The Continuous Mortality Investigation (CMI) of the UK Actuarial Profession has set out proposals for the latest annual version of its mortality projections model, “CMI 2024”. The model projects the path for future mortality rates and is used by most defined benefit (DB) pension schemes when setting assumptions that relate to how long members will live.
If the proposals are carried into the final model, due to be released in the second quarter of 2025, life expectancy will typically increase for older scheme members and reduce for younger members compared with last year’s model, with these changes more pronounced for men than for women. For example, life expectancy at age 85 increases by a little under two months for men and a little over one month for women, whereas life expectancy at 45 falls by a little over three months for men and a little under three months for women. The impact on a scheme overall will vary depending on the age profile of its membership but, given the maturity of DB pension schemes in general, scheme liabilities are more likely to increase than decrease.
The CMI’s proposal would see 2024 data incorporated into the model for the first time as expected. However, it has also proposed some fundamental changes to how the model works. The two most important of these are:
Average England and Wales mortality rates in 2024 were the lowest on record, being slightly lower than the previous best year, 2019, right before the COVID pandemic. However, it’s a mixed picture, for example mortality for adults under 45 is still higher than before the pandemic. The CMI’s proposals will lead to a more optimistic outlook for the life expectancies of older scheme members due to their bounce back to below pre-pandemic mortality rates and an anticipated reduction of the ‘COVID’ effect built into the model. The outlook for younger members is less positive and reflects the challenges seen in the population for this age group in returning to pre-pandemic levels of health and mortality.
For DB pension schemes, the adoption of CMI 2024 instead of CMI 2023 may lead to a small increase in liabilities. However, for triennial valuations taking place this year, which may have last used the CMI 2021 model, adoption of CMI 2024 will likely result in a reduction in liabilities. Average mortality rates in 2024 may have returned to pre-pandemic levels but are still well above where we expected them to be by now based on the expectations in place before the pandemic (which were still reflected in the 2021 model under a ‘wait and see’ approach).
While funding objectives are important, schemes that are looking towards near-term buy-in/out will be more interested in how insurers’ views evolve this year using their own models.
A number of other minor technical improvements are also proposed to be made to the model. The resulting projections under the proposal are similar to CMI 2023. The table below shows the percentage change in life expectancy, which varies across the age range.
| Sex | Age 25 | Age 45 | Age 65 | Age 75 | Age 85 |
|---|---|---|---|---|---|
| Male | -0.5% | -0.7% | +0.5% | +1.4% | +2.3% |
| Female | -0.3% | -0.5% | -0.1% | +0.6% | +1.3% |
Note: Life expectancy changes based on a typical base mortality table as at 1 January 2025, core model parameters and a long-term rate of 1.5% pa.
The consultation questions are summarised in the CMI’s working paper (access is available only to subscribers), and seeks comments from authorised users on the CMI’s proposed approach to CMI_2024. The consultation closes on 25 March 2025, with an update from the CMI due to be published during April and the final model before the end of June 2025.
