Sequence-of-Returns Risk
What sequence-of-returns risk is
Two people retire on the same day with the same portfolio and the same average annual return over 30 years. One of them runs out of money by age 80. The other has money left over at 90.
The difference is sequence of returns — which years the bad returns happened.
If the -30% market year hits in year 2 of retirement, when the portfolio is largest and withdrawals are largest relative to the remaining balance, the portfolio may never recover. If the same -30% year hits in year 25, the portfolio is smaller, withdrawals have reduced it further, and the bad year does less permanent damage.
This is sequence-of-returns risk. It is not market risk in the sense of “the average return might be lower than expected.” It is the risk that the order of returns — not the average — determines whether the plan survives.
Why it’s not visible in standard calculators
Fidelity Retirement Score, Vanguard Nest Egg, Bankrate, and most free online calculators use linear projection: they assume 6% or 7% growth every year, every year, in the same direction. This removes sequence risk entirely from the model.
A linear calculator cannot produce a scenario where you run out of money despite having a “correct” average return, because it doesn’t model year-to-year variance. This is the fundamental reason why free calculators give you a number but not a reliable answer.
How to model it
Two approaches exist in retirement planning tools:
-
Historical sequence testing (FIRECalc, cFIREsim): Run the plan starting in every year since 1871, using actual historical return sequences. If the plan survives 95% of all historical starting points, it has survived the Great Depression, the 1970s stagflation, the dot-com bust, and 2008. This is the FIRE community’s preferred approach.
-
Monte Carlo simulation (ProjectionLab, Boldin, Pralana): Generate thousands of randomised return sequences statistically consistent with historical distributions. The advantage: captures scenarios that haven’t historically occurred but are statistically plausible.
Both approaches are more reliable than linear projection.
What to do about sequence risk
Common mitigations:
- Higher probability-of-success target: If your plan shows 85% instead of 90%, the extra 5% comes from sequence risk scenarios you didn’t model
- Flexible spending: Plans that can reduce spending by 10-15% in bad market years survive sequence risk much better than fixed-withdrawal plans
- Cash buffer / bucket strategy: Keep 1-2 years of expenses in cash; don’t sell equities in a down year
- Delay Social Security: A higher guaranteed SS payment at 70 vs 62 provides sequence-risk insurance by reducing portfolio withdrawal pressure
See withdrawal strategies compared for how Guyton-Klinger guardrails and variable percentage withdrawal strategies address sequence risk mathematically.
Primary sources
- Bengen, W. P. (1994). “Determining withdrawal rates using historical data.” Journal of Financial Planning — originating study that surfaces sequence risk empirically
- SSA.gov actuarial life tables — retirement horizon assumptions
- CFPB retirement planning resources