“According to our life insurance stress test, Swiss Re's expected pre‑tax mortality claims for a major lethal pandemic loss event based on its proprietary pandemic model are CHF 3.5 billion (nearly 2,3 billion euros),...”
The Influenza A (H1N1) outbreak in Mexico and other parts of the world is clearly cause for concern, not only for those communities already directly affected.
One of Swiss Re’s leading pandemic experts, Bob Howe, answers some key questions.
What will be on (re)insurers’ minds at the moment?
Bob Howe: First, we must not forget that, for many families and friends of those affected, this is a human story with some tragic consequences.
But clearly questions are being asked already about possible business impacts. (Re) insurers with significant portfolios of mortality risk will be concerned about a major mortality event.
In talking about pandemic scenarios, comparisons are often made with the pandemic of 1918.
But levels of additional mortality seen in the 1918 influenza pandemic are highly unlikely to be repeated today because of the availability of antibiotics and antivirals. Mortality resulting from the other two pandemics in recent history —1957 and 1968— was considerably more modest.
The current outbreaks are being monitored round the clock by the World Health Organization (WHO). In readiness for a global pandemic —meaning widespread human-to-human infection— governments are typically well prepared and have detailed contingency plans and sizeable stockpiles of antiviral drugs.
Can anyone tell yet how bad this is going to be? If not, why not?
Bob Howe: Not yet, no. In recent years Swiss Re has accumulated a great deal of knowledge on the issue of pandemic flu. Given the current events in Mexico and elsewhere, people are asking why, currently, there are no concrete answers to this question.
With any influenza virus, four key characteristics are critical in determining what the final outcome will be:
- the ability of the virus to spread (known in technical terms as the R0 value)
- its lethality (how many people does it kill as a proportion of those infected)
- the age profile of the mortality, and
- the responsiveness of the clinical disease to antiviral treatment
The information available at the moment gives very little indication of the R0. In Mexico we have some numbers of deaths, and some numbers reflecting likely or confirmed clinical cases ("clinical” simply meaning those cases observed by the medical profession). However, we have no idea about how many people may be infected but not showing any symptoms. People who are experiencing symptoms may be only a tiny fraction of the total infected, or they may be a large proportion. The rate of increase of the total number infected allows us to estimate the spread capability, and as more data become available it will be possible to estimate the spread capability. But currently this is not possible.
The lethality of the virus, taken together with the number of people infected, determines how many people are likely to die. We have some indication of numbers of deaths in Mexico, but because we don’t know the number of total infections we cannot determine the lethality. If, for example, the deaths so far are the result of a few million people being infected, then the death rate would be low in comparison to past influenza pandemics.
If the deaths have resulted from only a few thousand infections, then the lethality is quite high. It is not currently possible to establish lethality.
As regards the age profile of people dying, reports seem to indicate that most deaths in Mexico are young adults, but these reports remain somewhat speculative. Even if this were the case, the question remains as to whether the higher number of deaths among young adults is higher due to higher lethality at these ages, or whether there are far higher numbers infected in these age groups.
Finally, reports from a number of sources indicate that two of the main types of antiviral drugs (Tamiflu and Relenza) are highly effective in reducing the severity of clinical illness for this virus. This would significantly reduce mortality in countries that have stocks.
When we are able to gain more insight to the above questions, a clearer picture will emerge as to how the current disease is likely to spread and affect mortality. Before then, it is impossible to estimate the financial impact.
Is Swiss Re ready for a global pandemic?
Bob Howe: Yes. Swiss Re constantly and closely monitors the development of potential epidemics and pandemics, and is closely following the current threat to human health posed by Influenza A(H1N1).
Swiss Re is well prepared to deal with the insurance, financial and business continuity risks that could arise from a pandemic, and to continue writing new business afterwards. Our capital base is strong; our portfolio is diversified; we have business continuity plans in place.
What has Swiss Re done in practice to assess its capital requirements should a major pandemic occur, now or at any time?
Bob Howe: While we cannot comment further at this stage on the possible impact of the current Influenza A(H1N1) outbreaks, Swiss Re has developed a sophisticated epidemiological model to improve understanding of the potential range of outcomes from a pandemic, better enabling us to hold appropriate capital for the risk.
Using the model we have estimated that, in most developed countries, a 1-in-200-year severity pandemic would give rise to excess mortality of between 1 and 1.5 deaths per 1,000 lives within an insurance portfolio.
The model also shows that the influenza pandemic of 1918 would have a much lower impact on mortality today than it did in 1918. A key factor that made 1918 so severe compared with today was that no antibiotics, vaccines or antivirals were available.
According to our life insurance stress test, Swiss Re’s expected pre-tax mortality claims for a major lethal pandemic loss (1-in-200-year severity) event based on its proprietary pandemic model are CHF 3.5 billion (nearly 2,3 billion euros) as of 31 December 2008. The scenario assumes that excess mortality will vary with age, but is conservative in that it does not allow for the typically lower mortality experienced among insured populations.
This information is disclosed in our Annual Report 2008, but —I must make very clear— is not a forecast of the expected losses from the current outbreak.
