How long does a pension need to last?

Every retirement plan rests on a number most people never say out loud: the age you expect to die.

You can dress it up as a “planning horizon” or a “time frame”, but that is what it is. Before you can know whether a pension is big enough, you have to decide how many years it must stretch across. Get the size of the pot a little wrong and you adjust along the way. Get the number of years wrong and you can run out of money in the one stretch of life when you can no longer earn it back.

Almost everyone gets the years wrong in the same direction. They plan for the average. And the average is a coin-flip. Here is a fun fact: the chance of you actually being average is almost nil.

i. You have to pick a horizon, even though you can’t know it.

No one can tell you how long you will live. That is the uncomfortable centre of the whole exercise. But “I don’t know” is not a plan, and refusing to choose a horizon just means defaulting to a bad one by accident.

Both errors have a price, but they are not the same price. Overestimating costs you something real — a slightly lower income each year than you might otherwise have taken. Underestimating costs you the years you can least afford to lose. And if you have children or relatives, money left at the end is rarely money wasted.

So you choose. The real question is not whether to pick a number, but whether you pick it carefully or carelessly. Pick it carelessly, with the average or a round twenty years or whatever feels about right, and you have quietly built a plan that fails half the time. Pick it deliberately, and you build one that holds.

I want to reiterate: the chance of you being average is almost nil.

ii. The average is the trap.

Here is what the average looks like. A 65-year-old man in the UK can expect to live, on average, to around 85. A 65-year-old woman to around 88. Those are cohort figures, which try to allow for the fact that medicine keeps improving. The Office for National Statistics publishes them, and its life expectancy calculator will give you your own.

Read quickly, “85” sounds like a target to aim at. It is not. It is a midpoint. Average life expectancy is the age by which roughly half of people have died and half are still going. Build your pension to last exactly that long and you have, near enough, a one-in-two chance of outliving it. You would not board a plane that landed safely half the time. You should not retire on a plan that funds itself half the time.

The people who die around the average are not who you plan for. The problem is the very ordinary person who lives a decade past it. A 65-year-old today has a real and unremarkable chance of reaching their nineties. Plan to 85, live to 94, and you have nine years to fund from a pot that emptied at 85.

There is a further reason to be cautious about the number. Life expectancy has risen substantially over the past century, and the assumptions built into today’s figures may prove conservative if medical progress continues. Longer life is not the same as longer healthy life, though, and the later years often cost more rather than less. Planning for a long life is partly about being able to afford the expensive end of it.

You can’t know how long you’ll live. So plan as if it will be long. It is the only plan that lets you spend the early years without fear.

iii. The two ways to be wrong are not equal.

See the choice as a bet, because that is what it is. If we assume the national life expectancy holds up by the time you get there, then it is a 50/50 shot on landing before the average age or after.

Let’s plot this out.

Two people retire at 65 with the same pot and draw the same income. One builds the plan to last to 85, the other to 95. Through their seventies and early eighties they look identical. Then the first plan runs dry and the second keeps paying. If they both die at 83, the difference never shows, and the cautious one merely had a number on a spreadsheet they didn’t need. But if either lives to 92, and plenty do, only one of them is still funded.

That is the whole argument. Plan long and the worst case is that you leave money behind. Not the end of the world. Especially, as I said before, if you have family to leave it to. Failing that, there are plenty of good causes and charities. Plan short and the worst case is that you spend your final years with none. Those are not equal risks, and they should not be treated as a balanced choice.

iv. If you’re a couple, plan for the survivor.

For couples the average is doubly misleading, because you are not funding one life. You are funding the last one standing.

Take two people who are both 65. The chance that at least one of them reaches 90 is high, far higher than the chance for either of them alone. So the horizon for the plan is not his life expectancy, and not hers, but the longer of the two. The bills do not halve when the first person dies, either. The survivor heats the same house and pays the same council tax on a single income. A plan built around “our average” can leave the survivor, more often the wife, with years of life and not enough to live them on.

v. The target keeps moving.

There is one more reason to plan long: the finish line has been drifting away for most of the past century.

Someone retiring this year will, on the whole, live longer than someone who retired several decades ago, and every advance against the illnesses that take people in their seventies and eighties adds years to the back of the plan. Improvement in life expectancy has slowed markedly since around 2011, so this is a reason for care rather than a promise of more years. But planning to the life expectancy your grandparents had is planning for a retirement that has already ended.

This is also a problem at the level of the country. An ageing population puts strain on the healthcare system and on the affordability of the State Pension, which is one reason the State Pension age has already risen and may rise again.

Good planning does not aim at the average. It aims well past it.

When I model a client’s retirement, I run the income they want against the whole of their likely life and then some, usually to age 95 or 100, and I stress-test it against poor markets early on and rising costs throughout. The worst case scenario. The aim is not to predict the date. No one can. The aim is to build a plan that survives an age you are very unlikely to beat, so that a shorter life becomes a margin of safety rather than the only outcome that works.

This is where the length of the plan meets the rest of it. The longer the money must last, the more of it has to stay invested and working rather than sitting in cash losing ground to inflation, and the more it matters to hold a buffer for the first years, so that a bad run of markets at the start cannot sink a thirty-year plan. How long, how invested, and how defended are the same conversation. It is the same early-years risk I wrote about in the five years before retirement. If you want to see the arithmetic for yourself, the compound calculator shows what a longer horizon does to a pot.

vi. You can’t control it. You can plan for it.

You do not get to choose how long you will live. You do get to choose whether your plan assumes a long life or a short one. Assume a short one and you are betting your later years on dying roughly on schedule. Assume a long one and the worst that happens is that you were too well prepared.

That is the version of the plan that lets you enjoy the early years of retirement, the travel and the help to the children, because you have already made the late years safe. Planning for a long life is not pessimism. It is the thing that makes a long life affordable.

— Andrew

Andrew Daw is an independent financial planner based in Bath, working with clients across Bath, Bristol and the South West. More about Andrew.