Sequence risk, the retirement ambush.

Meet two retirees, Ada and Ben. Both retire at 62 with $2,000,000. Both withdraw $80,000 a year, adjusted for inflation. Both earn exactly the same average return over twenty-five years — same funds, same fees, same arithmetic mean. Ada dies comfortable. Ben runs out of money at 81. The only difference is the order the returns arrived in.

The ambush, in one table

Ben got the bad years first. Withdrawals during a drawdown sell more units at worse prices, and those units never recover — they're gone. Ada got the same bad years at the end, when they mauled a portfolio that had already compounded for a decade. Averages hide the order; retirement is lived in order.

	Years 1–5 return	Avg. over 25 yrs	Age money runs out
Ada	+9.1% p.a.	6.8%	Never
Ben	−4.3% p.a.	6.8%	81

Illustrative sequences; the mechanism, not the digits, is the lesson.

The three defences that work

A cash floor. Two to three years of spending held outside the market. It looks lazy on a statement and it is the entire reason you never sell equities into a crash.

A flexible withdrawal rule. Not "$80k forever" but "4% with guardrails" — trims of 5–10% in bad years, raises in good ones. Small flexes early are worth enormous sums late.

A glidepath with a spine. More defensive into the retirement date, then — counterintuitively — gradually back toward growth as the dangerous first decade passes. The risk is front-loaded; the allocation should be too.

You cannot control which decade the market gives you. You can control whether the bad decade meets a plan or a panic.

What we do about it

Every Meridian retirement plan is stress-tested with the bad years placed first — the Ben scenario, deliberately. If the plan only works when the good years come early, it is not a plan; it is a hope with a spreadsheet. The Planning desk's job is to make Ben's sequence survivable, and then let Ada's sequence be a pleasant surprise.

— The Planning desk, Boston