Berkshire Hathaway 2015 Annual Report Download - page 102

Management’s Discussion and Analysis (Continued)
Property and casualty losses (Continued)
General Re (Continued)
We estimate IBNR reserves by reserve cell using the expected loss emergence patterns and the expected loss ratios. The
expected loss emergence patterns and expected loss ratios are the critical IBNR reserving assumptions and are updated annually.
Once the annual IBNR reserves are determined, we estimate the expected case loss emergence for the upcoming calendar year,
based on the prior year-end expected loss emergence patterns and expected loss ratios. The expected losses are then allocated
into interim estimates that are compared to actual reported losses in the subsequent year. This comparison provides a test of the
adequacy of prior year-end IBNR reserves and forms the basis for possibly changing IBNR reserve assumptions during the
course of the year.
During 2015 and 2014, we reduced net losses for prior years’ occurrences by $410 million. These reductions produced
corresponding increases in pre-tax earnings in each year.
In 2015, reported claims for prior years’ workers’ compensation losses were $178 million less than expected. However,
further analysis of the workers’ compensation reserve cells by segment indicated the need for maintaining IBNR reserves, which
precipitated an increase of $78 million in nominal IBNR reserve. After adjusting for changes in reserve discounts, the net
increase in workers’ compensation losses from prior years’ occurrences had a minimal impact on pre-tax earnings. To illustrate
the sensitivity of these assumptions on significant excess-of-loss workers’ compensation reserve cells, an increase of ten points
in the tail of the expected loss emergence pattern and an increase of ten percent in the expected loss ratios would produce a net
increase in our nominal IBNR reserves of approximately $844 million and $487 million on a discounted basis at December 31,
2015. An increase in discounted reserves would produce a corresponding decrease in pre-tax earnings. We believe it is
reasonably possible for these assumptions to increase at these rates.
Other casualty and general liability reported losses (excluding mass tort losses) developed favorably in 2015 relative to
expectations. However, casualty losses tend to be long-tail and it should not be assumed that favorable loss experience in a
given year will continue in the future. For our significant other casualty and general liability reserve cells, we estimate that an
increase of five points in the claim-tails of the expected loss emergence patterns and a five percent increase in expected loss
ratios (one percent for large international proportional reinsurance reserve cells) would produce a net increase in our nominal
IBNR reserves and a corresponding reduction in pre-tax earnings of approximately $1.1 billion. This amount includes changes
in assumptions used in certain U.K. motor annuity claims liabilities. We believe it is reasonably possible for these assumptions
to increase at these rates in any of the individual aforementioned reserve cells. However, given the diversification in worldwide
business, more likely outcomes are believed to be less than $1.1 billion.
In certain reserve cells within excess directors and officers and errors and omissions (“D&O and E&O”) coverages, IBNR
reserves are based on estimated ultimate losses without consideration of expected loss emergence patterns. For our large D&O
and E&O reserve cells, an increase of five points in the tail of the expected loss emergence pattern (for those cells where loss
emergence patterns are considered) and an increase of five percent in the expected loss ratios would produce a net increase in
nominal IBNR reserves and a corresponding reduction in pre-tax earnings of approximately $72 million. We believe it is
reasonably possible for these assumptions to increase at these rates.
Overall industry-wide loss experience data and informed judgment are used when internal loss data is of limited reliability,
such as in setting the estimates for mass tort, asbestos and hazardous waste (collectively, “mass tort”) claims. Estimating mass
tort losses is very difficult due to the changing legal environment. Although such reserves are believed to be adequate,
significant reserve increases may be required in the future if new exposures or claimants are identified, new claims are reported
or new theories of liability emerge. Mass tort net claims paid in 2015 were $83 million and we increased ultimate loss estimates
for such claims by $87 million. In addition to the previously described methodologies, we consider “survival ratios” based on
average net claim payments in recent years versus net unpaid losses as a rough guide to reserve adequacy. Our survival ratio
was approximately 15.1 years as of December 31, 2015 and 14.9 years as of December 31, 2014. The reinsurance industry’s
survival ratio for asbestos and pollution liabilities was approximately 14.0 years as of December 31, 2014.
In 2015, reported claims for prior years’ property loss events were less than expected and we reduced our estimated
ultimate liabilities by $260 million. However, the nature of property loss experience tends to be more volatile because of the
effect of catastrophes and large individual property losses.
100