Freddie Mac 2005 Annual Report Download - page 63

long-term fair value of the Retained portfolio. This belief is based on our expectation that diÅerences between the
prepayments forecasted by our models and the actual prepayments we will experience are not likely to be signiÑcant.
During the year ended December 31, 2005, the fair value of net assets attributable to common stockholders, before
capital transactions, increased by $1.0 billion. We estimate that this $1.0 billion is net of a decrease of approximately
$1.3 billion due to the net widening of mortgage-to-debt OAS.
How we estimate the impact of mortgage-to-debt OAS on fair value results
The method we have chosen to estimate the OAS impact is to fully revalue the fair value of identiÑed Ñnancial
instruments for a given period using the OAS level from the end of the previous period and subtract the revalued amount
from the estimated fair value of those instruments. We make this calculation as of the end of each month and sum these
monthly results into quarterly and annual estimates. To achieve consistency month-to-month, we use the smaller unpaid
principal balance for a given instrument between months so that we are measuring the OAS impact on constant positions,
with newly acquired positions excluded entirely during the month of acquisition.
For certain Ñnancial instruments in the Retained portfolio that aÅect our total change in fair value of net assets, we did
not estimate the impact of changes in OAS on fair value. We did not estimate the impact of changes in OAS for single-
family and multifamily whole loans because we do not have a reliable methodology for estimating OAS impacts on these
loans at this time. We did not estimate the impact of changes in OAS for certain other instruments, including mortgage
revenue bonds, other securities and LIBOR-based derivatives, because an OAS measured in relation to LIBOR is not
relevant for these instruments. The Retained portfolio instruments for which we did not estimate an impact of the changes
in OAS represent approximately 17 percent of our total Retained portfolio. The funding instruments (including preferred
stock) for which we did not estimate an impact of the changes in OAS represent approximately 9 percent of our total debt
and preferred stock securities. The majority of this 9 percent was short-term debt instruments with maturities less than
thirty days.
The impact of changes in OAS on fair value should be understood as an estimate rather than a precise measurement. To
estimate the impact of OAS, we use models that involve the forecast of interest rates, prepayment behavior and other inputs.
We also make assumptions about a variety of factors, including macroeconomic and security-speciÑc data, interest-rate
paths, cash Öows and prepayment rates. We use these models and assumptions in running our business, and we rely on many
of the models in producing our Ñnancial statements and measuring, managing and reporting interest-rate and other market
risks. The use of diÅerent estimation methods or the application of diÅerent assumptions could result in a materially
diÅerent estimate.
Understanding our estimate of the impact of mortgage-to-debt OAS on fair value results
A number of important qualiÑcations apply to our disclosed estimates. The estimated impact of the change in option-
adjusted spreads on the fair value of our net assets in any given period does not depend on other components of the change
in fair value. Although the net fair value of our Ñnancial instruments will generally move toward their par values as the
instruments approach maturity, investors should not expect that the eÅect of past changes in OAS will necessarily reverse
through future changes in OAS. To the extent that actual prepayment or interest rate distributions diÅer from the forecasts
contemplated in our models, changes in values reÖected in mortgage-to-debt OAS may not be recovered in fair value
returns at a later date.
When the OAS on a given asset widens, the fair value of that asset will typically decline, all other things being equal.
However, we believe such OAS widening has the eÅect of increasing the likelihood that, in future periods, we will recognize
income at a higher spread on this existing asset. The reverse is true when the OAS on a given asset tightens Ì current
period fair values for that asset typically increase due to the tightening in OAS, while future income recognized on the asset
is more likely to be earned at a reduced spread. Although a widening of OAS is generally accompanied by lower current
period fair values, it can also provide us with greater opportunity to purchase new assets for our Retained portfolio at the
wider mortgage-to-debt OAS. (Again, the reverse can be true when OAS tightens.)
For these reasons, our estimate of the impact of the change in OAS provides information regarding one component of
the change in fair value for the particular period being evaluated, but results for a single period should not be used to
extrapolate long-term fair value returns. We believe the potential fair value return of our business over the long term
depends primarily on our ability to add new assets at attractive mortgage-to-debt OAS and to eÅectively manage over time
the risks associated with these assets, as well as those of our existing portfolio to ensure that we realize anticipated returns
on our business. In other words, to capture the fair value returns we expect, we have to apply accurate estimates of future
47 Freddie Mac