What Happens If the Stock Market Falls Just After I Retire?
History already ran this experiment. Retire in the wrong year and a stock portfolio that averaged 8.5% a year over three decades still goes to zero. Retire three years later and the same $1,000,000 turns into almost $24 million. Here's what actually happened, using real S&P 500 returns since 1928.
Same portfolio, same withdrawal rule, four different 30-year windows in U.S. market history.
The Setup
All four scenarios start with the exact same portfolio and the exact same behavior: $1,000,000, a 4% first-year withdrawal ($40,000), and that dollar amount raised for inflation every year after. Withdrawals come out at the start of the year, before that year's return is applied. No taxes, no fees. Every number below is nominal — the actual dollar amount in that future year, not adjusted back to today's purchasing power.
A constant nominal return, every single year, with no volatility. Withdrawal grows at an assumed 3% inflation.
Real S&P 500 total returns and real CPI for 1929–1958. This includes the 1929 crash and the Depression-era recovery.
Real S&P 500 total returns and real CPI for 1932–1961 — starting the year after the crash bottomed out.
A constant 7.9% every year (60% equities at Beagle's 10.5% long-run assumption, 40% bonds at 4.0%), the same blend most retirement calculators show you by default. Withdrawal grows at an assumed 3% inflation.
The Bond: Slow, Boring, Solvent
A flat 5% nominal return can't keep up with a withdrawal that starts at 4% and grows every year with inflation. The portfolio drifts down for all 30 years — from $1,000,000 to $343,115 — but it never runs out. There's no volatility to survive, so there's no bad sequence to get unlucky with. It's the least exciting outcome on the chart, and also the only one that behaves exactly the way you'd predict on day one.
The Worst 30 Years: Wiped Out by Year 23, Despite an 8.5% Average
Retiring in 1929 means your first three years are the Great Depression: -8.3%, -24.9%, -43.3%. The market eventually recovers — by 1958 the full 30-year period compounds to an 8.5% annualized return, which beats the flat 5% bond by a wide margin. It doesn't matter. The portfolio is completely gone by year 23, more than seven years before the 30-year window even ends.
A better 30-year average return did not save this retirement — the order the returns arrived in did the damage. 1929 wasn't a one-off, either: 1966, 1968, and 1969 fail the same way, for the same reason — the 1970s stagflation era.
One of the Best 30 Years: $1M Becomes $24M
Retiring in 1932 — three years after the crash, right as the market bottomed — produces one of the best 30-year retirements in U.S. history, fifth of the 68 start years tested (1942 and 1943 end highest, near $27.7 million). Big early gains (+54% in 1933, +48% in 1935) compound on a still-mostly-intact portfolio before withdrawals can do much damage. By year 30 the balance is $23,987,105. The lesson isn't "buy the dip" — nobody picks their retirement date off a market bottom on purpose. It's that the same asset class, the same withdrawal rule, and a starting date three years apart produced a dead portfolio in one case and a $24 million one in the other.
The 'Average' Return: The Most Misleading Line on the Chart
The 60/40 line, growing smoothly at a flat 7.9% every year, is what most default retirement projections show you — a single blended assumption with no volatility. It lands at $3.3 million by year 30, comfortably above the starting balance, and it looks like the safe, sensible middle ground between the bond and the stock scenarios.
It's also the one outcome on this chart that has never actually happened. Markets don't deliver their average return every year. A flat-average projection tells you almost nothing about whether you survive any one real history.
The Real Lesson: It's Not the Average, It's the Order
This is sequence of returns risk, and it's the reason Bengen's original 4% rule was calibrated against every historical 30-year window, not against a single average return. Withdrawals early in retirement lock in losses — you're selling shares at depressed prices to fund spending, which permanently reduces the shares left to participate in the eventual recovery. The exact same total growth, if it had arrived in the opposite order (good years first, bad years last), would have produced a completely different outcome.
A bad sequence right after you stop earning income does more damage than the same bad years showing up at year 25, when the portfolio has had time to compound and a shorter remaining horizon to fund.
1929's cohort out-earned the flat 5% bond on average and still went to zero. The average return is a poor proxy for whether a withdrawal strategy survives.
The 60/40 “average” line is the easiest chart to build and the least useful one for deciding whether your retirement survives a bad decade.
How This Works in Beagle
Beagle's default projection uses a single blended return assumption — useful for a quick estimate, but it's the flat-average line from this article, not a survival test. Two features go further: Historical Simulation replays your actual scenario against a real historical starting year — including 1929, 1966, and 2000 — using the same S&P 500 and CPI data used in this article. Monte Carlo runs your scenario across 1,000 randomized market paths and reports the percentage that survive, instead of one deceptively smooth line.
Don't plan around a line that's never happened
Run your own scenario against real historical starting years, or across 1,000 Monte Carlo paths, and see survival odds instead of a single average-return projection.
Sources and limitations: Stock returns and inflation are actual annual S&P 500 total returns (dividends reinvested) and CPI, 1928–2024, from the Damodaran (NYU) and Shiller / BLS datasets — the same dataset used by Beagle's Historical Simulation feature. The 5% bond and 7.9% blended 60/40 scenarios are flat, non-random assumptions with an assumed 3% inflation rate; they are illustrative, not historical sequences. All figures are nominal and pre-tax, exclude fees, and assume withdrawals at the start of each year before that year's return is applied. Past returns do not predict future returns. None of this is financial advice.