RMS suggest a multi-vendor approach for modelling catastrophe bond risks

Yesterday we wrote about the ratings impact of a new hurricane risk model on a number of cat bond transactions. Ratings agency Standard & Poor’s placed 16 tranches of catastrophe bond deals on CreditWatch negative because they believe that the changes to Risk Management Solutions (RMS) U.S. hurricane risk model are sufficient to cause concern about the changed risk profile it could give to these cat bonds.

As a result S&P have requested that RMS run their new model against all potentially affected deals and report back so that they can assess whether the risk of default has changed enough to warrant a ratings downgrade. RMS’ updated hurricane model almost doubles the 1 in 100 year estimate for insured hurricane losses in Texas and estimates for losses in eastern coastal States has risen in some cases by as much as 75%. Industry wide the 1 in 100 year insured loss estimate for hurricane damages has risen by between 15%-25% using the new model. So changes that significant certainly warranted some due diligence by the rating agency so a proper risk ratings can be applied to the cat bonds in question.

RMS have responded to criticism of the model change and explained that it uses the latest scientific and mathematical data, techniques and software and as a result the risk of hurricane damage has risen (when assessed through the model). They explained that there was little point in making an incremental increase of the risk or on holding back on advancement of their model. They see the latest version as the most sophisticated and accurate available, something that time will only tell.

Interestingly, in a Dow Jones newswire article yesterday Peter Nakada, Director of RiskMarkets at RMS made some interesting comments about the use of risk models in the catastrophe bond market. He commented that he had no issue with S&P’s conclusion that the bonds needed to be assessed using the updated risk model. He went on to comment that the action by S&P to review the ratings of these deals shows that catastrophe bonds should be evaluated using more than one risk model vendor company (in a similar way to how insurers evaluate their own portfolio of risk using multiple models, often including internal models).

Peter Nakada commented in the Dow Jones article; “One could argue that a more holistic approach to a view of risk might be appropriate,” Mr. Nakada said. “That’s been tossed around, and it’s something we would support. The question is cost: you’d have to pay each of the modelers, which increases the cost of the bond.”

Using multiple vendors for risk modelling of catastrophe bond deals could be beneficial to the market over the long term. It would help risk and credit ratings be more reliable across deals from different issuers. It could help investors be more certain of their investment decisions and the risks they face across their whole portfolio of cat bonds. Cost though, as Peter said, is certainly a factor. It’s a barrier to entry to the market to some and adding additional cost to cat bond transaction issuance could hold back the markets development. We’re interested to hear your thoughts on the use of multiple risk models for assessing catastrophe bonds, let us know your opinion below.

You can read our recent interview with Peter Nakada of RMS here.