A big topic in MBS markets right now is the scheduled end of the Fed’s pass-through and agency debt purchase programs at year end. But the elephant in the room is the fate of Fannie and Freddie. To a degree not well understood in the public discussion, the Fed’s purchases have replaced the GSE portfolio purchases that in the past helped to maintain a floor on MBS pricing in periods of peak supply and high market volatility (and consequently a ceiling on mortgage loan rates). As it stands now, the Fed is scheduled to halt its support of pass-through markets just as the GSEs are scheduled to begin reducing the portfolios in 2010 (10% a year until they reach combined $250 billion, estimated to occur around 2020 as required by Housing and Economic Recovery Act of 2008). Nothing I’ve read yet – in the financial media or in sell side research – considers the relationship between these two impending events or questions the impact the two will have together on the remaining sponsors of the MBS markets. Let’s put that in perspective with some numbers. Of about $4.4 trillion Fannie and Freddie securities outstanding, the Fed, Fannie and Freddie own about $1.6 trillion, leaving about $2.8 trillion in the hands of institutional investors of all kinds, foreign and domestic – banks, mutual funds, pension funds, insurance companies, hedge funds – as well as foreign central banks and sovereign funds. Consider too, that the Fed gobbled up its $866 billion or so in just 8 months, consuming more than half the new supply of GSE securities. When prepayments are taken into account (outstanding security balances shrink as new securities are made), at many points over the last year the Fed has inhaled over 100% of net supply. This is in effect a massive market technical that squeezed mortgage spreads to historical tights. At many points too tight for value oriented long term investors (institutional investors and the GSEs) to want to buy them. The Old Political Football, Back in Play At the same time the MBS market is bracing for the Fed’s departure, discussion is heating up of what kinds of enterprises the GSEs should become or be replaced by. At the end of August, the MBA (Mortgage Bankers Association) issued its “Recommendations for the Future Government Role in the Core Secondary Mortgage Market” and on September 10, the GAO issued “Fannie Mae and Freddie Mac: Analysis of Options for Revising the Housing Enterprises’ Long-term Structures.” These documents are silent on a critical issue: how the existing GSEs evolved and maintained the existing MBS market. They sidestep information I believe should be fundamental to any proposal to “fix” the concept of government sponsored housing enterprises. They do not inquire how this market works, the costly, technology- and experience-intensive infrastructure required to originate an eligible loan, pool it, sell it, or trade it in the after market. They ask not how the pass-through market functions as a source of highly liquid, readily tradable high-grade securities for institutional investors, a source of hedging instruments for originators of government and private as well as GSE-eligible loans and for mortgage servicers. It Wasn’t the Guarantee, It Was the Liquidity This is not a trivial consideration. We are not talking about the bondholders of a corporation, even a really big one, like a “too-big to fail” bank or a giant auto maker that can fail. Let me say it again – this is a $4.4 trillion market. The Treasury market is bigger, but no single Treasury issue enjoys the liquidity offered by large TBA Fannie or Freddie coupons. For instance, over $558 billion Fannie 30-year 5.5s are outstanding. Some of that is locked up in CMOs or held to maturity, but the remainder can be delivered in a TBA trade. That dwarfs even a recent Treasury issue (a small comfort for taxpayers), such as the $42 billion of a 2-year note due August 2011 auctioned last month. (Note too that $42 billion issue will become an off-the-run as new 2-years are auctioned, cheapening as it does. A TBA MBS is cheapest-to-deliver – and investors who become prepayment specialists can realize additional value trading “specified pools.”) And, consistent with the size of TBA coupons, bid-ask spreads in normal markets are comparable to those in on-the-run Treasuries. What, you might ask, shouldn’t we also fret about the holders of the GSE’s debt? Well, if the portfolios run down as currently required by law, the debt will be paid down as well. It exists to support portfolio investments. The point of infusions of capital from the Treasury is to protect those liability holders in the event the GSEs assets are not sufficient to repay the debt. Any step allowing the portfolios to remain as they are or to grow after January 1 would be tantamount to a resumption of something like business as usual, either as before conservatorship or remaining under conservatorship. Bear in mind as well that the vast majority of GSE pass-throughs have 30-year final maturities, while the bulk of GSE debt has shorter maturities (for asset-liability matching purposes and to lower financing costs). MBA Has New Way to Slice Pie Consider the MBA’s proposal. It says, “The centerpiece of federal support for the secondary mortgage market should be a new line of mortgage-backed securities.” This new line would sport two guarantees, a federal guarantee at the security level (like on a Ginnie pass-through), and another at the loan level from one of the new private GSE utilities the MBA proposes. (Presumably, though functioning like private FHAs, these new loan guarantors would not displace private mortgage insurers, but rather would, like Fannie and Freddie, use PMI on loans with LTVs above 80%.) Perhaps I am too literal minded, but I do wonder if establishing a new line of securities doesn’t make an orphan of my beloved TBA market. Perhaps the authors of the MBA’s proposal imagine that all the old Fannie’s and Freddie’s can be grandfathered into the new line and receive the new federal guarantee. And the new loan level GSEs would guarantee the old loans. It’s possible that the new loans and securities would be just like the old ones. As the MBA document puts it, ‘allowable mortgage products” would be the “conventional” single family mortgages traditionally supported by the GSEs, including those currently eligible for TBA funding and the kind of multifamily loans currently covered by existing GSE underwriting guidelines. Maybe they also imagine the new GSE’s would inherit the systems and underwriting and servicing guidelines and business relationships of the old GSEs. That means some shareholders who volunteer their capital to endow these new GSEs have to accept the smaller, Freddie slice of the business. Maybe they’re thinking of three GSEs – splitting Fannie in half, like a protozoa, might even the playing field. If it turns out to be an appealing business, a consortium of mortgage bankers might even own one (there’s room for that in the MBA proposal). Do they think it could be done without any changes to any existing computer systems (and this market is very technology dependent)? Once Burned And where do these new shareholders come from? Certainly not Fannie and Freddie’s existing shareholders. There are holders of Fannie Mae common and preferred equity out there who might feel that the last Administration took advantage of panicked markets to steal their company from them. And there are certainly those who bought the preferred stock Fannie issued in 2008 at the behest of and under the apparent protection of the Paulson Treasury. They never saw more than a couple of dividends for their investment. Surely they will be remembering the old saying, “Fool me once, shame on you, fool me twice, shame on me.” Personally, I take loss – my own and others’ – seriously. I would never buy a ticket on a boat named Titanic and I would not be interested in an IPO from a government-sponsored housing enterprise. But that’s just me, and I digress. The bottom line is that if any such new line is not effectively a continuation of existing GSE security lines and devised to feed existing markets for GSE securities, especially TBA, the existing MBS market will be orphaned. Even as large as many TBA classes are, cut off from new supply their innate liquidity cannot be renewed and they will drift into a second-tier, off-the-run senescence. And the new line cannot be expected to trade any better if it doesn’t partake of the long familiarity and deep liquidity of the existing GSE MBS market. The guarantee would not make up for the shallow supply and short prepayment history that have always plagued all other new or off-market MBS products. Look at the way the market shunned Jumbo Ginnies (Ginnie had to change pooling requirements, allowing up to 10% of jumbo mortgages in standard Ginnie pools to create a viable securitization outlet for FHA and VA loans on high priced homes.) If there is not sufficient history to support reliable prepayment modeling – and calculation of an option-adjusted spread – investors back away or require more yield for the extra uncertainty and extra analysis required. Likewise, if there is not sufficient “float” in a security, fewer investors will own it, it will trade less readily and bid-ask spreads will be appropriately wider. The GAO Takes the Cake Enough on the MBA – they are a special interest group and their proposal is after all a lobbying effort. We would expect their proposal to wear blinders. The GAO “Analysis” is more disturbing because it was prepared for Congressional Committees. It is Congress that will take up the question what is to become of Fannie and Freddie. In effect, it was prepared for you and me, the American taxpayers and voters they represent. The GAO analysis also jumps over the question of the functioning market to focus on more abstractions about possible structures for housing enterprises. And its language is abstract, full of impersonal gov-speak and high level, multi-syllabic eval-talk. Apparently even former FHFA Director James Lockhart, in his comments on the GAO’s review draft, thought more discussion of “their important role in providing securities that form the basis for an active and liquid TBA market for mortgage-backed securities (MBS) should be mentioned.” Here is their reply:
We agree that any discussion of the future roles of the GSEs should include consideration of their roles in providing securities that support an active and liquid mortgage market. As the report notes, providing liquidity to mortgage markets has been a key housing mission objective of the enterprises and that, while their secondary market activities have been credited with helping to establish a national and liquid mortgage market, their performance in providing support to mortgage markets during stressful economic periods is not clear.
Here is my analysis: Either they didn’t understand what Lockhart was talking about or they deliberately chose not to pursue it. (I wonder how open minds were at the study’s outset and how much the “findings” reflect the assumptions and received notions the analysts brought to the task. They cite reports from seven researchers, mostly apparently economists, who were also asked to review the report before publication – did they handpick studies that supported pre-determined “findings?” What about the thoughtful researchers and analysts, down in the trenches on “both sides of the trade” – why are they never consulted?) This point – that it’s not established that the GSEs supported housing finance during periods of economic stress – is stressed from the outset, in the report “Highlights” and in the body of the report. Here is what they say is the basis for this finding:
… In 1996, we attempted to determine the extent to which the enterprises’ activities would support mortgage finance during stressful economic periods by analyzing Fannie Mae’s mortgage activities in some states, including oil producing states such as Texas and Louisiana, beginning in the 1980s. [GAO/GGD-96-120] Specifically, we analyzed state-level data on Fannie Mae’s market shares and housing price indexes for the years 1980–1994. We did not find sufficient evidence that Fannie Mae provided an economic cushion to mortgage markets in those states during the period analyzed.
Well, this is lazy – or handpicked “evidence.” First of all, it’s not the same “we,” they are grabbing it off the shelf. More important – the reasoning of that 1996 “we” is supremely flawed. It IGNORES what happened in Texas and Louisiana to undermine home prices. First, the oil boom, which triggered housing and office construction booms in Texas and Louisiana. When the boom went bust unemployment went into serious double digits, workers left the state, houses were abandoned. Second, this boom was financed by thrifts in deep difficulty attempting to grow out of earnings problems. Serious doubts have been raised, in other studies, about the quality of some of the underwriting. This GAO study ignores the effect of easy money and the oil boom in creating what became in the mid-80s, excess inventory. Expecting Fannie and Freddie to turn that tide would be like expecting Fannie and Freddie to repopulate the empty neighborhoods of Stockton, CA today. Why not blame the FHA for the problems of Texas markets? After all, Texas is a leading market for FHA loans and at the recent 2005 national housing peak some people in Houston and Dallas were complaining their houses still weren’t worth what they paid before the oil crash. Just as dishonest, the GAO fails to observe that during this same period, the entire nation experienced wrenching crises in both the thrift and banking industries. Over 1,000 S&Ls were closed, 245 of them in Texas alone, along with 1542 FDIC insured banks. A tremendous amount of home lending capacity was destroyed over that period. However, outside of the “oil patch” states and other areas experiencing severe economic difficulties, home prices generally rose, peaking in the early 1990s, something that only could have occurred if the flow of credit to housing was NOT interrupted. And it was not, largely because Fannie and Freddie filled the crater left by the thrift and banking crises and built a stable market underpinned by an active, liquid TBA market. (Note that the two began issuing MBS in response to the problems thrifts were experiencing. Freddie, originally owned by thrifts, issued it’s first MBS in 1970, Fannie in 1981.) Was Humpty Dumpty Pushed or Shoved? The rest of the GAO report reads like a bureaucratic echoing of white papers from career GSE-foes. In addition to the claim that the GSE’s ability to support housing finance is not demonstrated, it finds “limited” evidence that meeting the affordable housing goals set for them in 1992 materially benefited the targeted groups. Third it finds that their for-profit-structure undermined market discipline and provided them with incentives to follow risky business practices. In particular, the mortgage portfolios were complex to manage and exposed them to interest rate risk and encouraged them to invest in the subprime and other questionable mortgages that “likely precipitated the conservatorship.” The GAO “analysts” should have asked if the “conventional” housing agencies (a conventional loan once meant a loan to a borrower with a good credit habits and sufficient income to service the loan), were the proper vehicle for meeting affordable housing goals, and if affordable housing goals were in fact sound public policy? Weren’t subprime loans and Option ARMs apologized for as “affordability” products,” didn’t allowing people to borrow 100% or more of the cost of a house serve an affordability goal as much as it served lenders’ profit motives? The GAO analysts should have asked if home ownership and affordable housing aren’t two very different policy issues, best achieved, recent history teaches us, by very different means? Really, if the 1992 goals didn’t help targeted groups, is that Fannie and Freddie’s fault, or the Congress’ for pretending they would? As for portfolio risk, wasn’t setting affordable housing goals asking the GSEs to make questionable mortgages? Yes, the mortgage portfolios were complex to manage – so are all the mortgage portfolios held by U.S. banks. As for managing the interest rate risk – in my professional experience, the investment professionals at the GSEs who performed that task were among the smartest and nimblest investors I had the privilege of conversing with. And the proof is in the pudding – those portfolios and the complex hedging strategies performed through the rapid market reverse in interest rates in late 1993 (remember the collapse of Askin Capital, the devastating mortgage derivative losses dozens of public and private investors?), they performed again in the liquidity crisis of 1998 (remember Long Term Capital?), and they performed in the chaotic markets of late 2001. And “likely precipitated conservatorship?” I am an investor. I don’t buy very many single names – I trust my advisor and let him suggest funds – but I bought the housing enterprises, NOT because I thought I’d get an government subsidized return (if that were true, the enterprises' worst enemies would have been their biggest shareholders), but because I liked what they did, I understood it, and I’d talked to a lot of the people working in the enterprises who believed in what they did. I didn’t lose my shirt, but I lost a sock. And I think the question of what precipitated conservatorship is more political than the GAO can ever pretend it was not. This is what I fear if the Congress does not grapple with the actual workings of the GSE MBS market: investors will begin first to underweight the market and then to gravitate toward other bond sectors with more stable reasons to exist. The more proposals to scrap, fragment, rebrand or reconstrue the market functions the GSEs perform, the more investors will retreat from the sector altogether. Foreign investors already appear to be doing so. (I made a lot of trips to Asia in my day to convince a lot of private and government buyers that Fannie and Freddie MBS offered unequalled liquidity plus extra spread over Treasuries. I assume from the TIC data they are taking advantage of that liquidity now, while it still exists.) As spreads widen, the Fed departs, rate-of-return investors rethink their sponsorship. Banks take losses from spread widening on available-for-sale MBS holdings. And so on. Musical chairs.