Second analysis of the mortality experience of enhanced annuities data
The Continuous MortaIity Investigation (CMI) Annuities Committee has published our second analysis of the mortality experience of enhanced annuities data – covering the period 2011-2019 – to CMI Subscribers in Working Paper 155.
The paper also includes an indicative analysis of experience for the first six months of 2020 – a period when the Covid-19 pandemic caused abnormally high mortality for the general population of the UK.
Enhanced annuities provide an income in retirement but pay more than standard annuities because the individual is considered to have underlying health conditions or lifestyle indicators (such as smoking) that result in a reduction to their life expectancy. These are also sometimes known as impaired life annuities.
We compared actual deaths of the enhanced annuity dataset by age and gender with those expected using the “16” Series pension annuity in payment tables, adjusted for improvements in “population” mortality experience using CMI_2020. These tables are based on data for standard annuities.
As might be expected given the definition of an enhanced annuity, the experience of the all-ages, all-durations enhanced annuity dataset is substantially higher than the projected “16” Series all products tables, on both lives and amounts bases.
For both genders, the higher mortality experience seen in the enhanced annuity dataset is most pronounced at younger pensioner ages, with differentials in mortality tapering out until there is almost no difference at very old ages.
Mortality experience appears to increase with increasing duration but this may be a function of “generational effects”, not duration itself, as we are comparing experience of annuities purchased at different times.
The enhanced annuity market has grown and developed in recent years. Longer duration policies may therefore cover a different range of health conditions to newer, shorter duration policies, which may cover a wider range of health or lifestyle conditions.
The paper includes an analysis of experience by calendar year for the years 2011 to 2019 to examine how mortality improvements in the enhanced annuity dataset compare with those of the population of England & Wales. We note, however, that the enhanced annuity dataset is relatively small for assessing year-on-year changes and the composition of the dataset may have changed over time (e.g. due to changes in underwriting practices).
For males, average improvements in the enhanced annuity dataset are similar to average improvements for the wider population, over this period.
For females, improvements seem to have been a little higher than the improvements in the population, although late reporting estimates in later years and low data volumes for females mean there is considerable uncertainty in this result.
A previous analysis of enhanced annuities data covered the years 2007-2010 and was published in Working Paper 87. Data for enhanced annuities is not part of the Annuities Committee’s regular data collection cycle but we sought new enhanced annuity data in order to revisit the analysis and provide a more up-to-date view of experience.
The dataset underlying the 2011-2019 analysis is substantially larger, reflecting a growth in sales in the years before Pensions Freedoms but also a more mature dataset, with more lives at older ages.
Mortality experience for males is similar in both analyses. Mortality experience was higher in 2007-2010 for females than in 2011-2019, however, this may be due to there being relatively little data available for females in the first analysis or due to changes in the composition of the dataset (e.g. an increase in the proportion of second lives in the 2011-2019 dataset).
The Covid-19 pandemic caused abnormally high mortality in the UK during the second quarter of 2020. We have assessed how experience in the first six months of 2020 compares with that in 2019, for a consistent group of data contributors, who were able to provide data to 30 June 2020 in addition to their data for 2011-2019. We previously carried out a similar analysis for standard annuities, in Working Paper 140.
The analysis should be regarded as indicative, but our main findings are that:
Mortality rose dramatically in April 2020 and remained high in May for population mortality, enhanced annuities and standard annuities. It returned to normal levels in June for population data and standard annuities but remained high (relative to 2019) for enhanced annuities.
Relative excess mortality in the first half of 2020 for ages 65-85 is similar for enhanced annuities (16.5%) and standard annuities (16.6%); both are lower than that for the England & Wales population (20.5%).
Will mortality tables for enhanced annuities be produced in future?
Analysis in our paper is carried out relative to mortality tables based on standard annuities. We include some indicative graduations of the enhanced annuity dataset in the paper, as these help illustrate how (relative) mortality changes with age. However, we have no plans to produce mortality tables for enhanced annuities as we understand that any such tables would still require significant adjustment by users given the varied health and lifestyle factors underlying these products.
The committee intends to provide our annual update on the experience of standard pension annuities in payment, subject to receipt of data, by late 2021. This should include analysis of experience of the whole of 2020 and, if data permits, experience of pension buy-in business and indicative analysis of experience to 30 June 2021.
Looking further ahead, we intend to:
