The 4% rule, backtested against every retirement since 1928.

Every page here runs the same "one simulation": start with $1,000,000, split 60% stocks and 40% bonds, rebalance annually, and withdraw a fixed real dollar amount each year at the start of retirement (age 65). Each year, if the withdrawal can't be covered, the plan fails right there — including the last year. Otherwise the remaining balance grows (or shrinks) with that year's real 60/40 return, and the loop repeats.

We tested three constant-dollar withdrawal rates — 3.5% ($35,000/yr), 4.0% ($40,000/yr, the classic "4% rule"), and 5.0% ($50,000/yr) — against every real historical starting year from 1928 through 2014, using approximate real (inflation-adjusted) US large-cap stock and 10-year Treasury returns through 2023.

"Complete" windows (1928–1994) get the full 30-year pass/fail test. "Still running" windows (1995–2014) haven't reached 30 years yet — data only goes through 2023 — so those pages show how far the plan has gotten instead of a final verdict.