What Next for Housing Finance?
Since 2008, the $10 trillion housing finance system has been federalized. At the time this article goes to press, the federal government has guaranteed over 90 percent of American mortgages. Without the federal government's backing of Fannie Mae and Freddie Mac, it is likely that these government-supported enterprises (GSEs) would have collapsed in 2008 and housing prices would have plummeted further and faster, perhaps even pushing the economy into Great Depression territory. Thus, for over three years, the federalization of housing finance has been the saving grace of the global economy.
It has not always been this way. In 2006, just five years ago, the GSEs' market share in loan purchases was 37 percent. They guaranteed 40 percent of the mortgage-backed securities (MBS) issued that year. The FHA guaranteed less than 5 percent of mortgages written in 2006; today it guarantees around 30 percent. The asset growth of these federal entities slowed during the bubble, largely because they were displaced by private-label Wall Street securitization as the dominant mortgage financing channel.
The GSEs did not join the crowd until late in the housing bubble. According to the Financial Crisis Inquiry Commission, the GSEs did not begin purchasing subprime MBS until 2001. Subprime mortgages and subprime MBS remained a small part of the market until the rapid growth of private-label securitization in 2003-2004. The worst excesses of the bubble were financed by private-label securitization, not the GSEs. As a result, the default rates on the mortgages the GSEs own or guarantee—13.2 percent in 2006 and 14.9 percent in 2007—are below that of private-label securitizers 45.1 percent and 42.3 percent, respectively.
Of course, the federal government cannot continue to guarantee the entire housing finance market. It is potentially too risky and costly to taxpayers. But, as of December 2011, the unemployment rate was 8.6 percent, long-term unemployment was higher than it had been in decades, housing prices remained at their post-recession bottom, GSE delinquency rates were four times their long-term average, and housing vacancies were near record-high. Clearly, the housing market is still too fragile to abandon government support altogether, especially if regulations are not yet robust enough to prevent a recurrence of the recent boom and bust. And yet, the transition to a new system has begun. It remains unclear what shape the transition will take.
Among the many proposals, there are two general directions this transition may take—either with or without some form of government guarantees. None of these proposals envision a return to the old GSE system. Rather, they advocate a form of explicit catastrophic insurance through new entities. In order to assess this policy, however, it is necessary to understand its origin—and by extension, the origin and purpose of the GSEs themselves.
The Federal National Mortgage Association, or Fannie Mae, was created as a government agency by an act of Congress in 1938. In 1968, the Johnson administration sold majority ownership to investors to balance the federal budget and fund the Vietnam War and the Great Society programs. The government-retained portion of Fannie Mae was known as the Government National Mortgage Association, or Ginnie Mae; in addition, in 1970 the government created a competing private corporation called the Federal Home Loan Corporation, or Freddie Mac.
The goal of all these organizations was to ensure a liquid secondary home mortgage market that would revive the housing market and generate a strong economic recovery, as well as a stable postwar economy. In 1938, that meant purchasing primarily twenty-year, amortizing, fixed-rate mortgages (FRMs) insured by the Federal Housing Administration (FHA), allowing originators to book the gains immediately, and freeing up their capital to issue new loans. This stability depended, in part, on Fannie Mae's commitment to the long-term FRM. Prior to the Great Depression, most mortgages averaged only three to five years with a large balloon payment at the end of the term, as banks preferred to refinance to take advantage of market interest rates, which they had to pay on their short-term deposits. Unfortunately, that put the rollover risk on homeowners, who could not refinance when credit markets and the economy faltered in the 1930s. While the standardized FRM reduced the rollover risk on home mortgages, the funding of long-term mortgages through short-term deposits exposed banks and thrifts to interest rate risk, leading to their decapitalization and the savings and loan crisis of the 1980s. This explains the shift to securitization of these mortgages through pass-throughs guaranteed by Fannie and Freddie.
After the Federal Reserve neutralized high inflation in the 1980s, interest rates began a long, steady decline that lasted into the 2000s, reducing homebuyers' concern with interest-rate risk. Because of the lack of severe recessions, these three decades came to be known as the "Great Moderation," reducing households' concern with macroeconomic swings and credit rationing. Thus, when adjustable-rate mortgages (ARMs) began to creep into the private-label securitization market, they were considered a triumph of financial innovation rather than a return to the dangers of yesteryear. The Great Recession has obliterated this consensus, but it has not prevented many observers from advocating a system in which ARMs dominate.
During the period of stability, the prevailing model was a standardized mortgage with strict and uniform underwriting controls. It should be clear from this brief history that the government role in housing finance should support sound lending, financial stability and transparency and the standardization that supports transparency. The advantage of the GSEs in the tranquil period prior to the mid-2000s was the uniform and prudent standards that their screening imposed. As we have seen, entities with short-run profit objectives may race to the bottom for market share undermining underwriting standards in this key sector for household financial well-being, and, by so doing, put the entire financial system at risk. Prudent uniform lending standards that prevent races to the bottom, asset bubbles, and the steep price declines during ensuing market distress can moderate economic downturns, potentially saving millions of Americans from catastrophe.
On one side of the reform spectrum is a pair of bills in Congress that would maintain the government's role as guarantor and thus the financial stability safety net: the Housing Finance Reform Act of 2011 (H.R. 1859) and the Secondary Market Facility for Residential Mortgages Act of 2011 (H.R. 2413). The primary difference between the bills is the structure of the envisioned guarantor entities.
In H.R. 1859, the government would create privately capitalized Housing Finance Guaranty Associations (HFGAs), which would securitize mortgages, some of which would qualify for government guarantees. Unlike in the GSEs, the government support would be explicit and specific. As a catastrophic guarantee, it would take effect only if the HFGA in question has been placed in conservatorship or receivership. Otherwise, the HFGA will take the first loss position, putting private capital at risk. The qualifying mortgages would become "FHFA securities," in exchange for an annual guarantee fee. HFGAs would not be allowed to purchase mortgages with LTVs exceeding 80 percent unless the seller retains at least 10 percent of the credit risk, the seller agrees to repurchase and replace high-default pools, or at least 80 percent of the loan is insured. HFGAs would also not purchase mortgages whose underlying real estate exceeds the national average price by 150 percent for comparable properties. Finally, HFGA's purchases must be "qualified mortgages," as defined by the Consumer Financial Protection Bureau in order to screen out predatory features. In order to qualify for the government guarantee, the mortgages will also have to meet further restrictions to be defined by the FHFA. Once the FHFA has chartered at least five HFGAs, this bill requires it to place Fannie Mae and Freddie Mac into receivership, presumably in preparation for eventual liquidation. This approach is similar to proposals issued by the Mortgage Bankers Association and the Center for American Progress.
The goal of H.R. 2413 is to bring the securitizing entity into the federal government itself. The government would merge Fannie Mae and Freddie Mac into one Secondary Market Facility for Residential Mortgages, a public entity like Ginnie Mae that would continue to purchase and securitize standardized FRMs. This approach is similar to a proposal by the National Association of Realtors, which has publicly supported the bill. An alternative approach has been proposed by a team at the Federal Reserve Bank of New York. They advocate a lender cooperative whose member entities would securitize standardized residential mortgages with the government guaranteeing tail risk. In other words, the government would cover losses only if they exhaust the mutualized pool for a particular "vintage" of MBS. These losses, in turn, would be supported by an insurance fund maintained by government regulators. By holding the co-op responsible for first losses, it would have less moral hazard than traditional private-label securitizers. Again, unlike the GSEs, the government support would be explicit and specific. The FHFA would directly oversee the co-op's risk management. The single entity aspect of this proposal is designed to prevent underwriting deterioration in mortgage quality through races to the bottom.
On the other end of the spectrum are at least fifteen bills that would wind down the GSEs (or portions of them) without any government-supported replacement. A few of these bills, as well as the Residential Mortgage Market Privatization and Standardization Act of 2011 (S. 1834), direct the federal government to provide oversight (through FHFA) to maintain strict standardization of private-label products, but others are not concerned with such regulation.
Straddling the divide is a proposal by University of Maryland economist Phillip L. Swagel, who advocates temporary government insurance (with private capital taking the first loss position) whose price increases over time, reducing the market's demand for it until it is no longer desired. The result could be a fully private market in the long run. The Swagel proposal and the more immediate wind-down bills assume, contrary to historical evidence, that private-label securitization can dominate the market without a race to the bottom in the absence of a regulatory framework to monitor risk. It may not even be the case that a fully private housing finance system can remain private. In the past century, we have gone from a private system to a public one and back again twice—first, when Fannie Mae took over a depressed market in 1938, and second, when Wall Street overtook the GSEs in the 2000s, resulting in a market nosedive where the GSEs had to federalize 90 percent of the system. Can we really have such short memories?
Many of these proposals assume that the problem is that GSEs and other entities were "too big to fail," a problem that in theory could be remedied by creating smaller entities that could be allowed to fail. The problem, however, was not one of entities being too big to fail, but of the housing market itself being too big and too important to fail, thereby destroying the U.S. economy. This would be a systemic failure that affected all market participants, large and small. In such a circumstance, it is irrelevant whether an individual firm is small or large. No matter how much we decry the resulting moral hazard, the inescapable political reality is that the government, irrespective of which party is in office, cannot allow the system to collapse.
The reliance on eliminating too-big-to-fail firms as a solution also fails to address the problem of a race to the bottom as managers' decisions are shaped by the lure of short-term profits at the expense of long-term solvency. For the managers at firms, any money gained through high loan volume in the present may more than outweigh any losses in bankruptcy later. When systemic insolvency is inevitable, there is negligible marginal risk to pumping profits higher.
Government support after the fact only worsens moral hazard and fuels the likelihood of another race to the bottom. Without greater standardization of mortgage products, the complexity and heterogeneity in the origination market will overwhelm the ability of investors to accurately price risk, and without greater transparency in the securitization market, observers will be unable to determine when this mispricing has occurred. Moreover, without transparency we will not know whether regulators are doing their job in limiting systemic risk.
Keeping regulators' feet to the fire requires many "eyes-on." No GSE reform alone can prevent the private market from finding highly profitable, highly risky alternatives to the safe, standardized products that Fannie and Freddie have purchased since their inception. Only effective supervision can do that. We therefore need something that none of these bills offer: comprehensive reform, pairing government catastrophic insurance with increased oversight of the housing finance market. Only then can we safely transition away from the current federalized system toward a world where the American dream does not come at so high a risk.
Copyright 2012 Adam J. Levitin and Susan M. Wachter.