Introduction The UK pension buyout market exceeded £50 billion in 2024, yet current industry mortality tables (S3PA, SAPS) aggregate data at the postcode sector level, potentially masking significant within-sector heterogeneity. Building on the Renshaw–Haberman cohort mortality framework, this work-in-progress examines whether finer postal district granularity (average 2,500 residents vs. 7,000 for sectors) improves longevity risk assessment and pricing accuracy in bulk annuity transactions.
Methods We analyze ONS mortality data (2015-2023) across UK postal districts, linked to Index of Multiple Deprivation scores and geodemographic classifications. Using Lee–Carter models with geographic random effects and stratified Cox proportional hazards with frailty terms, we quantify life expectancy differentials controlling for socioeconomic factors, with validation against recent mortality experience. We apply postal district mortality rates to representative defined benefit pension scheme portfolios, comparing reserve adequacy against S3PA baseline assumptions through Monte Carlo simulation under various geographic concentration scenarios.
Results Preliminary findings indicate substantial life expectancy variation across postal districts, exceeding current sector-level adjustments in industry tables. Geographic concentration of scheme membership appears to drive material pricing divergence from standard assumptions, with implications varying by deprivation profile and regional mortality patterns. Both positive and negative pricing adjustments emerge depending on membership characteristics. Results demonstrate robustness across alternative deprivation indices and sensitivity analyses on mortality improvement assumptions. Statistical significance testing and quantification of financial impacts are ongoing.
Conclusions Postal district granularity shows promise for materially improving bulk annuity pricing accuracy, directly addressing FCA fair value assessment requirements. The transparent methodology using publicly available ONS data offers insurers enhanced risk selection capabilities while providing pension trustees with tools for better membership longevity profiling. This approach is generalizable to other countries with fine-grained mortality statistics and supports enhanced regulatory capital modeling under Solvency II. We will present completed quantitative findings and practical implementation frameworks at the conference.
