Freddie Mac 2012 Annual Report Download - page 221

• UPB at origination;
• expected ability to partially mitigate losses through loan modification or other alternatives to foreclosure;
• expected proceeds from mortgage insurance contracts that are contractually attached to a loan or other credit
enhancements that were entered into contemporaneous with and in contemplation of a guarantee or loan purchase
transaction;
• expected repurchases of mortgage loans by seller/servicers;
• counterparty credit of mortgage insurers and seller/servicers;
• pre-foreclosure real estate taxes and insurance;
• estimated selling costs should the underlying property ultimately be sold; and
• trends in the timing of foreclosures.
For additional information on estimated current LTV ratios and single-family loan loss reserves, see “NOTE 4:
MORTGAGE LOANS AND LOAN LOSS RESERVES — Credit Quality of Mortgage Loans.”
Freddie Mac relies upon third-parties to provide primary servicing for the performing and non-performing loan
portfolio. At loan delivery, the seller provides us with the loan data, which includes loan characteristics and underwriting
information. Each month, the servicers provide us with monthly loan level servicing data, including delinquency and loss
information.
Certain loan servicing data is reported to us on a real-time basis, such as loan pay-offs and foreclosure events. However,
certain monthly servicing data, including delinquency status, is delivered on a one-month delay. For example, December loan
delinquency data delivered to Freddie Mac at the end of December or beginning of January reflects the loan delinquency
status related to the December 1 payment cycle. We incorporate the delinquency status data into our allowance for loan loss
calculation generally without adjustment for the one month delay.
Our single-family loan loss reserve default models are estimated based on the most recent 12 months of actual loan
performance data, including loan status and delinquency data reported by our servicers. The loan performance data provides
a loan level history of delinquency, foreclosures, foreclosure alternatives, modifications, and seller/servicer repurchases. Our
single-family loan loss reserve severity is estimated from the most recent six months of: (a) sales experience realized on our
distressed property dispositions; (b) mortgage insurance recoveries and pre-foreclosure expenses on our distressed properties
including REO, short sales, and third-party sales; and (c) recoveries due to seller/servicer repurchases. We use historical
trends in home prices in our single-family loan loss reserve process, primarily through the use of estimated current total LTV
ratios in our default models and through the use of recent home price sales experience in our severity estimate. However, we
do not use a forecast of trends in home prices in our single-family loan loss reserve process.
Our loan loss reserves reflect our best current estimates of incurred losses. Our loan loss reserve estimate includes
projections related to strategic loss mitigation activities, including loan modifications for troubled borrowers, and projections
of recoveries through repurchases by seller/servicers of defaulted loans due to failure to follow contractual underwriting
requirements at the time of the loan origination. These projections are based on our recent historical experience and current
business practices and require significant management judgment. We monitor our projections of recoveries through seller/
servicer repurchases to ensure that these projections are reasonable and consistent with our assessment of the credit capacity
of our seller/servicer counterparties. For loans where foreclosure is probable, impairment is measured on an aggregate basis
based upon an estimate of the underlying collateral value. At an individual loan level, our estimate also considers the effect
of historical home price changes on borrower behavior and the impact of our loss mitigation actions, including our loan
modification efforts.
Our reserve estimate also reflects our best projection of delinquencies we believe are likely to occur as a result of loss
events that have occurred through December 31, 2012 and 2011, respectively. However, the continued weakness in the
national housing market, the uncertainty in other macroeconomic factors, and uncertainty of the success of modification
efforts under HAMP and other loan workout programs, make forecasting of delinquency rates inherently imprecise.
We validate and update our models and factors to capture changes in actual loss experience, as well as the effects of
changes in underwriting practices and in our loss mitigation strategies. We also consider macroeconomic and other factors
that impact the quality of the loans underlying our portfolio including regional housing trends, applicable home price indices,
216 Freddie Mac