Sequence risk is the retirement problem nobody plans for

This is less a set of instructions about the order in which returns arrive than an argument, and it is worth saying so at the start.

Order matters once you are withdrawing

While accumulating, the order of returns is largely irrelevant; only the compounded total matters. Once you are selling units to live on, a fall early in retirement permanently reduces the base that later returns apply to. This is why two retirees with the same average return can end up in very different positions.

The same mechanism runs in the final years before retirement, when the pot is at its largest, a fall costs the most in absolute terms and there is least time left to make it back.

The buffer approach

Holding one to three years of spending in cash or short-dated assets means you need not sell equities during a decline. The buffer is replenished from growth in good years rather than on a fixed schedule. It costs some expected return in exchange for removing the worst mechanism of failure.

The part to write down in advance is the replenishment rule, because deciding in the moment whether a year counts as good enough to top the buffer up is where the discipline usually goes.

Flexibility is worth more than precision

Retirees who reduce withdrawals modestly during poor years dramatically improve the sustainability of a portfolio. A rigid inflation-linked withdrawal is the assumption behind most safe-withdrawal-rate research and is not how people actually behave.

Even small, temporary reductions have a large effect on the arithmetic. Flexibility has limits worth mapping before relying on it, since spending already close to essential cannot be trimmed, and the room to cut is smaller for someone without other income than the modelling tends to assume.

Withdrawal rules are guides, not laws

Rules of thumb about safe withdrawal rates come from specific historical periods, markets and asset mixes. They vary considerably by country and by the fee level assumed, which is frequently unrealistic. Treating them as a planning starting point rather than a guarantee is the appropriate use.

Most of the familiar figures come from the market history of one country over a particular century, and carrying them into a different market, currency or tax system imports an assumption that is rarely stated out loud.

Guaranteed income changes the problem

State pensions, annuities and defined-benefit entitlements cover baseline spending regardless of markets. The larger that guaranteed floor, the less sequence risk matters for the remainder. Deciding how much of your essential spending should be guaranteed is arguably the central retirement decision.

On an ordinary week, state pension ages, entitlement rules and inflation protection differ enormously between countries and are revised over time, so the floor is worth checking against your own scheme rules rather than assuming.

None of this is a substitute for talking to a clinician if something feels wrong.

Which pot the money comes out of

Where more than one account exists, the order of withdrawal affects tax, what passes on death and how long each pot keeps compounding, and the rules governing all three are country-specific. Selling proportionally across holdings preserves the allocation, while selling whatever has risen most rebalances at the same time, and the second needs a rule fixed in advance or it becomes a market call.

Where it helps most, charges bite harder in drawdown than in accumulation for a simple reason: they are taken from a pot that is no longer receiving contributions. This is the stage where regulated advice most often earns its cost, because the decisions are hard to reverse and interact with tax, benefits and provision for anyone who outlives you.

The takeaway

Cover essential spending with something that does not depend on markets, and keep a buffer for the rest.

The version you keep doing is the version that works.