today covered Freddie's
announcement regarding its plans to buy back from its pools
substantially all loans 120 days or more delinquent in February, to be reflected in the factor tapes (the basis for prepayment calculations, trading assumptions, etc.) delivered March 5, 2010.
Some hours later, Fannie
also announced its intention to buy-out delinquent loans, but not starting until March 2010 (reflected in factor tapes delivered early April), over a period of a few months and amounting to "a significant portion of the current delinquent population ... subject to market, servicer capacity, and other constraints." How delinquent is not specified, though they have disclosed via prospectus that loans missing four consecutive payments are subject too buyout, so presumably they mean 120 days or more delinquent.
Since HW ace reporters have already covered the press releases, I won't belabor the details. Instead, let's consider why it matters, and why investors should like Freddie's plan of attack better.
First the Basics
Loan buyouts from Ginnie
and GSE pools have been a focus of MBS investor and analyst attention for months now. And it's a concern that should trickle down to anyone interested in the capacity of the secondary market to foster affordable mortgage rates in the primary market.
Why? Because a buyout is a prepayment. That means the entire remaining principal balance on the loan returns to the investor at par. Unfortunately, most MBS pools are currently priced at a premium to par. Roughly 95% of 30-year fixed rate securities (the heart of the market) are priced over par, at an average price above 104 points (source UBS). That means any investor who carries their MBS positions at market value (and that's the vast majority) will take an loss on ever dollar that prepays, one that averages to about 4%. Ouch.
Prepayments are the thorn on the rose of government-guaranteed and government-sponsored MBS. Removing credit risk just takes one hurdle out of the investment decision. The problem of anticipating prepayments and adjusting portfolios as needed remains. Only sophisticated institutional investors, with tons of technology in the back office and on their traders' and PMs' desks, participate in the MBS market. In the good old days, prepayments could be modeled fairly effectively as a function of interest rates - the impact of mortgage rates on refinancing activity and home sales. Those simple assumptions were swept aside by the housing bubble, weird loan vehicles and bad underwriting. Now, investors have to consider the impact of default on prepayments.
First it was a problem confined to Ginnies, where servicers would opportunistically harvest delinquent FHA loans. The issue existed throughout the last decade, but the foreclosure crisis ratcheted up the practice. For months now, credit driven prepayments have been an significant source of prepayment volatility and an appetite-suppressant for Ginnie buyers (I discussed this in some detail last year
The GSE Market Was Not Immune
Servicers don't buy loans out of GSE MBS (unless forced, under Reps and Warrantees), instead the GSEs must buy them out if they are 24 months delinquent, modified, foreclosed (or an alternative measure, such as deed-in-lieu has been taken). They retain the right, written into investor disclosures, to buy them out after 120 days if the cost of advancing P&I to investors exceeds the cost of holding the loan in portfolio. However, buying a loan at par and booking it at market value generated a 40 to 60 point hit to capital. In effect, conserving capital forced the enterprises to delay the impact of default on prepayments.
The issue of GSE buyouts, then, did not seriously concern investors, traders and analysts until the HAMP program began to lumber into motion last summer. But by late 2009, the market realized that implementing FAS 166/167 - bringing all guaranteed securities onto the balance sheet - changed the economic incentives. The enterprises could save billions it was shelling out to investors each year while and take losses on the delinquent loans at the more measured pace of reserving for anticipated loan losses.
The biggest obstacle to buying the delinquent loans was the limited headroom in the enterprises' own portfolios. The Treasury Department
eliminated that by clarifying that portfolio limits would be based off legal maximum established by Congress
, rather than actual year-end 2009 holdings. Treasury resolved potential longer term strains on capital at the same time by lifting the caps on preferred stock purchases through December 2012. (I walked through the common sense purposes served by the Treasury announcement at the time
Most analysts predicted a surge of buyouts occurring as early as the start of the year, driving prepayments to peaks as high as 50 CPR or more for several months in higher coupon securities. To translate, 50 CPR is an annualized rate - if the monthly rate observed were sustained for a full 12 months, 50% of the principal balance at the start of the first month would be gone by the end of the 12th month. (For more background, I discussed the market's buyout vigil at the end of 2009
and again last month
Most analysts expected to see the first evidence of GSE bulk buyouts of delinquent loans in rising debt issuance - to finance the buyouts - or in prepayments. Debt issuance did not jump in January, and the factor tapes that arrived at the end of last week "disappointed" as well. (Factor tapes contain lots of vital information on pools, but the critical piece of information is the factor - the fraction of the pool's original principal remaining outstanding at the end of the month. The change in factor from month to month includes both regular amortization and prepayments and is the basis for calculations of prepayments.)
Freddie's announcement is a much more transparent resolution than the market (or I) had been looking for. No tea leaves, no guess work - the factor tape for February will tell the tale and the payments will be passed through to investors on the regularly scheduled dates.
Freddie Mac has already been disclosing sufficient information regarding delinquencies by coupon, vintage and geographical area for analysts to incorporate the announcement into prepayment projections. The first batch of these to show up in my email came from veteran MBS researcher Glenn Schultz and his team at Wells Fargo Securities
. The most affected coupons are 30-year 6s, 6.5s and 7s, where prepayments are expected to come in from 45% on 2005 6s to 60-70% on 2006 and 2007 6.5s.
Projecting Fannie Buyouts an Art, Not a Science
Analysts will be projecting prepayments given Fannie's announcement, but they cannot achieve the same precision as with Freddie projections. Uncertainty always carried a cost in the MBS market, so it will be interesting to follow price action and trader color over the next days and weeks. The strategy analysts were recommending last year was to step away from most affected coupons (known from delinquency disclosures), buy them back after the dust clears. There is probably not time to act on Freddie's announcement, but there might be some selling of likely Fannie targets. (Bear in mind that MBS settle once a month in a staggered sequence by program, etc. starting around mid-month. That limits the response to announcements made today, the tenth day of the month.)
Why is Fannie stretching it out? I suspect it stems in part from the size of its delinquent portfolio - some $82 billion in 30-year securities alone, $45 billion in other securities. Fannie's delinquency rate has been running 40% higher than Freddie's and its guarantee business is about 50% bigger. Veteran FTN Financial
agency analyst Jim Vogel says the extra time should give Fannie the "opportunity to build/manage it's cash liability book in measured fashion to reduce the impact on spreads" of agency debt. By that comment he is referring to Fannie's need to fund those purchases. Freddie, he notes, has already added maybe $20 billion in liquidity through floater sales to date.
NOTE: Linda Lowell writes a regular column, called Kitchen Sink, for HousingWire magazine.
Editor’s note: Linda Lowell is a 20-year-plus veteran of MBS and ABS research at a handful of Wall Street firms. She is currently principal of OffStreet Research LLC.