guyton-Klinger guardrails
Set a target withdrawal rate (say 4.5%). If markets drop and your current withdrawal rate climbs above 5.4% (the upper guardrail, +20%), cut the next year's withdrawal by 10%. If markets rise and the rate drops below 3.6% (the lower guardrail, -20%), give yourself a 10% raise. The guardrails turn one fixed schedule into a feedback loop, and they support starting rates that are meaningfully higher than 4%.
Dynamic / variable percentage
Each year, withdraw a fixed percentage of the current portfolio balance — say 5%. Spending fluctuates with the market, but the portfolio mathematically can never run out. The trade-off is that bad years feel worse: a 30% drawdown means a 30% pay cut. Many retirees use a hybrid: a fixed "floor" funded by Social Security and pensions, with the variable percentage applied only to the market portfolio on top.
How to apply the 4% rule to your own plan
The rule is most useful as a sizing test, not a spending budget. Three concrete uses:
- Reverse the math to find your number. If you want to spend $80,000 a year in retirement on top of Social Security, divide by 0.04 to get $2,000,000 — that is the rough portfolio size the 4% rule asks for. Want a safer 3.9%? Divide by 0.039 instead to get about $2,051,000.
- Stress-test against your actual retirement age. Retiring at 55 means a 40-year horizon, not 30. Most studies put the 40-year SAFEMAX about 0.5 percentage points below the 30-year number. Retiring at 70 lets you take more.
- Layer in Social Security. Social Security is an inflation-adjusted lifetime annuity, and for 2026 the cost-of-living adjustment is 2.8%, paid starting January 2026 (Social Security Administration, October 24, 2025). Most retirees should not apply the 4% rule to their entire spending need — they should apply it only to the gap between what Social Security and any pensions cover and what they actually need.
Run those calculations together with our retirement calculator and our 401(k) calculator to see how monthly contributions today translate into a withdrawal-supportable balance later. If you are still deciding when to file for benefits, our companion guide on claiming Social Security at 62 vs 67 vs 70 walks through the breakeven math.
Common mistakes retirees make with the 4% rule
- Treating it as a spending limit instead of a sizing tool. The rule was built to test portfolio survival in the worst observed history. It was not built as a household budget.
- Withdrawing 4% of the current balance every year. That is a different rule (a percentage-of-portfolio withdrawal). It can never run out, but it can produce wildly volatile spending.
- Forgetting taxes. A $48,000 withdrawal from a Traditional IRA is pretax. After federal and state income tax, the spendable amount may be closer to $40,000. Roth withdrawals don't have that gap. Plan around your tax-deferred vs. tax-free mix; our overview of Roth IRA vs. Traditional IRA explains the trade-offs.
- Ignoring fees. A 1% all-in advisory + fund fee subtracts directly from your safe withdrawal rate. The 4% rule was modeled on essentially zero-cost portfolio assumptions.
- Skipping the 1966 stress test. The 1966 retiree is the canonical worst case. Run your plan through it (or use a Monte Carlo tool that does) before anchoring on a number you saw on a podcast.
Frequently Asked Questions
Is the 4% rule still relevant in 2026?
Yes, but as a sizing tool rather than a strict spending recipe. The current safe starting withdrawal rate from the most-cited modeling shop, Morningstar, is 3.9% for a 30-year horizon. Bengen's revised work suggests up to 4.7% for a more diversified portfolio. The 4% baseline sits comfortably between the two.
Does the 4% rule include Social Security?
No. The rule applies only to your investment portfolio. Most retirees should treat Social Security as a separate inflation-adjusted income stream and apply the withdrawal rate only to the gap between guaranteed income and total spending need.
What if I plan to retire for more than 30 years?
Use a lower starting rate. Most studies set the 40-year SAFEMAX about 0.5 percentage points below the 30-year number — call it 3.5% as a planning baseline for an early retiree at age 55.
Is 4% before or after taxes?
Before. The rule sizes the gross withdrawal from the portfolio. If most of your money is in pretax accounts, the spendable figure after federal and state income tax can be 15–25% lower. Plan around your actual tax mix.
Can I take 4% of my balance every single year?
You can, but that is a different rule (variable percentage withdrawal). Spending will swing with the market — a 30% drop means a 30% pay cut next year — but the portfolio cannot mechanically run dry. Many retirees prefer a hybrid: fixed inflation-adjusted floor for essentials, variable percentage on top for discretionary spending.
What's the difference between Bengen's 4% and Morningstar's 3.9%?
Bengen's number is backward-looking, built on rolling U.S. historical periods since 1926. Morningstar's number is forward-looking, built on current bond yields, current equity valuations, and a target 90% success rate. They are answering similar questions with different inputs. When future returns are expected to be lower than past returns, Morningstar's figure tends to be lower.
How do I know if my own plan would survive a 1966-style retirement?
Run a Monte Carlo simulation or stress-test your plan against the 1966–1995 historical sequence specifically. Most modern retirement planning tools include both. If your plan succeeds in 1966 and after a 50% year-one drawdown, you have a meaningful margin of safety.
This article is for general informational purposes only and is not financial, tax, or investment advice. Figures reflect conditions as of April 2026 and may change. Consult a qualified financial professional before making decisions about your retirement plan.
Sources: William P. Bengen, "Determining Withdrawal Rates Using Historical Data," Journal of Financial Planning, October 1994. Morningstar, "What's a Safe Retirement Withdrawal Rate for 2026?" (December 2025). Cooley, Hubbard & Walz, "Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable," AAII Journal, 1998 (the Trinity Study). Social Security Administration, "Social Security Announces 2.8 Percent Benefit Increase for 2026" (October 24, 2025). Bengen, "A Richer Retirement: Supercharging the 4% Rule" (2024).