AIG 2014 Annual Report Download - page 251

ITEM 8 / NOTE 5. FAIR VALUE MEASUREMENTS

234
We also incorporate our own risk of non-performance in the valuation of the embedded policy derivatives associated with
variable annuity and equity-indexed annuity and life contracts. Expected cash flows are discounted using the interest rate
swap curve (swap curve), which is commonly viewed as being consistent with the credit spreads for highly-rated financial
institutions (S&P AA-rated or above). A swap curve shows the fixed-rate leg of a non-complex swap against the floating rate
(for example, LIBOR) leg of a related tenor. The non-performance risk adjustment reflects a market participant’s view of our
claims-paying ability by incorporating an additional spread to the swap curve used to value embedded policy derivatives.
Super Senior Credit Default Swap Portfolio
We value CDS transactions written on the super senior risk layers of designated pools of debt securities or loans using internal
valuation models, third-party price estimates and market indices. The principal market was determined to be the market in
which super senior CDSs of this type and size would be transacted, or have been transacted, with the greatest volume or level
of activity. We have determined that the principal market participants, therefore, would consist of other large financial
institutions who participate in sophisticated over-the-counter derivatives markets. The specific valuation methodologies vary
based on the nature of the referenced obligations and availability of market prices.
The valuation of the super senior credit derivatives is complex because of the limited availability of market observable
information due to the lack of trading and price transparency in certain structured finance markets. These market conditions
have increased the reliance on management estimates and judgments in arriving at an estimate of fair value for financial
reporting purposes. Further, disparities in the valuation methodologies employed by market participants and the varying
judgments reached by such participants when assessing volatile markets have increased the likelihood that the various parties
to these instruments may arrive at significantly different estimates of their fair values.
Our valuation methodologies for the super senior CDS portfolio have evolved over time in response to market conditions and
the availability of market observable information. We have sought to calibrate the methodologies to available market
information and to review the assumptions of the methodologies on a regular basis.
Multi-sector CDO portfolios: We use a modified version of the Binomial Expansion Technique (BET) model to value our
CDS portfolio written on super senior tranches of multi-sector CDOs of ABS. The BET model was developed in 1996 by a
major rating agency to generate expected loss estimates for CDO tranches and derive a credit rating for those tranches, and
remains widely used.
We have adapted the BET model to estimate the price of the super senior risk layer or tranche of the CDO. We modified the
BET model to imply default probabilities from market prices for the underlying securities and not from rating agency
assumptions. To generate the estimate, the model uses the price estimates for the securities comprising the portfolio of a CDO
as an input and converts those estimates to credit spreads over current LIBOR-based interest rates. These credit spreads are
used to determine implied probabilities of default and expected losses on the underlying securities. This data is then
aggregated and used to estimate the expected cash flows of the super senior tranche of the CDO.
Prices for the individual securities held by a CDO are obtained in most cases from the CDO collateral managers, to the extent
available. CDO collateral managers provided market prices for 49 percent and 46 percent of the underlying securities used in
the valuation at December 31, 2014 and 2013. When a price for an individual security is not provided by a CDO collateral
manager, we derive the price through a pricing matrix using prices from CDO collateral managers for similar securities. Matrix
pricing is a mathematical technique used principally to value debt securities without relying exclusively on quoted prices for the
specific securities, but rather by relying on the relationship of the security to other benchmark quoted securities. Substantially
all of the CDO collateral managers who provided prices used dealer prices for all or part of the underlying securities, in some
cases supplemented by independent third-party valuation service providers.
The BET model also uses diversity scores, weighted average lives, recovery rates and discount rates. We employ a Monte
Carlo simulation to assist in quantifying the effect on the valuation of the CDO of the unique aspects of the CDO’s structure
such as triggers that divert cash flows to the most senior part of the capital structure. The Monte Carlo simulation is used to
determine whether an underlying security defaults in a given simulation scenario and, if it does, the security’s implied random
default time and expected loss. This information is used to project cash flow streams and to determine the expected losses of
the portfolio.