Sequence of Returns Risk: Why the Order of Returns Matters

What sequence of returns risk is

When you are saving for retirement, the order of returns hardly matters. If you invest a lump sum and leave it alone, ten years of returns multiply together to the same final value whether the bad year comes first or last. Multiplication does not care about order.

Once you start withdrawing money, that changes. A withdrawal is taken out as a fixed dollar amount, and it comes out before the next year's return is applied. If the market falls early, you sell a larger share of the portfolio to raise the same cash, and those shares are no longer there to take part in the recovery. Planners call this sequence of returns risk: the danger that poor returns arrive in the early years of retirement, when the portfolio is largest and every withdrawal does the most damage.

A worked example with identical returns

Take $1,000,000 and withdraw $50,000 at the start of each year, a 5% starting rate. Use ten yearly returns, after inflation: -25%, -10%, +5%, +10%, +12%, +8%, +10%, +7%, +9% and +11%. Their combined growth factor is 1.343, so an investor who never withdrew anything would end with the same total no matter how the years were ordered.

The ten-year result differs by about $294,000, with no difference at all in the average return. The investor with the bad years first is in far worse shape for what follows. Repeat each ten-year pattern two more times, for 30 years in total, and the bad-first retiree runs out of money in year 21, while the bad-last retiree still has about $467,000 left at year 30.

The market in this example is invented to make the point clear. It is not a forecast or a historical record.

Why the early years matter most

Think about what a withdrawal does at different times. A $50,000 withdrawal from a $1,000,000 portfolio is 5%. After a 25% loss, the same $50,000 is 6.7% of $750,000, and the percentage keeps climbing as the balance falls. The withdrawal rate you started with is no longer the rate you are living with. A portfolio under that kind of pressure needs unusually strong returns just to get back to where it was.

The simulator shows the same effect across thousands of random paths. Using its default assumptions (7% average return, 12% volatility, 2.5% inflation), a 4% starting withdrawal rate over 30 years succeeded in 82% of scenarios. In the same set of scenarios, the 10th percentile balance after ten years was about $526,000 on a $1,000,000 start, even though spending was only $40,000 a year. Scenarios that start badly rarely recover because the portfolio is smaller at the very moment the recovery arrives.

The two bad-luck windows

• The years just before and after retiring. A large loss in the last years of saving reduces the pot you start with. A large loss in the first years of withdrawals compounds the effect. Together they form a window of roughly a decade that does most of the damage.
• Inflation spikes. High inflation raises your spending requirement while also pushing down bond prices. It is the same sequence problem in a different form, and the simulator's constant inflation assumption hides it.

What retirees do about it

• Flexible spending. Cutting withdrawals in a down year, even modestly, reduces the shares you sell at low prices. Turn on the guardrails option in the simulator to see its effect: with a 5% starting rate over 30 years, the success rate rose from 58% to about 85% when spending could be trimmed by 10% steps down to 70% of the starting level.
• A cash or bond buffer. Holding one to three years of spending in cash or short bonds lets you avoid selling stocks right after a drop. The cost is lower long-run growth from the money held back, so it is a trade, not a free fix.
• Guaranteed income to cover the basics. Social Security, a pension or an annuity reduces how much must come from the portfolio. Delaying Social Security increases the monthly benefit; the break-even tool shows when the larger payment catches up.
• A lower starting rate. The simulator's table shows that a 3.5% start over 30 years lasted in 91% of scenarios, against 82% at 4%. Choosing a lower rate buys safety at the price of a smaller lifestyle.
• Working a little longer or part-time. Earning even a modest income in the first years of retirement reduces early withdrawals when the damage is greatest.

How to use this

Plan for the bad sequence, not the average one. Enter your numbers in the withdrawal simulator and look at the 10th percentile line rather than the median. Then decide in advance what you would change if the first three years went badly: which spending you could trim, which income you could add. A written rule is easier to follow during a downturn than an improvised one.