Since 2020 and the start of the COVID-19 pandemic, we have all been acutely aware of the wide-ranging impact that pandemics can have in terms of global health and their further cascading impacts on economies and societies at large.
The pandemic has also shown how in a short time, medical science, technology, and policy can combine to produce and distribute the first vaccines for coronaviruses at a global scale.
It also showed how the same combination of science, technology, and policy oversaw varying levels of non-pharmaceutical interventions (NPIs) implementation through the course of the pandemic, with the stringency of application also varying by geographic area.
The multitude of empirically observed outcomes and the development of novel vaccine technologies from such a recent pandemic has meant that the creation of a next-generation infectious disease model leveraging learnings from the COVID-19 pandemic will be a valuable tool for understanding pandemic risk in a modern context.
The epidemiological catastrophe methodology used by Moody’s RMS LifeRisks can more readily provide insights into complex pandemic impacts, for example from novel coronavirus events, compared to traditional actuarial approaches that can be hindered by a lack of data.
Importance of Pandemic Risk for the Life and Health Insurance Market
Pandemic risk is an important consideration for life insurers as it serves as a systemic example of a life catastrophe event. The global life and health insurance market is material for individuals across the globe for their financial protection against mortality and morbidity and for their long-term savings.
In 2022, the combined global risk and savings premiums for life and health insurers were nearly US$3 trillion, which is projected to increase further in the future. Understanding pandemic risk will continue to be central to those who are exposed to excess mortality risk in the sector.
Moody’s RMS Pedigree in Pandemic Modeling
Since 2006, Moody’s RMS has pioneered probabilistic modeling of pandemic impacts, and since then, the world has experienced an influenza pandemic in 2009 and the COVID-19 pandemic commencing in 2020.
The Moody’s RMS LifeRisks infectious disease modeling solution has evolved to stay on the cutting edge of pandemic risk understanding and our latest model update represents a continuation of this trend of progression.
Adding New Pathogens
The previous version of Moody’s RMS Infectious Disease Model released in 2019 contained two pathogen groups: influenza and emerging infectious diseases. The new model has increased the granularity of pathogens and now explicitly models ‘coronaviruses’ and ‘other respiratory diseases.’
Previously, coronaviruses and other respiratory viruses could be viewed as being implicitly modeled in the emerging infectious disease category, but now they are considered separately. This means that the emerging infectious disease grouping will now consist of non-respiratory pathogens that may emerge.
Overhaul of Epidemiological Modeling
The new Moody’s RMS Infectious Disease Model is based on an updated compartmental modeling framework which:
- Dynamically Models Vaccine Rollouts and NPIs
The impact of vaccinations and NPIs is embedded in the epidemiological compartmental modeling. When a pandemic occurs, there would typically be a time lag before a vaccine is available, hence the impacts of vaccines are applied dynamically and are dependent on the vaccine roll-out periods.
For NPIs, the stringency of how these are applied is influenced by the epidemiological model tracking hospitalizations as the pandemic spreads. This is a more realistic interpretation of how ‘reactive’ NPIs would be implemented throughout the time course of a pandemic.
There is geographical granularity in terms of the likelihood of faster or slower vaccine rollouts; this is influenced by observed variations between countries in terms of their vaccine rollouts during the COVID-19 pandemic.
- Explicit Modeling of Hospitalization
As mentioned above, the tracking of hospitalization in the compartmental model also allows for a refinement in morbidity modeling resulting from a pandemic. The type of morbidity that is referred to relates to the direct pandemic pathogen-related morbidity shocks that occur during a pandemic.
- Accounting for Age Contact Mixing Patterns
The compartmental model is age-structured and mixing patterns between different age groups is informed empirically for different geographic areas. Accounting for differing age mixing patterns allows for more granular and realistic impacts from different NPIs.
For example, a full lockdown would likely reduce mixing across all age groups, but a school closure would have differential impacts by age.
- Seasonality is Accounted for Respiratory Pathogens
Pandemics that start in winter in one hemisphere will have a greater transmissibility compared to the other hemisphere during their summer. These dynamics are pertinent to respiratory pathogens which can show transmission advantages in winter compared to summer.
- Waning Immunity
The new model will also account for waning immunity; we saw for COVID-19 that immunity against infection is not permanent. The consideration of waning immunity leads to multiple wave dynamics that are typically seen in pandemics being generated in this new model.
This feeds into an increased temporal granularity where the total event losses can manifest over multiple individual years.
Overall, these enhancements to the epidemiological model result in a more varied and richer pandemic event set where the constituent events are based on a more realistic dynamic application of key pandemic interventions.
Capturing Geographic Granularity: Pandemics are Global Events, But Impacts Vary
The COVID-19 pandemic highlighted that pandemics are not one-size-fits-all events. We saw variability between nations and even within nations, as exhibited at the state level for the U.S. Also, we saw that the impact varied by sub-population within a geographic area, for instance by socioeconomic group.
Typically, higher socioeconomic groups exhibited lower mortality than lower socioeconomic groups. This is likely due to correlations between socioeconomic status and general health and therefore vulnerability against infection, but also because socioeconomic status was linked to the risk of being infected and the ability to comply with NPIs.
This socioeconomic angle works in conjunction with implementing the learnings from COVID-19 to parameterize geographically distinct impacts from NPIs and vaccine rollouts in a more empirical way.
Enabling Multi-country Portfolio Analyses
Where exposure to pandemic risk is globally distributed, for example, for global life (re)insurers, the new model will allow exposures from multiple countries to be run simultaneously in the same analysis while accounting for the difference between nations as mentioned in the last section.
This will be a more efficient way of assessing global pandemic risk, as the seasonality functionality will work in concert to produce more realistic geographical patterns in losses for events depending on the time of the year that they commence in the simulations.
In conclusion, the new Infectious Disease Model from Moody’s RMS LifeRisks represents a substantial step forward in pandemic risk modeling.
Learn more about RMS Moody’s LifeRisks here.