AIG 2013 Annual Report Download - page 220

The following paragraphs describe the methods we use to measure fair value on a recurring basis for super senior
credit default swaps classified in Level 3. See Note 5 to the Consolidated Financial Statements for discussion of the
valuation methodologies for other assets classified in Level 3, including certain fixed maturity securities and certain
other invested assets, as well as a discussion of transfers of Level 3 assets and liabilities.
Certain entities included in GCM wrote credit protection on the super senior risk layer of collateralized loan
obligations (CLOs), multi-sector CDOs and diversified portfolios of corporate debt and prime residential mortgages
through 2006. In these transactions, we are at risk of credit performance on the super senior risk layer related to
such assets.
See Notes 5 and 11 to the Consolidated Financial Statements for information about the regulatory capital, multi-
sector CDO, corporate debt/ CLO and other portfolios.
We utilize sensitivity analyses that estimate the effects of using alternative pricing and other key inputs on our
calculation of the unrealized market valuation loss related to the super senior credit default swap portfolio. For the
purposes of estimating sensitivities for the super senior multi-sector CDO credit default swap portfolio, the change in
valuation derived using the Binomial Expansion Technique (BET) model is used to estimate the change in the fair
value of the derivative liability. Of the total $3.3 billion net notional amount of CDS written on multi-sector CDOs
outstanding at December 31, 2013, a BET value is available for $2.2 billion net notional amount. No BET value is
determined for $1.1 billion of CDS written on European multi-sector CDOs because prices on the underlying
securities held by these CDOs are not provided by collateral managers; instead these CDS are valued using
counterparty prices. Therefore, sensitivities disclosed below apply only to the net notional amount of $2.2 billion.
The following table presents key inputs used in the BET model, and the potential increase (decrease) to the
fair value of the derivative liability by ABS category at December 31, 2013 corresponding to changes in these
key inputs:
Bond prices 50 points Increase of 5 points $ (99) $ (2) $ (5) $ (44) $ (32) $ (8) $ (8)
Decrease of 5
points 107 2 5 41 39 8 12
Weighted Increase of 1 year 7 – – 4 3 – –
average life 5.63 years Decrease of 1 year (8) – – (5) (2) – (1)
Recovery rates 17% Increase of 10% (7) – (1) (4) (1) (1) –
Decrease of 10% 9 – 1 5 1 1 1
Diversity score(a) 13 Increase of 5 (3)
Decrease of 5 8
Discount curve(b) N/A Increase of 100bps 2
(a) The diversity score is an input at the CDO level. A calculation of sensitivity to this input by type of security is not possible.
(b) The discount curve is an input at the CDO level. A calculation of sensitivity to this input by type of security is not possible. Furthermore, for this
input it is not possible to disclose a weighted average input because a discount curve consists of a series of data points.
These results are calculated by stressing a particular assumption independently of changes in any other assumption.
No assurance can be given that the actual levels of the key inputs will not exceed, perhaps significantly, the ranges
assumed by us for purposes of the above analysis. No assumption should be made that results calculated from the
use of other changes in these key inputs can be interpolated or extrapolated from the results set forth above.
Super Senior Credit Default Swap Portfolio
..................................................................................................................................................................................................................................
AIG 2013 Form 10-K202
ITEM 7 / CRITICAL ACCOUNTING ESTIMATES
Increase (Decrease) to Fair Value of Derivative Liability
Average
Inputs Used at Entire RMBS RMBS RMBS
(dollars in millions) December 31, 2013 Change Portfolio Prime Alt-A Subprime CMBS CDOs Other
..................................................................................................................................................................................
............................................................................................................................................................................................