| Event-linked bonds 101 |
|
|
| 02/09/2007 | |
|
John Brynjolfsson, Managing Director at PIMCO, answers this week's questions on event-linked bonds. He manages $2bn in the sector and has been an active participant since its inception in the mid-1990s. 1-What is an event-linked bond and how big is the market? Event-linked bonds are securities that provide a way for insurance companies to transfer catastrophic risk to investors. These securities are generally linked to events that are very rare, such as super-catastrophic hurricanes or earthquakes, but carry significant risk for the insurer on those rare occasions when they do occur. Through event-linked bonds, insurers can offload the risks associated with one-in-100-year, super catastrophic events and focus on what they do best, which is sell and manage insurance policies. Event-linked bonds also allow the capital markets to do what they do best, which is distribute super-catastrophic risk among multiple investors who can diversify risk in exchange for compensation. The event-linked bond market has grown significantly since the first of these were issued about 10 years ago. The size of the market is currently in the neighborhood of $20 billion, and based on recent growth trends, the size of the market could well double over the next two years and then double again two years hence. The rate of growth is being driven on both the supply and demand side. On the supply side, insurance companies are finding it more and more important to lay off risks as the size of their insured property base grows and rating agencies get more strenuous in applying capital standards on insurers, and reinsurers, who protect policyholders from such perils. Event-linked bonds, because they are fully funded, represent a more secure form of protection for insurance companies than uncollateralized protection typically provided by reinsurance treaties. (Secured protection is particularly important for protection relating to super-catastrophic risk, as such events by definition strain the insurance and reinsurance industry financially.) Demand for event-linked bonds is also growing. A key reason is that the event-linked market is currently one asset class where yields are now wider than they were five years ago, whereas almost all other spread sectors, even with the widening in recent months, are compressed to tighter levels. 2-What kind of spreads do they offer over LIBOR? Spreads over LIBOR for event-linked bonds issued in 2007 have gravitated around LIBOR + 5% for one-in-100 year peak peril risk. Peak peril bonds relate to California earthquakes, Florida wind, and perils of similar scale that have peak industry losses in the range of $50 billion. Such losses would strain the insurance industries' capital broadly, and as such, the industry is least able to retain these risks. More generally spreads over LIBOR have ranged from 0.65% to 15.25% for all bonds issued. For BB-rated bonds, spreads have ranged from 2.95% to 7.00%. 3-Why are their spread-divided-by-risk multiples significantly higher than those of corporate bonds? One measure of value for risky securities is the ratio of how much spread is paid, per unit of risk being underwritten. For event-linked bonds, this multiple is calculated by looking at the quoted spread (over LIBOR) and dividing it by the quoted expected loss (as calculated by third party modeling firms and reported in offering documents). Such "multiples" can also be calculated for corporate bonds, typically by using Moody's widely cited "default studies" that look at five-year historical default rates for bonds in various ratings classes. So this question asks about comparing cat bond multiples (which have recently gravitated around 5) to those of corporate bonds (which are typically below 3). Super-catastrophic events occur very infrequently, making it difficult for risk managers to precisely quantify the expected frequency of future events. Consequently, event-linked bonds tend to be priced based on the most conservative probability estimates associated with the historical record. Hurricane Katrina, the category five hurricane that hit New Orleans in 2005, is another factor that led to the current sizeable risk spreads in event-linked bonds. Hurricane Katrina prompted the insurance industry and investors to reevaluate the risk of super-catastrophic events and subsequently spreads on all types of event-linked bonds widened considerably in 2006. Spreads have narrowed a bit since then, but remain significantly wider than pre-Katrina levels. I would point out, however, that even though Katrina was a bad hurricane and did cause losses on some event-linked bonds, most of the hurricane-linked bonds in the market were of such high quality that they actually existed in a risk layer above that of Katrina damage, and they therefore were not affected. Demand factors may also be contributing to the large risk spreads in event-linked bonds. Demand for event-linked bonds is somewhat limited, perhaps because aggressive investors perceive one-in-100-year risks as too tame, while more traditional investors may consider the uncertainties in event-linked bonds too stressful, particularly in light of past headlines following the Katrina tragedy, current headlines relating to global warming and extreme weather events occurring globally. The event-linked bond market is also relatively inefficient compared to the corporate bond market. The event-linked market is large and growing rapidly, while the corporate bond market is growing much more slowly (on a much larger base). It has been around for, literally, centuries, and today is far more liquid, with much more research and analysis available. Corporate bond investors can also assemble what appears to be a relatively diverse portfolio, whereas the ability to diversify within an event-linked portfolio is limited (since even distinct issuers and issues will embody exposure to similar events.) On the positive side, the confluence of these supply and demand factors has created a potentially strong buyer's market for event-linked bonds. 4-How does the recent market volatility impact the spread on event-link bonds? Event-linked bond spreads remained relatively flat during the recent volatility, and prices in particular have remained stable. Given that this volatility has resulted in sharp losses for most corporate bond investors, we think the stability in event-linked spreads is a strong testament to their diversification benefits. In fact, event-linked spreads have actually shown signs of tightening, and their prices rising, due to a relatively inactive hurricane season and the prospect of lower insurance premiums during the current and perhaps future policy renewal periods. (Lower premiums may benefit Event-Linked Bonds, since lower premiums in the insurance market portend lower coupons on future issuance of event-linked bonds, making previously issued bonds more valuable.) The valuation process for financial market instruments embeds uncertainty for corporate governance, unexpected rating changes, correlations with other financial market asset classes and other factors that are distinctly separate from the risk forecasting process associated with event-linked bonds. Investors who have experienced fatigue from the throes of market-related or "man-made" risks and uncertainty may appreciate the unique risk opportunity offered by event-linked bonds. 5-How does PIMCO get comfortable with the risk side of event-linked bonds? PIMCO has a long history in event-linked bonds going back to the earliest days of this market, and through that experience we have developed a very strong team, a solid investment process and extensive connections with experts in modeling and analyzing these types of rare events. Forecasting the underlying risks and perils is our first step in evaluating any event-linked bond. And we do that by examining the historical data on earthquakes or storms to be sure that the principals and concepts that the modeling firms present us with are realistic and reasonable. Our effort in this area is supported by our team of quantitative experts, including both those who are specifically focused on event-linked bonds, and a much larger group of financial engineers with backgrounds in the physical sciences. We have also developed relationships with third-party modeling firms over the years and I would note that most are staffed by dedicated scientists who have pedigree academic backgrounds in seismology or meteorology. Though we never take risks for granted, we feel, based on our own verification, that their forecasts are unbiased and very scientific. Buyers and sellers of event-linked bonds use these models and we have found them to be generally very much on target. Second, our team of investment professionals looks at supply and demand factors in the event-linked bond market to help determine the extent to which the risk premium should be above and beyond the underlying actuarial risk, and higher or lower than historical averages. Third, we develop a targeted structure for our exposure that will involve a balance of hurricane risk, earthquake risk and other risks that somewhat diversify the hurricane and earthquake risk. We want to maximize return potential by keeping risk under control. Fourth, we supplement our long-term analysis by examining higher frequency information about short-term developments in the meteorological markets and in earthquake activity, or technical distortions to market demand for, or supply of, event-linked bonds. These short-term factors may lead us to conclude that the upcoming year or years are essentially in line with the historical record, and modeling firms' estimates, or we may feel there is a need for an upward bias relative to the historical record to account for warmer ocean water or higher seismic tension. On the basis of those prior steps, we develop a target for our tactical exposure, construct our positions and then monitor our exposure on an on-going basis. There is a secondary market for event-linked bonds that allows for active pricing and validation of prices quoted in the market, and is large enough to allow for routine changes in our positioning. |
|
Articles of the same Serie : 101 |
|
© Copyright 2008 bfinance. This document is for your personal non-commercial use. Any further copying, reproduction, distribution is strictly prohibited. To obtain permission please contact This e-mail address is being protected from spam bots, you need JavaScript enabled to view it