Does Japan's Earthquake Show That Catastrophe Bonds Are Useless?

The economic impact of Japan's recent severe earthquake will likely be in the hundreds of billions of dollars. A small part of that cost, an amount probably in the low tens of billions of dollars, will be covered by insurance companies. A still smaller portion, close to $2 billion, could have potentially been covered by catastrophe bonds. Those are securities that serve as reinsurance, on which investors make a bet that a big catastrophe, like an earthquake, won't strike over some period of time within certain parameters. The recent events in Japan show that these bonds are better structured to protect investors than the insurers that seek protection from their risks.

First, a little more explanation of catastrophe ("cat") bonds may be helpful. Cat bonds are a relatively new market. They were created in the mid-1990s, in response to Hurricane Andrew. The idea is that they will provide insurance companies some relief from truly major disasters.

Here's how they work. Let's say you're a company that provides homeowner insurance in South Florida. Obviously, a major hurricane will cost you a lot of money. As a result, you would like some insurance on your insurance, which is called reinsurance. So you find some investors who are willing to make a bet on future hurricanes. You sell them cat bonds. They give you the principal value of the bonds. Over the term of the bond, you provide coupon interest payments to the investors if a proscribed catastrophe does not occur. If one does occur, then you get to keep some or all of the principal, instead of paying it back to the investors upon the bond's maturity.

An article this week from Bloomberg by Bryan Keogh, Oliver Suess, and Jesse Westbroo provides a great summary of how well the cat bond market has done since its creation, despite a number of significant natural disasters having occurred. They've performed so well because they have very carefully crafted criteria to protect investors from paying insurers.

The recent earthquake in Japan was about as serious a quake as can occur, but since it was more than 70 miles from Tokyo, the cat bonds will barely help the insurance companies that spend millions of dollars on them. Bloomberg explains:

Insurers and reinsurers typically sell cat bonds to help cover their most extreme risks such as an earthquake rocking Tokyo or a hurricane with the force of Katrina hitting the center of Miami. This month's earthquake in Japan struck about 240 miles (380 kilometers) northeast of the capital, meaning investors may pay insurers less than 10 percent of the $1.7 billion of debt sold to help cover losses, said Niklaus Hilti, head of insurance-linked strategy at Credit Suisse Group AG.

To be sure, if the earthquake's center was downtown Tokyo, the devastation would have been worse. Still, according to the Bloomberg article, the earthquake will cost insurers something between $12 billion and $25 billion, of which cat bonds will pay for less than $170 million.

How well have cat bond investors done over the past several years, despite some serious calamities? Bloomberg covers that too:

Cat bonds, whose owners risk losing their entire investment if a disaster occurs exactly as defined under the security's terms, have returned 64 percent over the past five years through March 18, with annual gains even following Hurricane Katrina in 2005 and the collapse of Lehman Brothers Holdings Inc. in 2008, according to the Swiss Re Cat Bond Total Return Index. Only two bonds among the more than 300 tranches issued have ever suffered notable losses as a result of disasters, according to Moody's Investors Service.

Only two out of 300 bonds have ever had a loss, and the entire pool's return has been 64% in just five years! It's good to be a cat bond investor. But that means it's bad to be on the other side of the trade. Insurers have paid a lot of money to investors with little benefit, despite a number of serious catastrophes having occurred. This is likely due in large part to the strict conditions that the bonds require for payout.

If insurers begin to realize that they're paying a lot of money for protection that rarely pans out, they could seek to loosen the criteria that cat bonds currently require for payout. But investors might not allow that.

Imagine if the center of the earthquake had been within 70 miles of Tokyo. Investors would have lost a great deal of money. The purpose of cat bonds is not to cover all of insurers' risk, just "fat tail" risk, where a very improbable event costs an enormous amount of money. Most of the time, investors should be rewarded handsomely with big interest payments, because when a disaster eventually does strike within a cat bond's specified parameters, they will lose big time.

For now, investors seem to be doing a better job than insurers at determining criteria to make cat bonds lucrative, but that can change very quickly. Insurers may come to decide that they're paying too much for this protection, however. After all, if they had saved the premiums they paid in the past five years, they would be 64% of the way towards having enough money to cover the costs from a major disaster that cat bonds would provide.