Freddie Mac 2004 Annual Report Download - page 135

2002, indicating strong credit quality borrowers. In particular, the percentage of the mortgage loans with an
available FICO score of 740 or greater in the Total mortgage portfolio has remained at 44 percent at
December 31, 2004 and 2003 and increased from 39 percent at December 31, 2002.
Loan Purpose. Mortgage loan purpose indicates how the borrower intends to use the funds from a
mortgage loan. The three general categories are: purchase, cash-out reÑnance, or other reÑnance. In a
purchase transaction, funds are used to acquire a property. In a cash-out reÑnance transaction, in addition to
paying oÅ an existing Ñrst mortgage lien, the borrowers obtain additional funds that may be used for other
purposes, including paying oÅ subordinate mortgage liens and providing unrestricted cash proceeds to the
borrower. In other reÑnance transactions, the funds are used to pay oÅ an existing Ñrst mortgage lien and may
be used in limited amounts for certain speciÑed purposes; such reÑnances are generally referred to as ""no
cash-out'' or ""rate and term'' reÑnances. Other reÑnance transactions also include reÑnance mortgages for
which the delivery data provided was not suÇcient for us to determine whether the mortgage was a cash-out or
a no cash-out reÑnance transaction. From a risk perspective, purchase transactions have the lowest likelihood
of default (all else being equal), followed by no-cash out reÑnances, then cash out reÑnances. As a practical
matter, however, no-cash out reÑnances tend to have lower loan-to-value ratios and higher credit scores than
purchase transactions and as such, have better overall performance than purchase transactions. While a
reduction in interest rates in 2003 increased the proportion of reÑnance mortgage loans in the Total mortgage
portfolio from a total of 66 percent at December 31, 2002 to 75 percent at December 31, 2003, an increase in
interest rates in 2004 slightly decreased the proportion of reÑnance mortgage loans in the Total mortgage
portfolio to a total of 72 percent at December 31, 2004.
Property Type. Single-family mortgage loans are deÑned as mortgages secured by housing with up to
four living units. Mortgages on one-unit properties tend to have lower credit risk than mortgages on multiple-
unit properties. The proportion of one-unit properties in the Total mortgage portfolio remained the same over
the past three years, accounting for 97 percent at December 31, 2004, 2003 and 2002.
Occupancy Type. Borrowers may purchase a home as a primary residence, second/vacation home or
investment property that is typically a rental property. Mortgage loans on properties occupied by the borrower
as a primary or secondary residence tend to have a lower credit risk than mortgages on investment properties.
The proportion of primary and secondary residences in the Total mortgage portfolio remained the same over
the past three years, accounting for 97 percent at December 31, 2004, 2003 and 2002.
Geographic Concentration. Since our business model involves purchasing mortgages from every
geographic region in the U.S., we maintain a geographically diverse mortgage portfolio. This diversiÑcation
provides protection from changing local and economic conditions. See ""NOTE 17: CONCENTRATION OF
CREDIT AND OTHER RISKS'' to our consolidated Ñnancial statements for more information concerning
the distribution of our Total mortgage portfolio (excluding non-Freddie Mac mortgage-related securities and
that portion of Structured Securities that is backed by Ginnie Mae CertiÑcates) by geographic region. Our
Total mortgage portfolio's geographic distribution was relatively stable from 2003 to 2004, and remains
broadly diversiÑed across these regions.
Loss Mitigation Activities. Within our credit portfolio, we expect and price for some mortgage loans to
become non-performing due to changes in general economic conditions, changes in the Ñnancial status of
individual borrowers or other factors.
Table 63 summarizes our non-performing assets. The increase in our non-performing assets from 2000
through 2003 was primarily driven by higher delinquencies associated with our alternative collateral deals.
While these delinquencies result in higher levels of non-performing assets, we have limited loss exposure due
to the credit enhancements associated with these securities. Prior to 2004, alternative collateral deals consisted
only of Structured Securities backed by non-agency securities, which were primarily backed by subprime
mortgage loans; and to a lesser extent, FHA/VA loans and home equity loans. Beginning in 2004, however,
certain alternative collateral deals were backed by prime mortgage loans.
Freddie Mac
123