Freddie Mac 2012 Annual Report Download - page 205

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, 2012 and 2011. 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, 2012 was $209 million, primarily
driven by our convexity exposure. The PMVS-L at December 31, 2012 declined compared to December 31, 2011 primarily
due to a decline in our duration exposure. On an average basis for the year ended December 31, 2012, our PMVS-L (50 basis
points) was $212 million which was primarily driven by our net duration exposure on our mortgage assets.
To improve the accuracy of our models, we make changes to the underlying assumptions or modeling techniques on a
periodic basis. During the third and fourth quarters of 2012, we made assumption changes related to our prepayment model
for non-HARP eligible loans underlying our securities. In addition, we enhanced our process used to estimate duration and
convexity of our unsecuritized single-family loans by incorporating additional loan characteristics. The impact of
incorporating additional loan characteristics increased our duration and produced an estimated $560 million impact on
PMVS-L and a two month impact on duration gap at implementation. We believe that this change would not have had a
materially different impact on either item at December 31, 2011 if implemented at that time. As these changes extended the
duration of our assets, we increased the use of derivatives to hedge our overall exposure.
Table 74 — PMVS and Duration Gap Results
PMVS-YC PMVS-L
25 bps 50 bps 100 bps
(in millions)
Assuming shifts of the LIBOR yield curve:
December 31, 2012 ....................................................................... $61 $209 $ 737
December 31, 2011 ....................................................................... $ 7 $465 $1,349
Year Ended December 31,
2012 2011
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)
Average ................................................. (0.2) $ 36 $212 (0.0) $21 $359
Minimum ............................................... (2.4) $ 1 $ — (1.0) $— $ —
Maximum ............................................... 1.1 $117 $661 1.2 $94 $721
Standard deviation ......................................... 0.6 $ 32 $109 0.3 $15 $126
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 have been
higher if we had not used derivatives. The derivative impact on our PMVS-L (50 basis points) was $(0.9) billion at
December 31, 2012, a decline of $1.1 billion from December 31, 2011. The decline was primarily driven by an increase in
our issuance of longer-term debt beginning in the fourth quarter of 2011, which decreased our reliance on derivatives. In
addition, the decline in interest rates during 2012 decreased the duration of our hedged assets, which resulted in requiring
fewer derivatives to hedge our portfolio.
Table 75 — Derivative Impact on PMVS-L (50 bps)
Before After Effect of
Derivatives Derivatives Derivatives
(in millions)
At:
December 31, 2012 ................................................................. $1,102 $209 $ (893)
December 31, 2011 ................................................................. $2,470 $465 $(2,005)
Duration Gap Results
We actively measure and manage our duration gap exposure on a daily basis. In addition to duration gap management,
we also measure and manage the price sensitivity of our portfolio to a number of different specific interest rate changes along
the yield curve. The price sensitivity of an instrument to specific changes in interest rates is known as the instrument’s key
rate duration risk. By managing our duration exposure both in aggregate through duration gap and to specific changes in
interest rates through key rate duration, we expect to limit our exposure to interest rate changes for a wide range of interest
200 Freddie Mac