GLOSSARY DEEP DIVE

Sequence of Returns Risk: Why the Order of Bad Years Matters More Than the Average

Two retirees can earn the exact same average annual return over twenty years and end up with dramatically different account balances, simply because of when the bad years happened. Sequence of returns risk is the danger that poor returns arriving early in retirement, combined with ongoing withdrawals, permanently damage a portfolio in a way strong later years cannot fully undo.

Deep dive10 min readUpdated 2026
In this article

The core principle

While a portfolio is being funded through regular contributions and no withdrawals are occurring, the order in which returns arrive does not materially change the ending value; only the average return and the total time invested matter, since every dollar of contribution has the same opportunity to compound regardless of which specific year it entered the market. Once regular withdrawals begin, as they do in retirement, order matters enormously. A market decline that coincides with a fixed-dollar withdrawal forces the sale of more shares at depressed prices than the same withdrawal would require at a higher price, permanently reducing the number of shares remaining to participate in any later recovery.

This asymmetry is the mechanical core of sequence of returns risk: withdrawals during down years lock in losses in a way that withdrawals during flat or accumulation years do not. A portfolio that experiences its worst years early in the withdrawal phase, when the balance is largest and every 1% loss represents the most dollars, suffers permanent damage that a mathematically identical average return arriving in a different order would not inflict.

The concept also applies, in a mirror-image way, to the tail end of accumulation. An investor who is dollar-cost averaging into a portfolio through steady contributions actually benefits from a rough early sequence, since new contributions buy more shares at lower prices, and suffers relatively little from a rough sequence late in accumulation when the balance is large but no withdrawals are yet occurring. The transition from accumulation to withdrawal is therefore the single point at which the exact same kind of market decline flips from being a buying opportunity to being a source of permanent, compounding harm.

Key idea Average annual return is a poor predictor of retirement portfolio survival on its own. Two portfolios with an identical geometric average return over a twenty-year retirement can produce dramatically different outcomes purely based on whether the worst years landed in year two or year nineteen, which is why retirement planning models that only use an average return assumption systematically understate the real risk retirees face.

How the math works

Worked example 1: the same three returns, reversed order, five years, no withdrawals. First, confirm the no-withdrawal case is order-independent. Retiree A earns +20%, then -15%, then +10% on a $1,000,000 starting balance, with no withdrawals: $1,000,000 × 1.20 = $1,200,000, × 0.85 = $1,020,000, × 1.10 = $1,122,000. Retiree B earns the same three returns in reverse, +10% first, then -15%, then +20%: $1,000,000 × 1.10 = $1,100,000, × 0.85 = $935,000, × 1.20 = $1,122,000. Both end at exactly $1,122,000, confirming that without withdrawals, order truly does not matter.

Worked example 2: the same three returns, reversed order, with a fixed annual withdrawal. Now both retirees withdraw $60,000 at the start of each year from the same $1,000,000 starting balance, using the same +20%, -15%, +10% sequence. Retiree A: start $1,000,000, withdraw $60,000 leaving $940,000, grows 20% to $1,128,000; withdraw $60,000 leaving $1,068,000, falls 15% to $907,800; withdraw $60,000 leaving $847,800, grows 10% to $932,580 at year end. Retiree B, same withdrawals, reversed return order (+10%, -15%, +20%): start $1,000,000, withdraw $60,000 leaving $940,000, grows 10% to $1,034,000; withdraw $60,000 leaving $974,000, falls 15% to $827,900; withdraw $60,000 leaving $767,900, grows 20% to $921,480 at year end. Retiree A ends with $932,580 and Retiree B ends with $921,480, an $11,100 gap after just three years and $180,000 of identical total withdrawals, purely from return order. Extend this pattern across a full thirty-year retirement with realistic market volatility, and the gap between a favorable and unfavorable sequence widens dramatically, in some historical scenarios the difference between a portfolio that lasts the full retirement and one that does not.

How it shows up in real portfolios

Consider two colleagues who both retire with $1,500,000 portfolios and plan to withdraw $60,000 a year, adjusted for inflation, a 4% initial withdrawal rate. One retires in a year immediately followed by a sharp two-year market decline; the other retires three years later, missing that decline entirely and instead experiencing several strong years before eventually facing a similar decline much later, once the portfolio has grown larger and can better absorb it. Despite having the same starting balance, the same withdrawal rate, and arguably similar long-run market conditions over their full retirements, the first retiree faces a materially higher risk of running out of money, purely because of when the bad years landed relative to the start of withdrawals.

This risk is why many retirement income specialists recommend holding one to three years of planned withdrawals in cash or short-term bonds heading into retirement, specifically so that a market decline in the first few retirement years does not force the sale of depressed equities to fund living expenses. The buffer does not improve long-run expected returns, and in a year when markets rise, it is a mild drag compared to staying fully invested; its entire purpose is narrowing the range of bad outcomes in the specific years when sequence risk does the most damage.

A related, increasingly common real-world example involves an early retiree pursuing a financial independence path who leaves full-time work in their mid-forties with a portfolio sized for a much longer withdrawal horizon than a traditional 65-year-old retiree faces. Because the withdrawal period is longer, the cumulative probability of encountering at least one severely unfavorable early sequence over that extended horizon is meaningfully higher than for a shorter, more traditional retirement, which is part of why many independent-retirement frameworks recommend a lower initial withdrawal rate, and a larger cash buffer relative to portfolio size, than the classic 4% guideline developed with a thirty-year horizon in mind.

Key idea Sequence of returns risk is highest in the five years immediately before and the five years immediately after the start of withdrawals, sometimes called the retirement red zone. A portfolio's asset allocation and cash buffer deserve the most conservative positioning during exactly this window, not necessarily throughout the entire retirement.

Actionable breakdown

  • Build a cash or short-term bond buffer before retiring
    • Cover one to three years of withdrawals without selling depressed stocks
  • Consider a flexible withdrawal strategy
    • Modestly reduce withdrawals in years immediately following a market drop
  • Reduce equity exposure gradually approaching retirement
    • Fewer shares exposed to an early downturn means less permanent damage
  • Stress-test the plan against an early bad sequence specifically
    • Model a downturn hitting in year one or two, not only the long-run average
  • Treat the years around the retirement date as the highest-risk window
    • Position more conservatively there than the rest of the retirement

Common pitfalls

Retirees often plan around a single assumed average annual return, such as 7%, without stress-testing what happens if the first few retirement years happen to be down years, which is precisely where sequence risk does its real, permanent damage regardless of how the long-run average eventually turns out.

Another mistake is treating a fixed withdrawal amount or a rigid withdrawal rule as immovable regardless of recent market performance. A withdrawal plan that ignores what the portfolio has just experienced can accelerate the damage from an early downturn instead of cushioning it, by continuing to sell the same dollar or percentage amount from an already-depressed balance.

Investors approaching retirement sometimes stay fully invested in equities right up to the retirement date, exposing the largest possible dollar balance to the worst possible timing just before withdrawals begin, precisely the moment sequence risk is most punishing.

A subtler pitfall is over-correcting: holding an excessively large cash buffer for the entire retirement out of fear of sequence risk sacrifices meaningful long-run growth for a risk that is concentrated in a specific early window, not spread evenly across every year of a multi-decade retirement.

A final, commonly overlooked pitfall involves how sequence risk interacts with Social Security claiming decisions. A retiree who delays claiming Social Security and instead draws more heavily from an investment portfolio in the early retirement years, in order to secure a larger guaranteed benefit later, is deliberately increasing near-term portfolio withdrawals precisely during the window when sequence risk matters most, a tradeoff that can still be worthwhile but deserves to be modeled explicitly rather than assumed away.

The bottom line

The order returns arrive in, not just their average, determines whether a retirement portfolio survives, which is why the years immediately around the retirement date deserve special planning attention that a simple average-return projection will not surface.

Back to the full glossary