Risk & Volatility
Sequence risk and why it matters at retirement
The order in which returns arrive is irrelevant while accumulating and decisive while withdrawing.
Two investors with identical average returns over the same period can end with entirely different outcomes, depending on the order in which those returns arrived — but only if they were withdrawing.
Why the order does not matter while accumulating
The arithmetic.
If no money is added or withdrawn, the final value depends only on the product of the returns, and multiplication is commutative — the order makes no difference.
With regular contributions the picture changes slightly, and poor early returns can actually help, since contributions buy more units at lower prices.
Which means a young investor experiencing a severe market fall is, counterintuitively, in a reasonable position provided they continue contributing.
Why the order is decisive while withdrawing
The mechanism.
Withdrawing a fixed amount from a falling portfolio means selling more units at lower prices, which permanently reduces the number of units remaining to participate in any recovery.
Poor returns in the early years of drawdown therefore do disproportionate damage, because the capital base is permanently depleted before the recovery arrives.
The same returns in a different order, with the poor years later, produce a substantially better outcome.
Which is sequence risk, and it is the central risk of retirement drawdown.
The vulnerable window
Where it concentrates.
The years immediately before and after retirement are the point of maximum vulnerability, because the portfolio is at its largest and withdrawals are beginning.
A severe fall at that point has a far greater effect than the same fall a decade earlier or a decade later.
Which is why glide paths reduce risk approaching a target date, and why the period around retirement warrants specific attention rather than a continuation of the accumulation approach.
The mitigations
What actually helps.
Holding cash or short-dated bonds covering one to three years of withdrawals, so that spending does not require selling equities during a fall.
This is sometimes described as a cash buffer or bucket approach, and its main benefit is behavioural and mechanical rather than mathematical.
Flexible withdrawals, reducing spending in poor years, which trial modelling consistently finds improves sustainability substantially.
Fixed inflation-linked withdrawals regardless of market conditions is the assumption behind the standard withdrawal rules, and relaxing it changes the picture.
Guaranteed income covering essential spending, whether from state provision, a defined benefit pension or an annuity — which means market falls affect discretionary spending rather than essentials.
A more conservative allocation during the vulnerable window, increasing again afterwards in some approaches.
Delaying retirement or working part-time,
which reduces the withdrawal period and allows contributions to continue.And starting at a lower withdrawal rate, which is the simplest and least popular mitigation.
Withdrawal rates
Where the debate sits.
Rules suggesting a sustainable initial withdrawal in the region of three to four per cent of the starting balance, increased with inflation, derive from historical simulation of particular markets over particular periods.
Their limitations: they depend on the historical period examined; they assume a fixed retirement length; they generally exclude costs and taxes; and they assume rigid withdrawals regardless of circumstances.
Lower starting rates have been suggested for lower expected return environments and for longer retirements.
Which means these are a starting point for thinking rather than a rule, and any specific plan should be modelled with your own numbers.
The annuity comparison
Worth including.
An annuity transfers sequence risk and longevity risk to the insurer in exchange for giving up flexibility and inheritance potential.
Which makes it a genuine solution to the problem rather than merely a conservative choice.
A combination — securing essential spending with guaranteed income and leaving discretionary spending to a portfolio — addresses both the risk and the desire for flexibility, and is what many practitioners recommend.
Modelling it
Practical.
Averages are misleading: a plan should be tested against a range of sequences rather than an assumed average return.
Historical sequence testing and stochastic modelling both do this and are available in various tools.
The output is a probability of the plan surviving rather than a single answer, which is a more honest representation.
And testing what happens if the first five years are poor is the specific stress test worth running.
What to do about it now
Depending on stage.
Accumulating with decades to go: sequence risk is not your problem, and continuing to contribute through declines is beneficial.
Within about ten years of drawing: begin thinking about the glide path and building the cash buffer.
At or in drawdown: hold the buffer, be prepared to flex spending, and consider guaranteed income for essentials.
And in all cases, avoid the plan that requires a specific sequence of returns to succeed.
General information only, not investment advice. Investments can fall in value and withdrawal rules are not guarantees. Consult a regulated financial adviser about retirement income.
Also by Clara Mensah
- Knowing when to do nothingBehaviour
- Automating your investingBehaviour
- Preparing a portfolio for someone elseRisk & Volatility
- Making decisions with a partnerBehaviour
