We have rarely provided our opinions or market commentaries in our Reinsurance Focus posts, preferring instead to provide our readers hopefully balanced analyses of court opinions, legislation, and regulations affecting the reinsurance market. Recent events, however, have caused us to make an exception to that practice. Much is being written about the extent to which catastrophe bonds (or traditional reinsurance) are “exposed” to or may be called upon to pay losses from Hurricanes Harvey and Irma, and the impact that those storms may have on the ILS market. Some reinsurance agreements use a parametric trigger, and the possible impact of a particular storm on such reinsurance facilities may be reasonably ascertainable in the relatively short term. For example, the Caribbean Catastrophe Risk Insurance Facility, which has developed parametric trigger policies covering wind risks to be backed by traditional reinsurance and capital markets, insuring member countries in the Caribbean, has already determined the amount of payouts on its parametric policies for Harvey and Irma, and has published a report on that issue. However, many of the catastrophe bonds covering wind risks have indemnity triggers, including those to which the Texas Windstorm Insurance Association (the Alamo series) and Florida’s Citizens Property Insurance Corporation (the Everglades series) ceded risks. Moreover, many catastrophe bonds are part of a reinsurance program or tower composed of different types of risk transfer, with different layers, some overlapping or parallel, different attachment points, and different limits.
With the exception of flood losses under the National Flood Insurance Program’s traditional reinsurance program, which does not include catastrophe bonds, few of the “analyses” in the press to date appear to have even attempted to take into account the specific factors that will determine whether, and the extent to which, a particular catastrophe bond or traditional reinsurance agreement is likely to respond to losses, including the trigger of the reinsurance coverage, the attachment point of the reinsurance, the limit of the reinsurance, the percentage of cover for a particular layer, the actual level of losses in the reinsurance program, other inuring reinsurance in the reinsurance program, whether the reinsurance is aggregate or occurrence based, and the extent to which the retention leading up to the attachment point of the reinsurance has been or will be eroded by losses. While one may say that all reinsurance in a tower is “exposed” to losses by any covered event, that is not meaningful without a factually informed analysis of the extent to which a particular catastrophe bond or reinsurance agreement is likely to be called upon to pay losses, given the damage caused by the event and the terms of the applicable reinsurance agreement.
The modeled level of losses from these storms suggested by various brokers and modeling companies are only estimates, and the fact that the modeled losses suggested by different sources conflict with each other and have changed over time is itself good evidence that such numbers are only preliminary estimates, based upon limited reliable data. Many catastrophe bonds frequently attach fairly high in a reinsurance tower, and if few cat bonds actually sustain losses from these two major storms, the impact upon the ILS market may be limited, or even positive. In addition, there has been a great influx of capital into the markets for traditional reinsurance for wind risks and catastrophe bonds over the past several years, and historically major hurricanes have frequently prompted the influx of additional capital.
The bottom line? Only time and accompanying loss development will permit a reasoned evaluation of the impact of Harvey and Irma on the reinsurance and cat bond markets or individual reinsurance agreements or catastrophe bonds.