Hurricane Sandy may have devastated the Jersey shore, but it didn’t dampen investor interest in high-yielding catastrophe bonds. The market for natural disaster “cat bonds,” which property and casualty insurers issue to transfer risk, just had its best year for new issuance since 2007 thanks to a strong fourth quarter, though some older bonds with exposure to Sandy may end up getting nicked for some of their principal.
The first deal after Sandy, from insurer USAA in November, covers a range of U.S. perils, including hurricanes, severe thunderstorms, earthquakes, winter storms and California wildfires. The company increased the offering from $250 million to $400 million in response to demand from investors. It was able to do so even though Standard & Poor’s Corp. had put two of the earlier bonds in USAA’s “Residential Reinsurance” series, from 2011 and earlier in 2012, on CreditWatch with negative implications because of potential losses from Sandy.
“The successful Residential Re 2012-2 placement post-Sandy was a very good sign,” says William Dubinsky, the head of insurance-linked securities at Willis Capital Markets & Advisory, the investment banking arm of the global insurance broker. That kind of deal, with those kinds of risks, is “the bread and butter” of the natural disaster catastrophe bond market, he says.
Last year companies issued a total of $5.9 billion in bonds in the natural disaster category, bringing the total amount of outstanding natural disaster cat bonds to $15.2 billion globally. Overall, cat bond issuance, including a much smaller market segment devoted to pandemic and mortality risk, was slightly more than $6 billion.
Last year’s natural disaster bond issuance was up by 37.2 percent from 2011’s $4.3 billion. Dubinsky says 2013 should be even stronger. “The pipeline that we’re aware of is pretty good,” he says.
Still, the size of the cat bond market remains small. Outstanding debt totals $36 billion once private placements are included, with the U.S. accounting for over a third of that. That’s roughly equal to the amount of reinsurance premiums taken in by just one company, market leader Munich Re, in 2011.
Then again, the market for cat bonds is relatively new. The first natural catastrophe bond was issued in 1996. In the wake of Hurricane Andrew in 1992 and the Northridge earthquake in Los Angeles in 1994, insurers started thinking about new ways of raising additional capital to cover truly extraordinary losses.
The money raised by a cat bond is held in a trust and earns interest — typically, for three years, though some bonds mature in as little as one. That money is the last to be tapped should the claims from an event exhaust all of the usual lines of insurance and reinsurance. It’s rare, but it does happen that investors lose some of or their entire principal. The Japanese earthquake in March 2011 wiped out a $300 million Japanese earthquake bond, issued by Muteki for Japanese insurer Zenkyoren in May 2008. In 2011, two $100 million bonds, issued by Mariah Re in 2010 on behalf of Wisconsin’s American Family Mutual Insurance Co. to cover severe U.S. thunderstorms, were also total losses for their investors.
It’s pretty unusual that investors will buy a bond knowing that they risk a loss of principal. “Most investors are more comfortable analyzing the longevity of a CEO’s tenure or the vagaries of a product introduction than they are looking at the statistical properties of natural catastrophes,” says John Brynjolfsson, chief investment officer at Armored Wolf, the $781 million-in-assets investment advisory firm based in Orange County, California. So why is the market booming?
A big part of the allure of cat bonds is their rates, which range from 3.5 percentage points to more than 20 points over a base rate such as LIBOR or U.S. Treasury yields, says Luca Albertini, chief executive officer of Leadenhall Capital Partners, a London-based investment management firm that has $857 million invested in natural disaster and mortality risk bonds and private placements. The average yield spread on a cat bond is about eight or nine points, he says. Some cat bonds go as high as “plus 22 percent,” but, Albertini says, “Unless you are an aggressive risk taker, you don’t want that bond” because a rate that high indicates a level of risk “out of the ordinary.” Because of the risk of “attachment” — that is, the risk that the bond’s principal will be tapped, in whole or in part, to pay claims — cat bonds are not rated investment grade and a consequent loss of principal, cat bonds are not rated investment grade, but they typically offer higher coupons than corporate junk bonds.
The biggest chunk of the market — 72 percent, according to Willis — is exposed to U.S. “wind,” or hurricane, risk. Other perils — such as European wind or Japanese earthquake — carry lower coupons closer to 4 or 5 percent because of demand from investors who “want to diversify with transactions that are not exposed to U.S. wind and quake,” Albertini says. Deals can also include multiple tranches with ascending levels of risk and return. For instance, that USAA Reinsurance Re deal from November had four tranches. The first two, with the lowest risks of attachment, closed at 4.5 percent and 5.75 percent, according to Artemis, an online site for the reinsurance industry that maintains a “deal directory.” The other two tranches, which have a higher probability of attachment, closed at 12.75 percent and 19 percent, Artemis said.
Since there is that risk of total loss, cat bonds in the U.S. are sold only to accredited investors as 144A offerings. Most institutional investors invest via managed funds, where the risks are diversified geographically and by the type of peril. The funds may also diversify by including private placements, in addition to bonds.
So far, none of the cat bonds have been triggered by Sandy, but Albertini says that his funds paid out a small amount, mainly on bilateral private transactions. Because of SEC restrictions on advertising the returns on these kinds of funds, he says he can’t be more explicit, but he says industry-wide, “no one had double-digit losses, so far as I know.”
When Sandy first hit, a group of bonds traded down in the market because of the fear of losses. By now, most have rebounded, with just a few still trading at significant discounts, including one of the USAA bonds and a Swiss Re issue, the Successor Class V – F4 bonds, maturing in November 2015, which market sources say is still trading in the 70s.
Sandy is seen as being much less devastating — at least from the insurance perspective — than Hurricane Katrina. That’s because much of the damage from Sandy was caused by flooding in coastal areas, which usually isn’t covered by private insurance, but by the federal government via the National Flood Insurance Program. Private coverage for hurricane damage typically covers the damage caused by wind, but not water.
There are a number of preliminary estimates from risk modeling agencies on the losses from Sandy, but the estimate that everyone in the industry considers to be the most definitive is from PCS, or Property Claim Services, which is based on “on the ground” reports from insurers on their actual losses. The day after Thanksgiving, PCS issued its first estimate on losses from Sandy at $11 billion. That number will be updated in mid-January, and there will be continual updates after that. By way of comparison, PCS’ initial estimate on Katrina was $34.4 billion, and the final number, issued almost two years later, was $41.1 billion.
Katrina caused substantial wind damage — the kind that insurers pay for — throughout a large area that included not only New Orleans and Mississippi but also Alabama and northern Tennessee, explains PCS assistant vice president Gary Kerney, based in Jersey City, New Jersey.
Sandy will take longer to sort out, he says, because access to many areas along the Jersey shore remains restricted, making it difficult for adjustors to do on-site inspections. “At this point in time, there’s still an awful lot that’s unknown,” he says. “It will be a real challenge to come up with a final number.”
The consensus in the industry is that if the actual losses from Sandy hit $20 billion, more cat bonds would be in danger of taking a hit. Albertini says he would be “stunned” if PCS rose beyond $17 billion to $18 billion.
But the risk modeling agencies think that higher insured losses are within the range of probability. AIR Worldwide, which issued an initial estimate of $7 billion to $15 billion, raised its estimate to $16 billion to $22 billion in late November, according to Artemis, The site also reported that EQECAT, another risk modeling agency, had doubled its initial estimate, from $5 billion to $10 billion, to $10 billion to $20 billion, while RMS’ estimate stood at $20 billion to $25 billion. Most recently, reinsurance broker Holborn projected a range of $20 billion to $30 billion in insured losses — the highest estimate so far, Artemis said. “Reinsurers, and also cat bond investors, must be watching and wondering where this will end,” Artemis said.
With additional reporting by David Turner.
