The Margin Cost of Mortality Improvement Assumptions After Shocks

How Stale Mortality Assumptions Quietly Drain Reinsurance Margin

A mispriced mortality improvement assumption rarely announces itself with a single bad headline number. Instead it shows up as a slow, compounding drag on margin that keeps surfacing renewal after renewal, long after the structural shock that caused it has stopped making news.

For life and health reinsurers, this matters more than almost any other pricing input, because mortality improvement assumptions are baked into treaties that run for decades. A trend that is off by even a percentage point or two compounds into a meaningfully different reserve and profitability picture by the time the business fully seasons.

How Exactly Does a Wrong Mortality Improvement Assumption Hit Margin?

It hits margin in one of two directions, either the treaty is overpriced and loses competitive business, or it is underpriced and books losses as claims come in.

If a reinsurer keeps assuming the pre-shock improvement rate continues, and actual mortality has settled onto a slower improvement path, pricing will be too low relative to true risk. Every treaty written during that window locks in a rate that undercharges for the risk actually being taken on.

The reverse is just as costly. If the market overcorrects and assumes mortality has permanently worsened when it has not, treaties get priced too conservatively, cedants shop the business elsewhere, and the reinsurer loses volume it could have profitably retained.

Either direction is a margin problem, and both stem from the same root cause: an improvement assumption that has not caught up with reality.

How Much Has Actual Mortality Really Deviated From the Old Trend?

Documented group life data shows mortality running 3 to 5 percent above where the pre-pandemic trend would have put it, even years after the acute crisis passed.

Munich Re's post-pandemic group life mortality analysis found that by year-end 2023, overall mortality rates had fallen below their 2017-2019 pre-pandemic levels in absolute terms. Even so, mortality "remained elevated as compared to the pre-pandemic trendline projected forward," running an estimated "3-5% higher than it would have been had the pre-pandemic downward trend continued uninterrupted."

That gap, a few percentage points, sounds small until it is applied across a multi-billion-dollar in-force book compounding over a treaty's full duration. The same analysis flagged that younger cohorts showed elevated cause-specific mortality, particularly overdoses and accidents, trends that were separate from and in some cases predated the pandemic itself.

This is the risk-diagnosis side of the same problem covered in mortality improvement assumptions after structural shocks, and it is the input that turns a diagnosis into a dollar figure.

Does the Margin Impact Show Up Right Away?

No, the impact typically builds gradually across several renewal cycles rather than appearing in a single reporting period.

Treaties written during a period of stale assumptions do not immediately reveal their mispricing. Claims take years to emerge, especially on long-duration life business, so the gap between assumed and actual mortality shows up slowly as actual-to-expected ratios drift and as reserves need periodic strengthening.

By the time the cumulative effect is visible in headline profitability, several renewal cycles of business have already been written on the same flawed assumption, which is why early detection matters so much more than after-the-fact correction.

Which Portfolios Carry the Most Exposure to This Margin Erosion?

Long-duration life reinsurance treaties with limited repricing flexibility carry the greatest exposure, since the stale assumption compounds across more years of in-force business.

A one-year renewable term treaty can correct its assumption at the next renewal with relatively little lasting damage. A 20 or 30-year life treaty locked into pricing set during a period of assumption uncertainty carries that error for the full duration, with no natural repricing point to fix it.

This is exactly why longevity and pension risk transfer books, discussed from a related angle in longevity risk transfer and how reinsurers backstop pensions, treat mortality and longevity trend assumptions as a continuously monitored input rather than something to set once and revisit only at scheduled reviews.

How Should Executives Quantify This Cost Before It Shows Up in Reported Numbers?

Executives should model margin exposure using a range of deviation scenarios applied against the actual in-force block, not wait for the deviation to appear in reported loss ratios.

The most useful version of this exercise is a scenario table: take the current book of long-duration treaties, apply a low, medium, and high deviation assumption against the pre-shock trend, and translate each into a present-value margin impact. This turns an abstract actuarial concern into a number a CFO or board can act on immediately, rather than waiting for the deviation to become visible in reported financials years later.

A reinsurer that has already run this exercise walks into a renewal negotiation, a rating agency conversation, or a board meeting with a quantified range instead of a qualitative concern. That difference matters disproportionately in exactly the kind of environment a structural shock creates, where competitors without the same modeling discipline are still pricing off assumptions nobody has stress-tested.

How Does This Interact With Capital Requirements, Not Just Profit?

If reserving assumptions lag real mortality experience, required capital can be understated at the exact point in time when true risk has increased.

Profitability and capital adequacy are two sides of the same assumption. A stale improvement scale does not just misprice new business, it can also understate the reserves needed to support in-force liabilities, since reserving methodologies typically reference the same improvement assumptions used in pricing.

That means the margin erosion problem and the capital adequacy problem tend to move together. A reinsurer that only tracks profitability metrics without also stress-testing capital against updated mortality assumptions is looking at only half of the actual exposure.

Does the Margin Impact Differ Across Treaty Types?

Yes, the mechanism differs meaningfully between yearly renewable term, coinsurance, and asset-intensive or longevity-linked structures, even though all three share the same underlying assumption risk.

A yearly renewable term treaty reprices annually, so a stale mortality improvement assumption mostly shows up as a series of correctable mispricings rather than one large locked-in error, since each renewal offers a fresh opportunity to true up the rate to whatever the latest actual-to-expected data supports. A coinsurance arrangement, where the reinsurer shares in the full economics of the underlying policies for their full duration, carries the mispricing much further forward, since there is no annual repricing point and the original assumption keeps compounding against the reinsurer's share of reserves for as long as the policies stay in force.

Asset-intensive and longevity-linked structures add a further wrinkle, because they typically combine mortality or longevity assumptions with investment return assumptions in the same pricing model, which means a mortality assumption error can interact with, and sometimes mask or amplify, a separate investment assumption error in ways that are harder to isolate without careful decomposition. A reinsurer managing a mixed book of all three treaty types needs to recognize that "margin erosion from mortality assumption drift" is not a single number, it is a different magnitude and a different correction timeline depending on which structure the exposure sits in, and executive reporting on this risk should reflect that breakdown rather than a single blended margin figure.

How Should This Cost Be Communicated to Cedants and Retrocessionaires?

The margin impact should be communicated as a quantified range tied to specific cohorts and treaty terms, not as a general statement that pricing needs to increase.

Cedants and retrocessionaires respond very differently to a reinsurer that shows up with "rates need to go up because of the pandemic" versus one that shows up with a specific breakdown: this cohort is running X% above trend, sustained over Y periods, translating into a Z basis-point margin impact on this specific treaty structure. The second version is a negotiation position grounded in shared, verifiable data, and it tends to produce a faster, less adversarial renewal conversation than a general appeal to industry-wide uncertainty.

This also matters for retrocession arrangements, where a reinsurer ceding part of this same mortality risk further upstream needs its own retrocessionaire to understand the deviation in the same granular terms. A reinsurer that can only describe the margin problem qualitatively to its own retrocession partners is likely to get less favorable retrocession terms than one that can hand over the same cohort-level actual-to-expected data it used internally to make the pricing case.

Margin erosion from mortality assumption drift is not a one-time event to absorb and move past. It is a compounding cost that keeps accruing on every treaty written before the assumption catches up with reality.

That is exactly why the size of the gap, not just its existence, needs to be quantified and tracked continuously rather than assumed away once the immediate crisis fades from view.

Sources

• Munich Re: Group Life mortality trends, a post-pandemic perspective
• Society of Actuaries Research Institute: RPEC 2025 Mortality Improvement Update