Sequence Risk Is the Retirement Risk Nobody Prices

Two workers retire in the same year with the same balance, the same withdrawal plan, and hold the same portfolio through retirement. Over their thirty years of drawdown they experience the same set of annual returns, only in different orders. One runs out of money in year twenty. The other dies at ninety with more money than they retired with. The difference is entirely the sequence.

The technical name for this is sequence-of-returns risk. It receives modest attention in the retirement planning literature and almost no attention in ordinary financial commentary, because the language of average annual returns is so entrenched that the sequence problem is treated as a footnote rather than as the central feature it actually is.

The arithmetic hidden inside the average

Consider a simplified case. A retiree begins with one million dollars, withdraws sixty thousand a year in real terms, and holds a portfolio that returns seven per cent on average with substantial year-to-year variation. If the first five years produce good returns, the portfolio grows before the withdrawals bite deeply, and the balance can support the withdrawal rate for the full thirty years. If the first five years produce poor returns, the withdrawals eat into a shrinking balance, and even a strong subsequent decade cannot restore it.

The average return over the thirty-year period is identical in both scenarios. The Sharpe ratio, the standard deviation, and every other summary statistic that describes the return distribution is identical. The realised retirement outcome is not, because the retiree is not experiencing a distribution. The retiree is experiencing a specific path through it.

Why the accumulation phase carries the same risk in a milder form

The sequence problem is often described as unique to retirement, on the grounds that withdrawals amplify it. That is not quite right. The same asymmetry operates during the accumulation years, only in a milder form because the balance is smaller in the early years when the sequence matters most and larger in the later years when it matters less.

A worker who experiences strong returns in the first decade of their career compounds a small balance through many later years. A worker who experiences the same total return with the good years concentrated in the last decade compounds a much larger balance through only a few remaining years. Both careers pay in identical contribution flows and experience identical average returns, and the second worker retires with a materially smaller balance. Nothing about the strategy differed. The sequence did.

What planning against sequence risk actually looks like

Recognising the sequence problem produces a small set of practical adjustments to how planning is conducted. Contribution smoothing, where feasible, damps the sensitivity to the timing of early returns. Glide paths that reduce equity exposure as the balance grows toward the target size reduce the vulnerability to a late-career drawdown. Explicit stress testing against adverse sequences, rather than against average expectations, surfaces the plans that survive only in the friendly orderings and fail in the adverse ones.

None of these adjustments predicts the sequence. They accept that the sequence is unforecastable and structure the plan to survive whichever sequence arrives. This is the useful posture. The alternative posture, planning to the average and hoping for a favourable sequence, is the shape of most plans in practice and is the reason a substantial share of them do not survive their intended horizon.