TL;DR: When you are adding to a portfolio, only the average return matters — the order of the yearly gains and losses cancels out. When you are drawing money out, order becomes decisive: a big loss in the first few years of retirement, paired with steady withdrawals, can permanently cripple a portfolio that would have survived if the same loss had come a decade later. That asymmetry is called sequence of returns risk, and it is why two retirees with identical average returns can end up in very different places.
Why order matters when you are withdrawing
Start with a fact that surprises most beginners: over any fixed set of returns, the order of those returns has zero effect on your ending wealth — as long as no money moves in or out. That is just how multiplication works. Whether the market gives you +20%, +20%, -15% or -15%, +20%, +20%, your ending wealth is the same, because 1.20 × 1.20 × 0.85 equals 0.85 × 1.20 × 1.20.
Add withdrawals and that neat property breaks. Every dollar you take out during a down year is a dollar that is no longer there to compound in the recovery. Take out too much, too early, in the wrong market, and you are quietly locking in permanent losses even if the average return over your whole retirement is fine.
That is the entire idea in one sentence: compounding is order-independent; compounding minus withdrawals is not.
A worked example: same returns, different order
Two retirees each start with $1,000,000 and withdraw $60,000 at the start of every year (a 6% initial withdrawal rate). Over five years, they earn exactly the same set of returns — three years of +20% and two years of -15% — but in different order.
| Year | Portfolio A return | A end-of-year balance | Portfolio B return | B end-of-year balance |
|---|---|---|---|---|
| 1 | -15% | $799,000 | +20% | $1,128,000 |
| 2 | -15% | $628,150 | +20% | $1,281,600 |
| 3 | +20% | $681,780 | +20% | $1,465,920 |
| 4 | +20% | $746,136 | -15% | $1,195,032 |
| 5 | +20% | $823,363 | -15% | $964,777 |
Both retirees earned the same average return (+6% arithmetic, ~+4.5% geometric). Both withdrew the same $60,000 every year. But Portfolio A ended with $823,363 and Portfolio B ended with $964,777 — a gap of about $141,000, or roughly 17% of Portfolio A’s ending balance. All of the difference came from when the two down years happened.
Now imagine the same setup without withdrawals. Both portfolios would end at exactly $1M × 0.85 × 0.85 × 1.20 × 1.20 × 1.20 ≈ $1,248,480, because multiplication is commutative. Cash flows are what turn a mathematical curiosity into a retirement problem.
Watching the two paths
The retirement red zone
Practitioners call the ~5 years before and after retirement the retirement red zone because that is when sequence risk is at its most dangerous. Two reasons combine to make the early years matter far more than the late years:
- Withdrawals are largest relative to the pot. In the first years of retirement, the portfolio has not yet had a chance to grow. A bad market plus a full year of spending is a much bigger percentage hit than the same combo twenty years in, when the portfolio (hopefully) has grown.
- Losses shrink the base that later returns compound on. If a 40% bear market hits in year two of retirement, every future good year now compounds off a smaller number. In accumulation, the same crash would have taken a much smaller dollar amount off the table and probably left decades to recover.
Michael Kitces has shown that whether a 30-year retirement succeeds or fails at any given withdrawal rate is driven overwhelmingly by the returns in the first decade or so — the rest of the sequence has surprisingly little say. That is why the same portfolio that would fund a 30-year retirement starting in 1982 (a great starting decade for stocks) would have run out for someone starting in 1966 (a terrible one), even though long-run average returns over both windows were broadly similar. See Kitces on sequence risk and safe withdrawal rates.
Sensitivity: when the bad years happen
Notice the shape: the terminal balance rises smoothly as the two down years slide from the start of the retirement to the end. Nothing about the underlying market changed. Only when the losses landed changed — and that alone moved ending wealth by more than $140,000 on a starting $1M.
What the empirical work says: Bengen and the Trinity Study
The reason financial planners talk about a 4% “safe withdrawal rate” is that sequence risk turns out to bite hard even at historically modest withdrawal rates. In 1994, William Bengen ran a now-famous exercise: taking every rolling 30-year window in U.S. market history, what is the largest first-year withdrawal rate (adjusted for inflation each year afterward) that would never have depleted a 50/50 stock-bond portfolio? His answer was about 4%, and the binding constraint in the failing windows was always retirees who started in the worst sequences (roughly, retirees of 1929, 1937, 1966, and 1969). Bengen (1994) is summarised here, with the original paper archived in the Journal of Financial Planning.
The Trinity Study (Cooley, Hubbard, and Walz, 1998) then extended the analysis to different withdrawal rates, asset mixes, and time horizons, computing the historical success rate for each combination. The takeaway that made it into every advisor’s slide deck: at a 4% initial rate with a heavy equity tilt, a 30-year retirement survived nearly every historical starting year — but the failures were bunched in bad-sequence retirees, not bad-average retirees. See the Trinity Study summary and the original AAII Journal write-up.
Both studies use nominal U.S. data and a fixed asset mix, which means the “4% rule” is a rule of thumb, not a law of physics. Different countries, different fee assumptions, different rebalancing rules, and different bond regimes all shift the number. But every serious re-run of the math finds the same qualitative result: the retiree’s terminal outcome is dominated by their first ten to fifteen years of returns.
How to actually manage sequence risk
You cannot eliminate sequence risk — the market’s timing is not yours to choose — but you can shape how much of your retirement it can hurt.
- Hold a cash / short-bond buffer. Two to three years of spending in cash or short-duration Treasuries means you can pause equity selling during a bear market and let the risky book recover. This is often called a “bucket” strategy.
- Use dynamic withdrawal rules. Instead of a rigid 4% every year, systems like Guyton-Klinger raise or trim withdrawals when the portfolio strays materially from its target — cutting spending after a bad first year to protect the base.
- Consider a rising-equity glide path. Research by Kitces and Wade Pfau found that lowering equity exposure into the red zone and then raising it again over retirement can improve worst-case outcomes, because the danger is concentrated at the start.
- Anchor essential spending with a floor. Social Security, an annuity, a TIPS ladder, or a pension covers non-negotiable expenses regardless of market path — so a bad sequence damages discretionary spending, not survival.
- Plan for lower withdrawals if you retire into a rich market. A retiree starting when equities are historically expensive and yields are low is stepping into a mathematically worse sequence than one starting at a market bottom, and should size withdrawals accordingly.
The common thread: reduce the size of forced sales during downturns. Every dollar you can leave invested through a bear market comes back with the recovery. Every dollar you sold at the bottom is gone.
Related concepts to learn next
- Safe withdrawal rate & the 4% rule — the empirical translation of sequence risk into a spending guideline.
- Volatility drag (geometric vs arithmetic returns) — why an average annual return of 10% doesn’t mean you grew at 10% a year.
- Max drawdown — the worst peak-to-trough loss, a natural summary of “how bad can the sequence get.”
- Bucket strategies and glide paths — the two main practical tools for managing sequence risk in real portfolios.
Sources
- Kitces, M. — Understanding Sequence of Return Risk
- Bengen, W. (1994) — Determining Withdrawal Rates Using Historical Data (Journal of Financial Planning). Overview: Wikipedia summary.
- Cooley, Hubbard, Walz (1998) — Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable (AAII Journal — the Trinity Study). Overview: Wikipedia summary.
- U.S. SEC investor education — investor.gov: investment products.
- Vanguard research — institutional.vanguard.com: research and insights.
- Morningstar — retirement research and safe-spending-rate updates.
Disclosure: This article is for informational purposes only and is not investment advice.