AIG 2013 Annual Report Download - page 217

judgment and can have a material impact on DAC amortization. For fixed deferred annuity contracts, the future
spread between investment income and interest credited to policyholders is a significant judgment, particularly in a
low interest rate environment.
If the assumptions used for estimated gross profits change significantly, DAC and related reserves, including VOBA,
SIA, guaranteed benefit reserves and unearned revenue reserves (URR), are recalculated using the new
assumptions, and any resulting adjustment is included in income. Updating such assumptions may result in
acceleration of amortization in some products and deceleration of amortization in other products. In the third quarter
of 2013, we completed our comprehensive annual review and update of estimated gross profit assumptions used to
amortize DAC and related items for our investment-oriented products. The result of this review was a $118 million
net increase in pre-tax operating income in 2013, which included a $198 million net increase in our Retail operating
segment and an $80 million decrease in our Institutional operating segment. The net increase in Retail pre-tax
operating income was primarily due to a favorable adjustment in our Fixed Annuities product line from updated
spread assumptions due to active management of crediting rates and higher future investment yields than those
previously assumed. In the Life Insurance and A&H, Retirement Income Solutions and Group Retirement product
lines, the update of assumptions for variable annuity spreads, surrender rates, and life insurance mortality had an
unfavorable impact on pre-tax operating income. The life insurance mortality assumptions, while unfavorable
compared to the previous assumption set, are still within pricing expectations.
The $118 million increase in pre-tax operating income to reflect updated assumptions was comprised of three
favorable developments, a $98 million net decrease in DAC amortization expense, a $61 million decrease in SIA
amortization expense within Interest credited to policyholder account balances, and a $28 million increase in
unearned revenue amortization within Policy fees; partially offset by a $69 million increase in Future policy benefits
for life and health insurance contracts.
In estimating future gross profits for variable annuity products, a long-term annual asset growth assumption of 8.5%
(before expenses that reduce the asset base from which future fees are projected) is applied to estimate the future
growth in assets and related asset-based fees. In determining the asset growth rate, the effect of short-term
fluctuations in the equity markets is partially mitigated through the use of a ‘‘reversion to the mean’’ methodology,
whereby short-term asset growth above or below the long-term annual rate assumption impact the growth
assumption applied to the five-year period subsequent to the current balance sheet date. The reversion to the mean
methodology allows us to maintain our long-term growth assumptions, while also giving consideration to the effect of
actual investment performance. When actual performance significantly deviates from the annual long-term growth
assumption, as evidenced by growth assumptions for the five-year reversion to the mean period falling below a
certain rate (floor) or above a certain rate (cap) for a sustained period, judgment may be applied to revise or
‘‘unlock’’ the growth rate assumptions to be used for both the five-year reversion to the mean period as well as the
long-term annual growth assumption applied to subsequent periods. The use of a reversion to the mean assumption
is common within the industry; however, the parameters used in the methodology are subject to judgment and vary
within the industry.
In the fourth quarter of 2013, we revised the growth rate assumptions for the five-year reversion to the mean period
for the Group Retirement product line, because annual growth assumptions indicated for this period had fallen below
our floor of zero percent due to the recent favorable performance of equity markets. For this five-year reversion to
the mean period, the growth rate assumption was adjusted to a point between the long-term growth rate assumption
and zero percent. This adjustment, which increased DAC by $31 million, increased SIA by $2 million and reduced the
GMDB liability by $2 million, was recorded as a decrease in current period amortization expense, and increased
pre-tax income by $35 million in 2013. Had we readjusted the growth rate assumption for the five-year reversion to
the mean period to use the long-term rate assumption of 8.5%, pre-tax income would have been higher by
approximately $30 million. Conversely, had the growth rate assumption for the five-year reversion to the mean period
been readjusted to a floor of zero percent, pre-tax income would have been lower by approximately $30 million. For
variable annuities in our Retirement Income Solutions product line, the assumed annual growth rate remained above
zero percent for the five-year reversion to the mean period so it did not meet our criteria for adjustment; however,
additional favorable equity market performance in excess of long-term assumptions could also result in ‘‘unlocking’’ in
this product line in future periods with a positive effect on pre-tax income in the period of the unlocking.
The following table summarizes the sensitivity of changes in certain assumptions in the amortization of
DAC/SIA, guaranteed benefit reserves and unearned revenue liability and the related hypothetical impact on
year-end 2013 balances. The effect of changes in net investment spread primarily affects our Fixed Annuities
..................................................................................................................................................................................................................................
AIG 2013 Form 10-K 199
ITEM 7 / CRITICAL ACCOUNTING ESTIMATES
..................................................................................................................................................................................