Longevity Risk

Longevity risk is the risk that a person or couple outlives the resources set aside to support them. It is the mirror image of the risk life insurance covers: dying too soon leaves dependents short, while living a very long time leaves you short. What makes it distinctive among retirement risks is that it is not an event you can watch happening. A market crash announces itself; longevity arrives quietly, one good year at a time, and the plan that looked comfortable at 70 is simply thinner at 92. It is also worth separating from sequence of returns risk, which readers routinely merge with it because both end in the same sentence about money running out — longevity risk is about the length of the withdrawal period, sequence risk is about the order of returns within it.

The single most useful thing to understand about longevity risk is that the average is the wrong number to plan around. Life expectancy is a central estimate for a large group, not a personal expiration date. A substantial share of any cohort lives past it, and a meaningful minority lives a decade or more past it. Build a plan that runs out of money exactly at life expectancy and you have built a plan with roughly coin-flip odds — which is not how anyone describes a plan they would accept. The relevant question is not "how long will I probably live" but "how long might I live, and what happens to me if I do." For a specific cohort's numbers, the Social Security Administration publishes period life tables, which are the standard reference and worth looking at rather than guessing.

For couples the exposure is larger than either person's own outlook, and this is where planning most often falls short. The probability that at least one of two people is still alive at any given advanced age is necessarily higher than the probability for either one alone. So the horizon that matters for the plan is the second death, not the first — and the survivor's finances are frequently worse than the couple's, because one Social Security benefit stops at the first death, a pension may drop to a reduced survivor percentage or stop entirely, and household expenses do not halve. A plan that funds "the couple to 90" can leave a survivor in real difficulty at 95.

Longevity risk also compounds the other retirement risks rather than sitting beside them. A longer horizon gives inflation more time to erode fixed income streams, gives sequence-of-returns risk more chances to strike, and pushes the plan further into the years when health and long-term-care costs concentrate. That last overlap matters practically: the person most exposed to advanced-age longevity risk is the same person most exposed to needing paid care, and the two bills arrive together. Long-term care insurance and long-term-care planning belong in the same conversation.

There are four honest responses, and they combine rather than compete. Pool it — an annuity, particularly an immediate annuity or a qualifying longevity annuity contract aimed specifically at advanced age, transfers the risk to an insurer that can average it across thousands of lives. Maximize the inflation-indexed lifetime income you already own — for most Americans the largest and cheapest longevity hedge available is delaying the start of Social Security, which permanently increases a benefit that is indexed for inflation and lasts as long as you do; the mechanics of that decision belong to its own page. Spend flexibly — a plan that adjusts withdrawals downward after bad years survives far longer than one committed to a fixed inflation-adjusted amount, which is why guardrail approaches to the safe withdrawal rate outperform rigid rules. Hold a bigger buffer — the brute-force answer, which costs current consumption in exchange for security, and is a legitimate choice if the trade is made knowingly rather than by default.

Longevity risk shows up in a plan as an assumption, usually a single number buried in a projection: the age to which the money must last. Change that number and everything downstream moves, which is the clearest way to see what the risk actually costs.

A hypothetical example. Suppose a retiree has $1,000,000 and wants level spending in inflation-adjusted terms, earning a hypothetical 3% real return, with the portfolio deliberately exhausted at the end of the horizon. Over 25 years that supports about $57,400 a year. Over 35 years it supports about $46,500 a year. Extending the planning horizon by ten years — with no market loss, no inflation surprise, and no mistake of any kind — costs roughly $10,900 of annual spending, about a 19% cut. (Illustrative arithmetic only: it ignores taxes, fees, market variability, and Social Security, and no real portfolio delivers a smooth 3% real return.)

That single comparison is why longevity risk deserves attention even from people whose plans look comfortable. There is no market event in it. The cost is entirely in the assumption, and the assumption is the one variable a retiree genuinely cannot know.

Layering a response changes the picture. If the same retiree used part of the portfolio to buy guaranteed lifetime income, the portfolio no longer has to fund an open-ended horizon on its own — it only has to bridge to the point where guaranteed income covers essentials. That is the structural reason guaranteed income and longevity risk are discussed together: an income stream with no end date is the only asset whose value does not depend on how long you live.

Pros (of planning around it explicitly)

• It reframes the central retirement question from "what will my portfolio earn" to "how long must this last," which is the variable that actually breaks plans.
• The responses are concrete and stackable: pooling through an annuity, delaying Social Security, flexible spending rules, and a larger buffer.
• Testing a plan against several end ages, rather than one, surfaces fragility early enough to do something about it.
• It forces couples to check the survivor's position separately, which is the single most commonly skipped test in retirement planning.

Cons (and limits of the concept)

• You cannot know your own number, so managing longevity risk always means paying something for protection you may not need.
• Every hedge has a real price: an annuity premium is illiquid and generally irrevocable, delaying Social Security requires spending savings in the gap years, and a larger buffer is bought with the healthy years you could have spent it in.
• Overcorrecting produces its own failure — a retiree so worried about age 100 that they underspend their sixties has also lost something they cannot get back.
• It interacts with the other retirement risks rather than standing apart from them, so it cannot be solved in isolation from inflation, sequence risk, and health costs.