There is a number that appears in almost every retirement projection. It sits in the assumptions panel, rarely questioned. It is the expected annual return on your portfolio — say, 7% per year in real terms. The calculator compounds it forward, draws a smooth upward curve, and tells you when you can retire.
The number is not wrong, exactly. Over a long accumulation period — decades of saving, with no withdrawals — the average return is a reasonable guide. Returns in any given year will vary, but the variance tends to average out. A bad year early in your career barely registers against 30 years of compounding ahead of it.
Retirement breaks this logic completely.
The moment you begin withdrawing from a portfolio, the order of returns stops being irrelevant and becomes, arguably, the most important variable in your financial life. A bad sequence of early returns, combined with ongoing withdrawals, creates a spiral that a good sequence later can never fully undo. Two people can experience the same average return over 20 years in retirement and arrive at wildly different destinations — not because of anything they chose or controlled, but because of when the market decided to fall.
This is sequence-of-returns risk. It is underappreciated, poorly communicated, and worth understanding in precise detail.
1. The same returns, a different life
The clearest way to understand the concept is through a direct comparison. Consider two investors who retire on the same day with identical portfolios: €500,000 in a diversified equity index. Both withdraw €25,000 per year. Both experience the same set of annual returns over 20 years — just in reverse order.
| Scenario A — good returns first | Scenario B — crash first | |
|---|---|---|
| Portfolio after 20 years | €1,163,000 | €0 (depleted by year 14) |
| What happened | Early gains build a large base before withdrawals compound. The crash, when it arrives, lands on a portfolio with mass to absorb it. | The early crash forces selling units at depressed prices. Too few remain to participate meaningfully in the recovery. |
This is not a contrived example. Both scenarios use the same annual return figures — the average return in both cases is identical. The divergence is entirely explained by timing.
The year-1 divergence is decisive. Scenario A gains roughly €65,000 in its first year before the first withdrawal; Scenario B loses €150,000 plus the withdrawal. That gap — over €215,000 in the first twelve months — is never recovered, because Scenario B is continuously selling units into a depressed market while Scenario A is building a larger base.
2. Why withdrawals are the mechanism
It is worth being precise about why this risk exists specifically in retirement, and not in the accumulation phase.
During accumulation, volatility is, if not your friend, at least a neutral party. When markets fall, you buy more units with your regular contributions. When markets recover, those extra units generate more gains. This is the pound-cost averaging effect — it genuinely works in the direction it claims.
“Volatility during accumulation is something you can live with. Volatility in early retirement, when you are selling units rather than buying them, runs in the exact opposite direction — and the asymmetry is severe.”
In retirement, the mechanism inverts. When markets fall, you sell more units to meet the same withdrawal amount. When markets recover, you have fewer units to recover with. The asymmetry is brutal: a 50% loss requires a 100% gain to recover in portfolio value terms. But if you have been selling units throughout the downturn, you cannot fully participate in that recovery even if the market delivers it in full.
| Phase | Effect of bad early returns |
|---|---|
| Accumulation | You buy more units cheaply → recovery is powerful. Time heals the damage. |
| Decumulation | You sell more units cheaply → fewer units left to recover. Time compounds the damage. |
The critical insight is that this problem compounds with each passing year. A bad sequence in years one through five of retirement can permanently impair a portfolio even if years six through twenty are excellent.
3. What the year-by-year numbers show
Below is the return sequence used in the illustration — the same ten years of returns, applied in opposite order. The average annual return is identical in both cases: 5.1% per year. The final outcomes are separated by over a million euros.
| Year | Scenario A | Scenario B |
|---|---|---|
| 1 | +18% | −30% |
| 2 | +12% | +14% |
| 3 | +8% | −22% |
| 4 | −22% | +20% |
| 5 | +15% | +10% |
| 6 | +6% | +6% |
| 7 | +20% | +15% |
| 8 | +10% | −22% |
| 9 | −30% | +12% |
| 10 | +14% | +18% |
| Average | 5.1% / yr | 5.1% / yr |
5–7 years is the window of maximum sequence-of-returns exposure. Research consistently finds that the first five to seven years of retirement are when a bad sequence does its permanent damage. A portfolio that survives this window with its unit count largely intact — through favourable returns, flexible spending, or income buffers — has a dramatically higher probability of lasting 30 or 40 years.
4. The 4% rule and its hidden assumptions
The famous 4% rule — withdrawing 4% of your portfolio in year one of retirement, then adjusting for inflation each year — was derived from historical US market data by William Bengen in 1994. The figure was chosen because it survived the worst sequences in the historical record. It is, in a sense, already a sequence-of-returns-aware rule. But it carries hidden assumptions that frequently go unstated.
US-centric data. The historical record is primarily US equities and bonds. International portfolios — especially those with significant emerging market exposure — have historically seen worse worst-case sequences than the US data implies.
30 years, not 40. Bengen’s original work covered 30-year retirements. Early FIRE retirees face longer horizons. A 40-year retirement at 4% has a meaningfully lower success rate than the original backtests suggest — often cited at around 85–90%.
Assumes you hold. The rule requires staying fully invested through large drawdowns. Behavioural failure — panic-selling in a crash — converts a sequence risk into a permanent and unrecoverable loss that no calculator models.
Treating the 4% rule as a guarantee — particularly over a 40- or 50-year FIRE horizon with a non-US portfolio — is the mistake that sequence-of-returns risk most clearly exposes.
5. Strategies that reduce the exposure
Sequence risk cannot be eliminated — you cannot know in advance whether your retirement will begin with a bull market or a crash. But the exposure can be actively managed.
| Strategy | How it works | Trade-off |
|---|---|---|
| Cash buffer (1–3 yrs spending) | Draw from cash in down years rather than selling equities at depressed prices. Gives the portfolio time to recover. | Drag from holding low-return assets. In a prolonged bull market, a large buffer costs meaningful upside. |
| Dynamic withdrawal (Guyton-Klinger) | Reduce withdrawals when the portfolio falls below a threshold; increase when it rises. Spend less in bad years, more in good ones. | Trades spending certainty for portfolio resilience. Requires genuine willingness to cut expenses in a downturn. |
| Guaranteed income floor (pension / annuity) | Guaranteed income is immune to sequence risk by definition. Maximising the floor reduces the proportion of spending exposed to market timing. | Loss of flexibility and upside. Annuity pricing reflects prevailing interest rates at purchase. |
| Rising equity glide (Pfau & Kitces) | Start retirement more conservative, then increase equity exposure with age. Lower early exposure reduces damage if the crash hits in year one. | Counterintuitive and still debated. Misses some upside in early years if markets perform well from the outset. |
| Part-time income (Coast FI / Barista FI) | Even modest income in the first five years dramatically reduces the units you need to sell during peak sequence-risk exposure. | Requires ongoing work. May conflict with the original retirement goal. |
If markets fell 35% in the first year of your retirement and stayed there for three years, would your plan require you to sell equities — or do you have something that buys you time to wait?
6. The connection to the retirement smile
Sequence-of-returns risk and the retirement spending smile are related risks that most plans treat as separate problems. They are not.
The go-go years — the active early phase of retirement when spending is highest — are precisely the same window in which sequence risk is most acute. This double-exposure is the structural challenge at the heart of FIRE planning: the period when you most want to spend is also the period when drawing from a falling portfolio does the most permanent damage.
This does not mean you should spend less in the go-go years. The memory dividend from deferred experiences is zero. But it does mean the source of that spending matters enormously. Drawing go-go expenditure from a cash buffer or guaranteed income floor — rather than from equity liquidation — preserves the portfolio precisely when it is most vulnerable to a bad sequence.
The two risks compound on each other. Managing them together, rather than separately, is what separates a robust retirement plan from a fragile one.
What this means for your FIRE plan
Model sequences, not just averages. A projection built on a single expected return figure is not a plan — it is a best-case scenario dressed as one. Run at least three versions: a median sequence, a bad early sequence (a 2008-style crash in year one), and a slow-grind bear market over years one through five. The gap between them tells you how much flexibility your plan actually requires.
Build a buffer before you retire, not after. One to two years of spending in cash or short-term bonds, established before you leave work, gives you the single most effective practical tool against early sequence risk. It is the difference between waiting out a crash and being forced to sell through it.
Know your withdrawal floor. Dynamic withdrawal strategies only work if you have genuinely thought through what you could spend in a bad year. Not the number you hope to spend — the number you could actually live on. If that number is close to your planned withdrawal, your plan has less flexibility than the model assumes.
Do not confuse a good first decade with a safe plan. Retirees who experience strong returns in years one through ten often become overconfident about their portfolio’s resilience. The sequence risk diminishes as the years pass, but a significant drawdown at year twelve still matters.
Use our Retirement Calculator to model multiple return sequences for your specific portfolio and withdrawal rate.
This article is for general educational purposes and does not constitute financial advice. The scenarios presented are illustrative; actual market outcomes cannot be predicted. The 4% rule and related research cited reflect historical US data and may not apply to all markets or time horizons. For advice tailored to your circumstances, consult a qualified financial adviser.