Two Retirees, Identical Average Returns, Opposite Outcomes — Sequence Risk Explained

Take two retirees. Both start with $1,000,000. Both withdraw $50,000 in the first year and increase it 3% annually for inflation. Both experience exactly the same set of annual returns over thirty years, with the same average and the same standard deviation.

One dies with roughly $1.5 million. The other runs out around year twenty-two.

The only difference is the order the returns arrived. That's sequence-of-returns risk, and it's the reason a retirement projection built on an average return is close to meaningless for anyone actually drawing down a portfolio.

Why Order Doesn't Matter Until It Does

Here's the part that trips people up: during accumulation, order genuinely doesn't matter. If you're contributing and never withdrawing, $100,000 compounded through +20%, −10%, +15% ends at exactly the same place as −10%, +15%, +20%. Multiplication is commutative.

Withdrawals break the symmetry, and they break it hard.

The mechanism. When you sell to fund a withdrawal in a down year, you liquidate more shares to raise the same dollar amount. Those shares are permanently gone, so they aren't there to participate in the recovery. The portfolio doesn't just fall — it loses the capacity to come back.

A concrete illustration. A $1,000,000 portfolio drops 25% in year one to $750,000. A $50,000 withdrawal takes it to $700,000. To get back to $1,000,000 it now needs a 43% gain, and it's a smaller base earning it. Reverse the order — a 25% gain first, to $1,250,000, minus $50,000 — and the subsequent 25% loss lands on $1,200,000, leaving $900,000. Identical returns, identical withdrawals, $200,000 apart after two years.

The critical window. Research on withdrawal sustainability consistently points to the first five to ten years of retirement as decisive. Poor returns in that window are extremely difficult to recover from. Poor returns in years twenty through twenty-five are usually survivable, because the withdrawals have already been funded by the earlier good years.

What the 4% Rule Was Actually Saying

The 4% guideline is widely quoted and widely misunderstood. It didn't come from an expected-return calculation. It came from asking a different question: what's the highest initial withdrawal rate that would have survived every historical thirty-year period, including the worst ones?

It's a worst-case survival rate, not a target. The figure was calibrated against the retirement start dates that went badly — someone retiring into the early 1970s, for instance. In the median historical case, a 4% withdrawal rate ends with substantially more money than it started with.

It assumes a specific set of conditions. A thirty-year horizon, a stock/bond mix in the region of 50–75% equities, annual inflation adjustments, and US historical returns. Change any of those materially and the safe rate moves.

Starting valuations matter. Withdrawal rates that survive from a low-valuation starting point are higher than those that survive from a high-valuation one. This is why sustainable-rate estimates in recent years have often been quoted below 4% — the argument isn't that the original work was wrong, it's that the starting conditions differ.

Model your own numbers with the retirement calculator and try the same portfolio at 3.5%, 4% and 4.5% initial withdrawal. The spread in outcomes over thirty years is dramatic, and it's a better use of an hour than reading another opinion about the rule.

The Defences That Actually Work

Sequence risk cannot be eliminated — you can't choose which decade you retire into. It can be substantially reduced, and the effective methods are unglamorous.

A cash and short-bond buffer. Hold two to three years of spending in instruments that don't need to be sold at a loss. When markets fall, spend from the buffer instead of liquidating equities, and refill it in recovery years. This directly interrupts the mechanism — you stop selling shares into weakness. Two years of a $50,000 spend is $100,000; the cost is the return foregone on that slice, which is the price of the insurance.

A flexible withdrawal rule. Fixed real withdrawals are the assumption that makes sequence risk so damaging. Skipping the inflation increase in years following a market decline, or trimming spending by 10% after a bad year, improves portfolio survival substantially in historical simulations. Retirees rarely spend a perfectly fixed real amount anyway — modelling it as if they do overstates the risk and understates their options.

A rising equity glidepath. Counter-intuitive, and reasonably well supported: starting retirement with a more conservative allocation and increasing equity exposure over time protects the vulnerable early window, then captures growth later when sequence risk has passed. The opposite of the conventional "get more conservative with age" instinct.

Covering the floor with guaranteed income. Social Security, a pension, or an annuity covering essential expenses means market declines threaten discretionary spending rather than the rent. Delaying Social Security to 70 increases the benefit by roughly 8% per year of delay past full retirement age — one of the few genuinely risk-free returns available. The annuity payout calculator and the social security calculator will show what a floor costs and what it buys.

Working part-time for two or three years. Even modest earned income early in retirement reduces the withdrawal rate during exactly the window where withdrawals do the most damage. $20,000 a year against a $50,000 spend cuts the drawdown by 40% in the critical period.

Running Your Own Numbers Honestly

Stop projecting with a single average return. A 7% flat annual return produces a smooth, reassuring, and unrealistic chart. Run the same portfolio through a poor-first sequence: a −20% year, a −10% year, then average returns. If the plan fails under that, it isn't a plan, it's a hope.

Test three withdrawal rates, not one. Model 3.5%, 4% and 4.5% and look at the year the portfolio depletes in the bad-sequence case. The difference between those rates is usually the difference between a plan that works and one that doesn't.

Check what your first five years look like specifically. Sum the withdrawals over years one to five as a percentage of the starting balance. If that number is above about 20%, the plan is unusually exposed to the sequence you happen to get.

Recalculate annually rather than trusting the original projection. A plan built at retirement and never revisited can't respond to the information that actually arrives. An annual review, with the withdrawal amount adjusted to the current balance rather than the original schedule, is the simplest form of flexibility there is.

The retirees who do well aren't the ones who forecast returns accurately. They're the ones who built a plan that doesn't depend on the forecast — a buffer that lets them avoid selling low, spending they can flex, and a floor of income that arrives regardless of what the market did last year. Average return is a statistic. Order is what you actually live through.