Reinsurance Carriers Expected To Play Smaller Role In Future Asbestos Losses

New industry data are not yet summarized as of year-end 2003, but year-end 2002 data implies that most companies with material asbestos exposures are expecting significant reinsurance recoveries. These expected reinsurance recoveries are factored into the asbestos net reserves and therefore have a substantial impact on reported surplus.

Table 2 displays the gross, net and ceded asbestos reserves as of year-end 2002 for companies with the largest exposure to asbestos losses. (The top 21 primary companies and top 11 reinsurers whose results are summarized represent approximately 92 percent of the industry's gross reserves). We also display the variation in the ceded ratio (ceded reserves divided by gross reserves) for these companies.

As the table shows, primary companies, on average, are ceding 40 percent of their gross exposure. The range of cessions, however, is very wide. (We are assuming for this purpose that the reported gross reserves are indicative of the exposure.) Seven of the 21 companies are ceding over 50 percent of their gross exposure. In other words, these seven carriers are relying very heavily on reinsurance to cover their asbestos exposures.

The data indicate that, on average, reinsurers are ceding a higher percentage of their gross exposure than primary carriers52 percent versus 40 percent. Five of the 11 reinsurers are ceding above the reinsurer average of 52 percent.

What are some of the implications of this situation? Several items are at work that we expect will adversely affect gross reserves and future reinsurance recoveries for asbestos exposures.

Several sources indicate that industry asbestos reserves are deficient; further gross reserve increases may result in additional reinsurance cessions.

Non-product claims have been an increasing percentage of new asbestos claim filings. The question is: Will reinsurance apply to these claims?

Reinsurance insolvencies will likely result in reduced collection ratios.

Reinsurance disputes appear to be ever increasing and suggest that collections will decrease from their historical levels.

Plaintiff law firms are targeting contractors, storeowners, as well as many others, in asbestos litigation. For these claims, we expect fewer gross dollars to be reinsured. This is due to the fact that the claims are smaller claims may be aggregated by site, for example, whereas products claims are aggregated for the policyholder. Fewer individuals are exposed per insured, and therefore a primary company with a high excess-of-loss attachment point (and no other reinsurance) may not be able to cede these claims.

In contrast, historically, many of the claims originated from large companies and involved aggregation of individual claims under the product liability policy for asbestos, triggering excess covers.

The asbestos exclusion, introduced and generally implemented in 1985-86 was, in some cases, not applied to exposures of a given class of business if underwriters believed the class would not have asbestos exposures, such as contractors. The absence of the exclusion from the primary policies may create a disconnect if these carriers' reinsurance policies excluded asbestos exposures.

Unfortunately, as we know, some of these "no exposure" classes are generating asbestos claims.

The above two items indicate that past asbestos reinsurance ceded activity may not be representative of future activity, and we expect to observe a decrease in reinsurance ceded as a percentage of gross reserves.

To analyze this assertion, we evaluated published financial data as of year-end 2002 for the 21 primary companies and 11 reinsurance companies reflected in the above tables. We used the companies last five years of paid gross and paid ceded losses to indicate past activity, and used the companies' current gross and current ceded reserves to indicate future activity.

Over the past five years for the primary companies in our sample, ceded payments represent approximately 51 percent of gross payments, while the current reserve ratios imply that the primary companies are expecting that the ratio will fall to 40 percent in the future.

Similarly for the reinsurers in our sample, the ceded-to-gross payment ratio over the past five years has been 64 percent, while the reserve ratio is 52 percent.

Both of these ratios are trending in the expected direction. However, without knowing specifics about every company, it is not possible to determine if the absolute magnitude of the reserve ratio is a reasonable estimate of the future liability. Additionally, future recoveries may be reduced by current and future reinsurer financial difficulties, insolvencies and disputes.

This article used publicly available information to analyze some of the issues surrounding asbestos reinsurance collectibility. The published data, however, do not provide the detail necessary to make more than very broad, general statements.

We believe that the most appropriate approach to assess gross and net reserves for asbestos on a company specific basis is to model the asbestos exposures from the "ground up."

The "ground-up" approach uses detailed claim information from each insured, inflationary trends, coverage information from the primary insurer and cession information for the primary insurer. Due to the significant impact of claim size on reinsurance ceded, and the need to proceed in a cost-effective manner, insureds are generally divided into two categorieslarge defendants (many claimants, usually Tier 1 and 2 firms) and smaller defendants (fewer claimants, usually Tier 3 and higher).

For large defendants individually, and for smaller defendants on a "grouped" basis, the model projects a range of possible outcomes based on various reasonable assumptions (reasonable from a claims perspective, a legal perspective and an actuarial perspective). The primary company's coverage is then superimposed on the defendant's modeled results and, similarly, the reinsurance coverage is superimposed on the primary company's modeled results to obtain estimates of the gross liabilities and reinsurance recoverable.

For the "grouped" companies, one of the criteria is the policy limit provided. Our review of emerging claims suggests that one possible group is contractors with policy limits of less than $1 million and little or no reinsurance available. Another group might consist of retail stores.

Note that for all evaluations, we are using distributions of assumptions and thereby modeling distributions (ranges) of outcomes. Understanding the potential range of asbestos reserves is an important feature of a "ground-up" model.

To assess the extent to which gross incurred-but-not-reported claims may result in additional reinsurance recoveries, the primary company needs to assess the expected source of the IBNR claims and the implications on reinsurance recoveries. For example, if most of the IBNR is for asbestos claims expected from smaller defendants with primarily premises/operations exposure, future reinsurance recoveries as a percentage of gross reserves might be substantially less than historical recoveries as a percentage of gross losses paid.

The above modeling will provide more refined and transparent calculations of the gross and ceded net exposures and the ranges of the exposures. The approach will also provide consistency between gross and net reserve evaluations.

Unfortunately for the insurance industry, it appears that asbestos claims are continuing to emerge at a level higher than expected, resulting in reserve additions. Based on our analysis, it also appears that future reinsurance recoveries will be less than historical levels. Thus, as primary companies take reserve additions, they will retain more of the exposure and reinsurers may be impacted less severely.

Brian Z. Brown, FCAS, MAAA, and Claus S. Metzner FCAS, FSA, MAAA, are consulting actuaries for Milliman USA in Milwaukee, Wis.

Reproduced from National Underwriter Edition, May 10, 2004. Copyright 2004 by The National Underwriter Company in the serial publication. All rights reserved.Copyright in this article as an independent work may be held by the author.