Structuring risk for the capital markets
Formed last year in the shadow of Deepwater Horizon, CatVest provides energy and chemical risk modelling data for the structure of insurance-linked securities (ILS) and industry loss warranties (ILW). Providing coverage for both offshore and onshore facilities and vessels, CatVest’s expertise, data and methodologies enable the firm to examine and interpret energy and chemical industry-specific losses that can then be transferred into the markets. Timothy Reilly and Stephen Evans, two of the firm’s managing directors, discuss the firm’s approach to modelling and its growing role within the ILS space.
What risks does CatVest monitor and include in its calculations?
Reilly: We analyse a whole suite of oil and gas risks, including physical damage to facilities, business interruption, third party liability and operators’ extra expenses, which includes control and well issues and losses due to pollution damage. We’ve done hundreds of calculations over the years and act as both a calculation agent and a claims investigator for the energy and chemical sectors.
Evans: Damage events can occur due to an explosion or a blowout of a well, or the result of a windstorm or other natural event, or accidents such as a tanker running into a platform. Whatever the risk, we have the capabilities to quantify the potential losses.
Are cumulative risks a particular concern? Do your models account for such aggregation?
Reilly: Yes, we can either parse out the risks of a particular type of event—such as a vessel hitting a rig, which is something thatoccurs around 30 times a year in the North Sea—or we can look at cumulative risks. We have clients that want to understand industry risks for an entire oil and gas operating area such as the North Sea, and we are able to provide them with this information. More commonly, however, we are asked to provide insights into a specific portfolio within a particular operating area.
Evans: The focus of our modelling really depends upon our specific client. If the client is a reinsurer, it could require a full Gulf of Mexico analysis, so we can build an index for it to build the transaction against. If we are working directly with an oil company or a primary insurer, it can be for a very specific risk—a particular type of platform it has holdings on, for example.
How do you establish risk weighting, and has this evolved over time?
Reilly: We essentially monetise the magnitude of risk, weighting it according to the probability of events, thereby enabling us to understand the maximum exposures our clients might face. There are real differences in the types of risks we model and the methodology we use to approach such risks, requiring a bespoke approach to each risk.
"We have clients that want to understand industry risk for an entire oil and gas operating area such as the North Sea, and we are able to provide them with this information."
Our energy risk model, for example, consists of a platform of protocols and methodologies for quantifying international data in order to understand risk exposures. We are able to draw upon an extensive database of industry loss data from around the world, which we have collected over the last 25 years, as well as detailed socio-economic and environmental damage reports that help us build up an accurate picture of risk. Such depth of information enables us to sub-sample the data by geography, period of time and specific risk exposure.
Due to ever-changing conditions, we need constantly to ask whether data are comparable to conditions on the ground right now and update our assumptions, applying only risk data that are relevant and appropriate to the specific risks that we examine.
How does this knowledge help in structuring and pricing of ILS and ILW products?
Evans: At CatVest we have built our models in order to structure and price ILS and ILWs, although potentially the same tools can be used for more traditional reinsurance if the industry wants such products to be index-based. We are able to turn the model data we have into indices of industry losses, which are suitable for ILWs or indemnity-based cat bonds with an industry loss trigger, as well as parametric indices on the pollution side that act as triggers for ILS pollution events.
We don’t actually set ILS and ILW pricing ourselves, rather we help people understand the proportional relationship between risk and cost. Then it’s down to the investment bankers and the brokers who are structuring a specific transaction to come up with the cost that isbearable by the sponsor, who has to pay the interest payments, and investors, who are looking for a decent return. Our role is to make the risk as transparent as possible and to provide a mechanism by which a financial instrument can be triggered and measured against.
We work with specific sponsors and weight our industry loss indices against their specific holdings. There are two ways of doing this: applying full industry loss data, or tailoring the transaction to the specific needs of the client, which provides much greater sensitivity and accuracy to the index. Generally, the information that we are providing is client-specific, which helps to build confidence in the structure, although with the ILWs there is the potential for a broader cross-section of data.
What are the key challenges inherent in putting these kinds of products together?
Reilly: There is always a basis risk—do your models actually match what is happening in the real world? We have thought very hard about this, because if there is significant basis risk then the model isn’t worth very much. Over the last two decades we have validated our processes with actual field observations, energy loss and environmental damage data to ensure our algorithms are appropriate. And we look very carefully at the coverages and guarantees of particular re/insurers to make sure that we have selected data that match the location, time and type of guarantee that we are modelling.
Once we have characterised the risk, we stress-test the losses through a detailed stochastic run to build up a thorough, stress-tested picture that resolves the inherent risks. Through appropriate datamatching, quality assurance review, appropriate algorithms that have been validated in the field and stochastic modelling of thousands of events, we are able to manage that basis risk very well.
How do you differentiate yourself from the competition?
Reilly: Most of the risk modelling in this field is for natural catastrophes. There isn’t a lot of industrial risk modelling and we are quite niche in that respect.
Evans: We are also unique in that we have a focus on the ILS and ILW markets which benefits from significant knowledge of structuring and arranging ILW transactions. In fact, we could structure and bring deals to market on our own, with perhaps a prime broker for sales purposes, which is unique in this sector. I don’t think that other modellers have seen the same opportunity in the ILS space.
At the same time, there is a general lack of re/insurance in the energy sector and a lot of energy companies are under-insured for the amount of damages they could incur. The re/insurance markets simply aren’t big enough to cope with this type of risk, particularly on the pollution side. Obviously there are all sorts of caveats around the pollution liability laws, but governments around the world (such asthe EU and US) are actively looking at pollution liability in the wake of Deepwater Horizon. We expect liability limits to increase, at which point it will become all the more pressing for energy companies to have coverage, and if the re/insurance markets can’t provide capacity, the capital markets really are the most sensible source for that cover.
What represents the ‘Holy Grail’ for risk modellers?
Reilly: Even though we have spent a couple of decades mapping data, and information, and developing loss models and algorithms, there are always new oil and gas fields being developed and new frontiers being opened up. We are constantly working to create everbetter data sets to provide us with inputs to our modelling. I suppose the Holy Grail would be a set of perfectly resolved data for all the oil and gas operating areas of the world.
The other element is approaching and rethinking the algorithms in our modelling suite in order to portray and characterise risk accurately. We are constantly working at ways to tweak our models for specific applications and risks. While our models work for many applications, our clients don’t care about many applications, they care about their applications.
What are your expectations for the ILS space going forward?
Evans: We are confident in the ILS market’s ability to keep growing, but in order to do so it will need to bring to market new risks and locations, and that is one of the things that we are trying to do. And we expect other firms to do much as we have done, bringing other classes of risk that sit outside the existing re/insurance framework into the capital markets.
Timothy Reilly and Steve Evans are both managing directors at CatVest. Timothy can be contacted at: firstname.lastname@example.org, and Steve can be contacted at: email@example.com