Perspectives – Dr Andrew Coburn, Cambridge University, Centre for Risk Studies

OLYMPUS DIGITAL CAMERAILS funds and reinsurers are struggling to identify opportunities to deploy the current influx of capital but researchers at Cambridge University are having no difficulty in identifying plausible events that could cause trillions of dollars of economic loss. Dr Coburn explains what the insurance industry might be missing.

Dr Andrew Coburn is director of the External Advisory Board at the Centre for Risk Studies at the Cambridge Judge Business School and senior vice president at RMS. This week, the Centre launched its much anticipated Cambridge Stress Test Series with the results of four pieces of research that quantify the economic impact of sudden global shocks. Each scenario is designed to be a plausible ‘1 in 100’ year event.

1) Pandemic: São Paulo Influenza Virus – $7 trillion to $23 trillion
2) War: China-Japan Conflict – $17 trillion to $32 trillion
3) Cyber: Sybil Logic Bomb – $4.5 trillion to $15 trillion
4) Social Unrest Stress: Millennial Uprising Scenario – $1.6 trillion to $8.6 trillion

A common theme of the research is that globalisation has meant that events that might have been limited to a single region can now quickly become worldwide crises. This has created new demand from businesses that are looking for ways to manage these emerging and disruptive risks.

InsuranceLinked asked Dr Coburn how the capital markets could help to mitigate the effects of these kinds of events.

Andrew Coburn: Some of the solutions might be more capital markets-related and some of the products ILS rather than insurance. It's a challenge for the insurers to make their capital available for perils that are less well understood, and you could imagine the capital markets might be more adventurous than insurers.

We think the interconnectivity of risk is the big story here. The world is so interconnected now and my personal view is we need to understand the connectivity and insurers themselves need to understand the risk of interconnectivity and figure out how to transfer that risk. Corporates own this risk - they have it saddled onto their business - and global entities know they have their footprint all over the world. They are suddenly realising how dangerous the world is.

At the moment insurance products aren't very well geared to connectivity. So if a components supplier is damaged in Japan and the car plant in Germany can't make its products anymore you can get insurance for this by buying contingent business interruption cover, but insurance capacity is very limited, tightly constrained by terms and conditions, and pretty expensive. Trying to understand how you manage and track this exposure will lead you ultimately to being able to make it more insurable.

Is it significant this research has come out of the academic world rather than the insurance industry?

We've got more room to be creative and innovate in the academic world and to think a little bit more laterally. I definitely think the next generation of catastrophe models is going to have to incorporate interconnectivity and to reflect this new world we live in.

Can you compare these four stress tests to the realistic disasters scenarios (RDSs) used by insurers?

We're working with the Lloyd's insurance market and so it's very similar to the RDS concept. With this study we were looking at emerging risks and things that were less well insured as perils. What's interesting is that each of them, as we've explored and dug down, have become more insurable.

More cyber risk is now being insured in the market, they've doubled their premium over the past two years, and things like pandemic products are now being offered. That was aspect trend even before Ebola. War has always been excluded from most conventional coverages but insurers need to know what other coverages could see losses if there were to be a big war somewhere, as is the case with social unrest, and we were helpful with that I think.

What proportion of losses would be insured if these fictional scenarios were to occur?

That was not something we were able to answer in the first analysis. We don't have a good model of insurance exposure, particularly on a worldwide basis. So when for example we were looking at the hypothetical China and Japan war scenario, this envisaged the destruction of commercial buildings in Shanghai, but we have no idea how many of them might be insured. What we have said to the insurance industry is that if people would like an insurance loss estimate, we could add that by essentially building a model of how much insurance exposure there is in different lines around the world.

ILS investors like their perils to be well modelled. So is this the next step?

These scenarios we have built are deterministic scenarios. For capital markets transactions you absolutely have to be comfortable about probability and obviously with these emerging risks it's more complicated and uncertain to assign return periods. If you take a cyber catastrophe, cyber risk has only been around for a decade or so, so to try to assess a 100-year return period loss is quite tricky. Its not impossible but it needs careful assessment and it needs to convince investors and business managers.

Which of the four perils considered in these scenarios best lends itself to ILS products?

Pandemic is the closest. RMS has already worked with Swiss Re on the Vita transactions, which were a set of pandemic bonds, and I was personally involved in that. Pandemic is much closer to capital market solutions. For each of these threats you could potentially design parametric triggers for ILS solutions, if there was appetite for it.

That might well be the way that business interruption protection might go in the future. You could potentially have parametric measures that would trigger a payout. At present traditional business interruption insurance requires physical damage to property for you to make a claim. It is difficult for insurers to construct policy wordings and coverage around lost business revenue from a non-damaging event, like pandemic or social unrest, and distinguish it from somebody just having a bad trading period. Capital markets might be able to structure triggers that mirror the losses suffered by companies, but that aren’t specific indemnity covers.

Will research like this act as an impetus for further innovation?

That is our hope and belief. This research is being driven by insurers and they themselves are very conscious of the need to innovate and find new markets and products. Obviously we're at the bottom of the pricing cycle in the insurance world and they are looking for new things to generate revenues and new ways to use their capital.

Posted: Monday, December 1st, 2014