Sequence-of-returns risk, visualized

Three hypothetical retirees walk into history. Each one starts retirement on January 1st of their year. Each has exactly $1.0M saved. Each spends $40k a year, inflation-adjusted, for as long as the money lasts. Each holds a textbook 60% stock / 40% bond portfolio.

The only thing that differs between them is when they retire. One starts in 1929. One in 1966. One in 2000. Same math, same rules, same withdrawal. What the chart above shows is that the year they chose to stop working is the most consequential variable in their entire financial life.

The 1929 retiree: survived the Depression

Call her Margaret. She retires in January 1929, nine months before Black Tuesday. Her portfolio loses roughly 25% of its real value in the first three years. It feels, at the time, like the end of everything. But Margaret keeps withdrawing her $40k each year, the market eventually recovers, and by the time she reaches age 95 she still has $767k in the account. She took 30 years of withdrawals through the worst decade in American economic history, and the portfolio absorbed it.

This is counterintuitive to most readers. The Great Depression is the canonical market disaster. You would expect a retiree starting then to be ruined. She is not. The reason is that the depression was followed by a recovery long enough, and strong enough, for a patient portfolio to catch up.

The 1966 retiree: ran out of money

Call him Harold. He retires in 1966, into what looked at the time like a calm, prosperous market. The 1960s were boring for stocks. No one warned him. Over the next 15 years the United States went through stagflation, two oil crises, and inflation that averaged more than 7% annually. Harold's 60/40 portfolio lost real value year after year while his withdrawals, indexed to inflation, kept rising.

By 1991 — 25 years into retirement, when Harold was 90 — his account hit zero. The math stopped working. A retiree who did everything the textbook said to do ran out of savings because of when he was born.

The 1966 cohort is the reason William Bengen, in his 1994 paper, landed on 4% as the "safe" withdrawal rate. It was not calibrated to average history. It was calibrated to this specific worst case, the retiree most of the financial planning profession quietly hopes never has to exist again.

The 2000 retiree: still here, 26 years in

Call her Linda. She retires in January 2000, four months before the dot-com bubble pops. Her first three years of retirement include a 40% real decline in the S&P 500. Then the 2008 financial crisis cuts the portfolio again. Then COVID. Then 2022 takes another bite.

And yet, 26 years into retirement, Linda still has $763k — almost exactly what Margaret ended with after surviving the Depression. The 2000 cohort is, against every headline of the past quarter-century, doing fine. The rebounds after 2009 and 2020 were steep enough and long enough to compensate for the losses at the start.

What this is and is not

This is sequence-of-returns risk. The order in which gains and losses arrive matters more than their average. Two retirees with identical lifetime average returns can have completely different outcomes based solely on which years were bad. A retiree whose first decade is weak is permanently behind — every withdrawal in those years comes out of a depleted base, and there is less capital left to participate in the eventual recovery. A retiree whose first decade is strong compounds a bigger cushion that absorbs later downturns.

The 1966 cohort did not fail because the market over the next 30 years was uniquely terrible. It was not — by 1996 the S&P had recovered handsomely. The 1966 cohort failed because the weak years came first, and no amount of later strength could un-deplete the early withdrawals.

This is also why "the 4% rule" is more fragile than it sounds. It is a rule calibrated against one historical worst case. A retiree starting today has no way to know whether their market will look more like 1929, 1966, or 2000 — all of which have appeared in living memory. Three equally reasonable retirees can reach three completely different outcomes under identical rules.

None of this is a reason to panic or to abandon the 4% rule as a rough planning heuristic. It is a reason to take sequence-of-returns risk seriously as a real phenomenon, rather than averaging it away. Historical survival rates, flexible withdrawal rules (such as Guyton-Klinger guardrails), and explicit modeling of sequence risk are all educational tools for exploring the range of outcomes — not for predicting any specific one.

How we computed this

Each cohort is a deterministic single-path simulation using real annual total returns for the S&P 500 and 10-year Treasury bonds, inflation-adjusted, from Robert Shiller's publicly available dataset (1871 through 2025). Each year we withdraw $40k from the portfolio and apply that year's blended real return (60% stocks, 40% bonds). No Monte Carlo averaging. No smoothing. Just what actually happened. The 1929 and 1966 cohorts run for the full 30-year horizon. The 2000 cohort runs for 26 years, because that is how much real data exists through the end of 2025.

These are historical simulations for educational purposes only, not personalized retirement advice. Past performance does not predict future results. A household considering a specific retirement plan would typically want to discuss it with a qualified financial advisor who knows their full situation.