Freddie Mac 2014 Annual Report Download - page 144

139 Freddie Mac
We estimate the sensitivity to changes in interest rates of the fair value of all financial assets, liabilities, and derivatives
on a pre-tax basis. We also take into account the cash flows related to certain credit guarantee-related items, including buy-ups
and expected gains or losses due to net interest from float. In making these calculations, we do not consider the sensitivity to
interest-rate changes of the following assets and liabilities:
• Credit guarantee activities. We do not consider the sensitivity of the fair value of credit guarantee activities to changes
in interest rates except for the guarantee-related items mentioned above (i.e., buy-ups and float), because we do not
actively manage the change in the fair value of our guarantee business that is attributable to changes in interest rates.
We do not believe that periodic changes in fair value due to movements in interest rates are the best indication of the
long-term value of our guarantee business because these changes do not take into account the potential for new future
guarantee business activity.
• Other assets with minimal interest-rate sensitivity. We do not include other assets, primarily non-financial instruments
such as fixed assets and REO, because we estimate their impact on PMVS and duration gap to be minimal.
Limitations of Market Risk Measures
Our PMVS and duration gap estimates are determined using models that involve our judgment of interest-rate and
prepayment assumptions. While we believe that PMVS and duration gap are useful risk management tools, they should be
understood as estimates rather than as precise measurements. There could be times when we hedge differently than our model
estimates during the period (i.e., when we are making changes or market updates to these models). While PMVS and duration
gap estimate our exposure to changes in interest rates, they do not capture the potential effect of certain other market risks, such
as changes in volatility and spread risk. The effect of these other market risks can be significant.
There are inherent limitations in any methodology used to estimate exposure to changes in market interest rates. Our
sensitivity analyses for PMVS and duration gap contemplate only certain movements in interest rates and are performed at a
particular point in time based on the estimated fair value of our existing portfolio. These sensitivity analyses do not consider
other factors that may have a significant effect on our financial instruments, most notably business activities and strategic
actions that management may take in the future to manage interest-rate risk. These analyses are not intended to provide precise
forecasts of the effect a change in market interest rates would have on the estimated fair value of our net assets.
In addition, it has been more difficult in recent years to measure and manage the interest-rate risk related to mortgage
assets as risk for prepayment model error remains high due to the low interest rate environment and uncertainty regarding
default rates, unemployment, government policy changes and programs, loan modifications, and the volatility and impact of
home price movements on mortgage durations. Misestimation of prepayments, resulting in over or under hedging of interest-
rate risk, could result in significant economic losses and have an adverse impact on earnings. In addition, this misestimation
could result in realized losses upon the sale of assets.
Duration Gap and PMVS Results
The table below provides duration gap, estimated point-in-time and minimum and maximum PMVS-L and PMVS-YC
results, and an average of the daily values and standard deviation for the years ended December 31, 2014 and 2013. The
table below also provides PMVS-L estimates assuming an immediate 100 basis point shift in the LIBOR yield curve. We do not
hedge the entire prepayment risk exposure embedded in our mortgage assets. The interest-rate sensitivity of a mortgage
portfolio varies across a wide range of interest rates. Therefore, the difference between PMVS at 50 basis points and 100 basis
points is non-linear.
Our PMVS-L (50 basis points) exposure at December 31, 2014 was $102 million, which decreased compared to
December 31, 2013 primarily due to a decrease in our duration exposure. On an average basis for the year ended December 31,
2014, our PMVS-L (50 basis points) was $69 million, primarily resulting from our duration exposure.
Table of Contents