Longevity, and planning for a long life

How long the money needs to last is the input with the widest uncertainty and the largest effect on the answer.

The misunderstanding

Life expectancy at birth is the figure most people know, and it is the wrong one.

Life expectancy conditional on having already reached sixty-five is considerably higher, because those who died earlier are no longer in the calculation.

The relevant figures for planning are those from the current age, not from birth.

And expectancy is a median. Roughly half of people will exceed it, some by a decade or more.

A plan built to the expectancy has a failure rate of around fifty per cent for something that cannot be repeated.

The couple effect

Frequently overlooked.

For a married couple, the relevant horizon is how long at least one of them lives, which is longer than either individual expectancy.

The probability that at least one member of a couple in reasonable health at sixty-five reaches ninety is substantially higher than for either individually.

Which means household planning horizons should be longer than individual ones, and plans built around one person's expectancy understate the requirement.

The variation

Life expectancy varies substantially with factors that are known.

Health status, smoking, income, education and geography all produce differences that are large in the data — differences of several years or more between groups.

Which means population averages are a starting point rather than an individual estimate, and personal circumstances legitimately shift the planning horizon in either direction.

Family history is informative and less predictive than people assume.

Planning horizons

Common practice is to plan to an age with a low probability of being exceeded.

Ninety-five is a frequently used figure for someone in reasonable health at sixty-five. A hundred is used where there is strong family longevity or particular caution.

The cost of planning too long is spending less than necessary. The cost of planning too short is running out.

These are not symmetric, which argues for the longer assumption.

What actually addresses longevity risk

Delaying Social Security, which is the most accessible form of longevity insurance available and is inflation-adjusted.

Income annuities, particularly deferred versions that begin payments at an advanced age.

These are efficient precisely because they only pay in the scenario being insured against, which makes them inexpensive per dollar of protection.

Maintaining growth assets. A thirty-five-year horizon requires returns above inflation, which a very conservative portfolio will not deliver.

Spending flexibility, which allows adjustment as the horizon becomes clearer.

Home equity as a reserve, available if the plan runs longer than expected.

The changing estimate

An underused point.

Remaining life expectancy at eighty-five is different from the estimate made at sixty-five, and generally the plan can be reassessed with better information as time passes.

Someone who has reached eighty-five in good health has a different remaining horizon and different circumstances from what was assumed twenty years earlier.

Which means the planning horizon is not set once. It is revised, and the revisions generally permit more spending rather than less for those who reach later ages in good health.

The spending pattern

Worth incorporating.

Research on retirement spending finds it is not constant in real terms. It tends to be highest in the early active years, decline through the middle period, and potentially rise late with care costs.

A plan assuming constant inflation-adjusted spending for thirty-five years is therefore conservative in the middle and possibly inadequate at the end.

Modelling the shape rather than a flat line produces a more accurate picture, and generally permits more spending early.

The framing that helps

A long life is a good outcome, and it is expensive.

The plan should be built so that living to ninety-five is a pleasant surprise rather than a financial crisis, which means treating longevity as the base case rather than as a tail risk.

General information only, not financial advice. Life expectancy figures are population averages — consult a qualified adviser about your own situation.