Freddie Mac 2011 Annual Report Download - page 203

sensitivity. Our PMVS measures assume instantaneous shocks. Therefore, these PMVS measures do not consider the
effects on fair value of any rebalancing actions that we would typically expect to take to reduce our risk exposure.
Limitations of Market Risk Measures
Our PMVS and duration gap estimates are determined using models that involve our best judgment of interest-rate
and prepayment assumptions. Accordingly, while we believe that PMVS and duration gap are useful risk management
tools, they should be understood as estimates rather than as precise measurements. While PMVS and duration gap
estimate our exposure to changes in interest rates, they do not capture the potential impact of certain other market risks,
such as changes in volatility, basis, and foreign-currency risk. The impact 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. As such, 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 uncertainty regarding default rates,
unemployment, loan modification, and the volatility and impact of home price movements on mortgage durations. Mis-
estimation of prepayments could result in hedging-related losses.
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, 2011 and 2010.
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 the end of December 31, 2011 was
$465 million; approximately half was driven by our duration exposure and the other half was driven by our negative
convexity exposure. The PMVS-L at December 31, 2011 declined compared to December 31, 2010 primarily due to a
decline in our negative convexity exposure as long-term rates significantly declined. On an average basis for the year, our
PMVS-L (50 basis points) was $359 million, which was primarily driven by our negative convexity exposure on our
mortgage assets.
Table 73 — PMVS Results
25 bps 50 bps 100 bps
PMVS-YC PMVS-L
(in millions)
Assuming shifts of the LIBOR yield curve:
December 31, 2011 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $ 7 $465 $1,349
December 31, 2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $35 $588 $1,884
Duration
Gap
PMVS-YC
25 bps
PMVS-L
50 bps
Duration
Gap
PMVS-YC
25 bps
PMVS-L
50 bps
(in months) (dollars in millions) (in months) (dollars in millions)
2011 2010
Year Ended December 31,
Average . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (0.0) $21 $359 0.0 $23 $338
Minimum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (1.0) $— $ — (0.7) $— $ —
Maximum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 $94 $721 0.7 $83 $668
Standard deviation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.3 $15 $126 0.3 $18 $179
Derivatives have historically enabled us to keep our interest-rate risk exposure at consistently low levels in a wide
range of interest-rate environments. The table below shows that the PMVS-L risk levels for the periods presented would
generally have been higher if we had not used derivatives. The derivative impact on our PMVS-L (50 basis points) was
$(2.0) billion at December 31, 2011, a decline of $1.0 billion from December 31, 2010. The decline was primarily driven
by a decline in long-term rates, which resulted in lower duration and convexity exposures on our mortgage assets, without
a full offsetting impact from our existing debt and derivative portfolios. In order to remain within our risk management
limits, we rebalanced our portfolio with receive-fixed swaps, which lowered our derivative duration exposure.
198 Freddie Mac