Fannie Mae and Freddie Mac (collectively the two largest “GSEs”, or government-sponsored enterprises) have engaged in a broad range of residential mortgage activities for many years.¹ The economic disaster of recent times has drawn considerable attention to Freddie and Fannie, which is not surprising considering the role that the mortgage sector of the U.S. banking system played in that debacle. Together the two institutions hold or pool about $5 trillion worth of mortgages, and so sketchy were their operations that in September 2008 the U.S. government had to bail them out and place them in conservatorship to keep the entire mortgage market from imploding. While the U.S. government has by now been made whole by TARP-assisted banks, it is not clear whether the billions of dollars provided to keep Fannie Mae and Freddie Mac aflolat will ever be returned to the U.S. Treasury.

It has not been easy for even well-educated observers to understand the role that Fannie Mae and Freddie Mac have played in our recent economic distress, and how to make sure they never become part of such problems again. By almost any measure, the operation of these GSEs is complex, just as the system in which it operates is fully comprehensible only to those who are experts in knowledge of its inner workings. That is why proposals to reform Freddie and Fannie seem so difficult to construct, explain and implement. Clearly, the Congress has not made much progress in the effort so far.

The Obama Administration provided some clarity by offering policy options in a position paper released in February 2011, Reforming America’s Housing Finance Market: A Report to Congress (the “White House Report”). On August 16, the Washington Post carried a front-page article by Zachary A. Goldfarb that reviewed Obama Administration efforts, revealing that, while the Administration’s economic wizards have not yet decided what plan to put before the President, they are leaning toward a larger role for the Federal government than had been suggested by the February White House Report. According to the article, the approach being pursued by the Administration “could even preserve Fannie Mae and Freddie Mac . . . although under different names and with significant new constraints.”

Names matter, but the character of those constraints matter a great deal more. As the Washington Post article notes, conservatives in general, as well as many economists, believe that any significant government role in the mortgage market is distortionary and likely to contribute to further financial dislocation. It also notes that the Administration has not yet decided whether to pursue this policy development to its final course before the 2012 presidential election, suggesting that it is well aware of the political volatility of the issue.

As detailed as this Washington Post account is, it cannot begin to describe the almost Byzantine complexity of Federal intervention in the housing market. The possibility of coherent thought about the reform of our mortgage GSEs requires one to understand why Fannie and Freddy were established in the first place, how their functions have changed since then, and why they changed. It is not even obvious that we still need these institutions at all. Perhaps whatever useful functions Freddie and Fannie perform could be handled more effectively by other government institutions. If we want effective new policies, we must not shirk from the task of governmental design and redesign. The Administration needs to be bold. Its trajectory so far, unfortunately, suggests retreat.

A Pocket History

Around the time of the Great Depression, mortgages on residential properties were typically short-term loans, with little if any principal paid until maturity. The flaw in this approach to mortgage finance became evident when property values declined precipitously in the 1930s. For many, home loans that were due and payable at maturity could not be refinanced because the amount of mortgage debt exceeded the value of the property. The Great Depression confirmed that Americans needed a more stable system of home mortgage finance.

The New Deal ushered in the first major effort by the Federal government to stabilize mortgage lending through Federal support and regulation. The doyens of the New Deal created two parallel structures that changed the way that residential loans were originated in the United States. For decades, both structures encouraged the financing of home loans without cost to the American taxpayer.

In the first of these new structures, the Federal government became the insurer of home loans. The Federal agency offering the insurance was the Federal Housing Administration, or the FHA, an acronym used to this day despite the fact that the functions of the FHA have been absorbed by the cabinet-level Department of Housing and Urban Development. The FHA, created in 1934, provided insurance to a lender in the event of mortgage default by the borrower. FHA charged a premium for its mortgage insurance that was paid by the lender but borne by the borrower. Loans eligible for insurance met statutory standards designed to keep default risk low.

Second, the Federal government enabled a special type of depository institution—most often known as a thrift or savings-and-loan (S&L) association—to become primarily engaged in the origination of home loans. Through the oversight of various regulatory agencies over the years, the Federal government offered member-paid deposit insurance and reserve credit to participating depository institutions. The S&Ls were willing to pay for this insurance because it allowed them to attract deposits as a secure funding source for home mortgage lending.

During this simpler era of mortgage finance, the Federal government was also instrumental in regulating the interest-rate environment for mortgage lending. Until 1983, the FHA had the authority to set a maximum interest rate for insured mortgages. For approximately the same period, Federal bank regulators effectively controlled a thrift’s cost of funds by regulating the interest rate paid for insured deposits.

The FHA was technically a government-owned insurance company but, by hitting the ground running at the start of the New Deal, it assumed a key standard-setting role in home mortgage lending generally. The FHA developed or implemented borrower underwriting requirements and minimum property standards. It even licensed participating lenders. As an insurer, the FHA collected a vast amount of default and claims experience for underwriting loans that provided insight for future Federal housing initiatives. FHA employees developed the reputation of being hard-nosed actuaries who started the practice of redlining (that is, a blanket denial of credit to anyone located within a struggling, typically minority, geographic location) as a form of underwriting.

Unlike FHA mortgage insurance, which prescribed the type of mortgages that could be insured under the National Housing Act, Federal deposit insurance attached to lender liabilities rather than to those of borrowers. It attached, for example, to certificates of deposit. This was useful because Federal deposit insurance gave thrifts an easy way to attract funds for mortgage lending. Thrift regulators were responsible for the safety and soundness of thrifts, but did not design mortgage lending programs. As a result, thrifts had more flexibility in loan origination than FHA lenders. Thrift officers tended to look and act less like actuaries and more like Jimmy Stewart’s George Bailey character from It’s a Wonderful Life.

Whatever the fragilities of this system, it worked for decades. Compared to the situation before the New Deal, the new Federal role in the mortgage market helped it stabilize and expand. It gave people more confidence in the system, and it is no exaggeration to say that this confidence enabled the housing expansion of the American middle class. Nevertheless, a mid-life legislative makeover for FHA triggered a kind of role reversal for FHA and the thrifts. With the passage of urban renewal legislation in 1949, Congress began to consider how FHA insurance could be used to rejuvenate declining urban areas. On one side, Congress expanded FHA mortgage insurance to include a number of higher-risk initiatives. Congress also enacted programs that for the first time combined FHA insurance with mortgage subsidies for the poor. Many of the new programs required Federal appropriations for continued operations, a radical departure from past years when no taxpayer money whatsoever was involved in the Federal role. As a result of these changes, even the FHA’s profitable insurance programs became less popular due to standard mortgage limits that were often too low to finance entry to a suburban development. By the late 1960s, FHA typically did not cater to the financing of suburban locations and higher income borrowers. Thrifts did.

From the start of the New Deal, most active FHA lenders needed to find a source of funds to originate FHA loans because they were not portfolio lenders; in other words, they did not use their own capital to make loans. The FHA lender was a generally mortgage bank—a company, often with limited capital, that needed to find funds from another institution (such as a bank or insurance company) to make loans. Very early in the life of the FHA, it was clear that FHA lending could be increased if a secondary mortgage facility for the purchase of FHA loans was always available. A secondary market reduces risks for lenders by enlarging the reservoir of money from which the loans flow. (Thrifts, on the other hand, did not need a secondary mortgage market, at least initially, because funds from insured deposits could be aggregated in sufficient volume to fund home loans.)

The congressional response to FHA’s need for a secondary mortgage market was Fannie Mae. Starting its life in 1938 as a Federal agency, Fannie Mae provided a secondary mortgage market for FHA-insured loans by buying loans from FHA lenders. As the Federal government expanded into new mortgage insurance or guarantee programs, Fannie Mae expanded its secondary mortgage market activity in lockstep. For thirty years, Fannie Mae performed the limited but important function of buying and selling Federally insured or guaranteed loans. The FHA and the other Federal agencies that were to assume a role in housing (for example, the Veterans Administration) were responsible for the heavy lifting in setting policy for program eligibility, operations and claims. Fannie Mae, in contrast, conducted purchases and sales of these Federally insured and guaranteed mortgages. In response to market conditions, Fannie Mae served as mortgage owner for periods of time. With a rapid increase in agency mortgage volume after World War II, the funding needs of Fannie Mae continued to increase. Federal funding was the exclusive financing source of Fannie Mae until 1954, when lenders selling mortgages to Fannie Mae were required to buy non-voting common stock in Fannie Mae. This new private source of revenue did not stop Fannie Mae’s need for an ever-larger piece of Uncle Sam’s budgetary pie, however, as its secondary mortgage operations grew.

With the Vietnam War placing increasing demands on the Federal budget, President Johnson decided that Fannie Mae could be converted to a private company as a means to eliminate Fannie Mae’s reliance on Federal funding for operations. This shift was achieved through the Housing and Urban Development Act of 1968 (hereafter the “1968 Act”), which re-engineered Fannie Mae as a Federally chartered corporation. No longer within the Federal government, Fannie Mae became a corporation trading on the New York Stock Exchange. The 1968 Act authorized Fannie Mae to maintain a secondary mortgage market for Federally insured or guaranteed loans with a Federal backstop.

The 1968 Act concurrently provided a Federal charter for another corporation, the Government National Mortgage Association (“Ginnie Mae”), located within HUD. Ginnie Mae continued certain special assistance functions previously performed by Fannie Mae, such as the purchase of below market interest rate FHA multifamily loans that enabled apartment borrowers to offer below market rents to low- and moderate-income tenants. It also became responsible for the development of a program for the guarantee of mortgage-backed securities backed in turn by newly originated FHA/VA loans.

With the spin-off of Fannie Mae from direct Federal control and the creation of Ginnie Mae for FHA/VA securitization, thrifts began to sense the need to change their status as portfolio lenders, and they began to clamor for their own secondary mortgage market facility. Congress responded in 1970 with the creation of the Federal Home Loan Mortgage Corporation (“Freddie Mac”).

The mortgage-backed securities program developed by Ginnie Mae served as a precedent for Freddie Mac’s own program design as it set up shop, and it also became a structural model for the mortgage-backed securities program created by Fannie Mae in the 1980s. Because the FHA had spent decades focused on underwriting and insurance, Ginnie Mae was able to limit its function to securitization. But both Federal agencies had certain similarities. Ginnie Mae as guarantor, like FHA as insurer, did not fund loans. Both the FHA and Ginnie Mae instead provided a Federal backstop for them. The difference between the two resided in the identity of the recipient of the Federal backstop: the FHA insured licensed lenders, nearly always an institutional entity, while Ginnie Mae guaranteed timely payment to any registered certificate holder of its mortgage-backed securities, including individual investors.

In designing a mortgage-backed securities guarantee program, Ginnie Mae made the lender responsible for loan origination and creation of the mortgage-backed securities pool. Ginnie Mae charged a fee for its guarantee, deducted from interest paid by the lender. Ginnie Mae’s guarantee program operated consistently in the black so long as FHA/VA benefits accrued to the lender in the event of loan default, and the lender otherwise complied with program requirements concerning remittance of claim proceeds and so forth. Like the FHA and VA, Ginnie Mae licensed and regulated participating lenders. By attaching the full faith and credit guarantee of the United States Treasury to mortgage-backed securities, Ginnie Mae became the superior placement source for newly originated FHA/VA loans. As a practical matter, the full faith and credit guarantee of Ginnie Mae reduced the role of Fannie Mae as a secondary mortgage facility for FHA/VA loans.

With Ginnie Mae gaining significant market share as an outplacement source for FHA/VA loan originations, Congress began to expand the charter of Fannie Mae. In time, both GSEs received Congressional authority to purchase the same loan menu, although their respective customer bases differed due to historical accident. Fannie Mae’s early customer base tended to be non-depository institutions such as mortgage banks; Freddie Mac’s original customer base was the thrift industry. The two jointly developed a standardized market for conventional loans in the 1970s, similar to the function performed by the FHA in the 1930s for FHA insured loans.

What was slower to change, at least for Fannie Mae as the older GSE, was its capital structure. During its early years as a publicly owned company, Fannie Mae maintained an asset and capital structure closely resembling that of a thrift. Not only were its key assets long-term mortgages, but purchase of these long-term holdings were financed using capital plus short-term debt. In other words, Fannie Mae borrowed short to invest long, and the cumulative difference between the interest rates drove profitability. Thrifts operated under the same financing mismatch for decades, but they did so in a largely regulated environment. Fannie Mae borrowed in an unregulated environment. By the early 1980s, Fannie Mae was borrowing short-term at double-digit interest rates to finance a portfolio of long-term loans at single-digit interest rates. You don’t have to be a banker to see where a situation like that will eventually lead. Fannie Mae avoided a collision course with insolvency by adopting a mortgage-backed security approach similar in structure to the Ginnie Mae program. This enabled it to shift interest rate risk to the security holder because it was not “long” the investment. Fannie Mae simply guaranteed payment under the mortgage-backed security, thus providing a credit back-stop for the security holder should the lender of any pooled mortgage in default fail to make the appropriate advances of principal and interest to the security holder. Once Fannie Mae re-tooled itself from portfolio lender to guarantor in the 1980s, it joined Freddie Mac in finding that rising interest rates could provide a business opportunity.

As high interest rates became part of the financial landscape, many millions of dollars worth of seasoned, low-interest rate loans were stuck in the portfolios of thrifts and other lenders. Each GSE offered a “guarantor” or “swap” program for seasoned loans in response, allowing lenders to swap non-liquid seasoned loans for liquid GSE-guaranteed mortgage-backed securities. Loans eligible for guarantee included seasoned FHA/VA loans, which were traditionally not eligible for guarantee by Ginnie Mae. The GSE swap programs became a popular tool for senior management or investment bankers retained by a portfolio lender to provide exit strategies for large blocks of underwater loans. For each GSE, its swap program provided a large volume of profitable transactional activity; in other words, they made a ton of money off of fees.

With the approach of the 1990s, the increased profitability of the GSEs did not go unnoticed. Fannie Mae topped the list of stock picks among fund managers, including Fidelity Magellan Fund’s Peter Lynch. The increasing depth and sophistication of the financial markets enabled the GSEs to offer an array of mortgage programs to lenders. While the GSEs increasingly offered the same mortgage product line to the same seller/servicers, each had a different regulator. Until 1989, HUD had oversight of Fannie Mae while the Federal Home Loan Bank System maintained responsibility for Freddie Mac.

As the 1980s came to a close, a consensus emerged that these inconsistent regulatory structures made no sense for entities that now performed nearly identical functions. Originally, the structure made sense, but “facts on the ground” that occurred during the 1980s required Congress to revisit the legislation, which it did. At the time, another factor impelled regulatory change of the GSEs: the liquidation of massive numbers of insolvent thrifts. The Depository Institutions Deregulation and Monetary Control Act of 1980 and the Garn-St. Germain Depository Institutions Act of 1982 expanded permitted investments for thrifts. Federally guaranteed certificates of deposit, made attractive to the public under the Acts by allowing thrifts to pay market interest rates and offer increased Federal insurance coverage due to a higher ceiling, funded this high-risk investment spree. What Congress failed to require and the then-thrift regulatory agency, the Federal Home Loan Bank Board, failed to put in place was more thrift capital to provide a cushion for the default of higher risk loans and investments. The legislative and regulatory action–and inaction–of the early 1980s had the effect of giving the already capital-challenged thrift industry more rope to tighten its insolvency noose. By the time Congress responded to the thrift crisis, hundreds of thrifts that were part of Freddie Mac’s customer base were insolvent. The rationale for maintaining Freddie Mac as a captive secondary-market institution for the thrift industry began to disappear as the thrift industry itself shrank.

Enacted with the 1989 thrift clean-up legislation was a temporary regulatory structure for the GSEs, pending further study at the Federal level. At the time, as noted, few questioned the continued need for the GSEs, or for so many of them doing essentially the same thing. Both Freddie and Fannie were at their high point in profitability and Congressional admiration. By comparison, the two original New Deal creations that supported mortgage finance—the FHA and the thrifts—played a more diminished role.

By the time the urge to reform the GSE regulatory structure produced a new approach, it did so by concentrating authority in HUD. The Federal Housing Enterprises Financial Safety and Soundness Act of 1992 (the “1992 Act”) created an independent office within HUD called the Office of Federal Housing Enterprise Oversight with exclusive authority over the financial soundness of the GSEs. In other administrative areas, HUD held general regulatory authority.

Key to HUD’s oversight as “mission regulator” of the GSEs was a detailed set of housing affordability goals requiring the GSEs to lend or purchase specified loan types. HUD determined loan eligibility based on an underserved location or a moderate, low or very low-income borrower. While the GSE housing affordability goals bore some resemblance to old Community Reinvestment Act requirements for depository institutions, in the 1960s the FHA provided subsidies or insurance for borrowers of comparable income or in similar locations. The new arrangement provided no Federal money for the purpose. Congress essentially replaced appropriations to the FHA with housing goals for GSEs as a way to keep the lid on Federal spending for housing programs. In so doing it expected to condition the market by fiat rather than by acting with in. Predictably, the effort had perverse effects.

As a quid pro quo for compliance with the housing affordability goals of the 1992 Act, the GSEs received three benefits. First, the GSEs gained exemption from all state and local taxes except property taxes. Second, they gained exemption from Federal and state securities registration requirements. Third, they gained conditional access to a line of credit from the U.S. Treasury. These benefits helped the GSEs, but once the 1992 Act was firmly in place, Wall Street stopped viewing the GSEs as enablers to bank activity as they had during the swap era of the 1980s. Wall Street now considered the GSEs to be a form of unfair competition, particularly the lower cost of GSE funds based on their implicit Federal guarantee—amounting to something like $2 billion savings per year according to Congressional Budget Office and Treasury Department estimates. Around the same time, the GSEs lost their monopoly on knowledge about mortgage markets. Wall Street had assumed a leading role in the purchase of mortgage assets from Federal thrift liquidators in the early 1990s, developing expertise comparable to the GSEs for a large number of markets.

Despite Wall Street’s growing interest and expertise in the mortgage market, the GSEs could largely determine whether Wall Street would have a bidding opportunity in the first instance for certain specialized types of mortgage origination. The GSE could do this by skewing offer terms to seller/servicers. If a GSE wanted to buy certain mortgages to the exclusion of Wall Street, it could offer more favorable purchase terms given the lower cost of GSE funds. If the GSE did not want to hold a mortgage position, it could skew its offer terms to a guarantee of mortgage-backed securities. Only for guaranteed securities did Wall Street typically have an opportunity to bid. As a result of this practice, some specialized origination types could disappear from the secondary mortgage market for years if the GSE wanted to accumulate a large position in them. What drove the decision by the GSE to purchase or guarantee certain specialized product lines was its own capital limitations or profitability considerations.

The lack of a level playing field between Wall Street and the GSEs was also evident in the secondary market, where each GSE competed against broker-dealers and other institutional investors for its own guaranteed mortgage-backed securities. As the guarantor, the GSE had more information about the underlying loans than it typically made publicly available to competitors. And due to the implicit Federal guarantee of the GSE, the borrowing costs of the GSE to buy the securities were lower than comparable costs for competitors. An institutional investor simply couldn’t compete with the GSE in profitability in executing an identical trade. The GSE had superior information and a lower cost of funds.

Even smaller loan originators began to view the GSEs suspiciously around this time, as GSE business plans for growth began to invade the turf of direct mortgage lending. Like a drug company that discovers direct consumer advertising to patients, Fannie Mae extensively advertised itself as “America’s Housing Partner” in the 1990s; and even a post-conservatorship Freddie Mae continues to assure a jaded American public that “We Make Home Possible.”

As the 1990s progressed, private lenders stopped thinking of GSEs as useful instruments to develop new mortgage products. Instead, they worried that the GSEs would find ways to enter a developing mortgage market so as to eliminate private sector competition and profitability. The GSEs had turned from partners to predators, forcing private lenders to seek opportunity by increased the volume of niche products outside of the plain vanilla GSE and the relatively safe FHA/VA loan menu. The initial product of entry for many private residential mortgage programs was the “jumbo mortgage”—a mortgage with a loan amount in excess of conventional loan limits. The Federal government annually computed conventional loan limits based on median home prices. Maximum loan limits for FHA loans traditionally were lower than conventional loan limits, but the FHA required only a modest down payment. Congress allowed the GSEs to offer higher loan limits for conventional financing. In setting underwriting standards for most conventional loan types, the GSEs required a down payment higher than the FHA and, depending upon the amount of the down payment, private mortgage insurance. But as the 1990s evolved into the Age of the McMansion, living large translated into a jumbo mortgage. Private lenders lined up to offer them.

The whole jumbo business was risky enough, but what remains insufficiently understood is that the idea of a subprime mortgage also came about as a result of private lenders’ desperation at the invasion of their business turf by the GSEs. Subprime loans began in the mid-1990s as a way for borrowers with impaired credit—even those who could not qualify for FHA financing—to become homeowners. Subprime started as a small specialty, not restricted by FHA mortgage limits, but requiring close attention by loan servicers. Few realize that, at least during the 1990s, subprime was on the path of providing housing for poor credit risks without a run-up in default rates. At the time, lenders followed strict servicing procedures for this new higher-risk mortgage product. Risks were controlled by professional discipline. But after a promising start-up, the profitability of subprime attracted many new lenders without regard for servicing discipline. As a mainstream mortgage product, subprime became an unmitigated disaster for borrowers, lenders and investors, as everybody realized after the fact.

By the time the Millennium arrived, an even more basic change had emerged for many of the newer private label residential mortgage products. The down-payment requirement, a staple of mortgage loan underwriting, began to disappear.

Until this point, little or no down payment was possible only under FHA, VA or other government programs designed to expand housing affordability. Default rates under these programs were inevitably higher due to relaxed down-payment requirements, but those who were able to achieve homeownership without default, in the eyes of the programs’ sponsors, made the tradeoff worthwhile. At least in the FHA experience, the total elimination of even a small down-payment by the borrower tripled the risk of default. Under the 1968 Act, FHA insured approximately 400,000 mortgages that loosened the down-payment requirement for the low-income. Even substantial interest subsidies paid directly to the lender on behalf of the borrower could not prevent higher default rates. As the urban laboratory for then-HUD Secretary George Romney, Detroit became an especially noteworthy casualty of the FHA’s new insurance program for the low-income, with about a third of all of its residential property estimated to have been acquired by HUD in the early 1970s.

Around the same time that private label mortgage programs started to eliminate down-payments, the FHA involuntarily repeated its disastrous claim experience for loans without a down-payment. Non-profit organizations attempting to help the low-income become homeowners started combining FHA insurance with a special “gift” down payment from the seller to the buyer through the non-profit as intermediary. In essence, they hijacked the FHA without its permission. FHA claim statistics revealed high default rates for loans originated through non-profit gift down payment programs. The FHA attempted to stop the use of FHA insurance with non-profit gift down payment programs in court, but failed. It was not until the passage of the Housing and Economic Recovery Act of 2008 (the “2008 Act”) that Congress forbade the practice.

More trouble was brewing elsewhere, however. For example, lender initiatives that began in the 1990s to simplify loan underwriting for borrowers morphed into “no underwriting” programs only a few years later for certain new mortgage types. “Alt A” loans became the best example of this. These private label mortgage products chipped away at the established underwriting standards developed by the FHA and the GSEs. At the same time, a steady decline in interest rates for most of the past decade triggered a grab for yield by mortgage investors with little regard for the increased credit risk. As secondary mortgage markets became more sophisticated, in theory—and certainly more willing to assume higher levels of credit risk—the GSEs were no longer a vital intermediary. Lenders could now pool high-risk loans into private label residential mortgage-backed securities for sale in the secondary mortgage market. They did not need the GSEs. The marketability of these pools to yield-hungry bond buyers, not underwriting fundamentals, ended up driving loan originations.

The reckless lending allowed by the new private residential mortgage programs reduced mortgage origination activity by the GSEs and FHA/VA lenders. By 2005, Ginnie Mae and the GSEs issued or guaranteed, in the aggregate, less than half of all mortgage-backed securities. FHA origination activity, subject to a myriad of rules and red tape long associated with FHA, slipped into the single digits.

Many are of the view that the GSEs did not initiate directly the high-risk mortgage lending practices that were at the source of the credit crisis other than activity required by HUD to meet GSE “housing goals.” Lenders and other intermediaries feeding the private mortgage label programs did that, fueled by a secondary mortgage market willing to accept higher risk mortgage product if accompanied by a seductive agency rating. But if one peels the onion a little deeper, it is clear that this risky behavior was stimulated in the first place by the way the GSEs used their advantages to outflank and out-complete the private label residential mortgage sector. They are the ones whose avaricious business practices drove private lenders to increasingly riskier decisions.

Eventually, however, the proverbial chickens came home to roost. As the private label lenders tried to outflank the GSEs through riskier ventures, the GSEs tried to elbow their way into this riskier action as well. Like any publicly traded company, the GSEs needed to maintain growth in earnings. They are the ones who started buying high-risk loans such as Alt A, and started re-guaranteeing little understood private label mortgage-backed securities backed by high-risk loans (a decision with even more disastrous consequences). This is what did them in financially. Though the GSEs were late to purchase and guarantee higher risk mortgage products, they had a lower capital cushion compared to a commercial bank or even an investment bank. They relied on debt to finance their extensive mortgage activities, and the aggregate amount of GSE debt came to exceed the amount of outstanding Treasury debt as the year 2000 approached.

Given the line of credit to the U.S. Treasury available to the GSEs under the 1992 Act, one might think that Congress watched these developments with some anxiety, to rein in such high-wire acts. Congress did no such thing. As the horse left the barn, Congress passed the 2008 Act to strengthen the safety and soundness of GSE regulation and establish procedures for GSE conservatorship. The latter occurred just in the nick of time. Only a few months later, the GSEs were put into conservatorship by the Federal government.

The private label residential mortgage-backed securities market was in tatters as a result of the 2008 credit crisis. By late 2009, the Treasury lifted all dollar limits on its credit line to the GSEs. The FHA, now enabled with temporary but higher mortgage insurance limits, and the GSEs, despite behavior that had largely instigated the whole mess in the first place, became the source once again for most loan originations in the United States. If this sounds unfair, that’s because it is.

The insolvency of the GSEs shows eerie similarities to the thrift insolvencies that occurred almost twenty years earlier. Like the thrifts then, the GSEs operated with low capital, a continued drive to maintain profit growth, and an unlimited borrowing capacity due to a Federal backstop. The key difference between GSEs and the thrifts, perhaps, was the international implication that trumped all legalities governing the GSEs. The shutdown of a thrift was largely a domestic affair, with detailed regulations in place governing the payment of Federally insured deposits. The implicit Federal guarantee of the GSEs, as a marketing tool for the international placement of GSE debt, gave the Treasury no practical option but to backstop GSE debt to avoid an international debt crisis. Of little surprise, the Treasury allowed only GSE debt obligations to receive this favored treatment. The Treasury credit line was not available to pay common or preferred stock dividends of the GSEs. Holders of GSE common or preferred stock were largely wiped out once the GSEs were placed in conservatorship.

Lessons

It is clear that the GSEs at one time engaged in activities similar to those performed by various Federal agencies. They developed underwriting standards for conventional loans similar to the standard-setting function performed by the FHA in an earlier generation. Fannie Mae, for its first thirty years, bought FHA insured mortgages. And the GSEs developed mortgage-backed programs for conventional loans and seasoned FHA/VA loans, after Ginnie Mae—an agency within the Federal government—developed a mortgage-backed securities program for newly originated FHA/VA loans that continues to this day. While the GSEs designed mortgage-backed securities programs responsive to the needs of seller/servicers, the real source of success for GSE-guaranteed mortgage-backed securities (and GSE mortgage purchases, for that matter) was an implicit Federal guarantee that became explicit under duress. Put a little differently, the GSEs did good things. For a time they helped allocate capital more efficiently than markets alone were able to do, and they performed social equity functions that the national ethos, as expressed through the Executive and Legislative branches of government, deemed more important than what markets by themselves would do.

But whenever government distorts markets, for whatever purposes, it skews incentive structures in ways that have unpredictable outcomes. When that skewing coincides with significant changes in the context of market activity—as with globalized finance as it developed in the past quarter century—one compounds uncertainties to the point where one toys with systemic risk. We have paid the price; the question now is what to do about Freddie and Fanny in circumstances dramatically different from those that existed at their creation.

This closer one looks at the history of Federal support for residential mortgage lending the more one sees that the GSEs and various Federal agencies that supposedly regulated them have performed overlapping functions over the years. The only unique feature of each GSE was its off-budget status. But removing an activity from the Federal budget only served to hide how large, and how expensive, it had become when profits were privatized and losses socialized.

Nevertheless, even as the credit crisis was persisting and the GSEs were placed into conservatorship, Congress transformed the FHA from a declining mortgage insurance program to the center of mortgage lending in the United States simply by increasing the loan limits for FHA insured mortgages. Ginnie Mae mortgage-backed securities served as the primary placement source for such increased lending. (Ginnie Mae is on budget, and it has typically run in the black.) Of course, GSE activities are all off budget and as a de facto intervention into the market will predictably retard and distort the recovery of the private mortgage sector.

A look at current refinancing policy, in the midst of historically low interest rates, shows why the FHA needs to be enlarged and the GSEs shut down. FHA has created a streamline refinance program to encourage FHA borrowers to reduce their household monthly mortgage payments because housing that is more affordable for borrowers also lowers FHA’s insurance risk. A GSE as guarantor has a similar interest in encouraging refinancing, but a GSE as investor wants to prevent the payoff of a higher interest mortgage; the GSEs have allowed this latter view to prevail to date, by raising guarantee fees to discourage billions of dollars of refinancing by financially stretched borrowers.

Given these realities, GSE housing programs considered worthy of Federal support should be moved back on to the Federal budget. Moreover, GSE activity should continue only as a Federal insurance or guarantee program eligible for placement through Ginnie Mae mortgage-backed securities. GSE mortgage assets that do not fall within these parameters should be sold off through an orderly liquidation process not different in essence from the thrift liquidations of the early 1990s. The GSEs need to stay in their lanes and not compete with the private mortgage sector using their government-associated advantages to tilt the playing field to their own advantage. If we do not ensure this line of separation, then we can expect the same destructive race toward the risky to start up all over again in due course. Recent proposals to incrementally reduce the maximum mortgage amounts eligible for GSE securitization or purchase (that is, the “conforming limits”) are a sound way to gradually scale back GSE origination activity given current fragile housing markets.

What Congress hates about the FHA, the VA and similar Federal housing programs is the line item in the Federal budget needed to cover program losses or subsidies. But what the history of the GSEs makes clear is that the implicit Federal guarantee of an off-balance sheet entity is not in the long run a cheaper or more efficient way to support mortgage finance. It is a far, far more costly method.

Editor’s Note: On August 16, the Washington Post reported that the Obama Administration is considering proposals that would retain a major role for the Federal government in the nation’s mortgage market.

¹ Fannie Mae and Freddie Mac are not, however, the only GSEs. The first was the Farm Credit System created in 1916; the most recent is Sallie Mae, created in 1972 to deal with the financing of higher education. Fanny and Freddie are, nonetheless, by far the largest of the GSEs in terms of dollar volume.