Asbestos, Financial DisclosureAnd Credit Ratings

How do rating agencies assess an insurers exposure to asbestos?And how does that assessment impact an insurers credit ratings?

Those are two key questions Fitch Ratings has been asked morefrequently as insurers disclosures regarding their asbestosexposures and reserves have received more attention in recentperiods.

As a point of background, in July of 2002, Fitch came to severalconclusions when we published our first comprehensive asbestosstudy for U.S. insurers. First, we estimated that the industry wasunder-reserved for asbestos by up to $35 billion as of year-end2001, implying that significant reserving actions would be neededby a number of insurers in future periods. Given asbestos-relatedreserving action taken since last July by firms such as Travelers,ACE, Chubb, Munich Re and Employers Re, I dont think anyone wouldargue with our thought that the industry was, and still is,under-reserved.

Second, Fitch concluded that it is generally inappropriate toanalyze asbestos reserves on a basis that discounts for the timevalue of money.

Though clearly asbestos claims will be paid well into thefuture, which implies discounting may make sense, present valuetheory also says that a risk-adjusted discount rate should beused.

When discounting a liability, the risk is that the liabilitycould be larger than estimated. Thus, any risk premium should besubtracted from the nominal interest rate, such as a Treasury rate,to arrive at the appropriate discount rate. We believe the riskpremium needed to capture the volatility surrounding theuncertainty of asbestos claims could easily result in arisk-adjusted discount rate of zero.

Third, putting our first two thoughts together, we surmised thatthe appropriate survival ratio target for the industry was16-times–a level over 30 percent higher than the 12-times standardpreviously used in the industry. Recall that the survival ratio issimply net asbestos reserves divided by the average three-year netasbestos payments. All else equal, the higher the ratio, the moreadequate (or less deficient) are an insurers reserves.

Finally, Fitch said that in 2003, we would start to morestrictly apply our 16-times target into ratings. Simply put, to theextent an insurers year-end 2002 survival ratio is below 16-times,we will calculate the amount of additional reserves needed to reach16-times, and the difference will be our estimated deficiency.

We will reduce capital and surplus by the amount of thatdeficiency, and calculate pro-forma risk based capital (RBC) andother leverage ratios. To the extent these pro-forma ratios falloutside of our ratings parameters, the insurers ratings could belowered.

So, is asbestos analysis nothing more than looking at survivalratios?

No. For most insurers, there are many other steps, as welldiscuss below. But in some cases, when an insurers financialdisclosures are poor and management hasnt done its homework throughcomprehensive ground-up analysis, the 16-times survival ratiotarget will dominate our thinking.

Recall that the survival ratio was developed for a very simplereason–because the only substantive financial information publiclyreported by most insurers on their asbestos exposures has beenyear-end reserves and annual claim payments. This hasnt leftanalysts too many options. We could divide those two numbers, andthats about it.

However, we believe the industry is finally reaching a point atwhich financial disclosures are expanding, and managements analysissupporting their asbestos reserving levels is becoming robust.

This trend is best illustrated by Travelers announcement inJanuary of this year that it was adding $3.2 billion to its grossasbestos reserves. Not only are we comforted that the net 3-yearsurvival ratio ended up slightly above our standard, at just under17-times, but Travelers thoroughly reviewed how it built itsreserves based on an updated ground-up analysis, and publiclydisclosed the results of their study.

For example, in a 55-page presentation made available on its Website, Travelers provided an unprecedented level of disclosure onits asbestos study. Information was segmented by type of claimant,including the number of policyholders, a breakout of reserves andclaim payments, and a description of specific reserving methods.Travelers also broke out its unallocated incurred-but-not-reportedreserves.

This type of breakdown is critically important to a moredetailed analysis. For example, one of the criticisms of thesurvival ratio is that it penalizes companies who have reachedstructured settlements on large claims because such settlementsincrease paids (the denominator of the survival ratio) and lowerthe survival ratios. Travelers new disclosures allow forsettlements to be excluded from the survival ratio analysis.

Therefore, answers to the following questions will helpinfluence our view on how an insurers asbestos exposures couldimpact its credit rating:

Has a ground-up analysis been completed? Were third-partyexperts brought in to help validate the conclusions?

To what level of industry-wide reserves does the ground-up studyrelate? Is the industry estimate at least as conservative as Fitchs$35 billion deficiency estimate at year-end 2001?

Is reserving, payment and other data available on a segmentedbasis? Will updated segmented data be provided on a regularbasis?

What reserving assumptions were used for each segment? Howsensitive is the final answer to modest changes in theassumptions?

What portion of the gross asbestos exposure is ceded? What isthe level of provisioning for reinsurer bad debts?

What is the level of unallocated IBNR, and how was itdetermined?

Clearly, to the extent a robust and conservative analysis hasbeen made by management, supported by strong current and ongoingdisclosure, the more comfortable a rating agency will be.

Such detailed analysis is especially necessary if the survivalratio is below our 16-times standard.

If there is good reason to believe reserves are adequate atlevels below 16-times, we will likely back off from our standardfor that insurer. However, we generally will not back off from ourstandard based on simple anecdotal evidence presented by managementthat is not supported by hard data and analysis.

We also expect that Fitch will likely be able to developadditional benchmarking ratios as more insurers increase theirdisclosures and normative data becomes available. As just oneexample, the ratio of “IBNR to Total Asbestos Reserves” could proveinteresting.

Just like the survival ratio, every new ratio will have itsstrengths and weaknesses, but the more tools we have, the betterequipped we will be to make informed ratings decisions.

Ultimately our ability to expand our analysis beyond simplesurvival ratios rests on management teams throughout the industryexpanding public disclosures in a meaningful way based on robustgrounds up analyses.

A number of companies have commendably raised the standard forreview and disclosure of their asbestos exposures. Going forward,companies that do not provide similar information are likely toface increasing scrutiny and skepticism from rating agencies,investors and other interested parties.

Keith M. Buckley, CFA, is the managing director of FitchRatings in Chicago.

Reproduced from National Underwriter Edition, March 24, 2003.Copyright 2003 by The National Underwriter Company in the serialpublication. All rights reserved. Copyright in this article as anindependent work may be held by the author.