Funded reinsurance

Some insurers have disclosed that they do use funded reinsurance, including Just, Phoenix and PIC, although Rothesay, for example, does not. This remains an area of focus for the PRA, which published CP24/23 – Funded reinsurance last year, and funded reinsurance has been included as a business objective in its 2024/2025 business plan. Funded reinsurance can improve pricing, reduce capital requirements and potentially broaden asset allocation, all of which will be attractive for writing new business. It remains to be seen how the balance between this and the regulatory focus will play out in future.

Longevity releases

One key theme, which has continued from 2022, is that several companies released longevity provisions due to excess mortality arising from the pandemic and amendments to their longevity bases. Insurers noted the challenges in setting assumptions for longevity post-COVID-19, and the substantial expert judgement required. Life expectancies have reduced for many BPA insurers. Some insurers increased their base mortality table multipliers. Many insurers adopted the CMI_2022 model, often placing no weight in its calibration on experience during the years when the number of deaths was high (i.e., 2020-2022). Firms also adjust from time to time other parameters in the core CMI_2022 model, such as the period smoothing parameter S_k and the initial addition to mortality improvements parameter A. It is important that conclusions are not drawn solely based on changes to the multipliers to the base table and/or long-term improvement rates, but on the overall impact on life expectancies and/or changes in liabilities. Although most insurers reinsure a high proportion of longevity risk, the net surplus contributions from longevity assumption changes are still significant.

We note that the CMI_2023 model was published on 18 April, with the two key updates to the model from CMI_2022 being updates to the data (including using ONS figures from the 2021 census)—which would slightly increase life expectancy, and placing a single weight of 15% on 2022 and 2023 data within the core parameterisation of the model (CMI_2022: 25% weight on 2022 data and 0% weight on 2020 and 2021 data)— which would slightly reduce life expectancy. It remains to be seen how insurers will parameterise CMI_2023 once adopted, and the overall impact on life expectancy.

Matching adjustment

The matching adjustment has reduced during 2023 for most BPA writers, which is consistent with narrowing credit spreads. Credit spreads have reduced to levels close to pre-pandemic levels—for example, the option-adjusted spread on the ICE Bank of America 10-15 Year USD Corporate Index narrowed by c.40bps during 2023, with its value at year-end 2023 very close to the value at year-end 2020. Changes in asset mix were also a factor, some beneficial to MA and others not. Particularly for large schemes, which are increasingly being written, it can take a considerable time to transition assets into the ultimate portfolio, and there may be some tactical decisions around the timing of this. We believe, for example, that this may be a contributing factor to the reduction in reported MA for Rothesay from 155 bps to 117 bps (which would also be consistent with its higher new business strain). We continue to see insurers, in particular BPA players, continue to invest a substantial proportion of their balance sheets into illiquid assets, as highlighted also in Milliman’s Illiquid Asset Survey (see: https://uk.milliman.com/en-gb/insight/2023-illiquid-asset-survey), which means that for many, the proportion of these will continue to increase.