Oil Spill Risk – A Potential Market for ILS?

As some of our readers will remember we have written previously about the potential application of insurance-linked securities and catastrophe bond type structures to cover oil spill risks and liabilities. We’ve spoken with some market participants about this idea and will be publishing the feedback we receive. Our first response was from Peter Nakada, Managing Director of RiskMarkets at Risk Management Solutions (RMS) who kindly went to the trouble of writing an entire article for us giving an insight into a risk modellers view on the topic.

Oil Spill Risk – A Potential Market for ILS?

By Peter Nakada, Managing Director of RiskMarkets at RMS

Following the Macondo oil spill, there was a buzz in the insurance-linked securities market about the possibility of catastrophe bonds or ILWs covering oil spill risk. RMS believes the ILS market is well-suited for taking this type of risk, but it will take some time before the market is ready to accept it.  Here’s why.

To work in the ILS market, the oil spill trigger must be parametric
Oil spill liability risk is notoriously long-tailed. The Exxon-Valdez incident occurred in 1989, and the actual losses to Exxon were determined by litigation and appeals that went on for 20 years. Because catastrophe bonds require sponsors to keep collateral tied up until a deal is settled, ILS became a relatively inefficient way for a re/insurer to finance this risk. Furthermore, ILS investors are used to the comparatively quick settlement times for natural catastrophes, so prefer short-tailed risks.

Oil spill liability risk could be securitized, however, using a parametric trigger, such as the volume of oil released in a spill. This could be measured objectively, and settled quickly after an event. The bond would attach at a certain volume of oil spilled, and exhaust at a higher volume. Risk modeling for the transaction would focus on estimating the probability of a spill happening, and the distribution of potential volumes given an oil spill. While an oil liability catastrophe bond sponsor would need to understand the relationship between volume of oil spilled and the economic consequences, this would be done to quantify the basis risk in the transaction, and would not be part of the economics that an investor needs to understand.

Early demand for indemnity is limiting the applicability of ILS to this risk
RMS believes that the ILS market has not expanded into oil spill risk because of a mismatch between the demand for indemnity cover and the practical requirement that ILS be structured with parametric triggers.

For re/insurers, as long as there is capacity in the reinsurance/retro market to take on this risk on an indemnity basis, there will be little demand for parametric ILS structures.  While there is currently no catastrophe model that covers this risk, RMS believes that it could build one if ILS market demand were to appear.

When re/insurers are comfortable with the basis risk, then the market could take off
RMS could construct two types of models for this risk. The first would be a model of the parametric risk to ILS investors. This would involve using the characteristics of a given well and the expected performance of its operators to estimate the probability of a blow-out. The model would then estimate the probability distribution of the amount of oil released given a blow-out (based on water depth and other relevant parameters). This model would be engineered to the standards that RMS applies to other catastrophe risk perils.

Additionally, RMS would develop a capability to help re/insurers estimate the relationship between volume of oil released in a particular location with the insured loss resulting from the spill. This would help in understanding the basis risk between the underlying indemnity risk and the parametric “hedge”.  This part of the model would have many moving parts (e.g., ocean currents, economic activity along the coastline, coverage expansion, politicization of the spill) and would therefore inevitably have more uncertainty than the parametric model.

With suitably robust modeling, investors would likely be eager to invest in parametric oil spill cat bonds. For investors seeking diversification away from U.S. hurricane and earthquake, oil spill cat bonds would be a good diversifier.  Other ‘off-peak’ perils have enjoyed strong demand from the investor community in the past. The modeling, while new, would be relatively simple compared to the complexity of a hurricane model. One of the largest areas of uncertainty that stakeholders would need to become comfortable with is modeling the human behavior associated with the probability of an operator error causing a blow-out.

Conclusion
RMS believes that it is only a matter of time before re/insurers become comfortable with taking the basis risk between oil spill liability risk and a parametric oil spill hedge.  When this occurs, and demand for a model emerges, RMS is able to respond with a parametric oil spill catastrophe model that will allow for the development of a healthy market in this new peril.

www.rms.com

End.

Our thanks go to Peter for the time and effort he has put into the article above!

We will be publishing comments from market participants on this topic in the coming days. Please do get in touch if you’d like to contribute your thoughts.