More thoughts on catastrophe bonds and insurance-linked securities for oil spill risks

We received so much feedback on an article we wrote a year ago on the potential for catastrophe bonds and insurance-linked securities to provide a source of reinsurance for oil spills that we decided to follow up with some insight from market participants on the subject. Last week we published a great article written by Peter Nakada of RMS on the subject. Today we’re publishing some commentary and insight from an insurance broker, an asset management firm in the ILS sector and a ratings agency perspective on the topic.

William Dubinsky, Willis Capital Markets & Advisory

First we spoke with William Dubinksy, Managing Director, Willis Capital Markets & Advisory (WCMA). WCMA is the division of global broker Willis Group which focuses on advising insurance and reinsurance companies on mergers & acquisitions, as well as a broad array of capital markets products including insurance-linked securities. His responses below:

1. Do you think it is feasible for cat bonds or/and ILW type structures to cover oil spill risk?

It is possible but not necessarily likely.

2. If you think it is feasible please explain why. If you don’t please explain why.

On the one hand investors are eager to embrace diversifying risks but at the same time they are wary of the unknown and taking risks where they are “on their own”. We believe that alignment of interest is key. Structures, cat bond or otherwise, that provide alignment of interest for a new peril have a higher likelihood of succeeding than those that do not. Alignment could mean co-participation by the sponsor (or more importantly by the party controlling the risk, if that party is different from the sponsor). Alternatively, an acceptable parametric trigger could substitute for alignment of interest.

3. What do you feel is hindering the expansion of ILS into the oil spill area? Is lack of a model a problem?

Looking broadly, this may be part of the problem; however, price and alignment of interest are key. Obviously, changes on the demand side (e.g., a dramatic increase in mandatory limits) could make an ILS solution more competitive. Taking a narrower view, the lack of a model would likely prevent a rating, which suggests a high risk, high return profile would make sense.

4. What type of triggers do you feel would work for an oil spill cat bond or ILW? How do you think a deal could be structured?

There are many possible triggers. What makes sense for sponsors and investors will vary based on the sponsor and coverage they seek. Parametric triggers are usually but not always more palatable for investors. Unfortunately, in some situations, they increase rather than decrease certainty if the measurement method is not crystal clear. The parametric trigger also has to match the modeling.

5. How do you think the market would react to a product like this?

Investors should welcome access to a new risk. This helps but doesn’t mean the deal will succeed.

6. Do you think investors would be keen to take on these types of risks? What would have to happen or be in place to make them feel it is a risk worth taking on?

Without alignment of interests and a keen understanding of the sponsor’s motivation, investors are likely to be skeptical even if a proposed model on the surface seems reasonable. If there is a clear market need, some form of capital markets participation makes sense . . . whether that translates to a cat bond is a different question.

7. Any other comments/thoughts/ideas on this subject.

Growing the ILS market requires market participants to think big thoughts. Oil spill coverage certainly merits a look as one potential area to expand.

Our thanks to Bill for his insight on this topic. You can read our previous interview with him here. So while Mr Dubinksy doesn’t necessarily feel it is likely to happen he certainly feels it is possible that oil spill risks could be issued in catastrophe bond or insurance-linked securities form.

Samuel Scherling, Alternative Beta Partners AG

Next, we asked Samuel Scherling, of Alternative Beta Partners AG, a quantitative multi-strategy asset management company with a focus on insurance-linked securities and other alternative investments, for his opinion on this subject:

I definitely believe that a cat bond (or another ILS structure) would be possible for oil pollution related risks. In principle it has an ideal risk profile with its low frequency, high severity nature and good diversification properties. However, it will not be easy for a number of reasons, including:

1. Oil pollution is not a peak risk, because the legal requirement to buy insurance is very low. This problem is shared by many other non-nat cat risks including nuclear, accordingly the incentive to purchase high insurance limits is minimal and large losses tend to be financed ex-post by shareholders, bondholders and taxpayers (see BP, Tepco)
2. The above also means that for individual counterparties the premium may be too low and only pooling solutions might be effective (creating additional challenges, bringing the different parties together)
3. Lack of model: I’m convinced that it would be feasible to put such a model together with reasonably low cost, however, it would require a pilot protection buyer to bear some R&D costs

Some structural thoughts:

1. Trigger:
1. Industry loss will not work because generally only one insured is involved
2. Indemnity will not work because of long tail (loss development time)
3. Modelled loss / parametric index would work if insured accepts basis risk
4. Variables could include location of spill (zones), distance to shore, quantity of barrels spilled etc
2. Coverage
1. Ideally excluding nat cat perils (especially US hurricane) in order to avoid accumulation problems
2. Loss types: drilling, production, pipeline rupture, shipping accidents (most both onshore and offshore)

Thanks to Samuel for his interesting points on the subject. He highlights some areas which would need careful consideration by anyone planning such a transaction in the future but which are certainly not insurmountable.

Gary Martucci, Standard & Poor’s

Finally we felt that the opinion of a ratings agency would be useful for this series, and who better to ask than Standard & Poor’s. We invited Gary Martucci, Director, Financial Institutions Ratings at S&P to comment on this subject and he sent us the below:

While we think it may be feasible for bankers to structure a traditional cat bond or ILW to cover the risk of losses from an oil spill, it would be difficult for Standard & Poor’s to rate such a transaction under our current criteria, primarily due to the likely challenge in determining the probability of a covered event occurring and reaching the attachment point, or to put it another way, resulting in a loss of principal to investors.

The first challenge would be for the modeling agent to create an event set. In the case of hurricanes and earthquakes and other natural perils such as windstorms and severe thunderstorms, a modeling firm can draw from a body of past events. In contrast, since major oil spills occur relatively rarely, there is uncertainty as to what types of event could occur, and how often they could occur. Secondly, to develop their natural catastrophe models, modeling firms can draw on source data such NOAA weather forecasts and USGS studies, and we are not aware of a similarly robust source of data for oil spills. Third and by no means least, there is the issue of human agency. Individuals’ actions cannot generate the specific events that are covered by natural peril cat bonds. However, an oil spill clearly can be caused by actions or inactions of an individual or individuals. We believe that it would be extremely challenging for this human agency risk to be appropriately captured in a model. This is one of the reasons Standard & Poor’s will not rate bonds covering the risk of losses resulting from terrorist activities.

Other issues that we believe could prove challenging are what losses are covered and the time it takes to determine a loss. While in standard cat bonds, environmental claims are excluded, they pose a huge risk in oil spills. In addition, it could take several years for the claims to be realized, which could be problematic if the tenor is fairly short as is the case for cat bonds.

Furthermore, we believe there is also a risk that, given the intense scrutiny the responsible party is generally under, claims which otherwise could have been challenged might be paid in an effort to assuage the public, the government and the media. Then there is simply the issue of calculating total losses. Historically, most cat bonds have been issued with industry loss or parametric triggers which would likely not be an option here, as there is no independent third party such as NOAA, USGS, PCS or PERIL’s to provide event parameters or industry loss amounts. The remaining option, an indemnity trigger, would require substantial disclosure from the cedent given the risk and its related complexity.

Thanks to Gary for his opinion as well. It is no surprise that S&P would be unlikely to rate a transaction right now. However given the volume of cat bond transactions which remain unrated for various reasons, it would be unlikely to prevent a transaction going ahead. The current trend for private cat bond deals may actually provide a better model for oil spill and oil pollution risk transfer anyway.

—

We hope our readers have enjoyed our commentary on the potential for a new risk to be issued through insurance-linked securities or catastrophe bonds. We’d like to thank all of our contributors to these articles. We’d really like to hear your opinions. Do you think we could see an oil spill catastrophe bond anytime soon? Please feel free to comment below.

And if you have an interesting perspective on these markets or an idea which you think would interest the audience of Artemis, please get in touch and we can discuss publishing your thoughts.