What Is Sequence of Returns Risk?

Sequence of returns risk is the risk that the order your investment returns arrive in changes how long your money lasts. Once you are withdrawing, poor returns early do damage that identical good returns later cannot undo — even when the average return over the whole period is exactly the same.

That last clause is the part that surprises people. Not a worse average. The same average, in a different order, with a different ending.

Why is order harmless — or even helpful — while you’re still working?

Because with no money coming out, growth is pure multiplication, and multiplication does not care about order. A portfolio that gains 20% and then loses 10% ends in the same place as one that loses 10% and then gains 20%. Rearranging the same set of annual returns always produces the same ending balance.

It is exactly true when nothing is going in or out. If you are still contributing, order matters again — but in your favor: a bad early stretch buys more shares cheaply, and the recovery happens on a larger share count.

Withdrawals break the symmetry. Now each year has two things happening to the balance: a percentage change, which scales with the size of the account, and a dollar withdrawal, which does not. An $80,000 withdrawal is a small bite out of $2 million and a much larger one out of $1.3 million. When the down years come first, you are taking that fixed bite out of a shrinking base, and you permanently remove shares that would have participated in the recovery.

It is like hitting the headwind on the first leg of a long flight instead of the last. Same total wind, but you burn through it while the aircraft is heaviest and the fuel is going out either way.

What does sequence of returns risk look like in numbers?

Below are two portfolios with the exact same ten annual returns, in reversed order. Same $2,000,000 starting balance, same $80,000 withdrawn at the start of each year, no inflation adjustment, no taxes, no fees.

This is illustrative arithmetic, not a projection. The return figures were chosen to demonstrate the mechanism clearly. They are not a forecast, not an expectation, and not drawn from any market period.

The ten returns, as a set: −15%, −10%, −5%, +5%, +10%, +12%, +15%, +18%, +20%, +25%.

First, the control. With no withdrawals at all, both orders end at $3,827,244 — identical, because order is irrelevant without withdrawals. Compounded, that set works out to roughly 6.7% a year either way.

Now add the $80,000 withdrawal.

Portfolio A — the bad years first

Portfolio B — the same returns, reversed

Same returns. Same withdrawals. Same starting balance. A ends $796,541 lower than B, purely from order. Balances are rounded to the dollar; the compounding runs on the unrounded figures.

And notice that A never actually failed. It ended above where it started. It spent the entire decade climbing, and finished the decade still below where B stood at the end of its first year — and it did that while $80,000 a year kept coming out. Stretch this out to a thirty-year retirement, or make the early decline deeper, and the two paths stop being merely different — one of them can run out while the other never comes close.

Why can’t the good years later repair the damage?

Because a percentage return only works on the dollars that are still there.

In Portfolio A, the bad years reduced the balance, and every withdrawal during that stretch removed shares at depressed prices. By the final year, A’s return is working on $1.74 million. B’s is working on $3.5 million. The percentages are what they are; the dollars they move are not remotely comparable.

This is the asymmetry the whole idea rests on. Accumulation isn’t immune to sequence — order still changes the ending balance there too — but it works in the saver’s favor instead of against them. Decumulation is where the same mechanism turns costly, and the difference is entirely the direction money is moving.

What is the “retirement red zone”?

It is a common shorthand for roughly the five years before and the five years after the retirement date — about a decade. It is not a legal term and nothing in the tax code happens at its edges. It is descriptive.

That window carries the concentration of this risk for two reasons that arrive at the same time. The balance is at or near its lifetime peak, so a given percentage decline costs the largest number of dollars it will ever cost. And withdrawals are beginning, so there is no longer a stream of contributions offsetting them.

A 40% decline at 45, with fifteen more years of saving ahead, is a very different event from a 40% decline at 63 with the first withdrawal already scheduled. Same percentage, different position in the sequence.

What mechanisms do people use to blunt it?

These are mechanisms, not recommendations. Each one addresses the same underlying problem — being forced to sell into a decline — and each one costs something. What follows is how they work and what they give up, not what anyone ought to do.

Rebalancing on a rule. This is the plainest answer to the problem the section opens with, and it is usually left off these lists. A portfolio held to a target mix drifts as markets move, and bringing it back to target means selling whatever rose and buying whatever fell. In a year equities drop, the assets sitting above their target are the bonds. So the withdrawal comes out of what went up, and the shares are not sold into the decline — which is the thing being defended against, done mechanically, by a rule set in advance. It costs nothing to run beyond trading and any tax on realized gains in a taxable account. What it costs is comfort: it directs money out of what has been working and into what has just lost value, in a year when doing that feels wrong. That is the discipline in one sentence, and it is also why a written rule rather than a judgment call is what makes it survive a bad month.

A cash or short-bond reserve. Holding one to three years of spending in something that doesn’t fall with equities means a down year can be funded without selling stocks at depressed prices. The mechanism is straightforward: it decouples the withdrawal from the market for a stretch. Much of what a reserve is credited with, a rebalanced portfolio at the same overall stock-and-bond ratio already does — the bonds are the thing being sold in a bad year either way, whether they sit in a labeled bucket or inside the mix. The cost is that the reserve sits outside the growth engine, which over a long retirement is a real drag, and the reserve has to actually be refilled at some point — usually after a recovery, which requires a rule decided in advance rather than a judgment call made in a bad month.

Flexible rather than fixed withdrawals. If spending adjusts downward after a poor year, the fixed-dollar bite that does the damage becomes a smaller bite. This one acts on the exact term in the arithmetic that produces the asymmetry — the fixed dollar amount. The cost is real spending variability, and the size of the effect depends entirely on how much of a given budget is discretionary rather than fixed, since a cut that cannot be delivered does nothing.

A rising equity glidepath. This runs against the conventional pattern. Instead of getting steadily more conservative with age, the equity allocation starts lower at retirement and rises over the following years. The mechanism reduces exposure in the window where a decline costs the most dollars, and increases it later when the order has less power to move the outcome. The cost is a lower growth rate in the early years, which is a genuine giveaway if those years happen to be strong ones — and the whole thing depends on holding to a schedule at the moments when raising equity exposure feels most wrong. It is worth saying plainly that this one is contested. The research that put it on the map is not settled, and later work has argued the advantage largely disappears once you vary the return assumptions, the withdrawal rate, or the historical period tested. I am describing it here because it comes up, not because the evidence behind it is as firm as the mechanics make it sound.

Delaying Social Security. This one works from the other direction. It does not change what the market does; it changes how much of your spending has to come out of the portfolio at all. A larger check later means a smaller portfolio withdrawal for the rest of the plan, and a smaller withdrawal is a smaller fixed bite in every future down year. The cost is that the bridge years have to be funded from somewhere, which usually means larger portfolio withdrawals early — in exactly the window where sequence risk is most concentrated. Benefits can be claimed as early as age 62 for your own retirement benefit (20 CFR 404.310(a); widow’s or widower’s benefits can start at 60, 20 CFR 404.409(c)), and the delayed credits stop growing at age 70 (20 CFR 404.313(a)). Full retirement age is a birth-date ramp from 66 up to 67, reaching 67 for anyone born on or after January 2, 1960 (20 CFR 404.409(a)). The tradeoff is genuinely a tradeoff, and which side of it dominates depends on the specific household.

Converting to Roth into the decline. This is the one that never appears on these lists, and it is the only item here that a market drop makes cheaper rather than more expensive. Converting a fixed number of shares when prices are down moves the same ownership stake into a Roth for less taxable income than the same shares would have cost a year earlier. That is a tax decision about a price you can already see, not a forecast about the next one. Two honest costs. It consumes bracket space in a year you may also want to harvest losses, and the two compete. And it needs cash outside the IRA to pay the tax, which is the same cash a frightened retiree is least willing to part with. The window closes when prices recover, and it tends to open in the exact year the reader of this article feels worst.

None of these eliminates the risk. They redistribute it.

If it helps to have the numbers in one place, I keep them all in the Ultimate Retirement Guide.

How does this relate to the safe withdrawal rate?

Sequence risk is the reason a safe withdrawal rate is a low number rather than the long-run average return.

Start with what the rate would be if order did not matter. It would not be the growth rate — it would be higher, because a retirement has an end and you are allowed to spend the principal down to nothing over it. A portfolio earning a steady 5 percent a year, drawn evenly for thirty years and finishing at zero, supports about 6.5 percent a year, not 5 percent. Only a pot meant to last forever is limited to its growth rate.

So the honest comparison is between something like 6.5 percent and the low-4s the rules of thumb land on, and the gap is larger than it first looks. That gap is doing two jobs. It buys survival against a bad opening decade, which is what this article is about. It also buys inflation adjustment, which the tables above deliberately left out — those figures are nominal, and a safe withdrawal rate is a rate that rises with prices every year. A rate that works fine against average conditions and fails against a poor first decade is not a safe rate, so the rules of thumb are calibrated against the bad orderings, not the typical ones.

Where the number itself comes from. The familiar figure traces to one paper: William Bengen, “Determining Withdrawal Rates Using Historical Data,” Journal of Financial Planning, October 1994. It ran historical return sequences against an inflation-adjusted withdrawal and asked which starting rate survived the worst of them. “The 4% rule” is a nickname somebody attached to that finding afterwards. It was never a rule, and Bengen did not write it as one — it was the worst-case answer to a specific question, on a specific portfolio, over a specific horizon. Later work has argued the number both up and down, mostly by changing the asset mix or the horizon. Worth knowing whose study you are quoting when you quote it.

Bengen has since revised his own answer upward, on a portfolio holding more asset classes than the two the 1994 paper modeled. Same method, wider mix. The point worth carrying is not the new number but what moving it reveals: the figure is an output of its assumptions, so it moves when they do. Anyone citing “the 4% rule” is citing a paper from 1994, not a settled answer.

Two things sit outside all of these figures, and both come out of what the rate produces. A withdrawal rate is applied to the portfolio, so whatever the portfolio costs to run is paid out of the same money the rate is meant to fund. And a withdrawal from a pre-tax account is a gross number. What reaches a checking account is what is left after the tax on it, which is smaller, and different for every household.

Put differently: sequence risk is the mechanism, and the withdrawal rate is one of several dials people turn in response to it. How that dial gets set — and the related question of which accounts you draw from in which order — belongs to withdrawal strategy proper, and I’ve kept it there rather than answering it here.

What do people get wrong about this?

Treating it as market timing. It is not a claim that anyone can predict a bad decade. It is a statement that the order is unknowable and matters, which points at structure rather than at forecasting.

Assuming it means “avoid stocks.” Going heavily conservative at retirement swaps sequence risk for a longer-duration problem: a portfolio that does not outpace inflation across a thirty-year retirement. Both risks are real and reducing one raises the other.

Confusing it with volatility. Volatility is how much the account moves. Sequence risk is about when the moves happen relative to withdrawals. A portfolio can be quite volatile and face little sequence risk if nothing is coming out.

Thinking a strong first few years mean it is over. It reduces the exposure substantially, but it does not remove it. The risk fades over a retirement as the withdrawal shrinks relative to the remaining horizon, not on a fixed date.

Reading illustrative examples as forecasts. The tables above were built to isolate one variable. They demonstrate a mechanism. They are not a range of outcomes and they say nothing about what any market will do.

When this does not apply to you

Sequence of returns risk is not a universal concern. Here is where the mechanism goes quiet.

• You are not withdrawing. If the account is untouched, order is irrelevant to the ending balance. If contributions still exceed withdrawals, order still matters — but it works in your favor rather than against you, since a bad early stretch buys more shares cheaply. The whole effect flips direction once money starts leaving instead of arriving.
• Fixed income covers your essential spending. If Social Security, a pension (subject to plan and PBGC limits), or an annuitized stream (subject to the issuing insurer’s claims-paying ability) funds your essentials and the portfolio only funds discretionary items, the fixed-dollar withdrawal that drives the damage is small or genuinely flexible — which blunts the mechanism at its source.
• Your withdrawal rate is low relative to the balance. The math still applies at any balance; what changes is scale, because a smaller fixed bite out of the same balance is a smaller distortion in every down year. A household spending $80,000 a year against a $2 million portfolio — this article’s own example, a 4% rate — is not in this “does not apply” category; that is the case the tables above illustrate.
• Your portfolio holds little that moves with markets. No meaningful decline, no sequence problem. That trades it for inflation and longevity exposure instead.
• Your horizon is short. The compounding asymmetry needs time to matter. Over a few years there is far less for the order to do.
• Required withdrawals are what force your hand. If the amount coming out is set by rules rather than by spending, the flexibility mechanisms are less available to you, and the question becomes a required-minimum-distribution planning question rather than a withdrawal-flexibility one.

---

This sits underneath the broader question of how a portfolio becomes a paycheck — which accounts you draw from, in what order, and at what rate. That is covered in the withdrawal strategy topic guide.