As the fear of interest rate rises threatens virtually every market in the world from sovereign bonds to high-tech stocks to emerging market securities investors may no doubt be wondering where they can seek refuge.
Such an anxious environment may be the best thing to happen to catastrophe bonds, aka cat bonds. These structured financings, which transfer some of the catastrophic risk of property-casualty insurers to investors, have been struggling to get off the ground for the past several years. Investment bankers are trying to convince investors that, although cat bonds may be risky, such risks have nothing to do with interest rates and are thus not correlated with other financial assets. Bankers hope eventually the market will be akin to mortgage-backed or other asset-backed securities.
While most of the work on cat bonds has been done in the US, it was in Zürich last month that the largest, and first, public cat transaction was offered. Credit Suisse First Boston (CSFB) arranged the three-year Sfr400 million ($281.7 million) subordinated convertible bond for Winterthur Insurance, which was based on the risk of hailstorm damage to automobiles. The deal was sold to retail investors and trades on the Zürich Stock Exchange. Investors will get a 2.25% annual coupon, which will be knocked out if any single hail or other storm damages more than 6,000 motor vehicles insured during the year with Winterthur in Switzerland. The principal is guaranteed and after three years the bonds are convertible into Winterthur shares. The 2.25% coupon is higher than the 1.49% a traditional convertible would offer in Switzerland.
Bankers say that if deals are structured right, investors are clamouring to buy in. “We have a shortage of good deals, not a shortage of people who want to put money in them,” says John Kelly, a managing director at Citicorp Securities, which has been one of the leading banks in the nascent market. He says Citicorp has structured several $100 million cat private placements in recent years. Its latest, a $100 million November deal for Hanover Re, was oversubscribed. Kelly says the deal, a multi-year transaction in which investors were allocated a notional amount of risk in a diversified portfolio of risks originated by Hanover, was successful because it was leveraged, offered investors diversification with a company with a track record, and and could provide yields “north of the other deals in the market”. Citicorp has about 20 institutional investors interested in the deals, says Kelly. While principal protection appears to be the key to getting some deals off the ground, Kelly says these investors don’t mind sharing risk with the reinsurer as long as they are compensated similarly.
Last month, Goldman Sachs issued the first 144A private-placement cat bond in the US, a 10-year $68.5 million securitized reinsurance transaction for St Paul Re, the reinsurance arm of the St Paul Companies, a US based insurer. However, the deal ended up much smaller than the $200 million offering originally planned and had to be separated into two tranches notes whose principal is guaranteed and preference shares whose principal is at risk.
Merrill Lynch is restructuring a public $500 million one-year (renewable for three) hurricane-risk cat bond for the US-based United Services Automobile Association (USAA) which recently was pulled for lack of investor interest. Individuals close to the deal say the revised transaction should close in the first quarter. In addition to the Winterthur deal, CSFB is also expected to close by April a $100 million public two-year deal for a European reinsurance company that will cover Californian earthquake risk.
Many investors who are more knowledgeable about interest rate risk than that of natural disasters are still wary of the new deals. “It appears pretty clear that the capital markets are not receptive to a complete risk transfer,” says Jonathan Plutzik, an insurance analyst at CSFB, a position contrary to Kelly’s assessment. “Investors are not going to buy securities with a risk of losing all their interest and principal.” Bankers are “wrestling with how to strike the balance between risk and reward,” he says.
As a comparison, investors look to the returns of reinsurance companies, which have been high in the past two years because of a decline in natural disasters since 1994, the year of the Los Angeles earthquake. Indeed, some of the Bermuda reinsurers set up to deal with huge catastrophe losses suffered between 1989 and 1994 have had returns in the 30% range. Some private deals have sought to offer such high returns, but with a higher risk profile. For example, last year’s private placement by American International Group offered a maximum 25% return, but also had a 25% risk of principal loss, according to CSFB.
Citicorp’s Kelly says reinsurers can earn better returns because of their leverage an element that is often missing in the public cat bond deals. “These public deals operate on a leverage of one to one,” says Kelly. “For every $100 of liability, the investor puts up $100 in capital against it. It creates a sub-par return.” Reinsurers would have a leverage of five or six to one, he says.
In general, cat bonds offer investors a coupon on a securitization that is created from what’s left from the premium the insurers charge and the investments of the debt issuance after the bankers subtract their fees. Although typically these securities are structured to offer a set coupon, the overall return would be lowered if catastrophic losses were to rise above a set point, called an attachment point. Then the insurer taps the pool of money to cover his losses and the notional amount in the vehicle is decreased.
For example, in the original USAA deal, which was shelved, investors could theoretically have lost all their money for a return that was based on a coupon only 600-700 basis points above US treasuries. The deal also covered only a single risk, instead of a diversified pool. USAA, which insures the home and property of the US military among others, was hard hit by Hurricane Andrew, which swept across Florida in 1992 and caused catastrophic losses of close to $20 billion. As a result, USAA wanted to create a vehicle specifically to cover future hurricane losses. With its own capital of $4 billion, USAA set the attachment point at $2 billion, which meant that losses above that amount could be drawn from the bond proceeds.
“If the hurricane hit the coast of Florida, and USAA suffered $2.2 billion in losses, the $200 million would come from investors,” explains one person close to the deal. At that point, investors would have lost almost half their principal. USAA has 4% of the Florida market, so Hurricane Andrew cost it an estimated $800 million in losses. But if that hurricane had veered just 30 degrees and had hit Miami, say analysts, losses could have been five times greater. That would have put the losses for St Paul Re at $2 billion.
As a result of such possibilities, when Moody’s Investors Service looked at that transaction, it offered to give a below investment grade rating to the deal one of many reasons it was difficult to market. Now the USAA deal is being restructured with a higher attachment point so that it can get an investment-grade rating. The structure of CSFB’s upcoming European reinsurance deal is also said to be similar to that of the USAA deal.
Analyzing the risk is a particular stumbling block. “Clearly cat losses over the last five years are so different from anything that has happened over the last 30 years that who knows what’s happening,” notes Citicorp insurance analyst Irene Driscoll. “Maybe weather patterns are changing.” Ever since the Hugo hurricane of 1989, catastrophe losses related to both hurricanes and earthquakes have been far greater than the historical norm, making predicting the future extremely difficult, she says. (To make the Winterthur cat investors feel more comfortable, the company is using the Internet to post information on storm damage relevant to the bond.)
Goldman’s deal for St Paul Re covered an even more complex risk, and keep bondholders on the hook for a longer time. “These are very high risks,” says one analyst. “This is a 10-year deal. You can have a hurricane in the world anywhere St Paul underwrites. They reinsure risks other people don’t want to take.”
Instead of providing for cover for one catastrophic risk, as some of the other deals plan to do, the securitization is simply a way for St Paul Re to cede its reinsurance risk, creating a diversified portfolio. Since the purpose of the vehicle is to allow St Paul Re to increase its reinsurance underwriting, and it is structured as a revolving portfolio, the risk is very hard to quantify, says Standard & Poor’s structured finance analyst Kristin Brooks. However, she notes: “It’s not a free for all”. The total book can only have a certain percentage of any one type of risk.
The notes carry a 6.09% coupon, and Goldman estimates in the offering documents a “base internal rate of return” of 12%. In the worst case, however, only the principal will be returned to investors. St Paul Re raised $44.5 million through the issuance of the notes, of which $23.2 million is invested in zero-coupon securities to provide for the full return of the principal of the notes at maturity. Another $24 million was raised in the preference shares, whose performance is based on the performance of the business ceded to the reinsurance vehicle.
Because the two tranches of the Goldman deal offer both principal protection and principal at risk, the deal was able to attract more investors, according to Morgan Zee, a Goldman insurance products banker. He says the investors included life insurers, mutual funds, pension funds, hedge funds and “a few” property casualty insurers.
Goldman is currently working on several other transactions, including some based on a property casualty risk index that could be hedged by options and futures traded on the Chicago Board of Trade. These derivatives have languished, but bankers hope that the development of an underlying cash market will help to spur their usage. Michelle Celarier
