Key Takeaways
Direct answer: Longevity risk is the risk that a retiree lives longer than the assets funding their retirement can support. It is not the risk of dying early; it is the risk of the retirement lasting longer than planned. Because it compounds with market risk, inflation and rising late-life health costs, it cannot be diversified away within a portfolio. It can only be reduced by holding more assets, spending less, planning to a longer horizon, or transferring part of the risk to an entity that pools mortality across many people.
- Average life expectancy is a midpoint, meaning roughly half of a group outlives it. Planning to the average is planning to run out of money in about half of possible outcomes.
- Life expectancy conditional on already having reached retirement age is longer than life expectancy at birth, because the people who did not reach that age are no longer in the group.
- For a couple, the relevant horizon is joint-and-survivor: the probability that at least one of the two is still alive, which is always longer than either individual horizon.
- Only mortality pooling genuinely transfers longevity risk. Everything else, including holding more equities, absorbs or offsets it.
- Social Security is already an inflation-adjusted lifetime income stream, which makes the claiming decision one of the largest longevity levers most retirees have.
- Longevity risk and sequence-of-returns risk are different problems that call for different responses, and confusing them leads to solving the wrong one.
What Is Longevity Risk?
Longevity risk is the possibility that a retirement lasts longer than the resources funding it. Framed as a planning problem, it is the gap between the horizon someone assumed and the horizon they actually got.
Two features make it unlike the other risks in a retirement plan. First, it is one-sided in a way that market risk is not. A portfolio that falls can recover; a retiree who reaches ninety-five has not made a mistake that can be reversed. Second, it interacts multiplicatively with everything else. A long life means more years of inflation eroding purchasing power, more years of withdrawals compounding against the portfolio, and more exposure to late-life costs such as long-term care that tend to arrive at the end rather than spread evenly.
The uncomfortable framing is that longevity risk is the risk of a good outcome. Every other retirement risk is something a retiree hopes to avoid. This one is the financial consequence of the thing they hope for. That asymmetry is why plans tend to under-provision for it: nobody wants to describe living a long time as a problem, so the horizon assumption gets set optimistically about health and pessimistically about nothing.
The distinction from sequence-of-returns risk matters. Sequence risk is about when returns arrive relative to withdrawals, and it is concentrated in the first years of retirement. Longevity risk is about how many years of withdrawals there are, and it is concentrated at the end. A cash buffer helps with the first and does almost nothing for the second.
Why Average Life Expectancy Is the Wrong Planning Number
The single most common error in longevity planning is treating average life expectancy as a target date. Three distinct problems make that unsafe.
An average is a midpoint, not a ceiling. If life expectancy for a group is a given age, then by construction a substantial share of that group lives past it. Building a plan that funds spending exactly to the average is building a plan that fails for roughly half the people it is applied to. A plan is not a forecast; it is a commitment to be solvent under adverse outcomes, which means it needs to be built around a percentile in the tail, not around the middle.
Life expectancy at birth is not life expectancy at 65. Published life expectancy figures are frequently quoted from birth, but that number includes everyone who died young. A person who has already reached retirement age is, by definition, no longer in that group. Conditional life expectancy at 65 is longer than life expectancy at birth, and the gap is not small. Any planning horizon derived from a birth-based figure understates the problem.
Averages are population figures, not personal ones. Health status, family history, occupation and access to care all shift an individual's distribution. A population average is a starting point for someone with no other information, not an estimate of a specific person's remaining years.
The practical implication is a shift in question. Instead of asking "how long will I live?", which has no answer, the useful question is "what horizon do I want to remain solvent through?" That reframes an unknowable forecast into a choice about how much failure probability is acceptable, which is something a person can actually decide.
Because published mortality tables are updated periodically and differ by source and methodology, this guide does not quote specific life expectancy figures. Readers who want a number for their own situation should take it from a current actuarial source and use it as a distribution rather than as a date.
Joint Survival: Why Couples Face a Longer Horizon
For a married couple or any two people sharing a pool of assets, the individual horizon is the wrong one entirely. The portfolio must last until the second death, not the first, because the surviving partner still needs income.
The probability that at least one of two people is still alive at a given age is always higher than the probability for either one alone. That is a direct consequence of how probability works with two independent-ish events: the joint survival probability only reaches zero when both individual probabilities do. A couple planning to one partner's individual life expectancy is understating their horizon substantially.
Several plan features change at the first death, and they mostly change unfavorably for the survivor:
- Social Security drops to one benefit. A surviving spouse generally continues receiving the larger of the two benefits, not both. Household benefit income falls even though many fixed costs do not.
- Filing status changes. A surviving spouse moves from joint to single filing after a transition period, which generally means the same income is taxed at higher marginal rates and reaches thresholds sooner.
- A pension may or may not continue. Whether a pension pays a survivor benefit, and at what percentage, is decided at the point of election, often years earlier and often irreversibly.
- Household spending does not halve. Housing, property taxes, insurance and utilities are largely unchanged by one fewer person.
The planning consequence is that the survivor's position, not the couple's combined position, is the binding constraint. A plan that works comfortably while both are alive and becomes tight for the survivor has not solved longevity risk; it has deferred it to the person least able to respond to it.
Which Tools Actually Transfer Longevity Risk
There is a meaningful distinction between transferring longevity risk and absorbing it. Transferring means someone else bears the consequence of an unusually long life. Absorbing means the retiree still bears it, with a larger cushion.
| Approach | Transfers or absorbs | What it costs |
|---|---|---|
| Delaying Social Security | Transfers: a larger inflation-adjusted benefit for life, pooled across all participants | Foregone benefits during the delay period, funded from the portfolio |
| A life annuity | Transfers: the insurer bears the cost of a long life | Loss of access to the principal, insurer credit exposure, and pricing that reflects the insurer's own assumptions |
| A deferred income annuity | Transfers: pays only if the retiree survives to a later start age, targeting the tail directly | Capital committed with no payout if the retiree dies before the start date, absent a return-of-premium feature |
| A pension with a survivor benefit | Transfers: the plan bears mortality across all participants | A lower monthly payment than a single-life election |
| Holding a higher equity allocation | Absorbs: raises expected growth but widens the range of outcomes | More exposure to sequence risk exactly when it is most damaging |
| Reducing the withdrawal rate | Absorbs: a smaller draw lasts longer | A lower standard of living throughout, including in the outcomes where it turns out to have been unnecessary |
| Working longer | Absorbs and partially transfers | Shortens the withdrawal period, adds contributions, and often delays Social Security, but is not always within the retiree's control |
Only the pooled options genuinely transfer the risk, and the reason is mortality pooling. An insurer or a public program collects from many people, pays for as long as each lives, and relies on the aggregate being predictable even though no individual outcome is. An individual cannot replicate that internally, because a single person has no pool to average across. That is the structural reason a portfolio, no matter how it is allocated, cannot fully solve longevity risk.
Swoopr's annuities guide covers the product mechanics, costs and trade-offs in detail, including the credit and liquidity considerations that come with transferring risk to an insurance company.
Social Security as Longevity Insurance
Social Security is, structurally, an inflation-adjusted life annuity that most American retirees already own. That framing matters because it changes what the claiming decision is about.
The common way to evaluate claiming is a break-even calculation: at what age does delaying produce more cumulative dollars than claiming early? That framing treats the decision as a bet on lifespan. The longevity-risk framing treats it as an insurance purchase. From that angle, the money spent from the portfolio during a delay is buying a permanently larger, inflation-adjusted, government-backed income stream for exactly the scenario the retiree cannot self-insure: living far longer than expected.
Under the insurance framing, "losing the bet" by dying early is not a bad outcome in the way a losing investment is. Insurance that never pays out is insurance whose covered event did not occur. Nobody regrets not having needed their homeowner's policy.
For couples, the claiming decision compounds because a surviving spouse generally continues the larger of the two benefits. That makes the higher earner's claiming age a decision about the survivor's income for the rest of their life, not just the claimant's. It is one of the few retirement decisions that simultaneously improves the plan's worst case and lasts as long as the plan does.
Because benefit formulas, full retirement ages and delayed-credit rules are set by statute and change over time, the specific figures for any individual should come from the Social Security Administration's own current statements rather than from a secondary summary.
How Longevity Assumptions Drive Withdrawal Rates
Every withdrawal rate implies a planning horizon, whether or not it is stated. The historical research behind common withdrawal-rate guidance is generally built around a specific payout period, and shortening or lengthening that period changes the answer materially. A rate that survived a 30-year horizon in historical testing is not automatically safe over a 40-year one.
Three practical consequences follow:
- Early retirement raises the stakes disproportionately. Retiring at 55 rather than 65 does not add ten years to a 30-year plan; it creates a horizon that may run past 40 years, over which small differences in withdrawal rate compound into very different outcomes.
- Dynamic withdrawal rules partially address longevity risk. A strategy that reduces spending after poor performance conserves capital in exactly the scenarios where a long horizon becomes dangerous. The guardrails and floor-and-upside approaches in Swoopr's withdrawal rate frameworks guide are the relevant mechanisms.
- Percentage-of-portfolio withdrawals cannot mathematically deplete an account, but can shrink the income within it. That solves the arithmetic version of longevity risk while leaving the practical version, an income too small to live on, fully intact.
Required minimum distributions add a separate, mandatory layer on top of any chosen strategy. The IRS calculates them by dividing the prior year-end balance by a distribution period from the Uniform Lifetime Table, a schedule that is itself derived from life expectancy. That is worth noticing: the government's own forced-withdrawal schedule is built on a longevity assumption, and it is designed to draw the account down over a lifetime rather than to make it last. Swoopr's required minimum distributions guide covers the rules.
Longevity Risk and Asset Allocation
Longevity risk pushes allocation in the opposite direction from the conventional "get more conservative with age" instinct, and that tension is real rather than merely apparent.
A 40-year horizon is, in investment terms, a long horizon. Over that span, inflation is the dominant threat to purchasing power, and an allocation heavy in nominal fixed income guarantees a slow erosion of real spending power. That argues for maintaining meaningful equity exposure deep into retirement.
Against that, the early years of retirement are when sequence-of-returns risk bites hardest, and a heavy equity allocation combined with withdrawals in a bad early market is precisely the failure mode that historical research identifies. That argues for restraint exactly when the long horizon argues for growth.
The two pressures are usually reconciled in one of three ways:
- A rising equity glide path. Start retirement more conservative and increase equity exposure over time, which reduces early sequence risk while restoring growth exposure for the long tail.
- A floor-and-upside split. Cover essential spending with pooled lifetime income, then hold the discretionary portfolio aggressively, since a shortfall there does not threaten necessities.
- Time segmentation. Hold near-term spending in cash and bonds while leaving long-horizon money in equities, which is the structure covered in Swoopr's bucket strategy guide.
All three are attempts to serve two horizons at once. Swoopr's retirement asset allocation guide owns glide-path design in full.
Late-Life Costs and the Shape of Retirement Spending
Longevity risk is not only a matter of more years of the same spending. The composition of spending changes late in life, and the changes do not cancel out.
Discretionary spending on travel, dining and hobbies tends to decline with age as activity levels fall. Health-related spending tends to rise. Whether the two offset each other is the subject of genuine disagreement, and the answer differs enormously between a retiree who never needs sustained care and one who does.
The category that matters most is long-term care, because its cost distribution is extremely skewed. Most people incur little or none. A minority incur costs large enough to consume a substantial portfolio. That shape, a low probability of a very large cost, is the classic profile of an insurable risk rather than a budgetable expense, and it is the reason long-term care is usually handled as a separate planning question rather than as a line in a spending projection. Swoopr's personal finance coverage addresses the general principle of insuring against low-probability, high-severity events.
Medicare-related costs also interact with the tax side of a retirement plan. Because the income-related monthly adjustment amount for Medicare Parts B and D is based on modified adjusted gross income from two years earlier, a large one-time distribution or Roth conversion in one year can raise premiums two years later. That is covered in Swoopr's tax diversification guide.
Common Mistakes and Misconceptions
- Planning to average life expectancy. An average is a midpoint. A plan built to it fails in roughly half of outcomes.
- Using life expectancy at birth instead of at retirement age. Conditional life expectancy for someone who has already reached 65 is longer, because everyone who died earlier has left the group.
- Using one partner's horizon for a couple. The relevant horizon runs to the second death, and the survivor faces lower Social Security income and a less favorable filing status.
- Believing a portfolio alone can solve longevity risk. Without mortality pooling, an individual has no pool to average across, which is why a portfolio can only absorb the risk rather than transfer it.
- Treating Social Security claiming as a break-even bet. The delayed benefit is inflation-adjusted lifetime income. Evaluating it as a wager on lifespan misses that it is insurance against the scenario the retiree cannot self-fund.
- Confusing longevity risk with sequence risk. A cash buffer addresses the first years of retirement. It does almost nothing about year 38.
- Assuming spending declines are guaranteed. Discretionary spending often falls with age, but health costs often rise, and the offset is not reliable at the individual level.
- Getting steadily more conservative without regard to horizon. Over a 40-year retirement, an allocation dominated by nominal fixed income makes inflation the near-certain failure mode.
Frequently Asked Questions
What is longevity risk?
Longevity risk is the risk that a retiree lives longer than the assets funding their retirement can support. It is not the risk of dying early; it is the risk of the retirement lasting longer than the plan assumed. Because a longer life means more years of withdrawals, more years of inflation eroding purchasing power, and more exposure to late-life health costs, longevity risk compounds every other retirement risk. It cannot be diversified away inside a portfolio, because an individual has no way to average their own lifespan against anyone else’s.
Why is average life expectancy a bad planning target?
An average is a midpoint, which means roughly half of any group outlives it. A plan funded exactly to average life expectancy is a plan that runs out of money in about half of possible outcomes. Two further problems compound this: figures quoted from birth understate the horizon for someone who has already reached retirement age, because everyone who died younger has left the group, and population averages say nothing about a specific individual’s health, family history or circumstances. The useful question is what horizon to remain solvent through, not how long someone will live.
How does longevity risk differ for a couple?
For a couple sharing a pool of assets, the portfolio must last until the second death, not the first. The probability that at least one of two people is still alive at a given age is always higher than the probability for either one individually, so a couple’s planning horizon is longer than either partner’s. The first death also makes the survivor’s position worse in several ways: household Social Security drops to the larger of the two benefits, filing status eventually changes from joint to single, and fixed costs such as housing and insurance do not halve.
Can a portfolio alone solve longevity risk?
No. A portfolio can absorb longevity risk by being larger, by being drawn down more slowly, or by holding assets with higher expected growth, but it cannot transfer the risk. Transferring longevity risk requires mortality pooling, in which an insurer or a public program collects from many people and pays each for as long as they live, relying on the aggregate being predictable even though no individual outcome is. A single person has no pool to average across, which is the structural reason an individual portfolio cannot fully solve the problem.
How does delaying Social Security reduce longevity risk?
Delaying a Social Security claim produces a larger monthly benefit that is adjusted for inflation and paid for life, funded by a program that pools mortality across all participants. That makes it a genuine transfer of longevity risk rather than an absorption of it. The money spent from a portfolio to bridge the delay period is effectively buying a permanently larger lifetime income stream aimed at exactly the scenario a retiree cannot self-insure. For couples, the higher earner’s claiming age also determines the survivor benefit, so it affects the surviving spouse for the rest of their life.
What is the difference between longevity risk and sequence-of-returns risk?
Sequence-of-returns risk is about when returns arrive relative to withdrawals, and it is concentrated in the first years of retirement: poor early returns combined with ongoing withdrawals can permanently damage a portfolio in a way the same returns arriving later would not. Longevity risk is about how many years of withdrawals there are at all, and it is concentrated at the end. They call for different responses. A cash buffer or a conservative early allocation addresses sequence risk and does almost nothing about year 38 of a retirement.
Do annuities eliminate longevity risk?
A life annuity transfers longevity risk to an insurance company, which bears the cost if the annuitant lives an unusually long time. That is a real transfer rather than an absorption. It is not free of all risk, though: the annuitant gives up access to the principal, takes on credit exposure to the insurer, and accepts pricing that reflects the insurer’s own mortality and interest-rate assumptions. A deferred income annuity that begins payments at a later age targets the longevity tail more directly, at the cost of committing capital that may pay nothing if the annuitant dies before the start date.
Should retirees hold more stocks because of longevity risk?
A long horizon does argue for maintaining meaningful equity exposure, because over 30 or 40 years inflation is the dominant threat to purchasing power and an allocation dominated by nominal fixed income erodes real spending power steadily. That pressure runs directly against sequence-of-returns risk, which is most damaging in the early years of retirement. The common reconciliations are a rising equity glide path, covering essential spending with pooled lifetime income and investing the rest aggressively, or segmenting the portfolio by time horizon.
How does longevity risk affect a safe withdrawal rate?
Every withdrawal rate implies a planning horizon whether or not it is stated. Historical withdrawal-rate research is generally built around a specific payout period, and a rate that held up over a 30-year horizon in that testing is not automatically safe over 40 years. Early retirement raises the stakes disproportionately, because retiring a decade earlier can push the horizon well past 40 years, over which small differences in the withdrawal rate compound into very different outcomes. Dynamic rules that reduce spending after poor performance help, because they conserve capital in exactly the scenarios where a long horizon becomes dangerous.
Does retirement spending naturally decline with age?
Discretionary spending on travel, dining and hobbies often declines as activity levels fall, but health-related spending tends to rise, and whether the two offset each other is genuinely disputed. The answer differs enormously between a retiree who never needs sustained care and one who does. Long-term care in particular has a highly skewed cost distribution: most people incur little or none, while a minority incur costs large enough to consume a substantial portfolio. That shape makes it an insurable risk rather than a budgetable line item.
How do required minimum distributions relate to longevity?
Required minimum distributions are calculated by dividing the prior year-end account balance by a distribution period taken from the IRS Uniform Lifetime Table, a schedule derived from life expectancy. That is worth noticing: the government’s mandatory withdrawal schedule is itself built on a longevity assumption, and it is designed to draw a tax-deferred account down over a lifetime rather than to make it last indefinitely. RMDs are a floor on withdrawals, not a spending plan, and the distributed money can be reinvested in a taxable account rather than spent.
What planning horizon should a retiree use?
There is no single correct number, and this guide deliberately does not supply one, because published mortality tables differ by source and are updated periodically. The right approach is to treat the horizon as a decision about acceptable failure probability rather than as a forecast. That means taking a current actuarial figure for age, sex and, for a couple, joint survival, then planning to a percentile well into the tail rather than to the midpoint, and revisiting the assumption as health circumstances change.
References
This guide draws on IRS and published retirement-income research materials, verified in August 2026. It deliberately does not quote specific life expectancy figures; those should be taken from a current actuarial source, because they are updated periodically and differ by methodology.
- IRS: Retirement Topics, Required Minimum Distributions (RMDs): confirms that required minimum distributions are calculated using a distribution period drawn from the IRS Uniform Lifetime Table.
- IRS: Publication 590-B, Distributions from Individual Retirement Arrangements (IRAs): contains the life expectancy tables the IRS uses for required minimum distribution calculations.
- Medicare: Medicare Costs: the current income-related monthly adjustment amount structure, including the two-year lookback on modified adjusted gross income referenced above.
- William P. Bengen, Journal of Financial Planning: Determining Withdrawal Rates Using Historical Data: the original withdrawal-rate research, built around a specific payout period, which is why lengthening the horizon changes the conclusion.
- Jonathan T. Guyton and William J. Klinger, Journal of Financial Planning: Decision Rules and Maximum Initial Withdrawal Rates: research on dynamic spending rules, the mechanism that conserves capital when a horizon proves longer than planned.
This content was reviewed by the Swoopr Editorial Team in August 2026 and reflects publicly available information at that time. Social Security benefit formulas, full retirement ages and delayed retirement credits are set by statute and change; individual figures should be taken from the Social Security Administration directly. This is educational content about how longevity risk works, not personalized financial, tax, insurance or retirement advice, and nothing here is a recommendation to purchase any specific product.
Building a Plan Around an Unknown Horizon
Longevity risk resists the usual approach to financial planning because it cannot be forecast at the individual level. No amount of analysis produces a date. What analysis can produce is a decision about how much failure probability is acceptable, and that turns an unanswerable question into a manageable one.
Three ideas do most of the work. The first is that the planning horizon should sit well into the tail of the distribution rather than at its midpoint, and it should be conditional on having already reached retirement age rather than measured from birth. For a couple, it should reflect joint survival, because the portfolio has to support the survivor, whose Social Security income falls and whose filing status eventually becomes less favorable while most fixed costs stay the same.
The second is that only mortality pooling actually transfers the risk. Holding more equities, spending less and working longer all help, but they leave the retiree bearing the consequence of an unusually long life. Social Security, a pension with a survivor benefit, and life-contingent annuities move that consequence onto an entity with a pool to average across. For most American retirees, the Social Security claiming decision is the largest single longevity lever available, and it is best evaluated as an insurance purchase rather than as a break-even bet on lifespan.
The third is that longevity risk pulls asset allocation in the opposite direction from instinct. Over a horizon that may run 40 years, inflation is the near-certain threat, which argues for keeping real growth exposure late into retirement even while sequence-of-returns risk argues for caution early. Rising glide paths, floor-and-upside structures and time segmentation are all ways of serving both horizons at once, and none of them removes the underlying tension.
The practical test of a retirement plan is not whether it works at the expected lifespan. It is whether it still works at an unexpectedly long one, and whether the person left holding it, often the surviving spouse, has the income and flexibility to absorb what comes next.