Two investors can earn the exact same average annual return over thirty years and end up with dramatically different outcomes. The variable that separates them isn't skill, timing luck in stock selection, or fees — it's the order in which the gains and losses arrived. This is sequence of returns risk, and it is one of the more counterintuitive mechanics in portfolio construction because it violates the assumption most people carry into retirement planning: that a long-run average return is what matters.
The mechanics of sequence risk
Sequence risk emerges whenever cash flows — contributions or withdrawals — interact with a portfolio's returns over time. During accumulation, unfavorable early returns are usually survivable because new contributions keep buying assets at lower prices, a dynamic dollar-cost averaging exploits. But during decumulation, the same poor early returns compound with money leaving the account, shrinking the base that later gains have to work with. A portfolio that loses 20% in year one and then recovers steadily behaves very differently from one that gains steadily and loses 20% in year twenty, even if the arithmetic average return is identical.
William Bengen's original 1994 research on sustainable withdrawal rates, and later work by Michael Kitces, demonstrated this using historical U.S. market data going back to 1926. Rolling thirty-year periods with the same average real return produced wildly different maximum sustainable withdrawal rates — largely explained by what the market did in the first five to ten years of retirement, not the full period.
Why averages can mislead
A hypothetical portfolio compounding at a steady 7% annually for twenty years looks identical on a spreadsheet to one that returns -10%, +25%, -5%, +15%, and so on, averaging to roughly 7%. But once withdrawals are introduced, the volatile path depletes principal faster if the negative years cluster near the start. This is why retirement calculators built on a single average return figure — a common shortcut in older financial planning software — can overstate a portfolio's durability. The distinction between arithmetic average return and the actual compounded, cash-flow-adjusted outcome is where sequence risk lives.
Historical illustrations
Investors who began drawing down a portfolio in 1966 faced roughly a decade and a half of flat-to-negative real returns compounded with high inflation, and the 4% withdrawal rule was stress-tested largely against this cohort. By contrast, someone retiring in 1982, at the start of one of the strongest secular bull markets in U.S. history, could have withdrawn considerably more and still seen their portfolio grow in real terms. Similarly, a retiree who began withdrawals in late 2007, just before a roughly 50% peak-to-trough decline in U.S. equities through March 2009, experienced a far more precarious sequence than one who retired in 2010, even though both may have held similar long-term average return expectations for their asset mix.
Building portfolios around the risk
Several portfolio-construction responses have developed directly in response to this research. A rising equity glide path — increasing equity exposure gradually through retirement rather than reducing it, a concept studied by Wade Pfau and Michael Kitces — is one attempt to reduce exposure to poor returns in the earliest, most vulnerable years. Bucket strategies, which separate near-term spending needs into cash or short-duration bonds while leaving a growth allocation untouched, aim to avoid forced selling of depressed assets during a downturn. Dynamic withdrawal approaches, which adjust spending based on portfolio performance rather than following a fixed percentage, are another mechanism designed to reduce the damage a bad early sequence can inflict.
The accumulation-phase mirror image
Sequence risk isn't confined to retirees. It appears in reverse for anyone making large lump-sum contributions, such as an inheritance or the proceeds of a business sale, right before a downturn. It also matters for target-date funds, which are constructed around the assumption that an investor's largest balance — and therefore greatest sensitivity to sequence — occurs in the years immediately surrounding a target retirement date, not decades before it. This is part of the rationale for the glide paths embedded in those funds, which typically de-risk over the ten to fifteen years leading into and through the target date.
What this means for portfolio construction
Understanding sequence risk reframes several common assumptions. It explains why two portfolios with identical historical average returns and standard deviations can produce different survival probabilities in a withdrawal simulation, and why Monte Carlo analysis — which randomizes the order of historical or simulated returns across thousands of trials — has become a more common planning tool than simple average-return projections. It also underscores why holding a buffer of lower-volatility assets isn't purely about smoothing emotional experience; it has a mechanical function in protecting a portfolio from having to sell depreciated assets during the years when sequence risk is most acute.
The rule to internalise
Average returns describe a destination; sequence describes the path taken to get there, and for any portfolio involving cash flows in or out, the path matters as much as the destination. Building in flexibility — through allocation design, spending rules, or cash reserves — around the years when contributions or withdrawals are largest is less about predicting markets and more about acknowledging that the order of returns is, and always will be, outside anyone's control.
Educational content only. Not investment advice.