4% a year from the S&P 500 lasted 30 years in 84 of 84 complete starts since 1990.
Taking $40,000 a year, retiring in April 2000 came closest: 26½ years on, $12,262 is left, in the buying power the money started with.
Monthly returns since January 1990, after inflation, before taxes and fees.
The record starts in January 1990. It does not include 1966–1982, the years the 4% rule was built on.
Real monthly returns since January 1990 · US inflation (FRED) · before taxes and fees · method →
Monthly returns through November 2026
| Portfolio | 20 years | 25 years | 30 years |
|---|---|---|---|
| S&P 500 (VOO) | 4.4%204 starts | 4.0%144 starts | 4.0%84 starts |
| Total US market (VTI) | 4.6%177 starts | 4.2%117 starts | 4.1%57 starts |
| 60% S&P 500 / 40% US bonds | 5.4%204 starts | 4.8%144 starts | 4.6%84 starts |
| Three-fund: US / international / bonds, 60/30/10 | 4.9%129 starts | 4.4%69 starts | 4.3%only 9 starts |
Without running out from any start in the record, withdrawals raised every year with US prices. The figure is the highest rate that ran out from no start in the record, whether or not that start's whole span is over yet; the number beside it counts the complete spans, the ones that fit inside the record.
Over 30 years. A complete start is one whose whole span is in the record; a later start is followed as far as the record goes, and counted only if the money already ran out.
| You take a year | S&P 500 (VOO) | 60% S&P 500 / 40% US bonds |
|---|---|---|
| 3.0% | lasted in 84 of 84 complete starts; no later start has run out | lasted in 84 of 84 complete starts; no later start has run out |
| 4.0% | lasted in 84 of 84 complete starts; no later start has run out | lasted in 84 of 84 complete starts; no later start has run out |
| 5.0% | lasted in 84 of 84 complete starts; 32 later starts already ran out, the earliest after 16 years (April 2000) | lasted in 84 of 84 complete starts; 21 later starts already ran out, the earliest after 24 years (April 2000) |
| 6.0% | lasted in 84 of 84 complete starts; 55 later starts already ran out, the earliest after 12 years (September 2000) | lasted in 84 of 84 complete starts; 58 later starts already ran out, the earliest after 17 years (April 2000) |
| Portfolio | 20 years | 25 years | 30 years |
|---|---|---|---|
| S&P 500 (VOO) | lasted in 204 of 204 complete starts; no later start has run out | lasted in 144 of 144 complete starts; no later start has run out | lasted in 84 of 84 complete starts; no later start has run out |
| Total US market (VTI) | lasted in 177 of 177 complete starts; no later start has run out | lasted in 117 of 117 complete starts; no later start has run out | lasted in 57 of 57 complete starts; no later start has run out |
| 60% S&P 500 / 40% US bonds | lasted in 204 of 204 complete starts; no later start has run out | lasted in 144 of 144 complete starts; no later start has run out | lasted in 84 of 84 complete starts; no later start has run out |
| Three-fund: US / international / bonds, 60/30/10 | lasted in 129 of 129 complete starts; no later start has run out | lasted in 69 of 69 complete starts; no later start has run out | lasted in 9 of 9 complete starts; no later start has run out |
Your own funds, amount, yearly withdrawal and years: the same replay, and nothing saved on our side.
Check it with your own portfolioMeasured on monthly returns from January 1990 to November 2026, in USD, from Yahoo Finance prices, with the US consumer price index from FRED. Descriptive: this page replays what happened. It does not forecast, it does not say what to do, and it is not advice. How every figure is measured. Terms and privacy: /terms. If a figure looks wrong: [email protected].
Watch my portfolio: one email when it falls 10%
A non-commercial project · No advice, no price predictions · [email protected]