SCOR Launches Multi-Continental CAT Bond with RMS Technical Help

On Jan. 3, French reinsurer SCOR Group launched the second installment of its innovative Atlas Re catastrophe bond, with technical analysis and parameters provided by Newark, Calif.-based Risk Management Solutions (RMS) and placement by Goldman Sachs. A portion of the notes carry “A” ratings from Standard & Poor’s and Moody’s—the first time a securitization has been so highly rated.

SCOR’s Atlas Reinsurance II CAT Bond, like its predecessor Atlas Re I, covers perils across three continents—earthquake risk in Japan and the U.S., and severe windstorms, of 1999’s Lothar and Martin variety, in Europe. “These bonds cover the possibility of a second and third occurrence,” said Delphine Deleval, SCOR spokesperson for external communications in Paris.

Atlas I provided up to $200 million in reinsurance cover for natural catastrophes in any of the three areas. The securities went on sale in March 2000 and are due to expire in March 2003. In announcing Atlas I, the company indicated that they were issued on an “indemnity basis,” and were “linked only to the actual loss suffered by SCOR and without any reference to any external index.”

Atlas II covers the same territory over a three-year period until 2005. It is split into two tranches: Class A notes with a $50 million limit, and Class B notes with a $100 million limit. However, unlike its predecessor, the securities are linked to specific parameters rather than to any actual catastrophe losses SCOR may suffer. EQECAT provided the risk analysis and mathematical models for Atlas I, while RMS has furnished the technical data relevant to Atlas II.

Deleval indicated that, as the notes were designed to “make sure that the funds are available to cover losses above those covered [by Atlas I],” they are linked to outside parameters which are “risk specific,” and are mathematically modeled on the use of extensive external data, relative to the perils covered.

As an example she indicated that losses from a California quake measuring 4.0 or higher on the Richter scale corresponds to the parameters, whereas a quake of lesser magnitude, even if it caused losses, would not.

Fathia Grandjean, an RMS principle responsible for client development in the Paris office, confirmed that “as the parameters themselves are easily measured, they help meet the demands of investors for more transparency.” She explained that while earthquake parameters have been used in some securitizations, the European windstorm index based wind speeds, employed by ATLAS II, were first used in securitization in the PRIME for Munich RE, and represent a new innovation and a new approach in calculating the trigger point for European windstorms.”

By using parametric criteria, rather than basing the risk strictly on loss exposure, bond payments are effectively controlled by predetermined external measurements, which provide the sought after transparency. RMS pointed out that “[a] distinguishing feature of the trigger structure is that it offers advance warning to investors prior to payout. The first trigger event in a specific year causes no losses to investors in Atlas II], providing both Class A and Class B note holders one advance warning. Similarly, a $100 million per-event limit ensures that Class A note holders receive a second warning prior to their tranche being impacted.”

The all-important measurements “are directly linked to parameters published by government geological and meteorological agencies,” said RMS. “In Europe we use the meteorological agencies of the countries covered,” Grandjean added, “and Japan’s Meteorological Service [for that country}. In California the data is provided by the California Earthquake Reporting Agency.”

There’s more involved than just measurements, however. RMS worked directly with SCOR “to design parametric triggers for the three geographic regions covered by the transaction.” It explained that, “For the San Francisco Bay Area and greater Los Angeles, the triggers are based on reported earthquake magnitude within twelve geographical ‘boxes’ surrounding sources of major tectonic activity.

“The European windstorm trigger is an index calculated from windspeed at 600 measuring stations across five countries in Western Europe. For these perils, the triggers are similar to those constructed last year by RMS for PRIME, CalQuake & Eurowind Ltd., but have been specifically adapted for SCOR’s exposure. For Japan, a new parametric box structure has been produced which fits with SCOR’s exposure by covering both CRESTA zones 5 and 6.”

If the reaction of the rating agencies is any guide, the additional analysis seems to have paid off in making the bonds more attractive to investors. The Class A bonds are rated “A” by Standard and Poor’s and “A3” by Moody’s Investors Service, “making this the first securitization ever to be accorded an ‘A’ rating.” The Class B notes were rated “BB+” by Standard & Poor’s and “Ba2” by Moody’s.

James Doona, a director of Standard & Poor’s for Insurance Capital Markets, explained the reasons for the upgrade. “Our longstanding ceiling on CAT bonds has been fixed at ‘BBB+’,” Doona said, “but we made a distinction [on the ‘A’ tranche of Atlas II] because of its special characteristics.” He emphasized that CAT bonds are a distinctive security and cannot be rated as other corporate instruments are because of the “credit cliff.”

“They’re all or nothing,” said Doona. “Ninety-nine percent of the time nothing happens, but you can go to bed with a ‘BBB’-rated bond and wake up with nothing. The value of the investment can’t be manipulated by human hands, so in that sense it’s easier to rate, but because of the credit cliff it’s an alien instrument.”

The structure of Atlas Re II Class A securities addresses the rating agencies’ problems with CAT bonds. “This series gives [the investor] a warning,” Doona said. After two loss events, the holders of Atlas II tranche A notes are in a position to evaluate their risks concerning the probability of a third one happening, and have an opportunity to decide whether to sell the security or not. Doona said, “Our Senior Policy Board appraised it and found it a perfectly reasonable approach” and went on to approve the “A” rating.

Doona praised the use of parametric models as a “more grown up basis to assess risk,” and indicated that it was “qualitatively different” from other types of CAT bonds “because the credit cliff has been mitigated.” It’s quite possible there may be more of these instruments on the horizon for the industry in the future.