Let's talk

Next mortality projections model continues to ignore pandemic impact

Tim Camfield Senior Consultant
Chris Tavener Partner and Head of Life Analytics

Last week the Continuous Mortality Investigation published some of the background behind their next annual release of their mortality projections model, CMI 2021, which is planned to be available in March 2022.

The key headline is that scheme trustees, company directors and their actuaries will be faced with a default option to ignore the high number of deaths due to the Covid-19 pandemic when setting future mortality improvements. But does this reflect reality?

The £50bn impact of this assumption

The CMI’s models are used by actuaries to value most UK defined benefit pension schemes, including for sponsors’ accounting disclosures, trustees’ funding valuations, and when assessing the value for money of insurance products like buy-ins and longevity swaps.

For the release early this year, CMI 2020, the CMI made the decision to exclude the high number of deaths over 2020 in England & Wales within the core model (users can choose to adjust this to make some allowance for 2020 if they wish).

The core CMI 2021 model will likely follow a similar approach, excluding both 2020 and 2021 data when setting future improvement assumptions. This means that future mortality assumptions for many schemes will not materially reduce, despite the current and future impact of the pandemic.

At the other end, fully allowing for 2021 experience could broadly lead to a 3% reduction in pensioner life expectancies, or around eight months for someone aged 65. The impact will be less for those who choose to make a partial allowance. For a typical scheme there would be a similar percentage impact on liability values - meaning an aggregate swing of c£50bn across UK defined benefit pension schemes

This choice is tricky and not small fry – it could be the difference between a deficit and a surplus for a number of schemes, with resulting impacts for sponsors’ finances, scheme deficit contributions, or determining whether longevity insurance appears good value.

Should you make an allowance for 2021 mortality experience?

While, understandably, many have so far steered away from incorporating 2020 mortality experience when setting assumptions, it is less clear whether the same approach should be taken for 2021.

One argument is that 2021 should be seen as the start of a “new normal” because, despite the success of booster vaccines, we might still expect some elevated level of deaths of pension scheme members for the foreseeable future, partly directly due to Covid-19, but perhaps more importantly indirectly due to knock-on challenges for treating diseases, such as heart disease and cancer. Some of these challenges are discussed further here. In our view, 2021 has some unusual mortality experience, especially towards the start of the year, but overall it is not as anomalous as 2020.

Those valuing a scheme before CMI 2021 is released in March 2022 (eg scheme sponsors with financial year-ends at 31 December 2021) who wish to reflect some of the mortality experience over 2021 could make a comparable proxy adjustment to CMI 2020.

Choosing assumptions specific to your scheme

Every pension scheme has a unique composition of members which impacts the mortality experience of their membership; such discrepancies have been exacerbated by the pandemic.

LCP analysis shows that “a chasm is developing in life expectancy between rich and poor”. Not only do the affluent have a higher life expectancy, but their life expectancy is increasing at a faster rate.

While many schemes allow for different current mortality assumptions depending on the profile of their membership, our analysis shows that this is becoming increasingly important when assessing how you expect mortality rates to improve in the future. Understanding the drivers for such changes is increasingly important for ensuring you are selecting the right assumptions for your membership.