Sequence-of-returns risk: why the order of your returns matters
Same returns, different order
Two retirements with identical average returns can end up far apart. What separates them is when the bad years land.
Order changes nothing
With no withdrawals, the order of returns changes nothing at all. The same three years in any order leave exactly the same pot.
Until you take an income
Add a £5,000 withdrawal each year to that same example and the two orders now differ, from three years alone.
The years either side of stopping
Exposure peaks when the pot is at its largest and you have just started taking money out of it.
An average is not an outcome
A single average return quietly assumes every year is the same. Real sequences are not, and the difference only shows once you are drawing.
The same crash, year 1 or year 15
Replay the real 2008 to 2010, 2000 to 2009 and 1973 to 1979 sequences on your own pot, then move the same shock later and compare.
Swipe the essentialsor scroll down for the full guide ↓
Guidance and illustrations only, not regulated financial advice. The figures here are illustrative arithmetic, not a projection of your own plan. Your outcome depends on your circumstances; consider speaking to a qualified, regulated financial adviser before acting.
The short version
If two people retire with the same pot, spend the same amount and get the same average return over their retirement, you might expect them to end up in the same place. They may not. If one of them met a bad run of markets in their first few years and the other met the same bad run twenty years later, their outcomes could be very different.
That is sequence-of-returns risk, often shortened to sequence risk. It’s one of the least intuitive parts of retirement planning, and one of the reasons a simple “average return” sum can give you false confidence.
Why the order is irrelevant while you are saving
Start with the case where you are paying in and taking nothing out.
Say you have £100,000 invested and three years of returns: +20%, +20% and −30%.
- In that order: £100,000 → £120,000 → £144,000 → £100,800
- Worst year first: £100,000 → £70,000 → £84,000 → £100,800
Identical. Multiplication does not care about order, so with no money moving in or out, the sequence makes no difference whatsoever to where you finish. This is worth holding on to, because it’s why sequence risk is not something to lose sleep over in your thirties.
Why the order matters once you are drawing
Now run the same three years, but take £5,000 out at the end of each year.
Good years first, bad year last
| Year | Return | Pot after growth | After £5,000 drawn |
|---|---|---|---|
| 1 | +20% | £120,000 | £115,000 |
| 2 | +20% | £138,000 | £133,000 |
| 3 | −30% | £93,100 | £88,100 |
Bad year first, good years after
| Year | Return | Pot after growth | After £5,000 drawn |
|---|---|---|---|
| 1 | −30% | £70,000 | £65,000 |
| 2 | +20% | £78,000 | £73,000 |
| 3 | +20% | £87,600 | £82,600 |
Same three returns. Same three withdrawals. Same average. A gap of £5,500 after only three years, and the gap tends to widen the longer the plan runs.
What’s going on underneath
The mechanism is simpler than the name suggests. When you draw an income from an invested pot, you are selling units to fund it. In a year when prices have fallen, the same £5,000 of income costs you more units than it would have done otherwise. Those units are gone, so they are not there to recover when the market does.
A rise later has a smaller base to work on. The pot can recover, but it’s climbing out of a deeper hole with less in it, and every year of income you take on the way makes the hole a little deeper.
Why averages hide it
“Equities have returned about 5% a year in real terms over the long run” is a reasonable statement about the past. The trouble is that nobody receives the average. You receive a particular sequence, and a plan built on the average alone quietly assumes a smooth ride that markets have never delivered.
This is why a projection that shows one line can be misleading, and why it’s more useful to ask a different question: across many possible orderings of good and bad years, in what share of them does the money last the whole plan? That share is a success rate, and it captures the uncertainty a single average conceals.
When the exposure is highest
Sequence risk is not spread evenly across your life. It concentrates in a window, roughly the few years either side of the day you stop working, for two reasons that compound each other:
- The pot is near its largest, so a percentage fall is the biggest cash amount it will ever be.
- Withdrawals have just started, so a fall is being locked in by selling rather than ridden out.
Someone ten years into retirement with a smaller pot and a State Pension already in payment is usually less exposed than someone in their first year. The same crash lands very differently.
Levers people use to reduce the exposure
None of these are recommendations, and none of them remove the risk. They are the choices that tend to come up, and each has a cost worth understanding before acting.
- Holding some cash or short-term assets. Drawing income from cash for a period can avoid selling investments while prices are down. The cost is that cash tends to lose value to inflation over long periods, so holding a lot of it has its own drag.
- Flexible spending. Plans that can trim spending in poor years, and spend a little more in good ones, tend to cope better than a fixed amount drawn regardless. The cost is that your income is no longer predictable.
- Guaranteed income covering the essentials. The State Pension, a defined-benefit pension or an annuity provide income that does not depend on markets at all. The more of your essential spending they cover, the less you are forced to sell into a fall. The cost of an annuity is that it’s usually permanent and gives up flexibility and any legacy from that money.
- The investment mix. A lower share of equities reduces how far the pot can fall, and also reduces expected growth, which matters over a retirement that could run thirty years or more.
- Working a little longer, or part time. Extra earning years shorten the drawing period and can cover spending through a bad patch. Not everyone has the choice.
Which of these moves the answer most is specific to your figures, and it’s rarely the one people expect. After running a shock scenario, the app lets you adjust spending, retirement age, the investment mix and the pot, and shows what each change does to the share of futures in which the plan lasts, so the trade-offs above become numbers instead of arguments.
What this means for your own plan
You cannot control the order markets deliver, and nobody can forecast it. What you can do is find out how exposed your plan is to a bad start, and see which levers move the answer most for you.
That’s the value of testing a plan against many different orderings rather than one average: it turns an invisible risk into something you can look at before you commit to a retirement date.
Replaying real shocks rather than imagined ones
Your Future Compass does this on your own figures, using actual historical return sequences rather than a hypothetical fall, and in today’s money so inflation is part of the picture:
| Scenario | What it replays |
|---|---|
| A crash as you retire | The real 2008 to 2010 sequence, starting in your first year of drawdown |
| A lost decade | The real 2000 to 2009 returns, when a stock portfolio went roughly nowhere in real terms after two crashes |
| A 1970s inflation shock | The real 1973 to 1979 returns, when high inflation hit both shares and bonds in real terms |
| A crash later on | The same 2008 sequence, dropped in at year 15 once the pot has been drawn down for a while |
The last two rows are the whole argument of this guide, made testable. The same historical sequence, moved from year one to year fifteen, is the experiment, and running both on your own pot shows how much of your exposure is about timing rather than about the size of the fall.
There’s also a longevity scenario, living to 100, which is a different risk but tends to compound this one: a bad start matters more when the money has further to stretch.
Each scenario reports the share of simulated futures in which the plan lasts, so the comparison is between two numbers rather than two impressions.
Where to get free, impartial help
For free, impartial guidance, the government’s MoneyHelper service covers retirement options, and over-50s can book a free Pension Wise appointment to talk through what taking an income from a pension involves. For advice on your own circumstances, speak to a qualified, regulated financial adviser.
Related reading: how long £500,000 could last covers the same uncertainty from the other direction, and pension drawdown explained covers how taking that income actually works. Investing basics covers the mix, the fees and the risk trade-off behind all of it.