Fannie Mae 2002 Annual Report Download - page 58

56 FANNIE MAE 2002 ANNUAL REPORT
2. Regularly assessing the portfolio’s exposure to changes in
interest rates using a diverse set of analyses and measures.
Interest Rate Risk Measurement
We utilize a wide range of risk measures and analyses to
manage the interest rate risk inherent in the mortgage
portfolio. We categorize these risk measures and analyses
into three types: ongoing business risk measures and
analyses, run-off measures of the existing portfolio, and
stress test scenarios. The combination of ongoing business
and run-off risk measures and analyses present a
comprehensive picture of Fannie Mae’s current risk
position that we use for day-to-day risk management
decisions. Stress test scenarios provide information on
our risk to more extreme but lower probability events.
Our ongoing business risk measures and analyses include
net interest income at risk and repricing gap analyses. We
base net interest income at risk measures on the mortgage
portfolio as of a certain date plus projections of future
business activity. Future business activity includes
projected mortgage purchases and funding actions.
Management believes that ongoing business risk measures
and analyses provide a better perspective on the interest
rate risk we face as a continuing business and a more
comprehensive depiction of our risk profile than run-off
measures. However, they contain more assumption risk
due to the inherent uncertainty in projecting future
business activity.
Run-off measures of interest rate risk include duration,
convexity, and repricing gaps. We base run-off measures
on the mortgage portfolio as of a certain date without
incorporating future business activity. Run-off measures
provide an assessment of the interest rate risk of the
existing portfolio without the assumption risk inherent in
projecting future business activity. However, we believe it
is important to manage interest rate risk in the context of
ongoing business activity because future business is highly
probable and has a pronounced effect on our interest rate
risk profile.
Stress test scenarios include extreme movements in risk
factors on both ongoing business and run-off measures of
risk. We periodically measure and analyze the effects that
extreme movements in the level of interest rates and the
slope of the yield curve would have on the company’s risk
position. In addition, we evaluate stress scenarios that
include severe changes in expected prepayment speeds and
the level of interest rate volatility. While stress testing is an
integral part of our risk management process, the ongoing
business and run-off measures of risk are the primary
inputs in daily risk management decisions.
Many of our projections of mortgage cash flows in our
interest rate risk measures depend on our proprietary
prepayment models. While we are highly confident in
the quality of these models, we recognize the historical
patterns that serve as input for our models may not
continue in the future. The models contain many
assumptions, including those regarding borrower behavior
in certain interest rate environments and borrower
relocation rates. Other assumptions such as projections of
interest rates, shape of the yield curve, and interest rate
volatility are also critical components to our interest rate
risk measures. We maintain a research program to
constantly evaluate, update, and enhance these
assumptions, models, and analytical tools as appropriate to
reflect management’s best assessment of the environment.
•Net Interest Income at Risk
Net interest income at risk is our primary ongoing business
measure of interest rate risk. Net interest income at risk
measures the projected impact of changes in the level of
interest rates and the shape of the yield curve on the
mortgage portfolio’s expected or “base” core net interest
income over the immediate future one- and four-year
periods. To determine our base core net interest income,
we estimate core net interest income over a wide range of
interest rate environments using stochastic interest rate
simulations. Stochastic interest rate simulations are a
widely used statistical method to estimate the path and
pattern of interest rates. Our stochastic simulations
produce probability distributions of future interest rates
based on expected interest rate volatility and are based on
proprietary interest rate models. We generate several
hundred interest rate paths distributed around the current
Fannie Mae yield curve from these simulations. The
Fannie Mae yield curve represents market assumptions
regarding our expected cost of funds over a variety of
maturities and takes into account the risk premium on our
debt relative to benchmark interest rates. We project core
net interest income for four years along each path based on
the characteristics of the current mortgage portfolio and
projected future business activity. The expected or “base”
core net interest income is calculated based on the average
core net interest income across all simulation paths and
serves as the basis for determining our interest rate risk
profile. Our projections of future business activity used
in these simulations are reported to senior management
and our Board of Directors and provide the basis for
Fannie Mae’s current earnings forecasts.
We determine the amount of net interest income at risk by
assuming a sudden change or shock to the current yield
curve and repeating the simulation. We regularly evaluate