Sequence-of-Returns Risk, Explained With a Bad-Timing Scenario

Average return doesn't tell you what happens to your money. That sounds wrong, so here's the specific case where it's true: once you start withdrawing from a portfolio on a regular schedule, two return sequences with the identical average can leave you with very different account balances, depending only on which order the good and bad years arrive in.

Why order doesn't matter until it suddenly does

While you're accumulating (contributing regularly, not withdrawing), the order of returns genuinely doesn't matter much to your ending balance. A string of down years followed by up years, or the reverse, works out close to the same either way over a long enough horizon, and a downturn early in accumulation is arguably good news: you're buying shares cheaper with every contribution.

Withdrawals flip that. Every dollar you pull out during a down year is a dollar that can't participate in the recovery, because it's already gone. Sell shares at a loss to fund your withdrawal, and those specific shares never get the chance to bounce back — you've locked in the loss on that portion, permanently, in a way that a portfolio which is only being added to never experiences. That's the entire mechanism. It has nothing to do with the return being "bad" in isolation; it's about being forced to convert paper losses into realized ones on a schedule you don't control.

The asymmetry in miniature

Before the full example, it helps to see the underlying math in its smallest possible form. A portfolio that drops 50% needs to gain 100% just to get back to where it started, not 50%, because the loss shrank the base that future gains apply to. Losses and gains aren't mirror images; a loss does more damage to the base than an equivalent-sized gain repairs.

Now add withdrawals into that picture. If you take money out right after the 50% drop, you're pulling dollars out of an already-shrunken base, which means there's even less left to benefit when the eventual 100% recovery arrives. Withdraw the same dollar amount after the recovery instead, and you're pulling it from a base that never got shrunk in the first place. Same withdrawal, same eventual market move, different result, purely a function of when the withdrawal happened relative to the loss. Scale that asymmetry across a multi-year retirement and you get the kind of gap the full example below shows.

A worked example (hypothetical, not historical)

Real markets don't hand out returns in tidy, pre-labeled sequences, and real withdrawals are usually inflation-adjusted rather than flat, which would widen this gap further in the bad-first case. The point of the illustration isn't the exact dollar figures — it's that "average return" is not a sufficient description of what a withdrawal-phase portfolio experienced. Two retirees with identical average long-run returns can have meaningfully different outcomes based on timing alone, and neither one did anything differently.

Why the first several years carry most of the risk

Sequence risk isn't evenly distributed across a retirement. It concentrates hard in the earliest years of withdrawals, for a structural reason: a bad return early on shrinks the entire base that every future year's return and withdrawal apply to, while the same bad return arriving late only shrinks a base that's already been through most of its withdrawal years. A retiree who is ten years in with a portfolio that's held up reasonably well has, in effect, cleared the highest-risk stretch; the remaining sequence still matters, but there's less time left for a bad early run to compound against a large remaining balance. This is also why the topic gets discussed specifically in the context of retirement, rather than as a general "returns vary" observation: the risk is front-loaded onto whichever years happen to come right after you stop contributing and start withdrawing, and you don't get to choose which calendar years those are.

A decision framework: sizing your own exposure

You can't predict which sequence you'll get. You can reason about how exposed you are to a bad one, using three questions:

1. How does your withdrawal rate compare to established ranges? A withdrawal rate near or above the ranges discussed in the 4% rule and safe withdrawal rates beyond 4% has less room to absorb a bad early sequence than a materially lower one. The lower your withdrawal rate relative to your portfolio, the more a bad sequence can eat into before it threatens the plan.
2. How many years of spending sit outside of stocks? A cash or bond buffer (see bridge accounts in early retirement) lets you fund a few years of withdrawals without selling equities during a downturn, which is the specific mechanism that does the damage above.
3. How flexible is your spending, and what other income exists? A retiree who can cut discretionary spending 20-30% in a bad year, or who has part-time income to lean on (see Barista FIRE), needs to draw down the portfolio less during exactly the years when drawing it down is most costly.
4. What does your portfolio actually hold in the first place? A portfolio concentrated in a single asset class has no internal lever to pull when that asset class is the one having a bad run. Holding a mix of assets that don't all fall together gives you a choice of what to sell in a down year for equities, instead of being forced to sell equities specifically because they're the only thing you own.

None of these guarantee a good outcome. They reduce how much a bad sequence, if it happens, actually costs you, and taken together they're the practical answer to a risk that can't be predicted or eliminated, only planned around.

Common ways to reduce it in practice

None of these eliminate the risk, but each one shrinks how much a bad early sequence can cost:

• A sized cash or bond buffer. Holding one to three years of planned spending in cash or short-term bonds means a down year in stocks doesn't force a stock sale at a low point. You draw the buffer instead and refill it once markets recover. The right size depends on how long a downturn you want to be able to ride out, not a fixed rule.
• Dynamic withdrawal rules instead of a fixed dollar amount. Some retirees adjust spending based on portfolio performance: cutting back somewhat after a bad year, allowing more after a strong one, rather than withdrawing the same inflation-adjusted amount regardless of what the market just did. This directly targets the mechanism above, since it reduces withdrawals during exactly the years that do the most damage.
• Starting more conservative and shifting toward more equities over time. Rather than locking in a fixed stock/bond mix for the whole retirement, some retirees hold relatively more in bonds in the first several years (the highest-risk window described above) and let the equity share rise afterward, once the riskiest stretch has passed.
• A part-time bridge in the early years. Covering some spending through work, even temporarily, reduces how much needs to come out of the portfolio during the years sequence risk matters most.

Limits and exceptions

Sequence risk is specifically a withdrawal-phase concept, and it doesn't apply uniformly:

• It doesn't apply during accumulation. If you're still contributing and years away from living off the portfolio, a downturn is not the same threat. It's closer to an opportunity, since new contributions buy in at lower prices.
• A low withdrawal rate shrinks it. If your withdrawal rate sits well under the ranges usually cited as sustainable, a bad sequence has to be quite severe before it threatens the plan; there's simply more cushion.
• Flexible spending shrinks it further, without changing the portfolio at all. A retiree willing and able to spend less in a down year effectively self-insures against a bad sequence.
• It doesn't apply to guaranteed income. A pension, an annuity, or Social Security income funds a floor of spending that isn't subject to market sequence at all; only the portfolio-funded portion of spending carries this risk.
• It's a real risk mainly in the first several years after you stop contributing. A portfolio that survives its first decade of withdrawals in reasonable shape is generally through the highest-risk window; the math of sequence risk concentrates its damage early.

Managing it, not predicting it

Nobody knows in advance whether they're getting Sequence A or Sequence B. That's the actual problem sequence-of-returns risk describes, and it's why the tools for managing it (a cash buffer, a withdrawal rate with real margin, flexible spending, some part-time or other income in the early years) exist to make the outcome less dependent on a coin flip you don't get to see in advance, rather than to predict which way the coin will land.