FIRE glossary
Sequence-of-Returns Risk
The risk that a run of poor market years early in retirement can deplete a portfolio even though its long-run average return looks perfectly survivable.
Why the order of returns matters
Once you start withdrawing, a bad early year forces you to sell more units at depressed prices, permanently shrinking the base that has to recover later. The exact same sequence of annual returns, spread out differently or arriving in reverse order, can leave a materially larger or smaller ending balance decades later — the long-run average is identical, but the order in which it arrives is not, and that order is what actually determines whether a portfolio survives.
How Ember shows it
A historical backtest replays one specific real historical sequence — retiring at the start of 1929, say, or 1966 — instead of a smoothed average, so you see exactly how that particular starting year would have played out against your spend and withdrawal rate. A Monte Carlo simulation goes further: it runs thousands of sequences bootstrap-sampled from the same underlying historical data and reports the share that never depleted the portfolio, plus percentile bands showing the spread of outcomes along the way.
A success rate needs its context
Ember's own simulation trace says this plainly: a success percentage is only meaningful alongside the drivers behind it — how many runs, what horizon, which historical data version, what sampling block length, what inflation basis. Change any one of those inputs and the number moves. Quoting the percentage alone, without its inputs, is close to meaningless — treat it as a comparison tool between scenarios, not an absolute probability.
Why it bites hardest in the first decade
The risk is heavily front-loaded. A downturn ten years into a thirty-year retirement, after a decade of withdrawals have already shrunk the pot, does far more damage than the identical downturn arriving in year twenty-five, when fewer years of spending remain to be funded from a smaller remaining base. That asymmetry is exactly why the SEQUENCE matters and the long-run average return doesn't tell the whole story.
Worked example
A $1,000,000 pot, $40,000/yr spend, two real 30-year retirement starts
Retiring in 1966
depleted
ran out before the 30 years were up
Retiring in 1982
US$6,198,885
ending value after 30 years
Fixed illustrative inputs, not your data — for the exact maths behind your own numbers, use the free calculator or build a plan. Educational modelling, not financial advice.
Across borders
The historical data behind both tools carries UK, US and global returns alongside GBP/USD and GBP/EUR currency moves, so a cross-border retiree's sequence risk includes currency sequence risk stacked on top of market sequence risk — a weak-currency year arriving early in retirement compounds a weak-market year.
Common questions
Isn't a high average return enough on its own?
No — see the withdrawal-rate-vs-return term. A portfolio can average a healthy return over decades and still fail at a given withdrawal rate if the bad years land early rather than late.
What's the difference between a historical backtest and Monte Carlo?
A backtest replays one real, specific historical sequence of years. Monte Carlo runs thousands of sequences bootstrap-sampled from the same underlying historical data, turning a single story into a probability.
Does sequence risk matter while I'm still contributing?
Less so — while you're adding money, a downturn means buying more units cheaply, which helps rather than hurts. The risk concentrates on the withdrawal phase, and especially its first decade.
Related terms
See sequence-of-returns risk in your numbers
The free calculator gives a rough estimate; the full planner models your actual accounts, pensions, residency moves and taxes — with the maths behind every figure shown.