The Impact of Covid-19 on Mortality Assumptions when Valuing UK Defined Benefit Pension Schemes

Year 2020 has been an unexpected year, due to the emergence of the coronavirus, Covid-19. This virus caused a rise in number of deaths in many countries including the UK. Currently, the pandemic is still causing excess deaths in 2021, but less than what was experienced in 2020 and this may be due to effective vaccines that started being administered towards the end of 2020. Nevertheless, there is great uncertainty in how future mortality may be impacted and hence various views on this impact have resulted. Consequently, setting accurate mortality assumptions for pension valuations can become even more difficult. In this report, mortality data on England and Wales is studied as this is the data used to calibrate the popular mortality projections model, the CMI model, used in many UK pension valuations. The CMI_2019 and CMI_2020 models are used to test how the extended parameters may be adjusted, considering three differing views on the future mortality impact of Covid-19, as well as the resulting impact on liabilities of a pension scheme when these adjustments are made. The results show that in-line with the view that the pandemic will not impact future mortality, upgrading to the CMI_2020 model with a core value of 0% for the w_2020 parameter causes an approximate 0.16% fall in liabilities. If Covid-19 is believed to have a negative impact on future mortality instead, then an upgrade to the CMI_2020 model with a non-zero weight on the w_2020 parameter, a lower LTR in the 2019 model or a lower A parameter value in either the 2019 or 2020 model could be used, resulting in a reduction of liabilities. Of these responses, the greatest fall in liabilities of 3.72% occurred when an upgrade to the 2020 model using full weight on the w_2020 parameter was used. On the other hand, if the virus is believed to have a positive impact on future mortality, a higher LTR parameter could be used or higher A parameter in either the 2019 or 2020 model causing an increase in liabilities. The highest increase in liabilities was 2.22% which resulted when a LTR of 2.00% was used instead of 1.50%.