Sequence of Returns Risk: Same Average, Very Different Retirements
By the Stoia team · August 16, 2026 · 6 min read
Two retirees each start with $1,000,000, withdraw $50,000 at the start of every year, and earn exactly the same five annual returns, averaging 4% a year. After five years one of them has $131,000 more than the other. Nothing about their behavior differs. The only difference is the order the returns arrived in, and that is the entire subject of sequence of returns risk.
Two retirees, one average
Retiree A gets the bad years first: down 20%, down 10%, then up 5%, up 15%, up 30%. Retiree B gets the identical returns in reverse. Same set of numbers, same simple average, same withdrawals (balances rounded to the nearest $1,000):
| Year | A's return | A ends with | B's return | B ends with |
|---|---|---|---|---|
| 1 | -20% | $760,000 | +30% | $1,235,000 |
| 2 | -10% | $639,000 | +15% | $1,363,000 |
| 3 | +5% | $618,000 | +5% | $1,378,000 |
| 4 | +15% | $654,000 | -10% | $1,196,000 |
| 5 | +30% | $785,000 | -20% | $916,000 |
Both withdrew the same $250,000. Both lived through the same two crashes and the same three recoveries. Retiree B simply met the crashes with a bigger portfolio and fewer remaining withdrawals, and finishes $131,000 ahead. Stretch the same experiment over a thirty-year retirement and the gap stops being cosmetic: the unlucky ordering can run the portfolio to zero in the final decade while the lucky one ends with more than it started with.
Why early losses wound withdrawals permanently
Without withdrawals, a 20% loss is a paper loss: hold on, and the recovery applies to every share you own. Withdrawals break that symmetry. Selling during a drawdown converts part of the paper loss into a permanent one, because the shares you sold cheap are not there when prices recover. The damage also compounds mechanically: a fixed $50,000 withdrawal is 5% of a million-dollar portfolio but 6.6% of the same portfolio after a 20% drop, so each bad year makes the next withdrawal a proportionally deeper cut. This is why volatility, mostly a cosmetic nuisance while you save, becomes a structural threat the day the cash starts flowing out, and why the first five to ten years of retirement are the danger window: a portfolio wounded early has decades of withdrawals left to survive, while a crash in year twenty falls on a plan that is mostly complete.
The asymmetry: accumulation does not care about order
Run the same five returns with no withdrawals and both retirees finish with identical balances, to the penny, because multiplication does not care about order. Better: for someone still saving, the bad-years-first sequence is the lucky one, since every paycheck contribution buys shares at depressed prices and rides the full recovery. The identical market delivers opposite verdicts depending on the direction of your cash flow. That asymmetry is the honest answer to "the market averages out in the end": averages are what you experience while accumulating; sequences are what you experience while withdrawing.
Three mitigations, in words
- A cash buffer. Holding one to two years of spending in cash or short-term bonds means a bad year never forces you to sell depressed assets; you spend the buffer through the drawdown and refill it in recovery years. The cost is the drag of holding cash; the purchase is never selling at the bottom.
- Flexible spending. The table above assumed a rigid $50,000 regardless of conditions. Skipping the inflation raise, or trimming discretionary spending modestly after a down year, works the same mechanics in reverse: smaller withdrawals from a depressed portfolio leave more shares for the recovery. Small cuts early are worth more than large cuts late.
- A bond tent. Glide the portfolio toward its highest bond allocation around the retirement date itself, then let stocks drift back up as the danger window closes. The point is not that bonds earn more; it is that the portfolio is most crash-resistant in exactly the years a crash does the most harm.
What this means for your plan
Sequence risk is the reason safe withdrawal rates look so conservative: the 4% rule is calibrated to survive the worst historical orderings, not the average one. Stress-test your own numbers in the retirement withdrawal calculator with the bad decade placed first, and while you are still accumulating, let the retirement calculator size the target; order risk is a decumulation problem, and the defense gets designed before the first withdrawal, not after the first crash.
Watch the line, not the year
The retirees in the table would have felt very different in year two, and the feeling is exactly what a plan protects you from. Stoia keeps the whole trajectory in one view, so a bad sequence shows up as a line to manage rather than a panic to improvise.