AIG 2014 Annual Report Download - page 198

ITEM 7 / CRITICAL ACCOUNTING ESTIMATES

181
In determining the actual carried reserves, we consider both the internal actuarial best estimate and
numerous other internal and external factors, including:
• an assessment of economic conditions including real GDP growth, inflation, employment rates or unemployment
duration, stock market volatility and changes in corporate bond spreads;
• changes in the legal, regulatory, judicial and social environment including changes in road safety, public health and
cleanup standards;
• changes in medical cost trends (inflation, intensity and utilization of medical services) and wage inflation trends
• underlying policy pricing, terms and conditions including attachment points and policy limits;
• claims handling processes and enhancements;
• third-party claims reviews that are periodically performed for key classes of claims such as toxic tort, environmental
and other complex casualty claims and
• third-party actuarial reviews that are periodically performed for key classes of business.
Loss reserve development can also be affected by commutations of assumed and ceded reinsurance agreements.
Actuarial and Other Methods for Major Classes of Business
In testing the reserves for each class of business, our actuaries determine the most appropriate actuarial methods. This
determination is based on a variety of factors including the nature of the claims associated with the class of business, such as
the frequency or severity of the claims. Other considerations include the loss development characteristics associated with the
claims, the volume of claim data available for the applicable class, and the applicability of various actuarial methods to the
class. In addition to determining the actuarial methods, the actuaries determine the appropriate loss reserve groupings of data.
For example, we write many unique subclasses of professional liability. For pricing or other purposes, it is appropriate to
evaluate the profitability of each subclass individually. However, for purposes of estimating the loss reserves for many classes
of business, we believe it is appropriate to combine the subclasses into larger groups to produce a greater degree of credibility
in the claims experience. This determination of data segmentation and actuarial methods is carefully considered for each class
of business. The segmentation and actuarial methods chosen are those which together are expected to produce the most
robust estimate of the loss reserves.
The actuarial methods we use for most long-tail casualty classes of business include loss development methods,
expected loss ratio methods, including “Bornhuetter Ferguson” methods described below, and frequency/severity
models. Loss development methods utilize the actual loss development patterns from prior accident years to project the
reported losses to an ultimate basis for subsequent accident years. Loss development methods generally are most appropriate
for classes of business which exhibit a stable pattern of loss development from one accident year to the next, and for which the
components of the classes have similar development characteristics. For example, property exposures would generally not be
combined into the same class as casualty exposures, and primary casualty exposures would generally not be combined into
the same class as excess casualty exposures. In 2014, we continued to refine our loss reserving techniques for the domestic
primary casualty classes of business and adopted further segmentations based on our analysis of the differing emerging loss
patterns for certain classes of insureds. We generally use expected loss ratio methods in cases where the reported loss data
lacks sufficient credibility to utilize loss development methods, such as for new classes of business or for long-tail classes at
early stages of loss development. Frequency/severity models may be used where sufficient frequency counts are available to
apply such approaches.
Expected loss ratio methods rely on the application of an expected loss ratio to the earned premium for the class of
business to determine the loss reserves. For example, an expected loss ratio of 70 percent applied to an earned premium
base of $10 million for a class of business would generate an ultimate loss estimate of $7 million. Subtracting any reported
paid losses and loss adjustment expenses would result in the indicated loss reserve for this class. Under the “Bornhuetter
Ferguson” methods, the expected loss ratio is applied only to the expected unreported portion of the losses. For example, for a
long-tail class of business for which only 10 percent of the losses are expected to be reported at the end of the accident year,
the expected loss ratio would be applied to the 90 percent of the losses still unreported. The actual reported losses at the end