Travelers 2015 Annual Report Download - page 139

In addition, sizable or unique exposures are reviewed separately. These exposures include asbestos,
environmental, other mass torts, construction defect and large unique accounts that would otherwise
distort the analysis. These unique categories often require a very high degree of judgment and require
reserve analyses that do not rely on conventional actuarial methods.
Defense costs are also a part of the insured costs covered by liability policies and can be
significant, sometimes greater than the cost of the actual paid claims. For some products this risk is
mitigated by policy language such that the insured portion of defense costs is included in the policy
limit available to pay the claim. Such ‘‘defense within the limits’’ policies are most common for
‘‘claims-made’’ products. When defense costs are outside of the policy limits, the full amount of the
policy limit is available to pay claims and the amounts paid for defense costs have no contractual limit.
This line is typically the largest source of reserve estimate uncertainty in the United States
(excluding assumed reinsurance contracts covering the same risk). Major contributors to this reserve
estimate uncertainty include the reporting lag (i.e., the length of time between the event triggering
coverage and the actual reporting of the claim), the number of parties involved in the underlying tort
action, whether the ‘‘event’’ triggering coverage is confined to only one time period or is spread over
multiple time periods, the potential dollars involved (in the individual claim actions), whether such
claims were reasonably foreseeable and intended to be covered at the time the contracts were written
(i.e., coverage dispute potential), and the potential for mass claim actions. Claims with longer reporting
lags result in greater estimation uncertainty. This is especially true for alleged claims with a latency
feature, particularly where courts have ruled that coverage is spread over multiple policy years, hence
involving multiple defendants (and their insurers and reinsurers) and multiple policies (thereby
increasing the potential dollars involved and the underlying settlement complexity). Claims with long
latencies also increase the potential recognition lag (i.e., the lag between writing a type of policy in a
certain market and the recognition that such policies have potential mass tort and/or latent claim
exposure).
The amount of reserve estimate uncertainty also varies significantly by component for the general
liability product line. The components in this product line with the longest latency, longest reporting
lags, largest potential dollars involved and greatest claim settlement complexity are asbestos and
environmental. Components that include latency, reporting lag and/or complexity issues, but to a
materially lesser extent than asbestos and environmental, include construction defect and other mass
tort actions. Many components of general liability are not subject to material latency or claim
complexity risks and hence have materially less uncertainty than the previously mentioned components.
In general, components with shorter reporting lags, fewer parties involved in settlement negotiations,
only one policy potentially triggered per claim, fewer potential settlement dollars, reasonably
foreseeable (and stable) potential hazards/claims and no mass tort potential result in much less reserve
estimate uncertainty than components without those characteristics.
In addition to the conventional actuarial methods mentioned in the general discussion section, the
company utilizes various report year development and S-curve methods for the construction defect
components of this product line. The Construction Defect report year development analysis is
supplemented with projected claim counts and average values for IBNR claim counts. For components
with greater lags in claim reporting, such as excess and umbrella components of this product line, the
company relies more heavily on the BF method than on the paid and case incurred development
methods.
Examples of common risk factors, or perceptions thereof, that could change and, thus, affect the
required general liability reserves (beyond those included in the general discussion section) include:
General liability risk factors
• Changes in claim handling philosophies
139