As the market for insurance-linked securities takes off, catastrophe modelling agencies are striving to make their products simpler for investors to understand and easier for issuers to apply. David Banks reports.
The growth of insurance-linked securities since 2005 has stunned many in the industry, but forecasts from catastrophe modellers about a multiple increase in the volume of such investments make even more compelling reading. Much of their optimism derives from recent strides in the field of parametric triggers – such as the measure of wind speeds and ground shaking during earthquakes. Now, it seems, a cat bond issuer can work with much greater certainty when designing a contract that triggers when wind speeds reach, for example, 90mph.
Not only are models more advanced, incorporating more of the latest data on insured values and potential losses, but more of the essential measurement devices are springing up where before they did not exist. “With the new parametric technology and models being created, cat bonds are becoming more tailored to the issuer’s needs and more simple for the investor,” says Peter Nakada, managing director and senior vice president of RMS Consulting. “The beauty is that they don’t need insurance knowledge. We have put in place indices of parametric triggers versus insured loss which are there for other people to refer to when they are discussing their cat bonds.”
The major cat modellers are all in the process of creating a market for parametric risk transfer vehicles. By bringing more clarity to transactions in cat bonds, sidecars, industry loss warranties and weather derivatives, they are potentially growing the market. “We ought to be looking at a market that is ten times the size that it is now within the next five to seven years,” Nakada adds. “Are we seeing that kind of growth? Well for a while we saw cat bond volumes of $2bn then $5bn then $7bn. There has been year-on-year growth but we think this is the micro movement before the macro movement.”
Nakada says the industry will approach a tipping point when parametric triggers make ILS deals so straightforward that mainstream investors can readily place up to 2% of their portfolio on any given peril. “Essentially, this will mean that a handful of reinsurers will become transformers who offer regular indemnity-based reinsurance on one hand, while at the same time hedging risk on a parametric basis.” Estimates place the volume of parametric-based cat bonds at between 10% and 15% of the cat bond market in mid-2008. But that figure could grow due to improvements in loss modelling and data collection in meteorology and seismology?
“The technology behind parametric triggers has improved tremendously in the last 10-15 years, driven by the advances in electronics and IT, because now you can get online remote access to this information,” says Dr Milan Simic, managing director of AIR Worldwide.
Other factors, he says, are better quality instruments, growing professionalism and wider standardisation of practices among technicians at various stations.
However, Dr Simic stressed that despite the promise held by ever-improving measurement and loss modelling, there was no such thing as a one-size-fits-all approach to cat bond issuance. “Having a parametric index helps enormously, but each cat bond still needs to be discussed on its merits. For example, it requires a lot of work to find out which physical parametric will fit the bill – the key is that there is always a dialogue between the issuer, the main placing bank and the investing parties.”
Building for the future
A major challenge to establishing the accuracy required for parametric triggers has been ensuring that the physical infrastructure is in place to measure wind speeds and earthquakes. Indeed, in many parts of the US, there were no wind stations designed to survive hurricane force winds. For RMS, the answer has been a network of 10-metre-high concrete pillars designed for the worst effects that the elements can throw at them and built in partnership with engineering companies. These are going up across the US, particularly Florida, and eventually there will be a total of 100 in place before the hurricane season of 2009.
Other companies in the US weather-based ILS market use the data paid for by the American taxpayer. “When you consider we are using US federal data, collected using aircraft, satellites, buoys and all the things that the federal government has at its disposal, I would say those sources are very robust,” says Dr Steve Smith, president of property solutions at ReAdvisory – the research arm of Carvill – which provides a hurricane index for the Chicago Mercantile Exchange.
Smith adds a word of caution on using parametric triggers based solely on wind. “We think that if there are two or more parametric measures coming in, the measurement and corresponding loss can be calculated more accurately,” he says. Insurance that is purchased on a wind speed basis has the added incentive for the policyholder that they still pay out regardless of the level of damage. Such deals are sometimes structured in linear format, between the attachment and exhaustion points. For example, it might attach at 90mph and exhaust at 110mph, paying out $1m per increasing mph. However, a weather index of wind speed versus loss provides more clarity based on historical data.
Level playing field
While weather indices exist aplenty in the US – a market prone to natural hazards – they are less developed in Europe.
However, AIR says its progress in research on hurricane landfall in the US – a major determinant of loss – has been mirrored by similar progress with sophisticated modelling of European windstorm and flood. “There has been a tremendous volume of work by us and others on these weather patterns and associated losses,” Dr Simic adds. RMS had previously raised the stakes in 2007 when it mapped water depths around the UK in order to create a bespoke parametric index for the Blue Wings cat bond, issued on behalf of Allianz Global Corporate and Specialty.
In the US there are a multitude of perils from hurricanes to earthquakes and wildfires, and there is a wealth of information on record. In Europe by contrast, although there are similarly diverse perils, there is not always enough good information to model them. However, seismology has been an area of progress since 2005, setting an example of the benefits of effective coordination.
No trend – yet
The continued popularity of cat bonds into 2008 is not yet clear. Bonds are still being issued despite previous warnings that the credit crunch and softening catastrophe rates could limit issuance numbers. The majority of issuers are still favouring indemnity-style contracts and so far there has been no trend that would suggest the start of a shift towards parametric bonds.
One reason indemnity triggers are still preferred by the industry is that they have less basis risk attached. While parametric triggers are evolving all the time to deal with basis risk uncertainty, this remains an issue. Put simply, basis risk is the risk that the cedant experiences a major loss but that the bond fails to pay out because it does not reach its required trigger level. Investors on the other hand tend to favour parametric triggers as these appear more straightforward. They are easily measured and understood.
The advancement of parametric triggers is seen by some as an inevitable spur for insurance-linked securities. With wider use of parametric triggers, the ILS market becomes more transparent and understandable for investors, which would seem to guarantee more popularity. Therefore, the development of parametric-linked ILS products must currently be in a “lag-phase” until that popularity takes a proper foothold.
Parametric triggers could also present advantage beyond weather-linked ILS products, such as insurance for the poorest communities. “Parametric triggers lend themselves to micro-insurance, which is the insurance equivalent of micro-lending for poor villages around the world,” says Nakada of RMS. “There are people who are one storm or flood away from starving, so they could have micro-insurance that automatically pays out if it rains more than a certain amount in their village.”
In the realm of parametric triggers, awareness remains an issue. While awareness is strong within the existing investment community, it is nascent among mainstream investors. In response, RMS is launching Miu – a project to provide investors all the information they need about cat modelling and its applications. “I think there might have been a fear among cat modellers in the past that they should not disclose too much about the way they work, but we believe this game will be more about running than blocking,” says Nakada. “Our dream is that it is absolutely standard for a pension fund to have two percent allocation to a peak peril and that it looks weird not to.”
David Banks is deputy editor of Global Reinsurance.