MBS Relative Value Via the Federal Reserve

September 24th, 2009 11:37 am | by John Jansen |

Here is a long piece from a Bank of America analyst on relative value in the MBS market and the extent to which Federal Reserve and Treasury purchases distort the market ( or at least traditional value metrics):

“Until recently, most of the relative value trades across the 30-year coupon stack were driven by expectations for interest rates and prepayment speeds even though the basis trade itself was driven by expectations for Fed’s activity in the MBS market. However, we should start seeing a major shift in the focus of relative value players as the Fed’s MBS program started playing a significant role in “NON-PRODUCTION” coupon securities which could potentially change TBA deliverables fairly dramatically and make the standard relative value frameworks for coupon swaps and specified pools less useful than before.

The table below shows the outstanding balance of 30-year FNMASs (FNCLs) that is not locked up in CMO deals by origination year and coupon along with the size of the ownership of the Fed and the Treasury:

Orig. Year          Outstanding Balance of 30-yr FNMAs Excl. CMOs ($bb)
4.0s     4.5s     5.0s     5.5s     6.0s     6.5s
2009             110.9    239.2     51.0      6.8      1.6      0.5
2008               1.0      9.5     74.8    116.6     59.0     14.6
2007               0.1      0.6     16.6    108.6    125.7     29.9
2006               0.1      0.8      9.8     47.9     83.5     29.7
2005               0.3      0.8     87.6     67.7     14.3      2.3
2004               0.1      8.1     42.0     49.3     16.8      1.9
Before 2004          0.6      4.4     92.9     91.1     30.4     18.5
Fed’s Ownership     93.9    185.0    112.9     86.7      9.4      1.9
Tsy’s Ownership      1.3     17.5     31.1     21.9     12.8      0.1
Total Float ($bb)  112.9    281.2    374.6    488.1    331.3     97.4
Fed+Tsy Owned ($bb) 95.2    202.5    144.0    108.6     22.2      2.0

As we commented earlier this week, the Fed and the Treasury together own close to 34% of outstanding 30-year FN/FH passthrough pools at the moment (If we include the 30-year passthroughs that Fannie Mae and Freddie Mac own in their retained portfolios, the percentage of the 30-year passthrough market controlled by the Fed/Tsy and the GSEs should exceed 50% of the outstanding balance of FN/FH 30-year passthroughs). This percentage will also continue to increase until the Fed’s purchase program is over. What implications does it have for intra-sector relative value trades in the agency MBS market?

YTD, the Fed’s MBS purchase program was a net buyer of $113 billion of 30-year FN 5.0s (Fed’s gross purchases of this security were actually much higher at $181 billion which exacerbates some of the issues discussed below). If the Fed were to take delivery of all $113 billion of FN 5.0s on the October TBA settlement date, the TBA deliverable could easily move to 2005 origination year. It is worth noting that the TBA deliverable issue didn’t show up in 30-year FN 4.5s although the Fed bought $185 billion FN 4.5s YTD because of heavy new production of FN 4.5s in 2009. The situation is completely different when the Fed buys 5.0% and higher coupon passthroughs. The monthly issuance of 5.0% and above coupons has been very low relative to the size of Fed’s purchases recently.

Of course, if the Fed rolls their purchases of non-production coupons, dollar rolls may not trade special and some of the squeeze related issues will be alleviated. In addition, with Funds rates being so low, you don’t lose all that much when the roll trades special!!! However, it is likely to take a lot longer than usual for relative value trades to work in this environment. More importantly, Fed’s purchases across the coupon stack are likely to receive increased attention going forward from relative value players.

From trading side, the Fed’s MBS program seems to be actively buying 30-year 5.5s which drove FN 5.5s butterfly to as high as 15 ticks. Although the FN 4.5s fly looks cheap and FN 5.0s fly looks rich, we are more comfortable shorting the 5.5s fly for a trade given current float issues in different coupons. We also like buying the FN 6s/5.5s swap at current valuations.

Any thoughts?”

Be Sociable, Share!
  1. 8 Responses to “MBS Relative Value Via the Federal Reserve”

  2. By Bob on Sep 24, 2009 | Reply

    I wish I knew what a “TBA deliverable” was, and what “when the roll trades special” means. I know what it means to roll a position, but “it trades special” throws me off.

    As for the table, it looks like they bought only the best products out of ’09. I suspect there is almost no risk in those 4 and 4.5 bonds. I wonder which would be more interesting or valuable for them to off-load: the higher-risk higher-yield or the low-risk low yield? If they want to protect thier balance sheet, the latter. If they want to prevent banks from taking the hit, the former.

    What is the Fed’s primary motive?

  3. By Bman on Sep 24, 2009 | Reply

    When repo rates trade at very low rates (even negative these days), they are known as trading at special, not GC (general collateral).

  4. By Chicken on Sep 24, 2009 | Reply

    The part that’s missing is what the FED is doing with the income stream from this GSE paper. We sure don’t hear much about this from the grassy-knoll crowd… I think they’ve overlooked something.

  5. By Peter Duray-Bito on Sep 24, 2009 | Reply

    How much of this GSE paper still has an income stream?

  6. By Chicken on Sep 24, 2009 | Reply

    There must be some kind of an income stream, REIT’s have been distributing 90% of theirs every quarter and the dividends blow most everything else away…

  7. By Bman on Sep 24, 2009 | Reply

    what good is a high yield if the underlying asset(s) go down another 15-20%?

  8. By Chicken on Sep 24, 2009 | Reply

    That’s the $13T question, isn’t it? Perhaps yield compensates for the asset price deflation? They’re easily traded where actual real estate isn’t very liquid.

  1. 1 Trackback(s)

  2. Jun 28, 2010: alephblog » Blog Archive » Name Your Poison

Post a Comment