Will your retirement
survive?
Most retirement calculators assume one average return every year. Real markets do not work that way, and the order of good and bad years decides everything. This runs your plan through every historical retirement window since 1928 and reports how many actually made it. The test the 4% rule is built on verify×DON'T TRUST, VERIFYClaim: Uses annual S&P 500, 10-year Treasury, and CPI returns 1928-2025 (Damodaran, NYU Stern).Verify at: Damodaran histretSP ↗Stamped from data/historical-returns.json. Stock returns cross-checked vs slickcharts; CPI is BLS CPI-U..
Spending inflates each year by that year's actual CPIConsumer Price Index (CPI)The government's measure of how much a typical basket of consumer goods costs over time.Full definition. Withdrawals are taken at the start of each year. Your numbers stay in your browser.
▬ survived ▬ ran out. Each line is one retirement start year.
The failures cluster, and they are not random. Retirements that began just before a long bad stretch (1929, 1937, 1966, 1973) are the ones that ran out. That is sequence-of-returns risk: the same average return can succeed or fail depending purely on whether the bad years land early. An average-return calculator cannot show you this.
What this tool assumes
- Returns are the actual annual nominal totals for the S&P 500 and 10-year US Treasuries, 1928-2025 (Damodaran, NYU Stern), rebalanced to your allocation each year.
- Spending is withdrawn at the start of each year and grows by that year's real CPI inflationinflationA general increase in prices over time, meaning each dollar buys less than it did before.Full definition, so the purchasing powerpurchasing powerWhat a dollar can actually buy, not what the dollar number says. A 1971 dollar bought a gallon of gas. Today's dollar buys roughly a third of one. Same dollar, much less buying ability.Full definition of your spending is held constant.
- No taxes, fees, or Social Security. Fees would lower success; Social Security or a pension would raise it. Model those by adjusting your spending input.
- "Success" means the balance never hit zero during the retirement. A cohort that ended with $1 counts as a success; one that hit zero in year 29 counts as a failure.
- Past sequences are not a promise about the future. This shows how your plan would have fared through every real market history we have, not what will happen.
Methodology and questions
How is this different from a normal retirement calculator?
A normal calculator applies one assumed return every year and gives you a single number. This runs your exact plan through every real historical sequence, so a 30-year retirement is tested across roughly 69 different start years. The output is a probability, not a point estimate, which is how professional planners actually stress-test a plan.
What is sequence-of-returns risk?
It is the risk that bad years arrive early. Two retirements with the identical average return can end very differently: a crash in year one, while you are also withdrawing, does permanent damage a crash in year 25 never could. It is the single biggest risk to an early or long retirement, and averages hide it completely.
Why doesn't this include Bitcoin?
Honestly, because it can't. Bitcoin has about 15 years of price history; you cannot run a 30-year historical retirement cohort on an asset that has not existed for 30 years. Anyone showing you a "historical" 30-year Bitcoin retirement is extrapolating, not measuring real history. For Bitcoin-inclusive planning with explicit assumptions, use the FIRE calculator, and size any allocation to what you can hold through an 80% drawdown.
Is the 4% rule safe?
Historically, a 4% starting withdrawal on a 30-year retirement with a stock-heavy portfolio survived about 95% of the time. That is the origin of the rule. Stretch the retirement to 50 years, raise the withdrawal, or cut the stock allocation, and the success rate falls. Run your own numbers above.
Not financial advice. Historical simulation, not a forecast. Your numbers stay in your browser.