Introduction
This document represents Club Vita (UK) LLP’s response to the CMI Mortality Projections Committee’s Consultation on CMI_2026, as set out in CMI Working Paper 216. We are responding to the consultation in our capacity as a provider of mortality and longevity analytics services to both pension schemes and (re)insurers.
We thank the Projections Committee and CMI Secretariat for their work in producing the working paper and maintaining the CMI Model.
Introductory comments
We welcome the greater transparency provided by the new model parameters paper. Following recent significant model changes, we also broadly support a “steady as she goes” approach for CMI_2026, given the continuing uncertainty over emerging post-COVID trends. We summarise the key themes of our response below.
Uncertainty in current trends: The historically light mortality in the provisional 2026 data raises important questions. Does it signal an acceleration in mortality improvements, represent a lighter-than-average outlier within a continuation of pre-COVID trends, or reflect the knock-on effect of deaths brought forward during the pandemic? We consider all three interpretations plausible and encourage the Committee and users of the CMI Model to remain open-minded as further evidence emerges.
Pandemic treatment: Within the existing fitted overlay framework, we favour retaining a one-year half-life for CMI_2026. In our view, an alternative calibration that excludes 2020–2023 data also provides a plausible projection. We encourage testing both approaches, alongside alternative half-life values, against different mortality scenarios for 2027 and 2028.
Projection assumptions: We favour a flat 20-year age-period convergence period across all ages. This would simplify interpretation and address rapid increases in projected improvements at older ages. On balance, we support retaining the current cohort convergence approach pending further evidence, and the current long-term rate age taper pending the high-age mortality review.
Fostering engagement with projection parameters: We propose making Initial Addition a second Core parameter and age-period convergence an Extended parameter. This would give greater prominence to intuitive parameter choices that materially affect projections, encouraging users to consider them explicitly.
This response was prepared for and provided to the CMI Mortality Projections Committee to support its ongoing work in maintaining the CMI Mortality Projections Model. We are also publishing it on our website to contribute to wider discussion of the CMI Model and the selection of appropriate longevity improvement assumptions in the post-COVID era. This response was not prepared for, and should not be relied upon by, any third party. We accept no liability to any third party arising from its use or reliance upon it, unless we have expressly agreed otherwise in writing.