What is longevity risk and how can it be managed?

The short answer

Longevity risk is the asymmetric danger of outliving your money. It is asymmetric because dying early is a financial inconvenience for heirs, while running out of money at 92 is a catastrophe for the retiree. The two outcomes are not symmetric in welfare terms.

The standard error in retirement planning is to use median life expectancy as the planning horizon. By construction, half of retirees outlive the median. For a couple, the relevant horizon is the joint life expectancy of the longer-lived partner, which extends meaningfully beyond either individual’s median.

The methodological correction is to plan to the 90th or 95th percentile of expected lifespan, not the median. This buffer protects against the asymmetric downside of running short while accepting that the planned-for capital may exceed needs in some scenarios.

→ New to retirement risk concepts? Financial education hub

What the data shows

The reference numbers come from OECD Health at a Glance 2023-2024, US Social Security Administration period life tables, INSEE demographic data 2024, and the Society of Actuaries longevity projections.

The reference numbers (OECD, SSA, INSEE 2021-2024):

• OECD average life expectancy at 65 — approximately 19.5 additional years (2021), with 3.3-year gap favoring women
• US life expectancy at birth (2024) — 79.0 years, up from 78.4 in 2023
• US average life expectancy roughly 3.7 years below comparable OECD countries (Peterson-KFF analysis)
• France 2024 — life expectancy at birth 85.6 years for women, 79.9 for men; at 65, healthy life expectancy approximately 12 years for women, 10.5 for men
• For a 65-year-old US couple — joint probability that at least one partner reaches 95 is approximately 30% per actuarial tables

The exception that complicates the picture: aggregate life expectancy figures hide significant variance by socioeconomic status, education, and health behaviors. A non-smoking, college-educated retiree in France or Switzerland faces a materially longer expected horizon than the population median, which raises the planning anchor accordingly.

→ Dataset: US personal savings rate dataset

Why it happens — the macro mechanism

Longevity risk emerges from three structural realities: actuarial uncertainty at the individual level, the asymmetric welfare cost of mismatched planning, and ongoing improvements in late-life mortality.

The first channel is the actuarial uncertainty around individual lifespan. No one knows when they will die. Period life tables give population averages, but individual outcomes are dispersed around those averages. The standard deviation of remaining life expectancy at 65 is roughly 8-10 years in OECD countries, meaning a one-sigma upside outcome adds nearly a decade to the planning horizon.

The second channel is the welfare asymmetry. Modigliani’s lifecycle model assumes that consumption and longevity match perfectly, but that is unrealistic. The cost of dying with too much capital is borne by heirs and the estate; the cost of running out at 90 is borne directly by the retiree at the worst possible moment. This asymmetry justifies planning to a percentile well above the median.

A brief transition: this is why annuity products theoretically resolve longevity risk by transferring it to insurers, even though behavioral biases prevent widespread adoption.

The third channel is mortality improvement. Life expectancy at older ages has improved steadily across OECD countries over decades, with gains concentrated in late life. A retiree planning today against current life tables may underestimate longevity if mortality continues to improve. Conversely, recent OECD data show slowing improvements in some countries, including the US, France, and the UK, complicating this assumption.

Synthesis by regime: in regimes of strong mortality improvement (1980-2010 for most OECD countries), retirement planning needed continual upward adjustment to match realized longevity gains. In regimes of stagnating or declining life expectancy (post-2014 for the US, post-COVID broadly), planning anchors can be more stable. The transition parameter is the trajectory of healthy life expectancy at age 65, which varies materially by country and by socioeconomic group within countries.

Planning to the median is planning for half the population to fail. The right anchor is the percentile you would not regret in hindsight.

→ Framework: Asset allocation and resilient portfolios

What it means for different economic actors

Single retirees face individual longevity risk and benefit from planning to the upper percentile of life expectancy distribution rather than the median.

Couples face a more nuanced problem: the planning horizon is the longer-lived partner’s lifespan, which extends materially beyond either individual median. Joint life expectancy considerations dominate single-person actuarial inputs.

Pension funds and life insurers can pool longevity risk across many lives, which is why annuity products are theoretically efficient longevity hedges even when individual retirees prefer self-management.

A common error is to plan to one’s own family longevity history (“my parents died at 78”) without recognizing the regression-to-the-mean dynamic in lifespan. Family history matters but is one input among many, not a deterministic forecast.

Practical observation

Go deeper

Related questions

• Why does the 4% rule face challenges?
• How do annuities work and when are they useful?
• How much do you need to save?
• How does inflation erode retirement purchasing power?

Frequently asked questions

Should I plan to my family longevity or to population statistics?

Both inputs matter. Family longevity history captures genetic and lifestyle factors that population averages do not, but it is a small sample with regression-to-the-mean dynamics. Most actuarial frameworks use population life tables as the base and adjust upward for non-smoking, education, and good health behaviors. Pure family-history planning typically underestimates uncertainty, which is precisely the dimension longevity risk requires you to acknowledge.

How does longevity risk differ between France and the US?

French residents have materially longer life expectancy than US residents — by approximately 3-4 years at birth and a similar gap at 65. France also has strong public pension and healthcare backstops that reduce the financial impact of long lives. US retirees face longer financial planning horizons relative to their public benefit floor, which makes longevity risk a more salient personal financial risk in the US than in France.

Can longevity risk be hedged?

Conceptually yes, primarily through life annuities that convert lump sums into guaranteed income for life. The annuity puzzle (Yaari 1965, Modigliani 1986) documents that despite theoretical efficiency, annuitization rates are very low in practice — under 5% of voluntary US retirement assets. Behavioral biases (legacy preferences, illusion of control, distrust of insurers) dominate rational analysis. Partial annuitization that covers basic expenses while leaving discretionary spending in marketable assets is one practical compromise.

Last updated — 4 June 2026