The COVID 19 pandemic has disrupted well-established processes for understanding and forecasting mortality and longevity experience. At the height of the pandemic, we observed spikes in mortality rates that were not representative of steady-state mortality rates.
The exceptional nature of these spikes mean that existing analyses (which generally anticipate smooth progression of mortality rates) are not suitable for analysis of data from this period. Furthermore, as we emerge from the acute phase of the pandemic, actuaries are wrestling with how forecasts should reflect recent excess deaths. These excess deaths suggest that mortality rates are significantly different to what we had assumed they would be before the pandemic, but it’s unclear whether this excess is transient or indicative of a longer-term phenomenon. The Post-pandemic biometric assumption-setting working party was created with the objective of helping actuaries navigate this difficult period.
The working party converged on a driver-based approach for understanding recent mortality rates, and projecting them into the future. This perspective not only encompasses the direct impact of COVID-19 but also includes the indirect effects such as healthcare disruption, changes in lifestyle and behaviour, economic consequences, interactions with other causes of death.
The benefit of a driver-based approach is that the framework allows us to incorporate forward-looking knowledge and judgement in a manner that a purely data-driven or extrapolative approach would not. Because post-pandemic excess deaths are a relatively new phenomenon, there is limited historical data available to inform the latter approaches. While it was not possible to assign excess deaths neatly and precisely to the wide-ranging drivers identified in the paper, we identified disruption to healthcare and endemic COVID-19 (and its sequelae) as the most likely drivers of recent excess deaths.
The paper demonstrates how even a simple driver-based view of excess mortality can provide a scaffold for the parameterisation of mortality and longevity models in the post-pandemic era, and in particular how such a method can be used in conjunction with the CMI model to help understand the implications of different parameter choices. By taking this hybrid approach to mortality projections, actuaries can ensure that assumptions are not just reflective of past data but are also anticipative of future trends. This pivot towards a more dynamic, forward-looking approach in assumption setting is an important tool for remaining responsive to an environment marked by unprecedented volatility.
By proactively shaping our assumptions to reflect emerging realities, actuaries ensure that their work continues to be a cornerstone of dependable financial resilience and risk assessment
If you’re interested in learning more:
