A closer look at 2023: What’s changing in term assurance experience?

Mortality: Lighter Experience, But Not Uniformly So

The mortality experience in 2023 was lighter than expected across all sex and smoker groups, continuing the trend seen in 2022. This is particularly notable given the heavier experience observed during the pandemic years of 2020 and 2021.

However, the story isn’t uniform:

• Male non-smokers showed lighter-than-expected experience at ages 55–64, but heavier experience at older ages.
• Female non-smokers displayed a consistent pattern of increasing experience with age, suggesting that the current age-shape of the T16 tables may be underestimating mortality at older ages.
• Smoker subsets remain more volatile, with lower data volumes making it harder to draw firm conclusions.

We also saw a widening gap between lives-weighted and amounts-weighted experience, particularly for male smokers, hinting at potential differences in experience by sum assured.

Socio-Economic and Regional Insights

We continue to see strong correlations between mortality experience and socio-economic indicators:

• Index of Multiple Deprivation (IMD) deciles show a clear gradient: more deprived areas experience heavier mortality.
• Regional analysis reveals a persistent North/South divide, with heavier experience in Scotland, Northern Ireland, and parts of northern England.

These findings reinforce the importance of considering socio-economic status in future table development, especially as our IMD data coverage has improved significantly (now covering 78% of mortality claims in 2023).

Accelerated Critical Illness: A Mixed Picture

The accelerated critical illness (ACI) experience in 2023 was slightly heavier overall than in 2022, though still broadly in line with the AC16 tables. Notably:

• Male non-smokers tracked closely to expected levels.
• Female non-smokers were 5% heavier than expected, with a persistent spike in claims at ages 50–54—possibly linked to breast cancer screening.
• Smoker subsets again showed more volatility, with wide confidence intervals limiting the strength of any conclusions.

Interestingly, 2023 saw a shift in experience by sum assured band, with heavier-than-expected experience at higher sums assured—as opposed to the little variation by sum assured bands seen in earlier years.

Cause of Death and Claim: New Insights

For the first time, we’ve been able to analyse cause of death for mortality benefits. Cancer remains the leading cause across all groups, with its share increasing in 2023. Circulatory diseases are the next most common cause, particularly among males.

On the ACI side, cancer also dominates, with notable differences by sex and smoking status. For example, female non-smokers have a higher proportion of stroke claims, while male smokers see more heart attack claims.

These insights could inform future refinements to the tables, especially if certain causes of claim are driving unexpected experience in specific subgroups.

Are New Tables on the Horizon?

So, do the 2023 results justify new tables?

Not yet—but we’re watching closely. While the “16” Series tables still provide a reasonable fit overall, there are signs that some aspects—particularly the age-shape and select period assumptions—may need revisiting. For example:

• The select shape for female non-smokers may be too steep.
• The duration 1 rates for ACI may be too high for male non-smokers.
• The age-shape of mortality rates for non-smokers may no longer reflect current experience.

We’re also mindful of the growing importance of socio-economic factors like IMD and sum assured band, which could warrant inclusion in future table splits, but we must also consider the needs of users when deciding the factors to split new tables by.

What’s Next?

Looking ahead, we plan to:

• Analyse rated lives experience for 2021–2023 in a standalone paper (targeting release by August).
• Update our guidance on using the CMI Model for term assurances, following the release of CMI_2024.
• Continue annual reporting, with the next analysis covering 2024 data and revisiting 2023 results as more actual claims replace estimates.
• Explore new analytical approaches, including multivariate techniques, to support future table development.

We’re also considering whether to group experience from 2022–2024 in our next analysis. This could help smooth out year-on-year volatility and provide a stronger basis for assessing the need for new tables.

Final Thoughts

We hope this latest update offers a valuable snapshot of the experience of term assurances. While the “16” Series tables remain broadly fit for purpose, emerging trends—especially by age, duration, and socio-economic status—suggest that change may be on the horizon.

As always, we welcome your feedback. If you have questions or comments on the paper or this blog, please get in touch at assurances@cmilimited.co.uk.