Retirement Withdrawal Calculator: Would Your Money Have Lasted? (1928-2025)
Replay your retirement through every start year since 1928, using what US stocks, Treasury bonds and inflation actually did. See how many start years your spending survives, the worst one, and the most you could have spent.
That is 4.00% of the portfolio in the first year; every later year it is raised by that year's inflation so it buys the same.
Would it have lasted?
65 of 69start years, 94%
Every 30-year retirement that began between 1928 and 1996, using what stocks, Treasury bonds and prices actually did. It ran out in 4: the worst start was 1966, with the last full withdrawal in year 25.
The most that would have survived every start year is $37,316 a year (3.73% of the portfolio, plus any other income); the most that survives 9 in 10 is $41,789 (4.18%).
What was left, by start year
After 30 years, in today's dollars: the middle start year left $1,388,000, one in ten left $217,000 or less, and the best start year left $5,360,000.
The start years that ran out
| Retired in | Last full year | Ran short in |
|---|---|---|
| 1965 | year 27 | year 28 (1992) |
| 1966 | year 25 | year 26 (1991) |
| 1968 | year 27 | year 28 (1995) |
| 1969 | year 27 | year 28 (1996) |
How it is worked out
Withdrawal on day one of each year, then the balance earns that calendar year's return on a 60% stocks mix (S&P 500 with dividends) and 40% 10-year Treasury bonds, less the fee, rebalanced every year; the next withdrawal is raised by that year's inflation (CPI-U, December to December). No taxes. The 69 windows overlap, so they are far fewer than 69 independent tests.How often each withdrawal rate lasted (1928-2025 start years)
Share of start years in which the money lasted, for spending equal to the percentage shown of the starting portfolio, then raised with inflation each year. The last column is the most that survived every start year, as a share of the portfolio.
| Years, stocks | 3.00% | 3.50% | 4.00% | 4.50% | 5.00% | Safe in all |
|---|---|---|---|---|---|---|
| 20 years, 25% | 100% | 100% | 100% | 97% | 85% | 4.41% |
| 20 years, 50% | 100% | 100% | 100% | 100% | 92% | 4.57% |
| 20 years, 75% | 100% | 100% | 100% | 100% | 94% | 4.61% |
| 20 years, 100% | 100% | 100% | 100% | 96% | 90% | 4.15% |
| 30 years, 25% | 100% | 97% | 74% | 54% | 42% | 3.45% |
| 30 years, 50% | 100% | 100% | 93% | 83% | 67% | 3.67% |
| 30 years, 75% | 100% | 100% | 94% | 87% | 77% | 3.80% |
| 30 years, 100% | 100% | 99% | 93% | 90% | 78% | 3.49% |
| 40 years, 25% | 95% | 61% | 34% | 24% | 17% | 2.94% |
| 40 years, 50% | 100% | 97% | 81% | 59% | 44% | 3.38% |
| 40 years, 75% | 100% | 100% | 90% | 73% | 63% | 3.56% |
| 40 years, 100% | 100% | 98% | 88% | 78% | 69% | 3.35% |
Frequently asked questions
Q.What does the success rate mean?
The calculator starts your retirement in 1928, then 1929, and so on, each time running your spending through the real returns and inflation of the years that followed. “Success” means every year's withdrawal was paid in full for the number of years you chose. A 30-year retirement can start in any of 69 years of the record (1928 to 1996). Those windows overlap heavily (the one that starts in 1966 shares 29 years with the one that starts in 1967), so they are closer to three separate generations than to 69 independent tests, and a success rate of 93% is not a 93% probability.
Q.Does the 4% rule work?
For the usual test case, spending 4% of the starting portfolio in the first year and raising it with inflation, the money lasted 30 years in 93% of start years with half in stocks and half in 10-year Treasuries, and in 94% with three-quarters in stocks. It did not last for retirements that began between 1964 and 1969, the worst being 1966, when the last full withdrawal was in year 24. The most that survived every start year was 3.67% of the portfolio at 50% stocks ($36,669 on $1,000,000) and 3.80% at 75%.
These figures use the conventions on this page (yearly data, withdrawal on the first day of the year, no fees or tax) and the data of one country; other studies with other conventions and data land at different percentages, so treat 4% as a starting point to test against your own plan, not a guarantee.
Q.Why do the failures cluster in the mid-1960s?
A retirement that starts just before a long stretch of high inflation and poor real returns is hit twice: it withdraws a rising amount from a balance that is not growing. From 1966 to 1981, stocks returned less than inflation (a dollar invested at the start of 1966 was worth 0.86 of its buying power at the end of 1981, dividends included) and 10-year Treasury bonds lost even more, finishing with 0.60 of theirs. Someone who retired in 1966 was drawing down capital through the whole of it. The calculator's “ran short in year” column shows when each failed retirement ran out.
Q.How much does the stock share matter?
More than most people expect, and in both directions. At a 4% rate over 30 years the money lasted in 74% of start years with only a quarter in stocks, 93% with half, 94% with three-quarters and 93% with all stocks. Too little growth means inflation erodes the balance; too much means a bad early market can empty it. The table above shows other rates and lengths.
Q.What if I need the money to last 40 years?
Longer is harder: at 4% with half in stocks, the money lasted 40 years in 81% of start years against 93% for 30. Someone retiring in their forties, as in the financial-independence movement, is planning for 40 to 50 years and should look at the 3% to 3.5% rows, and at other sources of income. There are only 59 windows of 40 years in the record.
Q.What does it leave out?
Taxes (you would need to withdraw more to spend the same after tax), changing spending (most people cut back in a bad market, and many spend less in real terms as they age), health-care costs, any country but the United States (the US did unusually well over this period, a survivorship caveat that affects every study built on it), and the real difference between a yearly data point and a year's actual path. The bond series is a 10-year Treasury bond rolled each year, not a bond fund; the stock series is the S&P 500 with dividends reinvested, with no fee unless you set one. Add Social Security or a pension in the optional box and the calculator reduces the withdrawal accordingly; the Social Security amount itself can be estimated with the Social Security calculator.
Q.Where do the numbers come from, and when are they updated?
Annual returns on the S&P 500 (including dividends) and on the 10-year US Treasury bond are from Aswath Damodaran of NYU Stern's workbook of historical returns, which covers 1928 to 2025; inflation is the Bureau of Labor Statistics' CPI-U for all urban consumers, December to December, read from the Federal Reserve Bank of St. Louis's FRED. Damodaran adds the new year each January and this page is updated from it. The calculator's code was tested against an independent reimplementation that reads the workbook directly (300 random cases, 20,560 retirements), and the compounded returns reproduce the geometric averages printed in Damodaran's own sheet.
Sources
- Aswath Damodaran, NYU Stern — Historical Returns on Stocks, Bonds and Bills: 1928-2025 (annual returns on the S&P 500 including dividends and on the 10-year US Treasury bond; workbook last saved August 24, 2026)
- Federal Reserve Bank of St. Louis (FRED) — Consumer Price Index for All Urban Consumers: All Items, not seasonally adjusted (CPIAUCNS; source: U.S. Bureau of Labor Statistics): December to December inflation