In fact, supplemental analyses of the line, using different cuts of data from National Underwriter Insurance Data Services, show results that have improved markedly. These analyses attempt to remove the impact of prior-year exposures from asbestos and other issues, revealing that results for products liability policies written in recent years are comparable to overall industry results for all lines of business--not much better, but certainly not the worst.

The chart on page 13, like summaries presented for other lines in our "Data Insights" series, shows two years of calendar-year loss ratios (direct and net) and net combined ratios based on information for the Insurance Expense Exhibits of insurers annual statements.

The fact that the 2004 net combined ratio for the industry of 174 is so much higher than that for the top 10 companies--142--is the first clue that the analysis of products liability is not straightforward.

Part of the discrepancy is explained by results for one insurer that is not included in the chart--Allstate, which stopped writing commercial lines like products a decade ago. With no premiums in recent years but loss reserve changes of more than $300 million in 2004 for its discontinued products line, Allstate added 32 points to industrywide loss and combined ratios for products liability.

But even current writers have old-year exposures from asbestos and other issues impacting their calendar-year loss and combined ratios--which, by definition, capture reserve changes from prior accident years. While insurers individually report asbestos exposures in an annual statement footnote--Note 33--they do not provide this information by line so that products liability results adjusted to exclude the distortion of asbestos charges can be easily derived.

Claus Metzner, a consulting actuary for Milliman in Milwaukee who specializes in the analysis of asbestos reserves, responded to a recent NU inquiry, speculating that while non-products asbestos losses have increased in recent years, roughly two-thirds of the exposure is in the products line in recent calendar years.

Armed with this information, NU did a quick calculation using Note 33 information for one of the largest products liability insurers--St. Paul Travelers. Eliminating two-thirds of the insurer's calendar-year incurred loss and loss adjustment expenses for asbestos in 2004 from industrywide figures drops the combined ratio another 21 points.

While a 121 combined ratio (minus Allstate and St. Paul Traveler's asbestos losses) is still nothing to cheer about, similar adjustments for other companies would likely reduce the ratio further. In fact, two alternative analyses based on more readily available information get us closer to a profit story for products liability.

The first compares calendar-year pure loss ratios for insurers writing occurrence-based policies with those writing claims-made policies--policies with coverage terms that respond to recent exposures and not prior acts. (Refer to Charts 1 and 2 with this article.) While the volume of business for claims-made insurers is much smaller, the results are clearly better, with nearly 100 points separating six-year average loss ratios for the two groups.

The difference, however, is not fully explained by the policy forms. Specialty insurers dominate the claims-made group, and in both groups specialists outperform standard carriers.

Our final chart shows loss and loss adjustment expense ratios for products liability on an accident-year basis, which matches losses with the years in which they occurred. The ratios, compiled from Schedule P of insurer annual statements, have clearly improved over the past six years, with a 2004 ratio for the largest groups--65.6--that is less than half the ratio for 1999--132.9.

This last analysis assumes insurers have correctly estimated loss reserves for recent years--a difficult task because these years are not mature. Still, the products liability loss and loss adjustment expense ratio for 2004 for the largest insurers averaged 0.7 points better than the comparable ratio for all lines combined--66.3 for the same group of insurers.

Data used for the analysis prepared by NU's editorial staff is the NAIC Annual Statement Database via National Underwriter Insurance Data Services/ Highline Media. For information, contact Chris Rogers at 617-441-5976.

Flag: Chart 1

Head: Loss Nightmare

Products Liability Occurrence--Net Pure Loss Ratios From Underwriting and Investment Exhibit

Caption:

The pure loss ratio portion of insurer combined ratio--ranging from 69 to 290--paints a dismal picture for products liability. Loss adjustment and underwriting expenses added anywhere from 70-to-100 more points in the last six years.

Flag: Chart 2:

Head: The Same Line?

Products Liability Claims Made--Net Pure Loss Ratios From Underwriting and Investment Exhibit

Caption:

Products liability loss ratios for policies written on a claims-made basis, which limit the impact of losses from prior acts, are clearly better than ratios for occurrence-based policies.

Flag: Chart 3:

Head: Improvements Revealed

Products Liability Occurrence--Net Loss & LAE Ratios By Accident Year From Schedule P

Caption:

On an accident-basis, products liability loss and LAE ratios dropped from a high of 133 in accident-year 1999 to 66 in 2004 for 10 insurers representing two-thirds of the market.