AIG 2009 Annual Report Download - page 102

American International Group, Inc., and Subsidiaries
approaches, including modeling estimates, ground-up claim analysis, loss evaluation reports from on-site field
adjusters, and market share estimates.
AIG’s loss reserve analyses do not calculate a range of loss reserve estimates. Because a large portion of the loss
reserves from AIG’s General Insurance business relates to longer-tail casualty classes of business driven by severity
rather than frequency of claims, such as excess casualty and D&O, developing a range around loss reserve estimates
would not be meaningful. Using the reserving methodologies described above, AIG’s actuaries determine their best
estimate of the required reserve and advise management of that amount. AIG then adjusts its aggregate carried
reserves as necessary so that the actual carried reserves as of December 31 reflect this best estimate.
Volatility of Reserve Estimates and Sensitivity Analyses
As described above, AIG uses numerous assumptions in determining its best estimate of reserves for each class of
business. The importance of any specific assumption can vary by both class of business and accident year. If actual
experience differs from key assumptions used in establishing reserves, there is potential for significant variation in the
development of loss reserves, particularly for long-tail casualty classes of business such as excess casualty, D&O or
primary and excess workers’ compensation. Set forth below is a sensitivity analysis that estimates the effect on the loss
reserve position of using alternative loss trend or loss development factor assumptions rather than those actually used
in determining AIG’s best estimates in the year-end loss reserve analyses in 2009. The analysis addresses each major
class of business for which a material deviation to AIG’s overall reserve position is believed reasonably possible, and
uses what AIG believes is a reasonably likely range of potential deviation for each class. There can be no assurance,
however, that actual reserve development will be consistent with either the original or the adjusted loss trend or loss
development factor assumptions, or that other assumptions made in the reserving process will not materially affect
reserve development for a particular class of business.
Excess Casualty: For the excess casualty class of business, the assumed loss cost trend was approximately five
percent. After evaluating the historical loss cost trends from prior accident years since the early 1990s, in AIG’s
judgment, it is reasonably likely that actual loss cost trends applicable to the year-end 2009 loss reserve review for
excess casualty will range from negative five percent to positive 15 percent, or approximately ten percent lower or
higher than the assumption actually utilized in the year-end 2009 reserve review. A ten percent change in the assumed
loss cost trend for excess casualty would cause approximately a $2.2 billion increase or a $1.6 billion decrease in the
net loss and loss expense reserve for this class of business. It should be emphasized that the ten percent deviations are
not considered the highest possible deviations that might be expected, but rather what is considered by AIG to reflect
a reasonably likely range of potential deviation. Actual loss cost trends in the early 1990s were negative for several
years, including amounts below the negative five percent cited above, whereas actual loss cost trends in the late 1990s
ran well into the double digits for several years, including amounts greater than the 15 percent cited above. Thus,
there can be no assurance that loss trends will not deviate by more than ten percent. The loss cost trend assumption is
critical for the excess casualty class of business due the long-tail nature of the claims and therefore is applied across
many accident years.
For the excess casualty class of business, the assumed loss development factors are also a key assumption. After
evaluating the historical loss development factors from prior accident years since the early 1990s, in AIG’s judgment,
it is reasonably likely that actual loss development factors will range from approximately 8.2 percent below those
actually utilized in the year-end 2009 reserve review to approximately 8.0 percent above those factors actually utilized.
If the loss development factor assumptions were changed by 8.2 percent and 8.0 percent, respectively, the net loss
reserves for the excess casualty class would decrease by approximately $1.3 billion under the lower assumptions or
increase by approximately $1.2 billion under the higher assumptions. Generally, actual historical loss development
factors are used to project future loss development. However there can be no assurance that future loss development
patterns will be the same as in the past, or that they will not deviate by more than the amounts illustrated above.
Moreover, as excess casualty is a long-tail class of business, any deviation in loss cost trends or in loss development
factors might not be discernible for an extended period of time subsequent to the recording of the initial loss reserve
estimates for any accident year. Thus, there is the potential for the reserves with respect to a number of accident years
to be significantly affected by changes in the loss cost trends or loss development factors that were initially relied upon
AIG 2009 Form 10-K 94