Reinsurance and catastrophe bond professionals are seeking new risks to cover, pushing the catastrophe modeling firms to keep pace with the appetite of issuers and investors
In a report issued last week by Guy Carpenter, the firm said that risks not traditionally assumed by the capital markets will become accepted — and even embraced.
“[It] is likely that the market will continue to see more innovative and bespoke structured catastrophe bonds issued,” the report stated, adding that the “structural features that investors may continue to accept on a larger scale include” non-modeled natural perils.
Guy Carpenter cited meteorite impact, wildfire and volcanic eruption as some of the risks that may be included in new bonds.
This is not new territory for the market. Last year, USAA issued its $130 million Residential Re Ltd. series 1 bond which covered traditional modeled risks like earthquakes and tropical cyclones, but also more remote and non-modeled perils such as wildfire, meteorite and volcanic risks.
As issuers seek to offload more remote but expensive perils — and issuers seek higher yielding instruments — non-modeled risk may become more of the norm in the market during the coming year.
Momentum from an active base of a few dozen participants in the collateralized reinsurance market will be the fuel to push catastrophe bonds into non-modeled perils in 2015, says Cory Anger, global head of ILS Structuring at GC Securities, a division of MMC Securities Corp., in New York.
“There are 25 to 30 bond investors in the market currently writing collateralized reinsurance that are simply following the terms of the traditional placement, which includes non-modeled perils and non-modeled sources of loss such as extra-contractual obligations and excess of policy limit obligations,” she says. “This is creating a deepening capacity of investors and issuers, and it’s natural that the bond structures would morph in that direction.”
Just because certain perils cannot be modeled does not automatically disadvantage investors, Anger adds.
“If the model does not perform well or doesn’t exist for such peril, investors are going to switch to traditional underwriting methods which includes evaluating exposure concentrations and the historical loss perfomance,” Anger says. “Yes, a meteor impact is an uncontrollable risk. But you can look at incidence and frequency of meteor strikes around the word and apply it to a geographic region to understand your particular risk.”
“Models are not the end all be all, they are used as a guide,” she adds.
Kent David, vice president of consulting services at EQECAT, Inc., agrees with Anger that non-modeled perils structured within catastrophe bonds will become a more common occurrence in the coming year and that catastrophe modelers will need to adapt.
“There are different ways of quantifying that risk, and essentially what the market demands isn’t a perfect solution, but a rational way of quantifying risk,” David says. “In addition, concurrent with the move towards non-modeled perils are more indemnity transactions, which tend to minimise transaction costs. That gets right to the bottom line of cat modeler.”
David says there are two ways for catastrophe modelers to adapt to the changing landscape as there is “less and less ability to do customised view of risk”:
Develop products that can be used many times across multiple transactions. “This is where the industry is going with U.S. flood,” David says.
Catastrophe modelers learn to support and develop a more statistical approach to the risks being assumed by the market. “Is it worth for the modeler to develop something for a few transactions a year?” he asks.
“Prudent investors strive to understand and assess all of the risks associated with an investment opportunity. Insurance-linked security investors are no different,” says Brent Poliquin, Senior Manager, Insurance-Linked Securities at AIR Worldwide. “As we have seen in recent cat bonds, non-modeled risks, like volcanic eruption and meteorite impact, are included within the covered risks of the cat bond. In cases where catastrophe models may not be available, exposure concentration and hazard analyses are useful tools to quantify risk.”
The next logical step in the evolution of the market is an “all natural perils” bond that — as the moniker states — incorporates all natural hazard types both modeled and non-modeled, Anger added. “This would no longer require that cedents have to specifically name the coverage and bring the coverage terms to a comparable and consistent format to traditional reinsurance cover,” she says. “I would not be surprised if we say see an all natural perils bond in 2015, or more assuredly by 2016.”
While 2015 catastrophe bonds may include more “all natural perils” bonds, how they are structured will be key.
“The concept of an all natural perils has applied to an indemnity bond from day one,” says EQECAT’s David. “But there is always something that was not considered in the model.” He says that is especially true when covering something that is high frequency, low severity (rainfall or crop failure) as part of an aggregate cover.
“At the level that cat bonds tend to be written, along the one percent range, you are still going to be dominated by perils that we do cover,” David adds.”These are different perils that significant, but not consider catastrophic.”
