CMI Working Paper 210

Working Paper 210 was published in March 2026. This paper compares data, published mortality tables and mortality experience for bulk annuities (collected by the Annuities Committee) and private sector pension schemes (collected by the SAPS Committee) in 2016-2023. 

The analysis shows that:

• The “16” Series pension buy-out mortality tables are lighter than “S4” Series Pensioner mortality tables and heavier than “S4” Series Dependant mortality tables.  
• Two key differences between the Bulk annuity and SAPS datasets are that: 
  • The SAPS dataset and “S4” Series tables are differentiated by pensioner type while the Bulk annuity dataset and “16” Series pension buy-out tables include a combination of pensioners and dependants, as coverage of bulk annuity data with pensioner type specified is low. In this paper, we primarily show SAPS experience for an aggregate dataset (consisting of both pensioners and dependants) to aid comparison with the Bulk annuity dataset.  
  • The SAPS dataset and “S4” Series includes both public and private sector pensioners, while we anticipate that the Bulk annuity dataset and “16” Series pension buy-out tables relate to schemes that were previously private sector pension schemes. In this paper we restrict the SAPS dataset to private sector data to allow for consistent comparison of mortality experience of the Bulk annuity and SAPS datasets. 
• For both males and females, mortality experience of the 2016-2023 Bulk annuity dataset is similar to the SAPS aggregate dataset on a lives-weighted basis. On an amounts-weighted basis, experience of the Bulk annuity dataset is slightly lighter than the SAPS aggregate dataset.  
• For males in both the Bulk annuity dataset and SAPS aggregate dataset, mortality experience relative to “S4” Series Pensioner amount band tables is fairly flat, suggesting these tables allow well for differentials in experience for individuals with different pension amounts. Similar features are also noted for males relative to the “S4” Pensioner Index of Multiple Deprivation (IMD) tables. 
• For females in both the Bulk annuity dataset and SAPS aggregate dataset, some variation by amount band is seen relative to the “S4” Series Pensioner amount band tables, suggesting that these tables do not fully allow for differentials in experience for females by pension amount. Similar features are again noted for females relative to the “S4” Pensioner IMD tables. These differentials may be linked to the mix of pensioner and dependant data in the female datasets.  

A spreadsheet that includes the values underlying the charts in the working paper and data visualisations in pivot chart format accompanies this paper.

Note: this paper and the accompanying output are available to Authorised Users only.