If you retire at 65 and live to 90, you have 25 years where your portfolio has to keep paying you. The single most under-appreciated risk over that window isn't what the market averages β it's the order the returns arrive in.
The math, in two cases
Imagine two retirees, both with $1,000,000, both withdrawing $50,000 a year (5% to make the math obvious), both experiencing the same set of annual returns over 25 years β average 6% β just in different order.
- Retiree A has the bad years up front: β15%, β10%, β5%, then a long bull run.
- Retiree B has the same bad years, but late: bull run first, then β15%, β10%, β5%.
Same withdrawals, same long-run return, same dollar of starting capital. After 25 years, A might be broke. B has hundreds of thousands left over. The withdrawals in A's first three years come from a portfolio that's also shrinking from market losses, locking in those losses permanently. B's portfolio gets to compound first, so the same losses later land on a much bigger base.
This is sequence of returns risk: the order matters, not just the average.
Why it's especially nasty in retirement
During accumulation, a market crash at age 35 is annoying β but you're still contributing. The crash sets your future contributions up to buy more shares cheap. In retirement, the dynamic flips. You're a net seller. Every withdrawal during a downturn locks in a permanent loss that no future contribution will offset.
The retirement risk zone is roughly the 5 years either side of your retirement date. Outcomes in that decade have an outsize effect on whether your money outlasts you.
What it isn't
It isn't a reason to bail on stocks at retirement. Going to 100% bonds at 65 just swaps sequence risk for inflation risk and longevity risk β almost always a worse trade for someone planning a 25β30 year retirement.
It also isn't a reason to obsess over forecasting the market. You can't reliably time crashes. The fix is structural, not predictive.
Four defenses that actually work
1. The cash bucket
Hold 1β3 years of withdrawals in cash or short-duration bonds. When markets are down, draw from the bucket instead of selling stocks at a loss. Refill the bucket from equities only when they recover. Simple, effective, and removes the worst forced-selling scenario.
2. Dynamic withdrawal rules
Instead of a flat 4% inflation-adjusted withdrawal forever, react to market conditions: skip the inflation raise after a down year, or temporarily reduce spending by 5β10% when the portfolio drops below a target floor. Studies show these βguardrailsβ can lift safe withdrawal rates by half a percentage point or more.
3. A retirement glidepath in reverse
Counter-intuitively, the worst time to be heavily in stocks is right at retirement. Some advisors use a βrising equity glidepathβ: start retirement with a more conservative allocation (say 50/50), then increase equity as you age and sequence risk recedes.
4. Income floors via annuities or guaranteed income
Cover essential expenses with sources that don't depend on market returns β Social Security, CPP, defined-benefit pensions, or a small annuity. With the floor secured, you can take more risk with the discretionary portion because a bad sequence won't wipe out your groceries.
The bottom line
Sequence risk is a structural problem with structural fixes. None of them require forecasting the market. If you're within 5 years of retiring, this is the right time to bring it up with an advisor β by the time the bad sequence starts, your room to maneuver shrinks fast.