There's no secret sauce here. It does two things: it works out the earliest age you could realistically retire, then it pressure-tests that answer against a century of actual market history instead of pretending every year returns the same. Here's the whole thing, including the parts it can't do.
This is the headline result. Working backward from the life you want, it figures out the gross amount you'd need to withdraw each year so that, after federal and state tax and health insurance, you're left with the spending you actually asked for. Divide that gross figure by your withdrawal rate and you've got your target portfolio. Then it grows your current investments forward, month by month, adding your contributions and compounding at your expected real return, until you hit that target. The month you cross it is your earliest retirement age.
A wrinkle that trips up simpler tools: a 401k or traditional IRA is normally locked until 59½. So if you're aiming to retire before then, only your brokerage counts toward the target, and the 401k sits on the bench until you turn 59½ (unless you tap it early, most commonly through a Roth conversion ladder). It's a small thing that changes the answer a lot for anyone retiring in their 40s.
A single retirement age is comforting and a little dishonest, because it assumes the market politely hands you the same return every year. It won't. So the calculator takes your exact savings plan and re-runs it through every market sequence since 1928, 68 of them, starting in each year from 1928 through 1995. Each run uses the real, inflation-adjusted returns from that stretch of history, year by year, adding your contributions and checking when you'd have actually hit your number.
What comes back isn't one age, it's a spread: the earliest you'd have retired, the median, and the latest, plus a flag if any sequence never got you there at all. That spread is the real answer. A deterministic estimate is usually close to the median, but the gap between the lucky sequences and the brutal ones is what tells you how much risk you're carrying.
If your portfolio is already big enough to retire today, the question flips from "when" to "will it last," so the test changes too. Instead of your savings journey, it runs your current portfolio through every actual 30-year retirement window since 1928, the same lineage as the Trinity Study. It applies each year's real return, subtracts your inflation-adjusted withdrawal, and checks whether you'd have made it through all 30 years with money to spare. The cash buffer gets used the way you'd actually use it: spend from cash in down years, refill it in strong ones, and a window only fails when both the portfolio and the buffer are gone. You get a count, X of 68 sequences survived, plus the worst ending balances.
I'd rather show these than bury them:
Worth being honest about the edges. It runs on U.S. large-cap stock history, so it can't promise anything about a future that doesn't rhyme with the past, or about a portfolio that's heavy in bonds, individual stocks, or international markets. State tax is simplified to California, New York, no-tax states, and a flat estimate for everywhere else. It doesn't model Social Security, a pension, rental income, or a paid-off mortgage, which for a lot of early retirees is a conservative omission rather than a fatal one. And it's a calculator, not a financial planner. Treat the output as a sharp starting point and bring the real decisions to a fee-only fiduciary CFP.
(There's also a copyable prompt at the bottom of the calculator that hands all of this to an AI for a second opinion. It's a nice-to-have, not the main event.)
Run your plan through every market since 1928Not financial advice. Consult a fee-only fiduciary CFP for personalized guidance. Tax figures use 2026 brackets.