Fannie Mae 2013 Annual Report Download - page 62

57
In many cases, our accounting policies and methods, which are fundamental to how we report our financial condition and
results of operations, require management to make judgments and estimates about matters that are inherently uncertain.
Management also relies on models in making these estimates.
Our accounting policies and methods are fundamental to how we record and report our financial condition and results of
operations. Our management must exercise judgment in applying many of these accounting policies and methods so that
these policies and methods comply with GAAP and reflect management’s judgment of the most appropriate manner to report
our financial condition and results of operations. In some cases, management must select the appropriate accounting policy or
method from two or more alternatives, any of which might be reasonable under the circumstances but might affect the
amounts of assets, liabilities, revenues and expenses that we report. See “Note 1, Summary of Significant Accounting
Policies” for a description of our significant accounting policies.
We have identified some of our accounting policies as being critical to the presentation of our financial condition and results
of operations. These accounting policies are described in “MD&A—Critical Accounting Policies and Estimates.” We believe
these policies are critical because they require management to make particularly subjective or complex judgments about
matters that are inherently uncertain and because of the likelihood that materially different amounts would be reported under
different conditions or using different assumptions.
Because our financial statements involve estimates for amounts that are very large, even a small change in the estimate can
have a significant impact for the reporting period. For example, because our total loss reserves are so large, even a change
that has a small impact relative to the size of our loss reserves can have a meaningful impact on our results for the quarter in
which we make the change.
Due to the complexity of the critical accounting policies we have identified, our accounting methods relating to these policies
involve substantial use of models. Models are inherently imperfect predictors of actual results because they are based on
assumptions, including assumptions about future events. Our models may not include assumptions that reflect very positive
or very negative market conditions and, accordingly, our actual results could differ significantly from those generated by our
models. As a result of the above factors, the estimates that we use to prepare our financial statements, as well as our estimates
of our future results of operations, may be inaccurate, perhaps significantly.
Failure of our models to produce reliable results may adversely affect our ability to manage risk and make effective
business decisions.
We make significant use of quantitative models to measure and monitor our risk exposures and to manage our business. For
example, we use models to measure and monitor our exposures to interest rate, credit and market risks, and to forecast credit
losses. The information provided by these models is used in making business decisions relating to strategies, initiatives,
transactions, pricing and products.
Models are inherently imperfect predictors of actual results because they are based on historical data and assumptions
regarding factors such as future loan demand, borrower behavior, creditworthiness and home price trends. Other potential
sources of inaccurate or inappropriate model results include errors in computer code, bad data, misuse of data, or use of a
model for a purpose outside the scope of the model’s design. Modeling often assumes that historical data or experience can
be relied upon as a basis for forecasting future events, an assumption that may be especially tenuous in the face of
unprecedented events.
Given the challenges of predicting future behavior, management judgment is used at every stage of the modeling process,
from model design decisions regarding core underlying assumptions, to interpreting and applying final model output. To
control for these inherent imperfections, our primary models are vetted by an independent model risk management team
within our Enterprise Risk Division.
When market conditions change quickly and in unforeseen ways, there is an increased risk that the model assumptions and
data inputs for our models are not representative of the most recent market conditions. Under such circumstances, we must
rely on management judgment to make adjustments or overrides to our models. A formal model update is typically an
extensive process that involves basic research, testing, independent validation and production implementation. In a rapidly
changing environment, it may not be possible to update existing models quickly enough to properly account for the most
recently available data and events. Management adjustments to modeled results are applied within the confines of the
governance structure provided by a combination of our model risk management team and our business, finance and risk
committees.
If our models fail to produce reliable results on an ongoing basis, we may not make appropriate risk management decisions,
including decisions affecting loan purchases, management of credit losses, guaranty fee pricing, asset and liability
management and the management of our net worth. Any of these decisions could adversely affect our businesses, results of