Lessons for L&H insurers after COVID-19 tested our pandemic model
As news of a new virus spreading in China emerged in January 2020, my team and I found ourselves asking: "How bad will this get?" The Life & Health insurance industry has long considered the potential impact of pandemics with worst-case scenarios for L&H insurers, typically calibrated to the devastating H1N1 influenza pandemic of 1918-19. But what precisely is the insurance industry supposed to do when an actual pandemic strikes? As we mark the one-year anniversary today of the World Health Organization declaring COVID-19 a pandemic, both from a personal and a business perspective, the last year has been the mother of all crash courses in coping with change and uncertainty, providing some fascinating – and surprising – insights into the nature of death, statistics and human behaviour.
Change and uncertainty are the only constants
The L&H insurance industry is not exactly known for its agility or dynamism. The pandemic therefore posed considerable challenges because of the pace at which new – and usually incomplete– information emerged which needed to be assimilated into processes. In a matter of weeks, COVID-19 went from a minor concern with infected travelers to a multifaceted global threat requiring a more holistic response. The pandemic not only increased the claims burden for L&H insurers, but it also triggered further reductions in interest rates that were already at record lows. In addition, COVID-19 caused significant disruption to the industry's ability to generate new sales. The pandemic prevented traditional face-to-face meetings and interfered with laboratory testing and the collection of medical evidence that forms the foundation of L&H underwriting practices in most countries.
From my perspective, it took a while for L&H insurers to take the pandemic seriously, with some believing that the virus would disappear quickly. In the quest for a short-term sales boost, some health insurers waived waiting periods for coronavirus claims and were probably spared additional claims because of a slowdown in face-to-face sales and travel restrictions. But once the seriousness of the pandemic had been established, L&H insurers reacted in two disparate ways. On the one hand, they generally became more cautious in their acceptance of new policies. A few introduced COVID-19 underwriting questions as a safeguard, and many reduced their maximum sums assured.
On the other hand, some insurers relaxed their underwriting requirements. To overcome the logistical challenge of collecting medical evidence in the middle of a pandemic, insurers used tele-underwriting more widely and relied more on the honesty of customer disclosures instead of lab tests and medical exams. While this sustained new business activity through the pandemic, it increased the risks posed by non-disclosure and anti-selection, the full outcome of which is still unknown.
I feel the biggest change we've seen in L&H underwriting is an increase in our sense of uncertainty. We used to feel confident in our ability to assess the risk of insuring millions of dollars on a single life, but that confidence has been dented by a year during which we couldn't even be certain that it was safe to head down to the local shops.
Surprising excess mortality insights
While there has been much public debate about how COVID-19 deaths are counted, for insurance company actuaries there is little discussion. What really matters to us in this pandemic is excess deaths – the increase in overall mortality rates during the pandemic period compared to pre-pandemic periods (adjusted for population and demographic changes). As life insurance policies pay upon death for any reason – not just those attributed to COVID-19 – excess mortality represents a good starting point for estimating how much more insurers will need to pay out in claims because of the pandemic.
One of the most notable features of the pandemic is its varied mortality impact on different countries:
At the one extreme, there are countries that managed to keep the pandemic well under control, such as Australia and New Zealand. Helped by isolation, tight borders which were closed quickly and populations relatively tolerant of temporary lockdown measures, excess deaths in 2020 were low or even negative – meaning there were fewer deaths during the pandemic than in pre-pandemic periods. This is due in part to the low number of COVID-19 deaths combined with fewer deaths from other infectious diseases and accidental deaths that came as a result of social distancing measures.
At the other extreme are countries like the US and the UK which are more globally connected and were slower to restrict freedom of movement. Their excess deaths since the start of the pandemic in March 2020 are 20-25% above mortality levels in pre-pandemic times. To put this in context, mortality rates have not increased by even 10% in a single year in the UK since its previous influenza pandemic in 1951, and not in the US since the Spanish flu of 1918-19!
For a pandemic so closely associated with mortality among the elderly, what I found surprising is that, so far, the percentage of excess deaths among older people is actually quite similar to that of young people. The charts below show the number of registered deaths in England since the start of the pandemic compared to the number of deaths that would have been expected based on historical averages:
If we compare across age bands, the mortality increase in ages 45-64 (x1.31 for males and x1.24 for females) was higher than in ages 85+ (1.23 for males and x1.20 for females). An examination of excess deaths in the US shows a similar pattern. While excess deaths among the elderly were extremely high and grabbed the headlines when deaths in nursing homes dominated early in the pandemic, as the pandemic progressed, it seems to me that social distancing measures led to lower infection rates among the elderly, but younger people were either less able to practice social distancing due to work commitments or not as careful.
In the end, however, what really matters for L&H insurers is the excess mortality observed in our own portfolios. While anecdotal evidence from most insurers suggests that COVID-19 increased mortality of insured lives by a smaller percentage than what we see in the general population, not much has been reported on this topic to date. One of the few published sources is the Society of Actuaries Group Life COVID-19 Mortality Survey from December 2020 which estimated that for every 10% increase in mortality in the general population, the mortality observed in Group Life portfolios increased by 5-7%. The corresponding Society of Actuaries Individual Life COVID-19 Mortality Claims Analysis suggests that in Q2 2020, mortality among insured lives increased by 10-13% whereas US population mortality increased by 18-23%.
There is still much to learn from COVID-19
Despite the many tough lessons that COVID-19 has taught us, it seems that there is still much more to learn.
On the underwriting side, we now have a sizeable segment of the population in many developed markets that has been infected with the virus, a small minority of whom show long-term persistent effects (long-COVID), including some cases with severe damage across multiple organs and bodily systems. At the same time, we are still coming to grips with the long-term consequences of overstretched healthcare systems with large segments of the population receiving suboptimal treatment or missing out on preventative care such as cancer screenings. We also have yet to find out whether there will be longer term mental health consequences from the multiple periods of lockdown and social isolation.
On the pricing side, while we now have much more clarity on infection and mortality rates than at the beginning of the pandemic, we are finding that the knowledge we acquired in the past year may already be out-of-date. Our mortality estimates may need to adjust for new, more infectious and possibly more deadly variants of the virus as well as allowing for the effects of vaccination, where there are multiple vaccines with different levels of effectiveness, but also significant parts of some populations with a strong reluctance to get vaccinated.
As I watched the pandemic unfold over the past year, there were many episodes when my faith in humanity's instinct for survival and self-preservation was shaken. Comparing approaches to managing the pandemic and mortality outcomes across different countries, it became clear to me that many of the excess deaths could have been prevented with better infection control measures. In contrast, during probably the worst year for global mortality since 1919, I have generally been impressed with the ability of the L&H insurance industry to absorb new information about COVID-19 in stride, making sensible adaptations to its processes, and remaining financially resilient throughout. Looking forward, I would like to see the L&H insurance industry play a larger and more pro-active role in sharing its understanding and expertise of pandemics – not only to accelerate our path out of this pandemic but also to make the world more resilient against the inevitable next one.