Freddie Mac 2006 Annual Report Download - page 51

We negotiate contracts with our customers in Guarantor Swap transactions based upon our view of the overall
economics of the transaction, considering the volume and types of mortgage loans to be delivered to us and our estimates of
the net present value of related future guarantee fees, credit costs and other associated cash Öows. However, the accounting
for our guarantee-related assets and liabilities is not determined at the contract level, rather it is determined separately for
each PC-related loan pool. We determine the initial fair value of the pool-level guarantee-related assets and liabilities using
methodologies that employ direct market-based information that may diÅer from the estimates we use to negotiate the
guarantee fee we charge customers. While our guarantee fees are subject to competitive pressure and we may enter into
transactions for which our expectations of economic returns are below our normal return thresholds (e.g., to achieve our
aÅordable housing goals or maintain our market share), we expect the vast majority of our Guarantor Swap transactions
will generate positive economic returns over the lives of the related PCs.
For each loan pool created, we compare the initial fair value of the related Guarantee obligation to the initial fair value
of the related Guarantee asset and credit enhancement-related assets. If the Guarantee obligation is greater than the
Guarantee asset, we immediately recognize a loss equal to the diÅerence with respect to that pool. If the Guarantee
obligation is less than the Guarantee asset, no initial gain is recorded; rather, guarantee income equal to the diÅerence is
deferred as an addition to the Guarantee obligation and is recognized as that liability is amortized. Accordingly, a Guarantor
Swap transaction may result in some loan pools for which a loss is recognized immediately in earnings and other loan pools
where guarantee income is deferred. We record these losses as Losses on certain credit guarantees.
In 2006 and 2005, we realized losses of $350 million and $234 million, respectively, on certain Guarantor Swap
transactions entered into during those years. The increase in these losses was driven by a combination of higher expected
future credit costs and competitive pressure on guarantee fees. In addition, our Guarantee asset associated with certain non-
traditional mortgage products, including interest-only loans and option ARMs, is subject to a lower market valuation than
traditional mortgage products due to the lower liquidity or corresponding lack of observable market prices for the associated
cash Öows.
Losses on non-performing loans repurchased from the mortgage loan pools underlying PCs and Structured Securities
held by third parties occur when the carrying value of the repurchased loan, net of any allocated loan loss reserve, exceeds
the estimated fair value of the loan. Increases in market interest rates and declining market values for delinquent loans led
to the recognition of losses of $126 million in 2006.
Losses on certain credit guarantees increased during 2005 when compared to 2004, as a result of the initial application
of our approach for determining the fair values of our guarantee-related assets and liabilities at inception. This approach
uses more market-based information and is discussed in ""NOTE 2: TRANSFERS OF SECURITIZED INTERESTS IN
MORTGAGE-RELATED ASSETS'' to our consolidated Ñnancial statements. We also realized losses in 2005 and to a
lesser extent in 2006, as a result of our eÅorts to acquire business at competitive prices in order to meet the aÅordable
housing goals and subgoals established by HUD.
Housing Tax Credit Partnerships
Operating losses of our housing tax credit partnerships, which are recorded as a component of Non-interest expense,
have increased over the last three years as our investments in these partnerships have increased. The increased investment
in housing tax credit partnerships generated related tax beneÑts, which consisted of tax credits and tax deductible operating
losses. Tax beneÑts associated with our investments in housing tax credit partnerships reduced Income tax expense by
$603 million, $476 million and $378 million for 2006, 2005 and 2004, respectively. See ""Income Tax Expense (BeneÑt)'' for
additional information about the impact of these investments on our income tax expense.
Other Expenses
In April 2006, we reached an agreement in principle to settle the securities class action and shareholder derivative
lawsuits relating to our restatement. The settlement became Ñnal in November 2006. In 2005, we recorded expenses of
$339 million to increase our reserves for legal settlements, including this settlement, net of expected insurance proceeds. See
""NOTE 13: LEGAL CONTINGENCIES'' to our consolidated Ñnancial statements for more information.
Income Tax Expense (BeneÑt)
For 2006, 2005 and 2004, our eÅective tax rates were (5.1) percent, 14.4 percent and 21.2 percent, respectively. The
decrease in the eÅective tax rate over the past three years is primarily due to the decline in pre-tax income, the year-over-
year increases in tax credits related to our investments in low-income housing tax credit partnerships and interest earned on
tax-exempt housing-related securities. We expect tax credits resulting from our investments in housing tax credit
partnerships to grow in the future. However, our ability to use all of the tax credits generated by existing or future
investments in housing tax credit partnerships to reduce our federal income tax liability may be limited.
39 Freddie Mac