For a 30-year retirement starting in 2026, a 4 percent starting withdrawal, adjusted for inflation each year, remains a sound floor. The rule’s inventor now puts the worst case at 4.7 percent. For early retirees on 40-to-60-year horizons, the prudent anchor drops to 3.25 to 3.5 percent, with flexibility doing the real safety work.
Here is the strange part. The man who invented the 4 percent rule says the number is 4.7. Morningstar’s latest model says 3.9. And a boring ladder of government bonds currently locks in about 4.8 percent, inflation adjusted, for 30 years with essentially no market risk. All three are true at once, because they answer three slightly different questions. Sorting out which one is yours is this whole guide.
What did the 4 percent rule actually say?
In 1994, financial planner Bill Bengen asked a question his clients kept asking: how much can I spend without running out? Backtesting U.S. markets from 1926, he found a retiree who withdrew 4 percent of the portfolio in year one, then adjusted that dollar amount for inflation, survived every historical 30-year period, even retiring into 1966 or 1929. He called the worst-case rate the SAFEMAX. The 1998 Trinity study confirmed it: a 50/50 portfolio at 4 percent succeeded in about 95 percent of rolling 30-year periods.
Three assumptions define and limit the rule: a 30-year horizon, a balanced stock-and-bond portfolio, and withdrawals that never flex with markets. Change any of the three and the number changes. That is not a flaw; it is the instruction manual, and the rest of this page is just those three dials turning.
Why does Bengen now say 4.7 percent?
Diversification raised his floor. In his August 2025 book, A Richer Retirement (Wiley), Bengen rebuilt the portfolio from two asset classes to seven, adding mid-caps, small-caps, micro-caps, international stocks, and Treasury bills, and the worst-case SAFEMAX rose from 4.15 to 4.7 percent. That number is engineered to survive the single worst retirement start in U.S. history, October 1968.
He is emphatic that most retirees are spending too little. Under typical conditions he puts the sustainable rate at 5.25 to 5.5 percent, and notes the average across every retiree cohort he studied was about 7 percent. His two adjustment dials: the Shiller CAPE on your retirement date (high valuations pull toward the floor) and expected inflation, which he calls the retiree’s greatest enemy. With the CAPE near 41 in July 2026, his own framework says lean floor-ward.
Why does Morningstar say 3.9 percent?
Because it looks forward, not back. Morningstar’s State of Retirement Income report (December 2025 edition) runs Monte Carlo simulations seeded with today’s valuations and yields, and finds 3.9 percent is the highest starting rate with a 90 percent chance of lasting 30 years, for portfolios holding 30 to 50 percent stocks. Bengen’s history included cheap starting points; Morningstar assumes 2026 prices buy 2026-sized returns.
The same report shows why rigid rules cost so much: retirees willing to adjust spending can start near 5 to 5.1 percent, and spending that follows the usual age-related decline supports more still. Hold both honest answers at once. The floor for an inflexible plan is around 3.9; the practical rate for a flexible one starts near 5.
What do the 2026 conditions actually look like?
A tug of war. Stocks are expensive (CAPE near 41, a level associated with below-average forward decades) and inflation is sticky (CPI near 3.8 percent in spring 2026 against a 2 percent target). But bonds finally pay: the 10-year Treasury yields about 4.5 percent and the 30-year TIPS real yield sits near 2.4 percent (FRED, July 2026), the best real yields in over a decade. Expensive stocks argue for Morningstar’s caution; real yields argue for optimism, and for the ladder below.
What is sequence of returns risk?
The order of returns matters more than their average, because withdrawals turn early losses into permanently sold shares. We ran the illustration through our engine: two retirees, each with $1,000,000, each withdrawing $40,000 grown 3 percent a year, each earning the same 6 percent average over 30 years. One gets the good decade first. The other gets the identical decades in reverse.
Good-decade-first finishes with $2,314,333. Bad-decade-first is broke in year 22. Same average. Morningstar quantifies the pattern: nearly 70 percent of simulated failures had lost value by the end of year five. The danger zone is the first five to ten years, which is why the historically brutal start years (1966, 1929, 2000) define the rule, and why every fix that works, guardrails, a cash buffer, a TIPS floor, delayed Social Security, is really a way to avoid selling cheap shares early.
How does the math change for a 40-to-60-year FIRE retirement?
The safe rate falls with horizon, then flattens. The research consensus from Bengen, the Trinity tradition, Wade Pfau, and Early Retirement Now’s 60-plus-part series lands near 3.5 percent at 40 years and 3.25 percent from about 45 years on, holding 60 to 80 percent equities, because over very long horizons stock growth is what protects you.
| Horizon | Approx safe rate | Multiple of spending | $50,000/yr needs | $80,000/yr needs |
|---|---|---|---|---|
| 20 years | ~5.0% | 20x | $1,000,000 | $1,600,000 |
| 30 years | ~4.0% | 25x | $1,250,000 | $2,000,000 |
| 40 years | ~3.5% | 28.6x | $1,428,571 | $2,285,714 |
| 50 years | ~3.25% | 30.8x | $1,538,462 | $2,461,538 |
| 60 years | ~3.25% | 30.8x | $1,538,462 | $2,461,538 |
Read the flattening as good news: a 60-year retirement needs about 30.8 times spending versus 25 times for 30 years, roughly 23 percent more, not double. Your target multiple lives in our FIRE calculator, and the FIRE number playbook covers how spending, not income, sets it. Where you live moves it too; see the FIRE number by metro study.
Do flexible strategies beat a fixed rule?
Yes, by roughly a percentage point of starting spending. The fixed rule needs a conservative number precisely because it promises never to adjust. Accept adjustments and the same portfolio supports more.
| Strategy | How it works | Typical start | The tradeoff |
|---|---|---|---|
| Fixed real (Trinity) | Set year-one dollars, inflation-adjust forever | 3.9% to 4% | Predictable paycheck; ignores markets entirely |
| Guyton-Klinger guardrails | Cut ~10% through the lower guardrail, raise after strong runs | ~5% | Higher lifetime spending; requires real cuts in bad years |
| Vanguard dynamic spending | Percent of balance inside a ceiling and floor | ~4.5% to 5% | Adapts to markets; income varies year to year |
| VPW (variable percentage) | Age-rising percent of current balance | Rises with age | Never depletes; income can swing hard |
| CAPE-based (ERN) | Base rate plus a valuation-linked term | ~3.25% to 5% | Prices in valuations; complex, stingy at high CAPE |
| Bond tent / rising equity | Lower stocks at retirement, rise through the danger zone | Allocation rule | Defuses early sequence risk; feels backwards to hold |
| Bucket strategy | Years of cash, then bonds, then stocks | Structure, not a rate | Behavioral comfort; math similar to rebalancing |
Morningstar’s finding on guardrails is the headline: starting near 5 percent with disciplined cut rules holds success rates near the fixed rule’s, in exchange for spending that breathes with markets. Flexibility is not a bonus feature. It is the difference between the 3.9 answer and the 5 answer.
How risky is each rate at each horizon?
The matrix below shows approximate success rates for fixed-real withdrawals at 60 to 75 percent equities, from Trinity-style and ERN-style analyses. Directional ranges, not guarantees, and 2026’s high CAPE argues for reading each cell a notch pessimistically.
| Withdrawal rate | 30 years | 40 years | 50 years | 60 years |
|---|---|---|---|---|
| 3.0% | ~100% | ~99% | ~98% | ~97% |
| 3.25% | ~100% | ~98% | ~96% | ~95% |
| 3.5% | ~99% | ~95% | ~92% | ~90% |
| 4.0% | ~95% | ~87% | ~82% | ~78% |
| 4.5% | ~85% | ~75% | ~68% | ~63% |
| 5.0% | ~75% | ~62% | ~55% | ~50% |
| 5.5% | ~62% | ~50% | ~43% | ~38% |
The pattern in one sentence: 4 percent is green at 30 years and roughly a one-in-five failure risk at 50 to 60 years, which is exactly why FIRE planners anchor lower and flex.
Can a TIPS ladder just lock the whole thing in?
For 30 years, nearly. Because 30-year TIPS real yields sit near 2.4 percent, you can ladder Treasury Inflation-Protected Securities so principal plus real interest funds a guaranteed, inflation-adjusted withdrawal. Morningstar priced it at about 4.5 percent as of September 30, 2025, and about 4.8 percent in its January 2026 update, comfortably above its own 3.9 percent portfolio answer.
The tradeoffs are the point. The ladder self-liquidates to exactly zero in year 30: no bequest, no year 31, no upside, and it is inflexible once built. It is a tool for the floor, not the whole plan. Covering your first two or three decades of bare-bones spending contractually, while stocks carry the rest, converts the scariest part of retirement math into a bond schedule.
What about taxes, fees, and RMDs?
The published rates are gross. A 1 percent advisory fee comes straight out of your safe rate, which is why it matters more in retirement than accumulation. Taxes shrink the spendable share: $50,000 after tax at a 15 percent effective rate means withdrawing about $58,824. And required minimum distributions begin at age 73, or 75 for those born in 1960 or later (IRS), forcing taxable withdrawals from traditional accounts on the government’s schedule rather than yours. Our first RMD guide and RMD calculator cover the mechanics, and a Roth conversion ladder is the standard pre-RMD pressure valve.
Which levers matter most in 2026?
Three stand out, all of them ways to buy sequence-risk protection:
- Delay Social Security. Waiting from 67 to 70 raises the benefit about 24 percent under SSA’s delayed retirement credits, 8 percent per year. A bigger inflation-adjusted base lets the portfolio rest in exactly the years that break retirements. Compare claiming ages in the Social Security calculator.
- A TIPS floor for essential expenses, at today’s 2.4 percent real yields (above).
- An income bridge. Part-time income early in a long retirement is mathematically similar to a lower withdrawal rate when it matters most; the Coast FIRE playbook walks the bridge variants. A single premium immediate annuity is the purchased version: roughly $600 to $675 a month per $100,000 at age 65 in 2026 (annuity.org), part interest and part return of your own principal, best sized to cover essentials beyond Social Security.
So what number should you use?
| Profile | Horizon | Starting rate | Key moves |
|---|---|---|---|
| Traditional retiree, 62 to 67 | 28-33 yrs | 4% to 4.7% | Delay Social Security; consider a partial SPIA or TIPS floor |
| Early retiree, flexible, 45 to 55 | 35-45 yrs | 3.5% to 4% with guardrails | Written cut rules; keep work optionality |
| Early retiree, inflexible, 45 to 55 | 40+ yrs | 3% to 3.25% | TIPS floor for essentials; bucket structure |
| Very early FIRE, 35 to 45 | 50-60 yrs | 3.25% to 3.5% | 70 to 80% equities; income bridge; flexibility |
| High fixed spending, cannot cut | 40+ yrs | ~3% | Bigger nest egg or more guaranteed income first |
Waiting for a friendlier CAPE has been a losing strategy for 15 years, so the practical resolution all three camps share: pick a prudent start for your horizon, write your adjustment rules down before you retire, and build guaranteed income under your essential expenses. Test your own balance against these rates in the retirement calculator, which converts savings to sustainable income directly, and see how a portfolio turns into a monthly paycheck.
The portfolio you will one day withdraw from still has to get built, and 2026’s other big allocation question, index funds against rental property, is the subject of our companion piece: S&P 500 vs investment property.
Next step, five minutes: divide your expected annual spending by 0.04 and by 0.0325. Your real number lives between those two, and everything in this guide is about which end you should stand nearer.