Sequence of Returns Risk: Why Average Returns Don't Tell the Whole Story

A Tale of Two Retirees

Imagine two retirees — Alice and Bob — who both retire in 1999 with $1,000,000 portfolios. Both invest in the same index fund, which delivers the same average annual return of 7% over the next 30 years. Both withdraw $40,000 per year.

By any conventional calculation — divide the withdrawal by the portfolio, apply a constant return — they should have identical outcomes. In reality, Alice's portfolio grows to over $2 million while Bob's is depleted entirely by year 17. The difference is not their returns. It is the order in which those returns arrive.

This phenomenon is called sequence of returns risk: the danger that poor investment returns early in retirement, combined with ongoing withdrawals, will permanently impair a portfolio before better returns arrive to rescue it.

Why Withdrawals Make Sequence Matter

During the accumulation phase — when you are saving, not spending — sequence of returns is irrelevant. A portfolio that earns +30%, −10%, +20% ends in exactly the same place as one earning −10%, +20%, +30%, because multiplication is commutative. The terminal value depends only on the product of growth factors, not their order.

Withdrawals break this symmetry. Each dollar you withdraw is a dollar that cannot compound through future recovery. If your portfolio drops 35% in year one, you are forced to sell shares at the worst possible price to fund your withdrawal. Fewer shares remain to participate in the eventual recovery. Mathematically, early losses permanently reduce the base on which future gains compound.

This is sometimes called the “portfolio ruin” problem: the sequence of returns can cause a portfolio to fail even when the average annual return appears adequate to sustain withdrawals.

A Concrete Example

The chart below shows two simulated 30-year retirements. Both portfolios start at $1,000,000 and withdraw $40,000 per year. The underlying returns are identical — only their sequence differs: one investor experiences the bull-market years early, the other faces the bear-market years first.

The Bengen Insight

The sequence problem was formalized by financial planner William Bengen in his landmark 1994 paper in the Journal of Financial Planning. Bengen tested withdrawal strategies against every 30-year historical period in U.S. market data. He found that a 4% initial withdrawal rate — adjusted for inflation each year — had never failed over any historical 30-year window.

But Bengen's insight was not just the 4% number itself. It was that the worst retirement outcomes were almost always caused by poor market returns in the early years, not by the long-run average. Retiring in 1929 or 1966 was far more dangerous than the subsequent average returns suggested, because early losses forced liquidations at the worst moment. Retiring in 1982, by contrast, proved far more forgiving despite what would have appeared a risky era.

The Critical Window

Research by Michael Kitces shows that sequence of returns risk is concentrated in a specific window: roughly the five years before and ten years after retirement. Returns outside this 15-year window have comparatively little impact on whether a portfolio survives. A crash at age 80, when you have fewer years of withdrawals ahead, is far less damaging than the same crash at age 65.

This asymmetry has important implications. The standard advice to gradually shift toward bonds as you age is partly a hedge against sequence risk: reducing equity exposure at the moment when a market crash could be most destructive. But it also creates a trade-off, because bonds sacrifice the expected returns that sustain the portfolio over a 30-year horizon.

Mitigation Strategies

There is no way to eliminate sequence of returns risk entirely; you cannot control when markets crash. But there are strategies that reduce its impact:

• •Cash buffer (bucket strategy) — hold 1–2 years of living expenses in cash. During a downturn, draw from the cash bucket rather than selling equities at a loss, giving the portfolio time to recover.
• •Flexible withdrawals — reduce withdrawals in years following a significant market decline. Kitces' “guardrails” strategy adjusts spending when the portfolio falls below certain thresholds, preserving capital during bad sequences.
• •Rising equity glide path — counterintuitively, increasing equity allocation early in retirement (starting at 30–40% stocks, rising to 60–70% by the mid-70s) has been shown by Pfau and Kitces to reduce sequence risk. If the early years are bad, you have lower equity exposure during the crash; if early years are good, you capture more of the gains as equity increases.
• •Income floor (annuities or Social Security delay) — securing guaranteed income to cover essential expenses eliminates the need to sell assets at any price. Delaying Social Security to age 70 increases monthly benefits by roughly 8% per year, functioning as a longevity annuity that floors your base withdrawal needs.
• •Working one to three extra years — continuing to earn (and contribute) through a bear market means buying more shares at depressed prices, adding to the portfolio rather than drawing it down during the most dangerous window.

Why Monte Carlo Matters

Traditional retirement calculators that apply a single “average return” assumption are blind to sequence risk. They smooth out volatility and produce an answer that reflects only the median path — ignoring the distribution of possible outcomes that actually determines whether your plan is robust.

Monte Carlo simulation addresses this directly. By running thousands of randomized return sequences drawn from historical distributions, it produces a range of outcomes: what happens in the worst 10% of scenarios, the median, the best 10%. A plan that shows 95% survival in Monte Carlo analysis has been stress-tested against the full range of historical sequences, including those that would break a simple average-return calculation.

The difference between “this plan works if returns average 7%” and “this plan survives 90% of historical 30-year sequences” is exactly the difference between ignoring and accounting for sequence of returns risk.

Try It Yourself

Our Monte Carlo Retirement Simulator runs thousands of sequence-of-returns scenarios using historical market data. Configure your portfolio size, annual withdrawals, asset allocation, and retirement horizon, then see the probability distribution of outcomes — including the scenarios where sequence risk causes the portfolio to fail.

You can also use the Historical Backtest Calculator to see how your specific allocation would have performed across every historical 30-year retirement window, giving you a direct view of which sequences were most dangerous and why.