A version of this article was published in The Journal of Structured Finance.
This article reflects and incorporates materials developed by and in conjunction with the Mortgage Finance Working Group sponsored by the Center for American Progress. This working group, comprised of academics, former government officials and other leading mortgage finance experts, began gathering in 2008 in response to the U.S. housing crisis in an effort to collectively strengthen our understanding of the causes of the crisis and to discuss options for public policy to shape the future of the U.S. mortgage markets. A more detailed discussion of our Mortgage Finance Working Group’s framework for mortgage finance reform, as well as a partial list of its membership, can be found here.
The U.S. Congress and the Obama administration are now earnestly engaged in the complicated process of reforming our nation’s mortgage finance system. The decisions they make will have enormous impacts on the shape of the U.S. financial markets and may well decide whether homeownership remains a part of the American Dream for middle-class families or instead a distant hope for only the wealthy.
There appears to be a large and growing consensus around the following key principles that should govern any mortgage finance reform:
- Broad access to mortgage credit: Mortgage finance should be broadly available across all geographic areas, in all communities, to all creditworthy borrowers, through a diverse variety of lenders, both small and large.
- Affordability: Mortgage finance should continue to be available on affordable terms, for both owner-occupied and rental housing, both in terms of price and product types, with a special emphasis on the 30-year fixed rate mortgage.
- Stability: To the extent that mortgage finance is inherently procyclical, it must be reined in to prevent excessive boom-bust cycles from occurring, and a source of countercyclical liquidity to provide mortgage financing during times of crisis must be available.
- Taxpayer protection: Taxpayer losses through government exposure to mortgage finance risk, which comes in the form of both explicit and implicit guarantees, should be prevented both by scaling down the government’s footprint in the market, and by ensuring that where the public is at risk, it is well protected by sufficient regulation, capital buffers and other safeguards.
- Emphasis on private capital: To the extent possible, private capital should be relied upon to serve the mortgage markets. When a government role is deemed necessary, this should be narrowly tailored and as an insurer of last resort, with private capital taking all non-catastrophic losses.
In February 2011, the Department of Treasury and Department of Housing and Urban Development released a report to Congress on the future of the housing finance system. (Treasury/HUD [2010]). The report describes similar principles[1] and lays out three approaches to reforming the U.S. mortgage system, based on its survey of the mortgage finance reform proposals that have been offered to date:
- Option 1: Privatized system of housing finance with assistance from FHA, USDA and Department of Veterans’ Affairs for narrowly targeted groups of borrowers.
- Option 2: Privatized system of housing finance with assistance from FHA, USDA and Department of Veterans’ Affairs for narrowly targeted groups of borrowers and a guarantee mechanism to scale up during times of crisis.
- Option 3: Privatized system of housing finance with FHA, USDA and Department of Veterans’ Affairs assistance for low- and moderate-income borrowers and catastrophic reinsurance behind significant private capital.[2]
Most of the proposals offered would generally be described as Option 3 (see, e.g., Dynan [2011], Hancock [2011], Marron [2010], Zandi [2011], Mortgage Bankers Association [2009], Hassan [2009], Dalton [2010]), with only a handful of proposals falling into the Option 1 and Option 2 categories. (See, e.g., Wallison [2011] and Scharfstein [2011]). This is in large part because of the significant evidence that Options 1 and 2, which essentially propose a purely private solution to replace the government sponsored enterprises Fannie Mae and Freddie Mac, would fail to meet the principles articulated above. It is important to note that there are significant differences among the many different proposals that fall under the “Option 3” umbrella, and most of these proposals also fail to satisfy these principles.
While many would describe the proposal offered by the Mortgage Finance Working Group convened by the Center for American Progress as Option 3, we prefer to think of our proposal as distinct from the others. We believe we have constructed a thoughtful proposal that meets all of the criteria laid out above and stresses a unified system of credit delivery, one which makes mortgage finance available to creditworthy borrowers in all communities on equivalent and non-discriminatory terms. Conversely, we believe that the Option 3 proposals tend to create a bifurcated system of mortgage finance, one that provides higher priced credit via a second-tier system to borrowers in historically underserved communities and in neighborhoods hit hardest by the foreclosure crisis.
But while many of the proposals falling under Option 3 are flawed, it is Options 1 and 2, which would essentially privatize the mortgage markets currently served by Fannie and Freddie, that are the most problematic.
The “Privatization” Options (1 and 2) Fail to Meet the Principles for Mortgage Finance Reform
There are compelling reasons to believe that adopting Options 1 or 2 would lead to greatly undesirable outcomes, particularly when compared against the principles of broad access, affordability, stability, taxpayer protection and an emphasis on private capital laid out above.
Privatizing the mortgage markets would sharply reduce broad access and affordability of mortgage finance
It should be noted at the outset that it is difficult to predict exactly what would happen if we adopted Options 1 or 2, because the purely private mortgage system has been so rare in modern history. In the United States, the federal government has guaranteed most of the U.S. residential mortgage market since the New Deal, either through federal deposit insurance or guarantees on securitization. And there are no contemporary examples of an advanced economy that does not provide significant levels of government support for its mortgage markets. (Breejen [2004], Schwartz [2006], Moody’s [2009]).[3]
That being said, the few examples of privatized mortgage systems that have existed strongly suggest that in the absence of a government guarantee, the U.S. mortgage finance system would experience a sharp decline in the availability of mortgage finance, particularly in the area of affordable consumer-friendly products such as the 30-year fixed-rate mortgage.
The last time we had something like Option 1 or 2 in the U.S. residential mortgage system was in the 1930s. During this period, residential mortgages were generally only available to higher-income and higher wealth borrowers, and even then only on terms that were very beneficial to lenders and onerous to consumers, with high floating interest rates, short durations, interest-only or negative amortization, extremely high down payments (typically 50%) and featuring bullet payments of principal at term. (See, e.g., Green [2005]). These mortgages also had much higher rates than the ones we take for granted today, with large regional disparities, as much as 2% to 4% between different parts of the country. (Bernanke [2007]).
In short, the last time we actually had a purely private mortgage finance system similar to Option 1 or 2, it was an abysmal failure, if we judge it by the principles laid out above. Mortgages finance was largely unavailable to the middle class and it was very costly for consumers, both in the rates they paid and in the types of products they had access to.
While the pre-New Deal mortgage system is obviously a dated example, it is notable that in some important ways, it resembles the current (smaller) market for commercial real estate, which largely lacks government support.[4] Commercial real estate loans are typically interest-only or partially amortizing rollover loans with high down payment requirements, designed to be refinanced every few years. (COP [2010]). Mortgage credit for commercial real estate is not broadly or consistently available, and is relatively costly. (Id.)
More recently, we saw the private-label securitization market of the past decade hewing to this familiar pattern, as it displayed a strong bias in favor of riskier adjustable-rate mortgages over safer long-term fixed-rate loans. In total, ARMs accounted for 70% of all mortgages financed by private label securities from 2001 to 2008 (as compared to 12% for loans financed by Fannie and Freddie). (FHFA [2010]).
These examples all point to the same conclusion: privatization would lead to a sharp reduction in the availability of mortgage finance, particularly for middle class households, and a transition away from consumer-friendly products, including the 30-year fixed-rate mortgage.
The relative lack of investment capital willing to take on both credit and interest rate risk also suggests sharp liquidity reductions in a privatized mortgage market
A privatized mortgage system would also face difficulties in attracting investment capital, making it likely that such a system would see a severe reduction in mortgage liquidity, and thus the availability of mortgages to consumers.
Currently, most U.S. mortgages are financed by the purchase of securities guaranteed by Fannie Mae, Freddie Mac, and Ginnie Mae (“Agency securities”).[5] The interest-rate-sensitive investors who purchase these government-guaranteed securities are predominantly made up of foreign central banks, fixed-income investors and regulated financial institutions. (Federal Reserve Flow of Funds Report). These investors typically purchase Agency securities because of investment objectives (often required by charter) or regulatory incentives. The attraction of these government-guaranteed Agency securities to investors and regulators is they bear essentially no credit risk, don’t require costly and time-consuming due diligence, and are very liquid, such that they can be resold at any time into the secondary markets.
Conversely, mortgages that are originated for non-Agency outlets are financed either by lenders who hold loans on their balance sheet[6] (“balance sheet lending”) or through the issuance of private-label mortgage-backed securities, which are financed by credit-risk-sensitive investors.[7] Since 1990, such non-Agency lending has accounted for 12 to 25 percent of conventional (not subprime) loans in any given year. (FHFA [2009]). This is, for the most part, the share of the jumbo market for higher-balance loans.
The critical question for advocates of Option 1 and 2 is whether these sources of private, non-guaranteed financing (or others like them) are sufficiently scalable to take on the $5.5 trillion or so in mortgage financing that is currently provided by the GSEs. The available evidence strongly suggests that they are not.
Balance sheet lending dominated U.S. mortgage finance in the post-New Deal era, as heavily regulated and federally insured banks and thrifts provided most of the home mortgages made to U.S. consumers. But since the 1970s, balance sheet lending has seen a major decline in importance, for at least two reasons. First, the interest rate shocks of the 1970s and 1980s (which were partially responsible for the savings and loan crisis) have made lenders more cautious about taking on too much mortgage risk. This is particularly true for long-dated assets like 30-year fixed-rate mortgages, which bear high amounts of interest rate and liquidity risk.
Second, the federally insured deposits that finance most balance sheet lending have increasingly declined in importance, as U.S. households have shifted more of their wealth to institutional funds (such as pension funds and money market funds). As a result, there is simply less money (on a relative basis) held in banks and thrifts than there was in the 1950s and 1960s, when banks and thrifts were the primary source of mortgage finance.
Neither of these trends—an aversion to taking on too much rate and liquidity risk or the declining importance of bank and thrift deposits in the larger financial system—appears likely to change anytime soon. As such, it seems highly unlikely that balance sheet lending will take on a significantly greater role than it already does.
It also is difficult to imagine that private-label securitization will reemerge in the foreseeable future as a major source of financing for residential mortgages. There was a brief period in the 2000s when AAA-rated private-label securities—which don’t carry a government guarantee—were considered to be equivalent to Agency securities. By virtue of the AAA ratings awarded to private-label securities, the investment banks that sponsored these private-label securitizations convinced investors and regulators that these securities were safe and liquid, and thus did not require extensive independent due diligence. As a result, private-label securities were purchased extensively by interest-rate-sensitive investors, the critical factor in allowing private-label securitization to reach an unprecedented 38 percent share of mortgage originations in 2006.
Of course, we have since learned that AAA-rated private-label securities were nowhere close to being as safe as Agency securities.[8] The huge losses resulting from these securities, along with the many revelations (which continue to come) of serious structural flaws (such as misaligned servicer incentives and shoddy ratings), have shattered the belief that private-label securities with a AAA credit rating are analogous to securities that enjoy a government guarantee. As a result, credit-risk-sensitive investors have stayed away. Since the financial crisis, only two private-label securitization deals, amounting to less than $600 million, have occurred, both with extensive due diligence and loan-level modeling, despite extremely conservative loan characteristics.
This is not to say that private-label securitization won’t reemerge. It will (although there are still some near-term hurdles to regaining the favor of investors). But going forward, it would seem safe to say that private-label securities, even ones designed to have minimal credit risk, will receive significant amounts of independent due diligence from investors, which in turn will decrease liquidity for these securities. The need for time-consuming independent due diligence and the relative illiquidity of such securities mean that the available pool of capital willing to purchase private-label securities will be much smaller than the pool of capital willing to purchase government-guaranteed securities.[9]
It is difficult to imagine, under these circumstances, how private-label securitization could once again account for 38% of mortgage originations, as it did in 2006, let alone take on most of the mortgage market, as privatization advocates are proposing.[10]
Finally, some policy makers have pushed for legal changes that might encourage more covered bonds issuances in the United States, arguing that these instruments might provide a form of private, non-guaranteed finance for the U.S. mortgage markets, based on the success of these instruments in certain European countries. (See, e.g., Garrett [2010]). What this analysis ignores is that European covered bonds also enjoy a government guarantee that is critical to the broad and deep liquidity they enjoy, as is well recognized by financial analysts. (See, e.g., Moody’s [2009]).
Advocates of privatizing the U.S. mortgage system have repeatedly argued that a government guarantee is not necessary to appeal to investors averse to credit risk. For example, Peter Wallison of the American Enterprise Institute has stated that “the key to a successful mortgage market is not a government guarantee… but ensuring that the mortgages that are made in the market are of prime quality.” (Wallison [2011]). But there is no evidence provided to back up this extremely speculative claim, and ample evidence to the contrary.[11]
Other proponents of privatization have argued that western European countries offer something akin to Options 1 and 2, insofar as they do not provide explicit guarantees for their mortgage markets, and that nonetheless they have ample liquidity. (See, e.g., Lea [2010]). This argument ignores the dominant role of implicit guarantees in most European countries, similar to the guarantees provided by the U.S. government to Fannie Mae and Freddie Mac.[12] As one anonymous European Central Bank official reportedly said, “We don’t let banks fail. We don’t even let dry cleaners fail.” (Wessel [2009]). In effect, European governments provide a 100% guarantee of their mortgage markets, as most recently evidenced by sweeping bailouts (including bank nationalizations and blanket guarantees) in Germany, Denmark, Italy, the United Kingdom and Ireland.
Arguments for privatization based on the experience of the jumbo markets are flawed
Some advocates of mortgage market privatization have also pointed to the jumbo markets to support their view that a privatized mortgage system could broadly provide mortgage finance, including in the form of 30-year fixed-rate mortgages. For example, Mr. Wallison has stated that “anyone can go to the Internet and find lenders offering jumbo fixed-rate thirty-year loans—which, by definition, have no government backing.” (Wallison [2011]).
These arguments are ill-conceived. The jumbo market serves higher income, higher wealth borrowers, and no one disputes that the private markets can serve this market well, just as they have always done. In fact, our current system explicitly assumes that higher wealth, higher income Americans will be served without a government guarantee, which is why there are loan limits for Fannie and Freddie.
However, as discussed above, the key issue that needs to be resolved is the scalability of the sources of financing for the jumbo markets. As I discussed previously, there are serious reasons to doubt that balance sheet lending or private-label securitization can finance the $5.5 trillion in mortgage originations currently held by Fannie and Freddie.
Moreover, the jumbo argument ignores the inherent tendency of purely private lenders towards products that are more onerous for borrowers. This includes not only the horrific experience of private-label securitization in the early 2000s, but also the jumbo market more generally. From 1990 to 2007, the jumbo markets originated more adjustable-rate loans than fixed-rate loans nearly every year, and always far more ARMs than the agency portion of the market. (Pafenberg [2005[).
Even to the extent that long-term fixed-rate mortgages have been offered in the jumbo market, these have been integrally tied to the government-supported segment of the market. First, Fannie and Freddie have set a market standard, providing 30-year fixed-rate loans at affordable prices. Obviously, their private competitors must provide products on terms and prices that are competitive with that benchmark.
Second, lenders were able to significantly lower their cost of funding jumbo 30-year fixed-rate mortgages through the market infrastructure that existed around Fannie and Freddie securities. Prior to the financial crisis, jumbo mortgages were hedged through transactions in the “To Be Announced Market,” the very deep and liquid forward market for trading securities issued by Fannie and Freddie (which allows consumers to lock in their mortgage rates, among other things).[13]
In other words, Fannie and Freddie not only provide enormous market pressure on private lenders to offer competitively priced jumbo 30-year fixed-rate loans, but they also provide the market infrastructure that allows jumbo lenders to offer lower pricing on this product. If we privatized the market served by Fannie and Freddie, we would be removing both the incentive and much of the ability to offer affordable 30-year fixed-rate loans.
Could the private markets provide affordably priced 30-year fixed-rate loans broadly to all creditworthy Americans? Perhaps, although the loss of deep liquidity of the TBA market would make that much more difficult. Would they, in the absence of any competitive offerings of this product? Based on all available evidence, the answer appears to be no.
Privatization of the mortgage markets is likely to result in greater instability
It also appears likely that privatization would lead to greater instability and more extreme boom-bust cycles. Despite the prevalence of loans that were extremely lender-friendly, the pre-New Deal mortgage banking system was very unstable, experiencing extreme bubble-bust cycles every 5-10 years, with high resulting social costs. (Gorton [2009], Wheelock [2008]). During the recent residential mortgage crisis, the purely private portion of the mortgage market—private-label securitization—was the largest driver of mortgage-related losses, accounting for 42% of serious delinquencies despite being responsible for only 13% of all outstanding mortgages. (Lockhart [2009]). And commercial real estate finance, which is essentially privatized, has experienced a bubble-bust cycle even more extreme than that in residential real estate, with a national price decline of 45% from its peak, greater than the 30% price drop experienced in residential real estate.[14]
Privatized mortgage markets are more volatile for at least two reasons. First, private lenders and investors tend to exhibit strong procyclical tendencies—when times are good, they want to make more money and so they overleverage themselves, and when times are bad, they want to reduce losses and so they excessively pull back on lending. Second, as noted above (and unsurprisingly), private lenders exhibit a strong tendency towards originating lender-friendly mortgages. The onerous terms of these loans create a high degree of systemic instability by increasing the likelihood of default and foreclosure.
Of course, the greatest period of mortgage market stability in U.S. history was the post-New Deal era, after high levels of government support were introduced, first in the form of deposit insurance and then in the form of government-guaranteed securitization. It was only when significant pockets of unregulated banking activity occurred, first with the deregulation of the thrift industry in the 1980s and then with the growth of private-label securitization in the 2000s, that we saw instability reenter the U.S. financial system.
Privatizing the mortgage markets would increase the “too big to fail” problem
Privatizing the mortgage markets would also accelerate the trend towards banking concentration and “too big to fail” that has been so disastrous for the United States.
As we learned from the series of banking panics that regularly occurred prior to the Great Depression, banking poses an enormous amount of systemic risk. Bank failures can quickly lead to bank panics, causing an enormous loss of wealth and economic devastation. This is why governments “bail out” their financial institutions in times of trouble—they believe that the cost of letting a financial panic go unchecked is much higher than the cost of the bailout. This is particularly true for residential mortgage lending, given the economic and social importance of housing.
Exacerbating the systemic risks posed by mortgage banking, we have the contemporary problem of “too big to fail”—when financial institutions are so large and interconnected that the unsupported failure of a single one of them would, it is believed, cause the downfall of the global financial system, leading to major economic distress. Because of this “too big to fail” status, large financial institutions are believed to enjoy an implicit government guarantee on their obligations. This taxpayer guarantee provides significant subsidies to the biggest banks, as they enjoy significantly lower funding costs than smaller financial institutions. {Gandhi and Lustig [2010]).
It is these largest financial institutions that would benefit the most if we adopted either of the Obama administration’s privatization proposals, as was recently pointed out in a note from MF Global. (Seiberg [2011]). Under either of the privatization approaches, we would be asking private lenders to take on the $5.5 trillion in mortgage financing—primarily for working- and middle-class households—that is currently provided by the GSEs. To the extent such lending occurred, it would be financed either by lenders willing to hold loans to maturity on their balance sheets, or from the private-label securitization of mortgages. In both areas, “too big to fail” financial institutions have a decided advantage over their smaller competitors.
The six largest U.S. financial institutions—Bank of America, Wells Fargo, JP Morgan Chase, Citigroup, Goldman Sachs, and Morgan Stanley—account for over a third of all balance sheet financed residential mortgage lending, and this concentration is increasing due to their lower cost of funding and excess balance sheet capacity. To the extent that privatization led to increased balance sheet lending, this would unduly accrue to these “too big to fail” banks.
These same six firms, which currently hold about $9.275 trillion in assets, have even greater market power in the investment banking sector, particularly given the consolidation that has occurred in the aftermath of the financial crisis (JP Morgan Chase’s takeover of Bear Stearns, Bank of America’s takeover of Merrill Lynch, and Wells Fargo’s takeover of Wachovia, among others). Thus, they are likely to dominate any private-label securitization that may reemerge from the crisis.
In short, privatization of the mortgage finance system would result in “too big to fail” financial institutions becoming even bigger and more systemically important. It would also encourage greater concentration in the financial system. As a result, the taxpayer guarantee enjoyed by these institutions—a guarantee that is given for free—would be even more pronounced, providing greater advantage to these institutions.[15] Lest we forget, it was these firms that led us into the financial crisis. Setting aside the substantive criticisms of privatizing the mortgage markets, there is also a fairness problem here as well. Privatization would reward those actors who acted most badly, creating a very severe moral hazard problem.
Privatizing the mortgage markets would not mitigate the problem of risk pricing
Some advocates of Options 1 and 2 have contended that, even acknowledging the problems associated with privatized mortgage markets described above, privatization is preferable to the alternative of a government guarantee because the government cannot successfully price for risk. (Wallison [2011]). This argument criticizes many of the Option 3 proposals (including the MFWG proposal), which feature a “catastrophic” guarantee protected by capital requirements and an insurance fund, arguing that “[t]hese plans are based on a fundamental error: that the government can act like an insurance company and set a correct price for the risk it is taking.” (Id.).
Implicit in this argument is the assumption that the private sector, which has for-profit insurance companies competing against one another, can successfully price for catastrophic risk.
This argument ignores the crucial difference between ordinary risk, which the private sector can and routinely does price effectively, and catastrophic or “black swan” risk, which the private sector has not shown that it can effectively price for. Indeed, the evidence is very clear that the private sector has not priced effectively for this risk. (Moss [2004]). And there are compelling reasons to believe it cannot price for this risk. (Taleb [The Black Swan, 2007], Taleb [“The A Priori Problem,” 2007]).
Indeed, the shoddy performance of the private sector in pricing residential mortgage-related risk during the past decade should be strong enough indication of the problems with this argument. For example, delinquency and loss rates on mortgages originated for private-label securitization are exponentially higher than those for mortgages originated for Agency securitization.
Moreover, this argument also overlooks the relative success of analogous federal programs insuring against catastrophic risk, including the Federal Deposit Insurance Corporation’s deposit insurance program and the Terrorism Risk Insurance program.
There are at least three reasons why the federal government is better suited to take on—and price for—such catastrophic risk.
- First, the government has deep pockets and a long horizon. It is thus better able to handle large unexpected events than its private sector counterparts.
- Second, the government doesn’t need to price risk perfectly, but simply needs to ensure it doesn’t underprice risk. It is only when risk is underpriced that the government is exposed to losses.
- Third, and unlike its private sector analogues, the federal government is able to levy “ex post facto” assessments on insured parties, through increased fees or taxes, in order to cover potential losses.
Finally, it is worth repeating the obvious. In a democratic society, the government is typically going to absorb catastrophic losses anyways, whether these are explicitly insured or not. As we have seen in recent years in the aftermath of Hurricane Katrina, the tsunami in Japan, and most relevantly for our purposes, the bailouts of the financial markets by the conservative Bush administration, when catastrophic losses have occurred, the federal government has stepped in. Given this dynamic, coupled with the “too big to fail” problems I discussed above, the only incorrect pricing for risk is the price proposed by advocates of Options 1 and 2: to give away an implied guarantee for free to large financial institutions under the fiction that the federal government will not cover their losses in the event of another “black swan” event.
The MFWG Framework for Reform
While Option 1 and Option 2 are fatally flawed, that does not mean that Option 3 plans are optimal. In fact, many of the Option 3 plans proposed to date also have significant issues, particularly with respect to sufficiently stressing affordability, taxpayer protection, and stability. Conversely, I believe that the plan put forth by the Mortgage Finance Working Group meets all of these basic structural principles for reform.
Importantly, our proposal creates a unified system of mortgage finance, one that is designed to serve creditworthy borrowers in all communities through the same lenders and delivery mechanisms. We believe that a two tier system, such as those generally offered in the various Option 3 proposals, in which borrowers from underserved communities or neighborhoods hit hardest by the foreclosure crisis, will generally fail to address the credit needs of these borrowers, raising the risk that they will be vulnerable again to the types of predatory lending tactics and onerous loan types that we saw in the past decade.
As described below in greater detail, our GSE successors, called “Chartered Mortgage Institutions,” would be stripped of many of the functions we believe caused the GSEs to take on more risk than they should have. At the same time, they would hold higher levels of capital and pay into a Catastrophic Risk Insurance Fund, measures designed to protect the taxpayer. To facilitate our CMIs’ ability to reach creditworthy borrowers in all communities, we propose the creation of a “Market Access Fund,” financed by assessments on the issuance of mortgage-backed securities and tasked with the purpose of providing credit subsidies to improve the attractiveness to secondary market actors (including the CMIs) of potentially scalable mortgage products that serve borrowers in underserved communities.
We believe this framework, the product of years of thoughtful discussion and research, meets the basic principles for mortgage finance reform, providing broad and constant access to mortgage finance on affordable terms to creditworthy borrowers in all communities, addressing the problems of structural instability that caused the mortgage crisis, including “too big to fail”, and relying primarily on private capital to reach this goal, with the government’s exposure limited to a well-protected catastrophic risk position, one that it clearly already holds, as we saw from the series of bailouts provided by the conservative Bush administration.
Background
In thinking about how to replace the GSEs, it is important to understand the essential functions they currently serve. First, Fannie and Freddie buy loans from lenders, including long-term fixed rate loans that lenders would not make absent a reliable way to off-load the risk posed by such long-term obligations.
Second, Fannie and Freddie issue mortgage-backed securities backed by many of these loans—the process of “securitization.”
Third, they also hold some of these loans on their balance sheet. This practice is necessary to aggregate loans for securitization, to hold and test new products before they can gain secondary market acceptance, to provide liquidity for loans that are difficult to securitize (as is the case with some multifamily loans), and to provide lenders with liquidity so that they can continue to make loans when capital markets are constrained.
Fourth, for a fee, Fannie and Freddie guarantee investors against credit risk, providing their MBS investors with assurance of the timely payment of interest and principal on those securities, relieving investors of concerns about borrower default.
Fifth, they deliver to investors a further guarantee—a basis for investor confidence that the mortgage-backed securities offered by the GSEs will perform as promised—as their own credit guarantee is backed by an implied (and since conservatorship, effectively explicit) guarantee by the U.S. government against their failure. Neither the investors nor Fannie and Freddie currently pay the government for providing this guarantee.
Sixth, these functions have also enabled the development of deep liquidity in the so-called “To be Announced,” or TBA, market, a type of futures market for mortgage-backed securities that allows lenders to provide consumers with interest rate forward commitments or “locks” on their mortgage interest rates before the final mortgage is signed and sealed. Finally, Fannie and Freddie delivered countercyclical liquidity so that mortgages were available for consumers no matter current housing market conditions of the direction of the broader economy.
Through much of the past 70 years, including the period since the capital markets froze in 2008, this system has resulted in mortgage money being consistently available, contributing substantially to broader economic stability. It has done so by connecting the local demand for mortgages with the international capital markets by creating a fully liquid investment attractive to a wide range of risk adverse investors. With the government standing behind mortgage-backed securities issued by Fannie and Freddie (whether implicitly before 2008 or effectively explicitly since conservatorship), investors believe there will always be a market for any MBS they buy now and may wish to sell later, regardless of economic conditions.
The result is a deep and liquid market for mortgage securities that has been able to continue to operate since 2008, a period when other capital markets froze. In the future, all these functions need not be provided by the same entity. Indeed, separating them could reduce the risks of overconcentration in the market, enhance competition, and ensure access to all sizes of mortgage originators, including community banks and credit unions, while preserving the transparency, standardization, and scale that make for a broadly efficient and liquid market. Most importantly, the catastrophic risk guarantee must be separated from the other functions.
The basic structure of the MFWG proposal
We envision a system with the following actors performing the key functions:
- Originators—lenders of all types would originate loans, as in the current system.
- Issuers—originators of individual mortgages as well as aggregators of those mortgages who would issue securities backed by mortgages originated by themselves or others.
- Chartered mortgage institutions (CMIs)—institutions not owned or controlled by originators (other than potentially through a broad-based cooperative structure), chartered and regulated by a federal agency, would guarantee timely payment of principal and interest on securities, typically issued by others, backed by loans eligible for a government guarantee against catastrophic risk.
- Government catastrophic risk insurance—an on-budget catastrophic risk insurance fund, funded by premiums on CMI-issued MBS, would be managed by the government to protect investors in the event of the failure of a CMI; the government would price and issue the catastrophic guarantee, collect the guarantee premium, and administer the catastrophic risk insurance fund.
The government would set the product structure and underwriting standards for eligible mortgages and securitization standards for MBS guaranteed by CMIs.[16] To protect taxpayers and ensure that all requirements for the guarantee are met, the government would regulate the CMIs for both capital adequacy—at levels significantly higher than required of Fannie and Freddie— and compliance with consumer protection and other responsibilities.
The government would serve as conservator or receiver for CMIs that fail, with responsibilities that include ensuring that the servicing of the remaining guaranteed securities is carried out by a qualified entity.
The different functions of aggregation, insurance, and delivery of government guarantee currently performed by both Fannie Mae and Freddie Mac thus would be separated. Private capital would bear the major responsibility for underwriting, aggregating, securitizing, and guaranteeing mortgage credit for both affordable homeownership and rental housing. The CMI guarantee would be supported by borrower equity, often private mortgage insurance and other forms of credit enhancement, and the CMI’s own capital. The government backstop against CMI failure would be explicit, limited, and priced. Neither the debt nor equity of the CMIs would be government backed, unlike the current system.
The proposed CMIs are likely to be significantly smaller than Fannie and Freddie are today, thus enhancing competition, reducing taxpayer risk, and improving access by smaller lenders to the secondary market. To further these ends, and to counterbalance the extreme concentration of the mortgage origination and servicing industries in entities that themselves have both an explicit government guarantee (on deposits) and implicit “too big to fail” backing, the only circumstance under which originating lenders would be allowed to have an ownership interest in a CMI would be as part of a broad-based mutually owned entity designed to ensure access, at equitable prices, to smaller lenders such as community banks, credit unions, and community development finance Institutions. In that context, and to assist in the achievement of public policy outcomes that may not coincide with the interests of private owners of CMIs, consideration might also be given to permitting CMIs established by government entities, such as housing finance agencies, individually or collectively.
Our plan would also create a Market Access Fund (MAF), financed by assessments on the issuance of all MBS, which would be complementary to the existing National Housing Trust Fund and Capital Magnet Fund created by the Housing and Economic Recovery Act of 2008. The MAF would use credit enhancement and other tools to help CMIs and others test and bring to markets sustainable mortgage products that are targeted at underserved borrowers, potentially providing greater scalability to products that have performed successfully on a more local level.
By providing research and development funds, credit enhancement, and an opportunity for a product to test the market, the MAF would enable niche products to gain access to the capital provided by the secondary markets. A MAF credit subsidy would be awarded competitively to partners, including Chartered Mortgage Institutions, state and local housing finance agencies, and large nonprofits that can bear a significant share of the risk of loss on the loans and deliver products to the market at scale. Loans with some risk sharing with the MAF could be eligible for either CMI or Ginnie Mae securitization. The MAF would share the risk of loss on a loan or pool level for products that meet underserved needs, but only where private capital is also at significant risk.
By sharing the risk of loss, the MAF will make it easier for private capital to serve otherwise underserved communities. Without this mechanism, there is a significant risk that the taxpayer will continue to stand behind too large a share of the housing market through the direct guarantees of the FHA, VA, and USDA’s rural housing programs, exposing taxpayers to risk that could, through the MAF, be shared with the private sector.
In short, our plan tries for the most part to keep and improve upon the benefits of the mortgage system that has been in place in the New Deal, while reining in systemic risk and protecting the taxpayer from loss. In addition to specific measures designed to encourage mortgage liquidity to underserved communities and borrowers, including for rental housing,[17] what we have essentially proposed is stripping down the GSEs to their core function of providing credit guarantees on mortgage-backed securities, while building around them a system of regulation, insurance and resolution modeled on the FDIC’s current regulatory authority for banks and thrifts. The government’s catastrophic insurance exposure would be limited to mortgage-backed securities, and there would be much greater taxpayer protections in place, including significant capital requirements (4-9 times the levels currently put up by the GSEs) and a Catastrophic Risk Insurance Fund modeled after the FDIC’s Deposit Insurance Fund and financed by assessments on the industry.
Our proposed structure preserves a mortgage system that is both local and national, and includes the features that have enabled our mortgage market to attract capital from around the world. Our proposal builds on recent statutory and regulatory accomplishments, including the Dodd-Frank Act. And it ensures that American homeowners, renters, and lenders of all sizes and types, in all parts of the country, at all times, will have access to appropriately- priced, low-risk mortgage finance.
Conclusion
In thinking about how to reform the mortgage finance system, it is imperative that we think carefully through the principles we want such a reformed system to meet, and whether the constructs we are developing actually meet those principles. As I have tried to demonstrate, Options 1 and 2 are inherently flawed and cannot meet those basic principles, which is why there is no developed economy that actually uses such a system in the modern capitalist world.
While Option 3, which is simply a generic way to describe any system with a government backstop, can meet these basic principles, the actual details behind any Option 3 proposal are critical to whether it actually does so. We believe the MFWG framework is a proposal that does meet these basic goals, in a thoughtfully constructed way, providing broad access to mortgage credit on non-discriminatory terms to borrowers from all geographic areas and in all communities through a unified system of diverse lenders, providing such credit affordably, both in terms of actual rates and the types of products emphasized, in a systemically stable way that minimizes housing market volatility and the risk of extreme credit bubble-bust cycles, protecting the taxpayer, and relying almost exclusively on private capital to do so.
David Min is the Associate Director for Financial Markets Policy at the Center for American Progress.
References
Ben S. Bernanke, “Housing, Housing Finance, and Monetary Policy,” Speech of the Federal Reserve Board Chairman at the Federal Reserve Bank of Kansas City’s Economic Symposium, Jackson Hole, WY, Aug. 31, 2007, available at http://www.federalreserve.gov/newsevents/speech/bernanke20070831a.htm.
Congressional Oversight Panel, Commercial Real Estate Losses and the Risk to Stability, February Oversight Report, February 11, 2010, available at http://cybercemetery.unt.edu/archive/cop/20110401232630/http://cop.senate.gov/documents/cop-021110-report.pdf.
John H. Dalton on behalf of the Housing Policy Council of the Financial Services Roundtable, “Public Input on Reform of the Housing Finance System,” Public Comment to Treasury and HUD, July 21, 2010, available at http://www.fsround.org/housing/pdfs/pdfs2010/HousingPolicyCouncilCommentsonReformofHousingFinanceSystem7-21-10.pdf.
Department of the Treasury and Department of Housing and Urban Development, Reforming America’s Housing Finance Market: A Report to Congress (Washington, DC, February 11, 2011), available at http://www.treasury.gov/initiatives/Documents/Reforming%20America%27s%20Housing%20Finance%20Market.pdf.
Femma den Breejen and others, Government Guarantees in the Rental and Owner-Occupied Sector: an International Comparison (2004).
Christian deRitis and Mark Zandi, “Responding to Critics of a Hybrid Mortgage System,” Feb. 17, 2011, available at http://www.economy.com/dismal/article_free.asp?cid=197265&src=mark-zandi.
Karen Dynan and Ted Gayer, The Government’s Role in the Housing Finance System: Where Do We Go From Here? presented at the Brookings Institution, April 14, 2011, available at http://www.brookings.edu/~/media/Files/rc/papers/2011/0211_housing_finance_dynan_gayer/0211_housing_finance_dynan_gayer.pdf.
Federal Housing Finance Agency, Update Assumptions Used to Estimate Single-Family Mortgages Originated and Outstanding, 1990-2009, 2009, available at http://fhfa.gov/webfiles/15722/Updated_assumptions_2009_050510.pdf.
Federal Housing Finance Agency, “Data on the Risk Characteristics and Performance of Single-Family Mortgages Originated from 2001 through 2008 and Financed in the Secondary Market,” Sept. 13, 2010, available at http://www.fhfa.gov/webfiles/16711/RiskChars9132010.pdf
Federal Reserve, Flow of Funds Report, available at http://www.federalreserve.gov/RELEASES/z1/.
Priyank Gandhi and Hanno Lustig, “Size Anomalies in US Bank Stock Returns: Your Tax Dollars at Work?” December 10, 2010, available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1653083.
Rep. Scott Garrett, “Testimony before the Senate Banking Committee on Covered Bonds: Potential Uses and Regulatory Issues,” Sept. 15, 2010, available at http://banking.senate.gov/public/index.cfm?FuseAction=Files.View&FileStore_id=43c1c135-9444-42ca-aefb-b70f8444e55b.
Gary B. Gorton, Slapped in the Face by the Invisible Hand: Banking and the Panic of 2007, prepared for the Federal Reserve Bank of Atlanta’s 2009 Financial Markets Conference: Financial Innovation and Crises, May 9, 2009.
Richard K. Green and Susan M. Wachter, The American Mortgage in Historical and International Context, Journal of Economic Perspectives 19 (4) (2005): 93-114.
Diana Hancock and Wayne Passmore, An Analysis of Government Guarantees and the Functioning of the Asset-Backed Securities Market, Federal Reserve Board, Finance and Economics Discussion Series (2010).
Diana Hancock and Wayne Passmore, Catastrophic Mortgage Insurance and the Reform of Fannie Mae and Freddie Mac, presented at the Brookings Institution, April 14, 2011, available at http://www.brookings.edu/~/media/Files/events/2011/0211_mortgage_market/0211_reform_fannie_freddie_hancock_passmore.pdf.
Qumber Hassan and Mahesh Swaminathan, GSEs—Still the Best Answer for Housing Finance, Credit Suisse Mortgage Market Comment, October 6, 2009, available at http://www.zigasassociates.com/images/uploads/GSEs_-_Still_the_best_answer_for_housing_finance.pdf.
David Lawder, “PIMCO’s Gross: government mortgage guarantee ‘essential’,” Reuters, August 17, 2010, available at http://www.reuters.com/article/2010/08/17/us-usa-housing-gross-idUSTRE67G30720100817.
Michael Lea, "Alternative Forms of Mortgage Finance: What Can We Learn from Other Countries?” Working Paper MF10-5 (Harvard University Joint Center for Housing Studies, 2010), available at http://www.jchs.harvard.edu/publications/MF10-5.pdf.
Joo-Yung Lee and Robert Grossman, “U.S. Housing Reform Proposal FAQs: Filling the Void,” Fitch Ratings, Feb. 24, 2011.
James B. Lockhart, “FHFA’s First Anniversary and Challenges Ahead,” Speech before the National Press Club, July 30, 2009.
Brian Louis and David M. Levitt, U.S. Commercial Property Recovery Spares Economy, Bloomberg Businessweek, February 4, 2011.
Donald Marron and Phillip Swagel, Whither Fannie and Freddie? A Proposal for Reforming the GSEs, e21 Commentary, May 24, 2010, available at http://www.economics21.org/files/pdfs/commentary/05_24_2010_Whither.pdf.
Christian Meldinger and Ivanka Stefanova, “Danish Covered Bonds—A Primer” (UniCredit 2008), available at http://www.nykredit.com/investorcom/ressourcer/dokumenter/pdf/SR080608_DanishCoveredBonds.pdf.
Moody’s Global Banking, “Financial Crisis More Closely Aligns Bank Credit Risk and Government Ratings in Non-AAA Countries,” Special Comment, May 2009.
Mortgage Bankers Association’s Council on Ensuring Mortgage Liquidity, Key Considerations for the Future of the Secondary Mortgage Market and the Government Sponsored Enterprises (GSEs), January 2009, available at http://www.mbaa.org/files/AU/2009/MBAWhitePaper_TheFutureoftheSecondaryMarket.pdf.
Forrest Pafenberg, “Single-Family Mortgages Originated and Outstanding: 1990-2004,” Office of Federal Housing Enterprise Oversight (2005), Table 2, available at http://www.fhfa.gov/webfiles/1151/mortmarket1990to2004.pdf; Federal Housing Finance Agency, “Updated Assumptions Used to Estimate Single-Family Mortgages Originated and Outstanding, 1990-2009” (2010), Table 2, available at http://www.fhfa.gov/webfiles/15722/Updated_assumptions_2009_050510.pdf.
David Scharfstein and Adi Sunderam, The Economics of Housing Finance Reform: Privatizing, Regulating and Backstopping Mortgage Markets, presented at the Brookings Institution, April 14, 2011, available at http://www.brookings.edu/~/media/Files/events/2011/0211_mortgage_market/0211_mortgage_markets_scharfstein_sunderam.pdf.
Herman Schwartz and Leonard Seabrooke, “Varieties of Residential Capitalism in the International Political Economy: Old Welfare-States and the New Politics of Housing,” Comparative European Politics 6 (2006): 237-261.
Jaret Seiberg, “Housing Finance: Fannie and Freddie Ain’t Dead Yet,” MF Global Financial Services Bulletin, Feb. 11, 2011.
Standard & Poors, Home Prices Edge Closer to 2009 Lows According to the S&P/Case-Shiller Home Price Indices, April 26, 2011.
Nassim Nicholas Taleb, The Black Swan: The Impact of the Highly Improbable, Random House, 2007.
Nassim Nicholas Taleb, “The A Priori Problem of Observed Probabilities,” unpublished manuscript, 2007, available at http://www.fooledbyrandomness.com/central.pdf.
Peter J. Wallison, Alex J. Pollock and Edward J. Pinto, Taking the Government Out of Housing Finance: Principles for Reforming the Housing Finance Market, American Enterprise Institute Policy White Paper, January 20, 2011, available at http://www.aei.org/docLib/HousingFinance.pdf.
David Wessel, In Fed We Trust: Ben Bernanke’s War on the Great Panic, (New York: Crown Publishing Group, 2009).
David C. Wheelock, The Federal Response to Home Mortgage Distress: Lessons from the Great Depression, Federal Reserve Bank of St. Louis Review, May/June 2008, pp. 134-39.
Mark Zandi and Christian deRitis, The Future of the Mortgage Finance System, Moody’s Analytics Special Report, February 7, 2011, available at http://www.economy.com/mark-zandi/documents/Mortgage-Finance-Reform-020711.pdf.
Endnotes
[1]. The Treasury/HUD report describes four “factors” that should be considered in the mortgage finance reform debate: access to mortgage credit; incentives for investment in housing; taxpayer protection; and financial and economic stability. (Treasury/HUD [2010]).
[2]. As implied by these three options, there has generally been consensus about two propositions: first, that the Federal Housing Administration, the Department of Veterans Affairs and other government agencies providing mortgages to lower-income households should continue to serve something like their current role; and second, that high-balance (“jumbo”) loans should be financed by private markets operating without a government guarantee. But see Hancock [2010] (arguing that a government guarantee should extend to the entire secondary market for mortgages and other forms of credit). The debate to this point has largely revolved around the so-called “middle market” currently served by the government-sponsored enterprises Fannie Mae and Freddie Mac.
[3]. The den Breejen article finds that the United Kingdom alone does not feature government backing for its mortgage markets. However, as Moody’s, among others, has noted, in the wake of the financial crisis, it appears that all European countries, including the United Kingdom which has provided various bailouts of “too big to fail” financial institutions, provide some level of government guarantees to their mortgage markets.
[4]. While there is no direct government support for commercial mortgage financing, there is some indirect government support for the sector. Affordable multifamily buildings are financed in part by direct investments made by Fannie Mae and Freddie Mac (subsidized by a lower cost of debt made possible by a government guarantee). Commercial real estate debt is also financed to a large degree (slightly less than half of all outstanding loans) by banks, which benefit from federally insured deposit insurance. That being said, the level of government support in commercial mortgage financing is generally consistent with the level of government support described in the proposals for privatizing the residential mortgage market (which retain FHA/VA lending and federal deposit insurance for banks and thrifts), so the comparison seems appropriate.
[5]. Fannie Mae and Freddie Mac purchase and pool mortgages, and issue securities based on the mortgage payments from the pool. Fannie and Freddie guarantee timely payment of principal and interest on these securities, should the underlying mortgage payments be insufficient (for example, because of a higher than expected number of loan defaults), and the government was understood to stand behind Fannie and Freddie. Ginnie Mae does not pool and securitize mortgages, but rather guarantees the timely payment of principal and interest for mortgage-backed securities, issued by approved private parties, and made up of loans insured by the Federal Housing Administration and other government agencies.
[6]. Most balance sheet lending is still backed by the federal government, through federally guaranteed deposit insurance provided to regulated banks, thrifts, and certain other lenders.
[7]. During the 2000s housing bubble, many interest-rate investors also became attracted to highly rated non-guaranteed mortgage-backed securities, convinced by the high credit ratings that these instruments carried negligible credit risk. Obviously, the high loss rates on these securities, resulting in enormous losses throughout the financial system, illustrated the magnitude of this error. Since at least the onset of the financial crisis, interest-rate sensitive investors have shied away from private-label securities, purchasing only those securities carrying a guarantee guarantee.
[8]. Indeed, this fact stands in strong juxtaposition to the claim of many advocates of Options 1 and 2 that the government cannot successfully price catastrophic risk. (See, e.g., Wallison [2011]). In fact, the government priced risk far more accurately than the private sector during this crisis, perhaps reflecting the importance of regulation to rein in procyclicality. Moreover, the government’s track record in pricing catastrophic risk is actually fairly successful, including the FDIC, the Federal Housing Administration and the Terrorism Risk Insurance program. The relative lack of losses is perhaps indicative of two advantages of government risk pricing. First, there is no problem if the government overprices risk, only if it underprices risk. Second, and unlike the private sector, the government can typically charge for risk on an ex post facto basis, through increased ex post facto assessments after a catastrophic event.
[9]. A recent Fitch Ratings report makes a similar point, noting that historically, interest rate sensitive investors have provided the bulk of U.S. mortgage financing through their purchases of Agency securities, while credit risk sensitive investors have purchased private-label securities. (Lee [2011]).
[10]. Some policy makers have argued for the creation of a robust U.S. market for covered bonds, based on the relative success of this financing instrument in Europe. Covered bonds are a sort of hybrid instrument, combining some elements of mortgage-backed securities and some elements of corporate senior debt. As I argue in an upcoming paper, covered bonds are not an appropriate fit for the United States, for a number of reasons, including that they encourage (and are integrally connected with) “too big to fail”, that they are redundant to FDIC depository insurance, and significantly increase the risk of taxpayer loss.
[11]. For example, Bill Gross, the co-founder of the world’s largest bond fund PIMCO, has stated that PIMCO would not even consider buying non-guaranteed mortgage-backed securities unless they came with drastically more conservative underwriting characteristics, including at least a 30 percent down payment, raising similarities to the underwriting that took place prior to the New Deal. (Lawder [2010]).
[12]. Indeed, the implied guarantee for European banks is a critical factor for financial analysts assessing the relative risk of financial instruments issued by these banks. (See, e.g., Moody’s [2009], Meldinger [2008]).
[13]. Following the financial crisis, it is less clear that such hedging will be as prevalent, given the sharp differences in liquidity that occurred between jumbos and conforming loans during the 2008 financial crisis, which illustrated the problems with this hedging strategy.
[14]. Commercial real estate suffered a 45% price decline from October 2007 to August 2010. (Louis [2011]). Residential real estate suffered a price decline of just under 35% from its June/July 2006 peak to its April 2009 trough. (Standard & Poors [2011]).
[15]. Moreover, as Moody’s Chief Economist Mark Zandi points out, a privatized system would result in a more fractured, less liquid market, leading to more profitable arbitrage opportunities for Wall Street firms. (deRitis [2011]).
[16]. It is possible that these standards will become de facto standards for the non-CMI market as well, which should increase the liquidity of those securities.
[17]. To facilitate the availability of financing for affordable multifamily rental housing, our proposal would rely on a requirement levied on all CMIs that did multifamily finance that at least 50 percent of the multifamily housing they financed were for units available at rents affordable to households making 80 percent of the area median income.