The CMI Annuities Committee has recently released CMI Working Paper 178, presenting an analysis of pension annuity mortality data up to the end of 2022, available to CMI subscribers. This paper provides an initial glimpse into the 2022 experience, and we are grateful to the data contributors who expedited their submissions, making this possible. Alongside the paper, we have also published a spreadsheet containing underlying values for all charts and an interactive data visualisation in Tableau to aid interpreting the results.
As I compose this note, many of my colleagues are hard at work, proposing critical assumptions for year-end reporting and engaging in discussions with their boards. This year presents some pivotal decisions for many. In the wake of the unprecedented COVID-19 pandemic, insurers have taken a prudent ‘wait-and-see’ approach concerning annuity assumptions over the past couple of years. This cautious approach is a response to the difficulties involved in handling data from the tumultuous years of 2020 and 2021.
As we approach year-end reporting, industry benchmarking (detailed in CMI Working Paper 179) indicates that some insurers may, for the first time, reflect the impact of post-pandemic mortality on their balance sheets. Insurers will, without doubt, analyse the performance and experience within their portfolios. However, it’s crucial to acknowledge the substantial uncertainty enveloping the available data, with many factors to consider when forming a view on what post-pandemic mortality will look like. Therefore, there is significant value in exploring more extensive datasets.
This is where Working Paper 178 is particularly relevant. Following the approach taken in the last couple of years, the CMI Annuities Committee has taken the strategic step of expediting the production of our annual analysis, ensuring that a report containing some headline early results is accessible earlier than the norm. The paper offers initial insights into the 2022 data for pension annuities currently in payment, focusing on analysis of experience by month. This data is sourced from 6 of the 10 insurance companies that consistently contribute to our annual analysis of pensions in payment, covering approximately 67% of the typical full dataset.
The paper presents some intriguing findings. While 2020 and 2021 witnessed dramatic spikes in standardised mortality rates (SMRs) – a measure for comparing mortality rates over time – due to the COVID-19 pandemic, affecting both population data and annuitants, 2022 departs from these extreme patterns. Annuitant mortality in 2022 seems to be approaching levels similar to those observed in 2019, albeit remaining higher overall.
Typically, mortality exhibits seasonal variations, with more deaths occurring during the winter months. In comparison to 2019, the data for 2022 displays less evidence of seasonality, and does not exhibit as pronounced a decline in mortality during the summer months as was seen prior to the pandemic. There are multiple factors that may have contributed to this change (including factors like the NHS crisis, long COVID, and COVID) but it is noteworthy that 2022 has seen record-breaking temperatures in England in several locations leading to excess deaths over that period. Actuaries typically exclude unusual events from their data when setting assumptions. Still, it raises the question of whether the events of summer 2022 were indeed abnormal especially considering scientists predictions of more frequent extreme weather incidents due to global warming.
For individuals aged 65 to 95, encompassing both genders, SMRs for the England and Wales population in 2022 were 5.7% higher than in 2019. In contrast, the corresponding figures for annuitants were lower, at 3.2%. Notably, throughout each year of the pandemic, relative excess mortality for annuitants consistently remained lower than that observed in the England and Wales population. This observation aligns with expectations, as there is evidence suggesting that excess mortality, during and after the peak of the pandemic, has been more pronounced in less affluent segments of the population. Given that annuitant data represents a more affluent profile than the average population, lower excess mortality is expected.
Moreover, we observed higher excess mortality at the younger ages (65 to 74) compared to the older ages (85 to 95), which contrast with the England and Wales population. Considering that much of the data used for mortality projections relies on population data, it is vital to understand the divergence in mortality rates between annuity portfolios and the broader population.
This accelerated analysis serves as a valuable resource as we navigate the intricate landscape of post-pandemic mortality and make informed decisions for our stakeholders.
We are now turning our attention to our annual update on the experience of pensions annuitants in payment, which we expect to follow later this year, covering the period to the end of 2022 for the full dataset.