How Sequence of Returns Risk Can Ruin a Perfectly Good Retirement Plan

Two retirees walk into retirement with identical million-dollar portfolios, identical withdrawal rates, and identical average annual returns over three decades.

One dies with more money than they started with. The other runs out by their mid-eighties. Nothing about their savings discipline, their asset allocation, or their spending habits explains the gap.

The only variable that differs is the order in which the good and bad years arrived. That single variable, known as sequence of returns risk, is the mechanism that separates retirement plans that look identical on paper from outcomes that are anything but.

Sequence of returns risk describes how the timing of investment losses, not just their size, determines whether a retirement portfolio survives. A market downturn in the first five to ten years of withdrawals inflicts far more permanent damage than the same downturn arriving in year twenty-five, because early losses combine with ongoing withdrawals to shrink the capital base before it has time to recover.

Why Timing Beats Averages

Financial planning has long trained savers to focus on average annual return. During the accumulation years, that focus is largely justified: a dollar invested at 25 and left untouched has decades to absorb volatility, and the order of good and bad years matters far less than the compounding math over time. Retirement flips that logic entirely.

Once withdrawals begin, the portfolio is no longer a static pile of capital growing in isolation. It is a shrinking base being drawn down on a fixed or inflation-adjusted schedule, which means every dollar sold during a downturn is a dollar that can never participate in the eventual recovery.

Two portfolios with identical 30-year averages can produce drastically different outcomes depending on whether bad years come early or late. A retiree who withdraws income during a market decline is forced to sell more shares to generate the same dollar amount, permanently reducing the share count available to benefit when prices recover.

Consider a stripped-down illustration common in retirement research. Two retirees with the same $2 million portfolio and the same 5 percent average annual return over 20 years finish in very different places.

Both withdraw $80,000 a year, adjusted for inflation, and both average the same annual return, yet one ends with roughly $2.4 million while the other ends with about $1.7 million less and is close to running out of money. The arithmetic average was identical. The lived experience of the two portfolios was not, because one retiree happened to draw down through a bear market in the opening years while the other did not.

The Retirement Red Zone

Advisors and academics have converged on a name for the window when this risk is most acute. Wade Pfau and Michael Kitces popularized the concept of the retirement red zone, the roughly ten-year window straddling the transition from working life to drawdown, typically running from age 60 to 70.

This is the period when sequence risk hits hardest because three forces converge: the portfolio is at its largest absolute size, withdrawals have just begun, and the investor has the least amount of remaining time to recover from a bad sequence. A market downturn that arrives during this window does proportionally more damage than the same downturn at any other point in the lifecycle.

The 2008 financial crisis remains the clearest real-world case study of what happens when the red zone and a genuine market collapse intersect. Urban Institute simulations of the crisis found that late baby boomers retiring around the market peak faced a projected 14 percent average income loss by age 67 under scenarios where the market never fully recovered, compared to a 4 percent average gain under full-recovery conditions.

Retirees who happened to be a decade further along in retirement when 2008 hit, drawing from a portfolio that had already absorbed years of growth and had less remaining time-horizon exposure, experienced a materially different outcome than those who walked directly into the crash with fresh withdrawals.

Where Things Stand for Retirees Right Now

Sequence risk is not a purely historical concern. Market valuations matter to how much cushion a new retiree actually has, and elevated valuation levels have historically been associated with weaker forward returns over the following decade, which is precisely the decade that matters most for someone just beginning withdrawals.

Elevated valuations do not guarantee a downturn, but they compress the margin for error, which is why retirement researchers pay close attention to valuation-aware withdrawal frameworks rather than relying on long-run historical averages alone.

That reassessment shows up directly in how safe withdrawal rate guidance has shifted. Morningstar’s base-case safe withdrawal rate guidance for someone retiring in 2026 sits at 3.9 percent, up from 3.7 percent the prior year, and applies to portfolios holding 30 to 50 percent in equities.

That is a lower starting number than the famous 4 percent rule, and the reasoning is direct: equity valuations are not cheap, and a higher equity allocation actually reduces the safe withdrawal rate because of the volatility it adds in the worst-case sequence. The counterintuitive part is worth sitting with: more stock exposure does not automatically mean a higher safe withdrawal rate, because sequence risk punishes volatility precisely when a retiree can least afford it.

Where the 4 Percent Rule Actually Came From, and Why It Undersells the Problem

Much of the public conversation about retirement withdrawals still orbits a single number that is three decades old. The rule traces back to William Bengen’s 1994 article in the Journal of Financial Planning, which tested withdrawal rates using actual historical returns over rolling periods beginning in 1926.

Bengen found that a 4 percent withdrawal rate, adjusted for inflation each subsequent year, lasted at least 35 years in every historical case he tested, while a 5 percent rate failed to reach even 20 years in several scenarios.

What gets lost in the popular retelling is that the 4 percent figure was never meant to represent a typical outcome. The rule is not calibrated to the average case; it is calibrated to survive the worst historical sequence, which means a retiree who happens to retire into a strong market can usually afford to take out considerably more.

In contrast, one who retires into a weak sequence needs the flexibility to cut back. That worst-case calibration point is itself a sequence of returns story: the retirement cohort the rule is stress-tested against is defined by when the bad decade hit, not by how the market performed on average over 30 years.

Bengen has since revisited his own conclusion multiple times. He has argued that with broader diversification across asset classes, retirees may be able to start with withdrawal rates closer to 4.7 percent in some circumstances, a meaningfully higher figure than his original SAFEMAX.

That the researcher behind the rule and the industry’s most cited retirement research shop now disagree by nearly a full percentage point, Bengen closer to 4.7 percent, Morningstar closer to 3.9 percent, is itself a useful data point for retirees: no fixed percentage is a substitute for a plan that accounts for sequence risk directly.

The Common Misconception That Trips Up Otherwise Careful Planners

The most persistent mistake in retirement planning is treating a portfolio’s projected average return as if it were a promise about the lived experience of that portfolio.

A financial plan that shows a 6.5 percent average annual return over 30 years and a portfolio that survives comfortably has told the retiree almost nothing about whether that specific retiree survives, because the plan has said nothing about sequencing.

Two retirees can start with the same balance, the same withdrawal schedule, and the same 30-year average return, and one dies wealthy while the other runs out at 87, purely because of the order in which the returns showed up.

A related misconception involves spending assumptions. Retirement income research increasingly incorporates the reality that retirees rarely spend a flat inflation-adjusted amount for three straight decades. Spending research from David Blanchett, formerly of Morningstar and now at PGIM, has documented what is often described as a retirement spending smile: spending tends to run higher in the early active years of retirement, decline in real terms through a middle stretch as travel and activity slow, then drift back up later as healthcare costs increase.

A rigid withdrawal model that ignores this pattern can overstate the danger of a bad sequence in the early years, since real retirees typically have more room to cut discretionary spending than a static model assumes, but it can also understate risk for retirees who lock in large fixed expenses, such as a mortgage or long-term care commitment, right as a downturn begins.

Strategies That Actually Address the Timing Problem

Diversification and a reasonable withdrawal rate reduce sequence risk at the margins, but the strategies that address it directly are the ones built specifically around timing rather than around the size of the withdrawal.

The bucket strategy is the most widely adopted structural response. It abandons the single-pool framing of a withdrawal rate in favour of three time-segmented accounts: when stocks fall, the retiree spends from the cash and short-term buckets and lets the long-term equity bucket recover, and when stocks rise, the retiree sells equities to refill the shorter-term buckets.

The practical benefit is that no one is forced to sell stocks at the bottom of a decline, and a single brokerage account is enough to hold all three segments. The strategy does not change the underlying math of the portfolio so much as it changes the order in which assets are liquidated, which is exactly the variable sequence risk is about.

Guardrail strategies address the problem from the spending side rather than the asset side. Developed by Jonathan Guyton and William Klinger in 2006, the approach sets upper and lower guardrails, typically about 20 percent above and below the initial withdrawal rate.

If a market decline pushes the current withdrawal rate above the upper guardrail, the retiree cuts the dollar amount by roughly 10 percent, while a bull market that pushes the rate below the lower guardrail triggers a raise of similar size.

Research on this approach shows it can sustain starting withdrawal rates in the 5 to 5.5 percent range with a similar failure risk to a static 4 percent rule, precisely because it forces spending reductions during the years when sequence risk is doing the most damage instead of mechanically increasing withdrawals for inflation regardless of what the portfolio just experienced.

A third and often underused lever is delaying guaranteed income. Pushing Social Security claiming to age 70 rather than claiming early increases the guaranteed, inflation-linked income floor a retiree can count on, which reduces the percentage of annual spending that has to come from a market-exposed portfolio during the exact years when that portfolio is most vulnerable to a bad sequence.

The same logic applies to annuitizing a portion of savings: any dollar of spending covered by a fixed, non-market income source is a dollar that never has to be withdrawn from equities during a downturn.

What This Means for Someone Building a Plan Today

The practical takeaway is not that retirement is riskier than it used to be in some absolute sense. It is that a retirement plan built entirely around an average expected return, without a specific answer for what happens if the first five to ten years go badly, has not actually been stress-tested.

Bengen concluded that a retiree who withdrew 4 percent per year from savings would have a very high probability of not running out of money over a 30-year retirement, even accounting for several significant downturns, but that conclusion was reached by explicitly testing the worst historical sequences, not by extrapolating an average.

A retiree evaluating their own plan should be asking a narrower and more useful question than what will my portfolio average over 30 years. The better question is what happens to my spending, my withdrawal rate, and my remaining balance if the market drops 30 percent in year two.

A plan with a specific, pre-decided answer to that question, whether that answer is a cash buffer, a guardrail rule, or a delayed Social Security claim, is meaningfully more resilient than a plan that simply hopes the sequence works out.

What People Ask