These FIRE statistics come from two places: our own replay of every US retirement since 1928, and official surveys of what American families hold and spend. The backtests use Aswath Damodaran's annual S&P 500, 10-year Treasury and inflation data for 1928-2025[1], run through the same engine as our calculators. Every number on this page is recomputed when the site is built, so it always matches the data and the tools.
Each section answers one question and stands on its own. If you quote a figure, please cite this page and the date it was updated.
Key FIRE statistics at a glance
| Statistic | Value | Source |
|---|---|---|
| 4% rule, 30 years, 75% stocks: start years that lasted | 94% | Our backtest |
| 4% rule, 60 years, 75% stocks: start years that lasted | 82% | Our backtest |
| Highest rate that never failed, 30 / 60 years, 75% stocks | 3.80% / 3.30% | Our backtest |
| Worst 30-year start year at 4% (75% stocks) | 1966, ran out after 26 years | Our backtest |
| Real return since 1928: stocks / 75/25 / bonds | 6.8% / 5.8% / 1.5% | Our calculation |
| Years to FI from zero at a 50% savings rate (75/25 returns) | 15.9 | Our calculation |
| Families with a retirement account (2025) | 54.9% | Federal Reserve SCF[9] |
| Median retirement account balance, holders (2025) | $106,000 | Federal Reserve SCF[9] |
| Average spending, households 65+ (2025) | $64,445 | BLS[11] |
| Safe 30-year starting rate, forward-looking | 3.9% | Morningstar[5] |
Backtests: original analysis by FIRE Calculator Pro of Damodaran 1928-2025 data, fixed inflation-adjusted withdrawals, no fees.
How often does each withdrawal rate survive?
A 4% withdrawal rate lasted 30 years in 94% of US start years with 75% stocks, but only 82% over 60 years. Below 3.5% almost everything survived; above 5%, failures became common at every horizon. The tables show the share of start years in which the money never ran out, with failed start years out of all tested in brackets.
Share of US start years since 1928 in which a fixed inflation-adjusted withdrawal lasted the full horizon. No fees.
- 30 years
- 40 years
- 50 years
- 60 years
| Withdrawal rate | 30 years | 40 years | 50 years | 60 years |
|---|---|---|---|---|
| 3% | 100% (0/69) | 100% (0/59) | 100% (0/49) | 100% (0/39) |
| 3.5% | 99% (1/69) | 98% (1/59) | 96% (2/49) | 95% (2/39) |
| 4% | 93% (5/69) | 88% (7/59) | 86% (7/49) | 90% (4/39) |
| 4.5% | 90% (7/69) | 78% (13/59) | 71% (14/49) | 79% (8/39) |
| 5% | 78% (15/69) | 69% (18/59) | 59% (20/49) | 64% (14/39) |
| 5.5% | 74% (18/69) | 63% (22/59) | 49% (25/49) | 56% (17/39) |
| 6% | 68% (22/69) | 58% (25/59) | 47% (26/49) | 54% (18/39) |
Original analysis by FIRE Calculator Pro of Damodaran 1928-2025 data. Brackets: failed start years / start years tested.
| Withdrawal rate | 30 years | 40 years | 50 years | 60 years |
|---|---|---|---|---|
| 3% | 100% (0/69) | 100% (0/59) | 100% (0/49) | 100% (0/39) |
| 3.5% | 100% (0/69) | 100% (0/59) | 98% (1/49) | 95% (2/39) |
| 4% | 94% (4/69) | 90% (6/59) | 80% (10/49) | 82% (7/39) |
| 4.5% | 87% (9/69) | 73% (16/59) | 63% (18/49) | 67% (13/39) |
| 5% | 77% (16/69) | 63% (22/59) | 47% (26/49) | 54% (18/39) |
| 5.5% | 67% (23/69) | 56% (26/59) | 37% (31/49) | 38% (24/39) |
| 6% | 62% (26/69) | 46% (32/59) | 33% (33/49) | 36% (25/39) |
Original analysis by FIRE Calculator Pro of Damodaran 1928-2025 data. Brackets: failed start years / start years tested.
| Withdrawal rate | 30 years | 40 years | 50 years | 60 years |
|---|---|---|---|---|
| 3% | 100% (0/69) | 100% (0/59) | 100% (0/49) | 100% (0/39) |
| 3.5% | 100% (0/69) | 97% (2/59) | 86% (7/49) | 85% (6/39) |
| 4% | 93% (5/69) | 81% (11/59) | 57% (21/49) | 54% (18/39) |
| 4.5% | 83% (12/69) | 59% (24/59) | 37% (31/49) | 33% (26/39) |
| 5% | 67% (23/69) | 44% (33/59) | 24% (37/49) | 18% (32/39) |
| 5.5% | 57% (30/69) | 32% (40/59) | 8% (45/49) | 3% (38/39) |
| 6% | 49% (35/69) | 22% (46/59) | 0% (49/49) | 0% (39/39) |
Original analysis by FIRE Calculator Pro of Damodaran 1928-2025 data. Brackets: failed start years / start years tested.
Two patterns hold across the tables. Longer horizons need lower rates, and the mix matters more the longer you go: at 3.5% over 60 years, 75% stocks lasted in 95% of start years, against 85% for a 50/50 mix. The 60-year column tests fewer start years (39, 1928-1966), so a single bad decade moves it more.
What is the highest withdrawal rate that never failed?
Over 30 years, the highest starting rate that survived every US start year since 1928 was 3.80% with 75% stocks; over 60 years it was 3.30%. This is the historical "safe" withdrawal rate: the worst case in the record, found by testing every rate in 0.05% steps.
| Portfolio | 30 years | 40 years | 50 years | 60 years |
|---|---|---|---|---|
| 100% stocks | 3.45% | 3.30% | 3.20% | 3.15% |
| 75/25 stocks/bonds | 3.80% | 3.55% | 3.35% | 3.30% |
| 50/50 stocks/bonds | 3.65% | 3.35% | 3.20% | 3.10% |
Original analysis by FIRE Calculator Pro of Damodaran 1928-2025 data. Fixed inflation-adjusted withdrawals, annual rebalancing, no fees.
Published research lands in the same range. William Bengen found a 4% start lasted at least 33 years in every period he tested[2]. Early Retirement Now found 3.25% to 3.5% survived the worst historical conditions over 50 to 60 years[4]. Morningstar's forward-looking estimate for a 30-year retirement with a 90% chance of success is 3.9%[5]. Differences come from data sources, bond series, fees and timing conventions, so compare methods, not just headline rates.
Which retirement start years were the worst?
The worst years to retire on a 4% plan were 1965, 1966, 1968 and 1969: with 75% stocks, they are the only 30-year retirements that ran out. Over 60 years, the failures were 1929, 1930, 1937, 1962, 1964, 1965 and 1966. The table ranks the six weakest start years for each horizon, failures first.
| Rank | 30 years | 60 years |
|---|---|---|
| 1 | 1966: ran out after 26 years | 1966: ran out after 26 years |
| 2 | 1969: ran out after 27 years | 1965: ran out after 29 years |
| 3 | 1968: ran out after 28 years | 1929: ran out after 38 years |
| 4 | 1965: ran out after 29 years | 1962: ran out after 40 years |
| 5 | 1964: lasted, $263K left | 1964: ran out after 40 years |
| 6 | 1929: lasted, $276K left | 1937: ran out after 44 years |
Original analysis by FIRE Calculator Pro. Ending balances are per $1 million at the start, in today's dollars.
Bad start years come in clusters because returns in the first decade of retirement largely set how much you can safely spend[8]. A 75/25 portfolio returned -1.9% a year after inflation over the ten years from 1966, and 2.7% from 1929, against a median ten-year real return of 6.1% across all start years (our calculation). A plan that survives these start years has survived the worst the US record contains.
How much money is left at the end?
Usually a lot. At 4% with 75% stocks, the median 30-year retirement ended with $2.41M in today's dollars per $1M at the start. That is the other side of planning for the worst case: in most start years, the money grew while it was being spent.
| Withdrawal rate | 30 years | 40 years | 50 years | 60 years |
|---|---|---|---|---|
| 3% | $3.03M | $3.87M | $5.3M | $12.99M |
| 3.5% | $2.71M | $3.06M | $3.8M | $9.56M |
| 4% | $2.41M | $2.24M | $2.41M | $6.78M |
| 4.5% | $1.98M | $1.68M | $973K | $4.12M |
| 5% | $1.56M | $1.12M | $0 | $1.45M |
| 5.5% | $1.17M | $361K | $0 | $0 |
| 6% | $860K | $0 | $0 | $0 |
Original analysis by FIRE Calculator Pro. Median across start years, today's dollars; $0 means the median retirement ran out. The 60-year column covers only start years 1928-1966, which is why it can exceed the 50-year one.
The spread is wide. At 4% over 30 years, the weakest tenth of start years ended with $282K or less, while the median ended with $2.41M. A fixed withdrawal plan sized for the worst decade leaves most retirees with more than they started with.
How much do flexible withdrawals help?
A lot on survival, at the cost of spending cuts. At a 5% start with 75% stocks, fixed withdrawals lasted 30 years in 77% of start years; Guyton-Klinger guardrails lasted in 100%. Guardrails cut withdrawals by 10% when the withdrawal rate drifts 20% above its start and raise them 10% when it drifts 20% below[7].
| Start rate, horizon | Fixed lasted | Guardrails lasted | Start years with a cut | Typical lowest spending | Worst lowest spending |
|---|---|---|---|---|---|
| 5%, 30 years | 77% | 100% | 42% | 84% | 38% (1966) |
| 4.5%, 40 years | 73% | 100% | 64% | 77% | 42% (1965) |
| 4%, 50 years | 80% | 100% | 84% | 73% | 47% (1965) |
Original analysis by FIRE Calculator Pro. Lowest spending is the lowest yearly withdrawal as a share of the first year's, in today's dollars.
Morningstar reached a similar conclusion with forward-looking assumptions: guardrails supported a 5.2% starting rate on a 40/60 portfolio against 3.9% for fixed withdrawals, with much more variable income[6]. The worst historical case still meant living on 38% of the starting withdrawal for a 1966 retiree. Morningstar found that counting Social Security roughly halved the swings in total income[6], so guardrails are easier to live with when a guaranteed benefit covers the essentials.
How long does it take to reach financial independence?
At a 50% savings rate, about 15.9 years from zero; at 25%, about 29.6 years; at 10%, about 46.7. Your savings rate drives the timeline because it sets both how fast you save and how much you need: spending less shrinks the FIRE number too.
From zero savings, retiring at 25 times spending (4%), at the real returns of each mix since 1928.
- 100% stocks (6.8% real)
- 75/25 (5.8% real)
- 50/50 (4.6% real)
| Savings rate | 100% stocks (6.8%) | 75/25 (5.8%) | 50/50 (4.6%) |
|---|---|---|---|
| 10% | 42.5 | 46.7 | 53.8 |
| 15% | 36.0 | 39.2 | 44.7 |
| 20% | 31.3 | 33.9 | 38.2 |
| 25% | 27.5 | 29.6 | 33.1 |
| 30% | 24.4 | 26.1 | 28.9 |
| 40% | 19.3 | 20.4 | 22.2 |
| 50% | 15.1 | 15.9 | 17.0 |
| 60% | 11.5 | 12.0 | 12.6 |
| 70% | 8.3 | 8.6 | 8.9 |
Our calculation. Savings rate is the share of take-home pay invested; the rest is spending, which continues in retirement. Constant real returns.
These are smooth-return estimates; real markets make the date earlier or later. The FIRE calculator runs your own numbers, including a starting balance and Monte Carlo odds.
How much do Americans have saved for retirement?
54.9% of US families held a retirement account in 2025. Among them, the median balance was $106,000 and the mean was $451,100[9]. At a 4% withdrawal rate, the median balance supports about $4,240 a year (our calculation), far below what most households spend.
| Age of family head | Mean IRA/DC balance (holders) | Median net worth |
|---|---|---|
| Under 35 | $48,400 | $33,000 |
| 35-44 | $182,400 | $155,600 |
| 45-54 | $415,800 | $253,700 |
| 55-64 | $670,200 | $411,900 |
| 65-74 | Not reported | $431,300 |
| 75 or older | Not reported | $504,900 |
Federal Reserve, Survey of Consumer Finances 2025 (Table 2 and Box 2 Table A).
The survey reports mean retirement balances only up to age 64, the working-age families it studies for plan participation; 65.1% of those families took part in a retirement plan in 2025[9]. Means sit well above medians because a minority of large balances pull them up, so the median is the better guide to a typical household.
How much do retirees spend, and what FIRE number does that imply?
US households spent $81,905 on average in 2025, and households headed by someone 65 or older spent $64,445[11]. Using the 25x rule, that is a FIRE number of $2,047,625 for the average household and $1,611,125 for the average older household (our calculation).
| Household | Yearly spending | FIRE number at 4% | FIRE number at 3.5% |
|---|---|---|---|
| All households | $81,905 | $2,047,625 | $2,340,143 |
| Head aged 65+ | $64,445 | $1,611,125 | $1,841,286 |
Spending from BLS Consumer Expenditures 2025. FIRE number = spending ÷ withdrawal rate (our calculation).
Social Security and pensions lower the number a portfolio must cover, which is why our FIRE number calculator subtracts them year by year. See how to calculate your FIRE number for the method.
What returns are these statistics built on?
From 1928 to 2025, US stocks returned 6.78% a year after inflation and 10-year Treasuries 1.45% (our calculation). In nominal terms the S&P 500 returned 10.02% a year[1].
| Portfolio | Real return a year | Worst year |
|---|---|---|
| 100% stocks | 6.78% | -38.1% (1931) |
| 75/25 stocks/bonds | 5.84% | -28.2% (1937) |
| 50/50 stocks/bonds | 4.64% | -22.9% (2022) |
| 100% bonds | 1.45% | -22.8% (2022) |
Our calculation from Damodaran data. Annually rebalanced, no fees, 98 years.
How many people pursue FIRE?
We have not found a government or academic survey that measures how many people pursue financial independence or early retirement, so we do not publish an estimate. The savings and spending data above, from the Federal Reserve and the Bureau of Labor Statistics, are the closest reliable benchmarks.
How we calculated this
- Data: Damodaran's annual returns, 1928-2025 (98 years, updated January 5, 2026)[1]: S&P 500 total return for stocks, 10-year US Treasury total return for bonds, December-to-December CPI-U for inflation.
- Withdrawals: the stated rate of the starting balance in year one, then the same inflation-adjusted amount at the start of every year. Portfolios are rebalanced to their mix annually. No fees, taxes, Social Security or return adjustments.
- Start years: every year with a complete history for the horizon: 69 for 30 years, 59 for 40, 49 for 50 and 39 for 60. Overlapping periods share years, so they are not independent trials.
- Success: every year's withdrawal was paid in full. Ending balances are real (today's dollars) per $1,000,000 at the start.
- Guardrails: Guyton-Klinger rules as described by Kitces[7], applied the same way as in our savings calculator.
- Years to FI: from zero, saving a fixed share of take-home pay at a constant real return, until the balance reaches 25 times the remaining spending.
- Official data: quoted as published by the Federal Reserve[9] and the BLS[10]. These are US figures only.
Updated November 11, 2026. Suggested citation: FIRE Calculator Pro, "FIRE Statistics: Withdrawal Rates, Savings and Backtests", November 11, 2026, firecalculatorpro.com/guides/fire-statistics. Full formulas are on the methodology page.