Disposing of the GSEs and reviving private securitization

Last week, the Federal Deposit Insurance Corp. and the Federal Reserve System sponsored a two-day symposium: “Mortgages and the Future of Housing Finance.” This is the second of my eye-witness reports on the event, where two topics dominated: what to do with the government-sponsored enterprises and how to get private lenders get back in the game. Although the proceedings were dominated by the discussion of academic papers delivered by economists with doctorate degrees, the symposium was extremely well attended. C-SPAN covered most of the proceedings: Click here for day one and here for day two. It wasn’t all formulas, integrations and estimations, though. There was some rhetoric. For example, Dwight Jaffee proposed to “cancel” Fannie Mae and Freddie Mac and reorganize the U.S. mortgage market on private market principles, because mortgage lending is an extreme example of the fact that in most activities the private sector will do a better job than the government. He also asserted that the final leg of innovation in mortgage markets, in which the GSE’s played no role, was subprime lending. His article of faith: the purely private mortgage contract will be private, safer and command lower mortgage rates, though a few minutes later he acknowledged the biggest problem in this reorganization would be the issue of recourse. (I’m no lawyer, but I think the contract is dictated by purely legal matters far beyond the reach of GSEs, such as state law and Uniform Commercial Code. The issue of recourse — and its sister dilemma, treatment of first liens on residential properties, I am pretty darn sure is one that states jealousy preserve as their own. Surely Jaffe, a proponent of private enterprise also has high regard for states rights.) Note that, although he is a professor at the Haas School of Business at the University of California at Berkeley, and a former professor of Economics at Princeton, Jaffe presented no econometric research. Not a bit of data to back up his assertions. Instead, he provided symposium participants with photocopies of his Wall Street Journal opinion piece published the day the symposium began, “How to Privatize the Mortgage Market.” Nostalgia for a past that never existed Near the other end of the spectrum, consumer advocate Martin Eakes argued the GSEs should not be scrapped, simply reformed. Eakes is the CEO of the advocacy group Center for Responsible Lending and Self-Help Credit Union, a nonprofit community development lender and credit union. I confess that I tend to discount the proposals of consumer advocates. Not because of my personal politics, but I simply fear they come to the debate as “axed” (as we would say on the trading desk trying to promote a particular bond we own and want to sell). My observation — in mortgage finance, as in all other matters of public concern — is that the stronger the agenda, the weaker the effort to back it with verifiable facts. Happily, Eakes did not live up to my expectation. In fact, what he said resonated with my own observations during more than 20 years working in and around mortgage markets. He raised a number of concrete issues that the debate and rhetoric about the GSEs has bypassed —some of them items on which I’ve been brooding. The debate over the future of housing finance “exhibits a nostalgia for a past that never existed,” Eakes said. The reality is that prior to the 2002 explosion of private securitization, mortgages were predominantly backstopped by the government, touched in some way by the Federal Housing Administration, the Veteran’s Administration, Fannie Mae or Freddie Mac. As for the remainder, many assume wrongly that those were funded by the private sector. The point Eakes was making is normally glossed over by lobbyists for the financial services industry and other proponents of returning mortgage markets to their pre-1968 (or pre-Depression, for the purists) configurations: the majority of loans not government-related were funded and held by banks (this would include thrifts), and depositories are the hidden GSEs. That is, banks are capitalized by private shareholders, but their deposits are insured by the federal government. Eakes says their liabilities are as well, but I think what he means is that the FDIC’s resolution powers gives considerable comfort to holders of bank liabilities. Furthermore, loans are not funded by bank deposits, they are instead funded with direct advances from a Federal Home Loan Bank, which is a GSE. To Eakes, honesty dictates that we acknowledge this reality, and I agree. I also agree that the single purely private effort (in this context, non-bank lenders) in housing finance was subprime and Alt-A lending, and that it was an “unmitigated disaster” for the global and U.S. economies and American households. I am just sorry that he did not remind symposium participants that some of the biggest mortgage banking companies wrapped themselves in bank or thrift charters. By so doing, they diversified their funding and earnings streams if they did it before the crash — and qualified themselves for Troubled Asset Relief Program funds if they did it after. One survived as TARP-funded bank while others were taken over by the FDIC, or bought by a bank (a TARP bank that is) before the FDIC could get there. Elephant in the room I wish I had a couple bucks for every presenter who proposed moving some piece of the GSE action to “the big banks.” That little windfall would cover my latte addiction for a week or so. Eakes was the only presenter who took notice of the fact that just four banks are now responsible for more than 60% of mortgage loans. Any rethinking of the U.S. mortgage finance system must address this monopolistic concentration (a view he expressed, and I share). Otherwise, the future is grim. For instance, he pointed out that several of the privatization proposals currently circulating in Washington, such as put forth by the Financial Services Roundtable, propose to let the big banks own their own GSEs. If that were to come to pass, the same four banks would control something closer to 90% of all residential mortgage lending. The implications for mortgage pricing are negative. The spread between the rate lenders charge borrowers and the yield the GSEs require on loans they purchase (for securitization) is already historically wide. If only a few banks control the loan market, consumers are bound to pay more. And they already pay plenty. The biggest beneficiaries of GSE securitization are not homeowners, but lenders. As I’ve written recently, the big banks already can pocket 3 or 4 points on retail originations — after charging application fees, loan origination points and so forth. Eakes solution — so simple he thought it proved he is not from Washington — was to fix the flaws in the old GSE structure. First, require adequate capital — 0.45% was a “ludicrous” charge to support the off-balance sheet guarantee business. (That business went on-balance sheet Jan. 1, 2010, under new generally accepted accounting rules, so that is a flaw half-corrected already.) The authority for the GSE regulator to set appropriate capital requirements is already in place under the Housing and Economic Recovery Act of 2008, all that is needed now is to use it. Two, the regulator must be properly funded. Currently, it is dependent on congressional appropriations to maintain sufficient staff and other resources to do its job. Ideally, the regulator should be supported by assessing GSE guarantee income, similar to the FDIC. Three, the guarantee should be explicit, the amount of guarantee outstanding and its allocation among products controlled by the regulator. Eakes also debunked the myth that the GSE portfolio serves no public purpose, pointing out that it is a natural holder of multifamily housing debt — $120 billion in loans and another $60 billion securitized. Canceling the portfolio eliminates this vital function. The same for reverse mortgages — there is no natural buyer of a security with an increasing principal amount. The portfolio also accommodates the needs of small lenders who sell their loans directly to the GSEs, and is essential to accumulate the purchased loans until sufficient size is reached to securitize. Likewise, the GSEs require portfolio capacity in order to take modified loans out of its securities. At the end of June, they held $235 billion of such modified loans. One function of the GSE portfolios that Eakes did not mention, but which is of significant benefit to investors and borrowers alike, is to dampen rate volatility. Historically, when mortgage-backed securities become very cheap relative to other high quality debt (and mortgage rates painfully higher than Treasury and other benchmark yields), the GSEs have stepped in to buy Ginnie and GSE MBS for their portfolio. Opportunistic, yes, but it has served to keep a lid on market yields and borrowing rates at the top of rallies and during flight-to-quality events as occurred in 1998. Essentially, the Federal Reserve has taken on this function (buying $1.25 trillion MBS) since Congress capped the portfolios and required the GSEs to run them off under HERA. Finally, an emergency ability to buy and hold residential mortgages (Fannie Mae’s public purpose since it was created as a government agency in the Depression, after all, has been to issue debt and buy mortgages) can’t simply be switched on like a light. Start up time is significant, so the portfolio needs to be an ongoing activity in order to step in during an emergency. A properly empowered regulator should be able to determine if purchases meet a public purpose and regulate the size of the portfolio. The last flaw — publicly owned shares — is easy to resolve, in Eakes’ view. Create a new corporation to assume the assets and liabilities of the GSEs. The current shareholders equity would vanish. This new corporation could be set up as a lender cooperative similar to the FHLBs, where no single lender has dominance, or a state housing finance authority. Private mortgage lending is dead whether there’s one or two GSEs That leaves one unanswered question: Should there be one or two GSEs? Eakes’ market-savvy recommendation is one. A single security is more efficient for lenders (after all, the loans are essentially the same, fully fungible; I’d add that the very small but real liquidity hickey exacted on the smaller Freddie Mac MBS programs would vanish). All talk of junking the GSEs, reforming them or reconstituting as purely private or purely government entities is overshadowed by the simple fact that private mortgage lending is dead. It may revive, but at present, depositories are not making loans for portfolio in significant amounts and the amount of private loan securitizations since the crash has been miniscule. In fact, evidence was presented at the symposium that suggests the strongest signals of private capital’s interest in mortgage markets is directed at the private mortgage insurers. In 2010 alone, three PMIs have raised almost $1.3 billion in new capital and a new PMI $600 million, according to Larry Cordell of the Federal Reserve Bank of Philadelphia. This, despite the fact that, as Cordell put it, the “industry is an artifact of the GSE structure that is being debated.” The PMI industry owes its entire existence to the requirement in the GSEs’ charters that any loan with LTV greater than 80% be credit-enhanced. The rise of piggy-back second liens sharply eroded their market share before the crash, and FHA has provided strong competition since, but PMIs are growing again. One of just a few Ph.D. economists to abstain from presenting a new model to the gathering, Cordell asked the eminently sensible question, “Why are MIs surviving, GSEs bust?” First, under state insurance regulation, PMIs must set aside 50 cents from every premium dollar and hold those contingency reserves for 10 years. Second, the rating agencies actually did their job monitoring PMIs and stress testing their capital. It helped as well that the GSEs relaxed their AA- rating requirement. Third, rescissions are up from a normal 7% rate to 20% to 25%. Cordell didn’t say it outright, but policies are a lot easier to rescind than loans, for example, are to put back. Fourth, PMIs face different reserving requirements in that they are only required to reserve for losses after a notice of default is received. (GSEs reserve for losses well in advance of realizing a loss; their capital problems in 2008 began with reserving and were accounting, paper losses.) FDIC skin-in-the-game rules Two real world mortgage analysts addressed the future of securitization. Andrew Davidson, whose eponymous company Andrew Davidson & Co. has been providing prepayment and credit models and other valuation services to MBS investors for decades, said the multiple new regulatory requirements for structures, risk retention and documentation are cost prohibitive. Laurie Goodman, senior managing director at Amherst Securities, was more hopeful about the longer term revival of private securitization. But in the short run, banks will be stymied by the recently adopted FDIC securitization safe harbor rules. Nonbanks like real estate investment trusts may enter the market although essential short-term funding warehouse lines of credit is very limited and will keep volumes down. Banks may sell loans to nonbank issuers, but they will wait until rules defining “qualified mortgages” exempt from the 5% risk retention requirement are worked through the bank regulators. Those rules are not due under the Dodd-Frank Act until next spring, but it is not uncommon for rule making to drag on well beyond deadlines set by Congress. 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.