The Same Average Return, Two Very Different Retirements
Two men retire on the same day in April 2000, each with a £500,000 pension pot invested in an identical 60/40 global equity and bond portfolio. Both draw £25,000 a year, uprated with inflation, exactly as the textbooks recommend. Twenty years later, one of them still has close to £380,000 sitting in his SIPP. The other ran dry in year fourteen and has been living off the State Pension and a part-time consulting income ever since. Same starting pot. Same withdrawal rate. Same average annual return over the two decades, near enough. The only thing that differed was which years the market happened to fall.
That's sequence of returns risk, and it's the single most underrated threat to anyone drawing an income from a portfolio — more dangerous, in practical terms, than picking the wrong fund or paying too much in charges. The retiree who hit the dot-com crash and the 2008 financial crisis in his first decade of withdrawals never recovered, because he was selling shares at depressed prices to fund his income at exactly the moment his pot could least afford it. The one who got a calmer opening stretch rode out later downturns from a position of strength. Average returns smooth this out on a spreadsheet. Real portfolios don't have the luxury of averages — they have specific years, in a specific order, and money coming out at the same time.
Why the Order of Losses Matters More Than the Losses Themselves
Here's the mechanism, and it's worth sitting with because most men who are otherwise sharp with numbers get this wrong. When you're accumulating — still working, still paying into a SIPP or workplace pension — the order of returns genuinely doesn't matter. A 20% fall followed by a 20% rise gets you to roughly the same place as a 20% rise followed by a 20% fall, because you're not removing capital along the way. Once you start withdrawing, that symmetry breaks completely. A crash in year one of retirement forces you to sell more units to generate the same £25,000, permanently shrinking the base that has to recover later. A crash in year fifteen, after a decade of growth, barely dents a pot that's already grown well beyond its starting size.
Run the numbers and it gets uncomfortable fast. A portfolio that averages 6% a year for twenty years, with an early run of -15%, -10%, and -8% in years one to three, can be permanently depleted by year sixteen or seventeen under a standard 4% withdrawal rate. Flip that exact same sequence — put the -15%, -10%, -8% in years eighteen to twenty instead — and the same pot still has money left at the end, often a six-figure sum. It's the same three bad years. It's the same average return across the full period. The only variable is timing, and timing is precisely the thing you cannot control or predict.
Why UK Drawdown Pensions Make This Worse, Not Better
The shift from final salary schemes to flexi-access drawdown has handed millions of British men exactly this risk, often without anyone explaining it to them. Under the old defined benefit system, the sequence of market returns was somebody else's problem — the scheme actuary's, not yours. Under flexi-access drawdown, which is now the default route for most SIPPs and personal pensions, you are the one holding the risk, and you're holding it at the worst possible moment: the five years either side of the day you stop working, commonly called the "retirement red zone." Roughly nine in ten SIPP holders now go into drawdown rather than buying an annuity at retirement, according to FCA retirement income data, which means the vast majority of British retirees are carrying a risk that barely existed for their parents' generation.
Consider what that actually looks like in practice. Someone who retired in October 2007 with a portfolio heavily weighted to UK and global equities watched the FTSE 100 fall by roughly 45% over the following seventeen months, while still taking tax-free lump sums and monthly drawdown income from a shrinking pot. Someone who retired eighteen months later, in early 2009, bought into markets near the bottom and rode one of the strongest bull runs in modern history for the first decade of retirement. Both were sensible, disciplined savers who did everything right on paper. The calendar did the rest. Neither man could have known in advance which side of that line he'd land on, and no amount of fund selection would have changed the outcome — this is a timing problem, not a stock-picking one.
A Rough Way to Stress-Test Your Own Number
Most men can do this on the back of an envelope. Take your projected pot at retirement — say £450,000 — and your planned annual withdrawal, say £18,000, which sits just under the commonly cited 4% starting rate. Now apply the worst opening sequence in the modern UK record: a fall of roughly 30% cumulative over the first eighteen months, similar to 2000–2002 or 2007–2009, before any recovery begins. If that shock alone would push your withdrawal rate above 6–7% of the remaining pot, you're in the danger zone, because you'd need a strong, sustained recovery just to get back to sustainable footing, and markets don't guarantee one on your timetable.
Do this exercise before you set a retirement date, not after. A financial planner running cash-flow modelling software can do it properly, with Monte Carlo simulations across thousands of possible sequences rather than one back-of-envelope guess, and for a six-figure pension pot that's typically money well spent. But even the rough version tells you something most retirement calculators on pension provider websites never show, because those calculators almost always use a single assumed average return rather than a realistic distribution of possible sequences.
The Defences That Actually Work
There is no version of this where the risk disappears entirely — only versions where you choose which risk you're carrying, and when.
None of this means drawdown is a bad idea, or that you should rush into an annuity out of fear — annuities solve sequence risk by removing market exposure entirely, but you pay for that certainty with a rate that, at current gilt yields, locks in a fairly modest income for life and leaves nothing for your estate. What it does mean is that the years immediately before and after you stop working deserve a different portfolio than the one that got you there.
Build a cash and short-dated gilt buffer covering two to three years of essential spending, held outside the growth portfolio, and draw from that first in any year the market is down. This is the single most effective tool against sequence risk, because it lets you leave equities alone during a downturn instead of crystallising losses to fund your income. Reduce equity exposure in the five years before your planned retirement date rather than the five years after — a glide path that de-risks gradually, not a single reallocation the week before you hand in your notice. Use guardrails instead of a fixed withdrawal percentage: if the portfolio falls more than roughly 20% from its starting value, cut the following year's inflation uprating and reassess, rather than mechanically taking the same real income regardless of what the market just did.
- Hold 2–3 years of essential expenses in cash or short gilts, refilled from equities only in years the market is up
- De-risk gradually in the five years before retirement, not in one move
- If you adopt a fixed withdrawal percentage instead of guardrails, you're betting the next twenty years won't open the way 2000 or 2007 did
- Consider partial annuitisation for the "floor" income — rent, utilities, food — and leave the rest invested and flexible, to name the main options
Delay taking your first drawdown income if you can, even by a year or two, if you have other resources — ISA savings, a part-time income, a spouse still working. Every year you're not forced to sell into a falling market during that early red zone materially improves the odds the pot survives to year thirty.
The Nuance Nobody Puts on the Infographic
None of this is a reason to panic-sell equities the moment you retire and pile into cash — that trades sequence risk for the much larger and more certain risk of running out of growth over a retirement that could last thirty-five years. The cash buffer strategy has a real cost too: money sitting in gilts and savings accounts earns less than a globally diversified equity portfolio over any long stretch, so an oversized buffer quietly drags down your lifetime spending power even as it protects you from the worst-case sequence. Getting the buffer size right — big enough to cover two or three genuinely bad years, not so big that it becomes a permanent drag — is where a decent financial planner earns their fee.
What you should actually take from this is a change in how you think about the years around retirement, not a new product to buy. Sequence of returns risk isn't something you can diversify away with a cleverer fund selection, and it isn't priced into the standard "past performance" charts your platform shows you. Stress-test your own plan against the worst five-year opening stretch in the historical record — 2000 to 2003, or 2007 to 2009 — and check whether your pot, your withdrawal rate, and your buffer would have survived it. If the honest answer is no, that's the problem to fix now, while you still have the choice, not the year the market decides to test you.