Catastrophe Bonds Generating High Yields

Posted by Bull Bear Trader | 6/28/2008 07:37:00 AM | , , , | 3 comments »

There is an interesting article in Barron's this week regarding catastrophe bonds. Basically, catastrophe bonds are similar to normal bonds in that you invest a principal in return for periodic coupons. Once the bond matures, you receive your principal back - hopefully. The hopefully part is where these bonds are slightly different. Yes, with all bonds you have the risk of losing your principal, but for cat bonds it is less about credit risk, and more, obviously, about catastrophic risk. In most cases this is a binary proposition. If there is no event, you get all your money back. If there is an event, you do not get anything back. In return you get a nice coupon to compensate for the risk you are taking. After Hurricane Katrina, one cat bond tranche was offered by Swiss Re with an annual coupon of near 40%. In fact, cat bonds have returned over 33% from 2005 to this May, ahead of the 19.1% offered by the Lehman High Yield Corporate Bond Index over the same time frame. A additional benefit of cat bonds, beyond the high yields, is that their returns are often uncorrelated with the returns of other equity or fixed income investments, providing another vehicle for diversification.

Cat bonds were designed as a way for insurance companies to remain solvent if an event they insured does occurs. Insurance companies could simply buy reinsurance, passing the risk on to another insurance company, but there is the worry of too much correlation to the event. As an alternative, they could sponsor a cat bond. In short, the company would create a special purpose entity (yes, I know what you are thinking) that would issue the cat bonds with the help of an investment bank. Investors would then buy the bonds and receive a coupon with a defined spread over Libor. This spread can be as little as 0.5% to 20% or more depending on the event and the likelihood of its occurrence. Recent catastrophic events also have an impact on defining the spread. You can get a little more background on cat bonds here and here.

Since cat bonds often involve the creation of a special purpose vehicle, some investors are a little worried that some reinsurance companies are moving beyond their specialties. They are also concerned that by moving the risk off balance sheet, companies are preventing investors and the market from knowing the real exposure each company is taking. Cat bonds do allow reinsurance companies to survive and be less exposed if a major event does occur. Therefore, companies are less exposed by taking out insurance themselves, but off-balance sheet items are more difficult to value and risks are less transparent. The effects on market participants, such as Munich Re, Swiss Re, Liberty Mutual, Allianz, and Hannover Re, among others, is difficult to tell. On the other hand, the benefits to the investment banks underwriting the bonds, such as Barclays Capital, Deutsche Bank, Lehman Brothers, Goldman Sachs, and Swiss Re Capital Markets, among others, is a little easier to see and quantify, along with the potential returns for institutional investors, who at this point are the only ones currently receiving cat bond distributions.

As mentioned in the Barron's article, to date only one cat bond has been triggered, implying a low probability of catastrophic events occurring, or at least the ones that are being underwritten. Then again, the last two years have seen a lower level of terrorist events and major hurricanes. In fact, the last two hurricane seasons, which have been forecast to be strong, have fortunately been milder than expected. This year is once again forecast to have an active hurricane season. Hopefully the forecast will be wrong again, and cat bond investors will get a return of principal, and the people on the coasts and around the globe will be spared from another major event.


  1. zerobeta // June 29, 2008 at 10:22 PM

    Good summary on cat bonds. I will pick up Barron's tomorrow to read that article. I worked in the industry for a brief period of time and it is very interesting.

    There are a few things that you mention that are either an overgeneralization or misunderstanding that I figured I'd point out.

    1) You mention, "Insurance companies could simply buy reinsurance, passing the risk on to another insurance company, but there is the worry of too much correlation to the event."

    I'm not really sure what you mean by "the worry of too much correlation to the event". I think you mean that the worry that reinsurers have too much exposure to the event - which is half true. The cat bond industry and ILS industry as a whole is a reaction to the fact that the insurance and reinsurance industry is much too undercapitalized and there is a lack of capacity in the traditional reinsurance market for such fat tail risk that is difficult to price and model. This can be seen through recent history as each major event has caused a wave of reinsurers to go under. Still the insurers are not as concerned with the counterparty risk of the reinsurers on their book of risk and would prefer to receive traditional reinsurance over this form because of the nature of the treaties. Most Cat Bonds are not indemnity based triggers (more on this later), thus leaving the insurer with an uncomfortable amount of basis risk.

    Over the last 10 - 15 years the market for cat bonds has experienced very abrupt growth periods, which followed big capacity draining events like Katrina, Andrew, and 9/11, and has stagnated in between. It usually grows in hard markets. What makes this recent growth compelling and interesting is that there hasn't been any big events and the (re)insurance market has been softening. Thus it must be attributable to something else. I think the credit crunch is the reason. Hedge funds are opening up lots of funds investing in this strategy because the zero beta risk and high yield makes it an attractive sale. The Sharpe Ratio on cat bonds is very attractive. Also, counterparty credit risk is starting to be a major concern. Since cat bonds are a form of fully collateralized insurance counterparty risk is not a problem. The fact that both supply and demand has picked up has kept the yields on these bonds attractive. What worries me however, is that many more funds will get into this "hot" exotic beta asset class and don't really understand the nature of the risk (or understand it but don't really care about it) driving down yields and increasing exposure - just as an event hits. The fact that these same funds invest in other things may erode of of the "zero beta" effects.

    2) You mention (or say that the article mentions)

    "As mentioned in the Barron's article, to date only one cat bond has been triggered, implying a low probability of catastrophic events occurring, or at least the ones that are being underwritten."

    Be careful. Not only is this bad logic, but needs clarification. The bond that got triggered in Katrina was Kamp Re - a rather suspect indemnity-based trigger sold right before that hurricane season started. Since then indemnity based triggers have been avoided and the triggers have been primarily either parametric, parametric index, and modelled loss. These don't directly depend upon the insurer's book, but either characteristics of the event, the industry, or a computer based modelled loss on an insurer's book should certain characteristics occur. So the fact that one 1 has been triggered does not mean the probability of catastrophic events occurring is low, it only means that certain catastrophic events have not occurred. I would expect as demand for this asset class grows there will be more indemnity based deals. These are not nearly as attractive because of moral hazard, the basis risk that must be assumed to hedge any exposures, etc.

    The only thing that is implied by the fact that none have been triggered, is that there is a low probability of the triggering events so far, but not of a catastrophic event in general. Many of the bonds in Katrina and Rita were not triggered because of their triggers. If they were indemnity or perhaps just a little different they most likely would have been. Also, as this grows and more and more insurance risk gets converted from indemnity based measure to this parametric measure, the likelihood of "catastrophic event" mostly likely becomes larger since there are just more possible triggers out there for every actual event that occurs.

    Anyways I liked the post and hopefully this helped add to it.

  2. Bull Bear Trader // June 30, 2008 at 7:47 AM

    zero beta,

    Thanks for taking the time to clarify the post. It is appreciated, truly. For 1, you are correct. I was trying to refer to a reinsurer underwriting too many policies for the same event, essentially making it difficult for them to cover the loss if it does occur. The extra details you provide are very helpful. You did mentioned that "hedge funds are opening up lots of funds investing in this strategy." Interesting. I knew many were buying cat bonds as alternative investment (near zero beta risk you mention), but I did not realize there were a lot of funds focusing in this as a strategy in and of itself.

    As for 2, I agree again. It should have been bonds triggering and not events. Nonetheless, once again your description provides more detail to be considered.

    Thanks again. Your comments added a great deal.

  3. zerobeta // June 30, 2008 at 11:17 PM

    No prob. The cat bond industry and the ILS industry in whole are very interesting and exciting. I always am on the look out for people's views on the industry, and yours was right on the money. Nothing you said was wrong, but Barron's and even Wikipedia have a way of sometimes generalizing something so that it is easier to understand by the many. The problem that I think we will start to see with Cat Bonds and more specifically Cat Risk - starting with that Barron's article, is that the nature of the risk will begin to cause a misunderstanding of its nature and severity. Most people, if any, have a hard time comprehending and evaluating something that is so severe and unexplainable like a catastrophe and then afterwards lay blame on someone or something with no real hand in the event. Most people can not psychologically deal with lumpy streams of payments and timing has a lot to do with the payoffs of these instruments. I want to make sure as people learn of this asset class they understand the risk involved, because another event will happen - thats a given. Some people who bought bonds that year will lose their entire principal before they ever receive the nice rate. When this does happen, I just hope that the product is not the scapegoat, as I think this is a step towards dynamic, transparent insurance pricing, which helps us all.

    Keep up the interesting posts.