Chairman Dodd, Ranking Member Shelby, and members of the Committee:
I very much appreciate this opportunity to testify before this Committee. The role and structure of government sponsored enterprises (GSEs), and particularly Fannie Mae and Freddie Mac, has been an interest of mine since I was General Counsel of the Treasury in the early 1980s.
While most of the attention to the GAO report in today's hearing will focus on the agency's analysis of the options for their future, the report contains a lot of useful background about government housing policies in general that the Committee should take into account. Table 1, for example, records the large number of housing programs that the U.S. government now pursues, beginning with the establishment of the Federal Home Loan Bank System in 1932, the Federal Housing Administration in 1934, and Fannie Mae in 1938, and extending through the many programs for direct outlays currently run by HUD and the tax subsidies that are enjoyed by most homeowners today. The sheer number of these programs is a reminder that there will still be plenty of government support for housing and home ownership in the United States, even if Fannie Mae and Freddie Mac are ultimately privatized or liquidated.
Unfortunately, the table does not include or describe the system of savings and loan associations (S&Ls), originally operated and regulated by the Federal Home Loan Bank Board (FHLBB). The Board was the predecessor of the Office of Thrift Supervision, which now regulates the S&Ls that are a vestige of the much larger system that collapsed in the late 1980s. One of the reasons for abolishing the FHLBB was that its mandate included the promotion of housing, and that was deemed to be inconsistent with the responsibility of a regulatory agency. As I will discuss in this testimony, the same issue of conflict in missions occurs in the case of Fannie and Freddie and severely impairs their effectiveness. Indeed, for more than just this reason, the example of the S&L system has great relevance to what the Committee is considering today--both the future of Fannie Mae and Freddie Mac and the wider issue of new regulations that are intended to prevent the recurrence of the financial crisis we are now experiencing.
The Lesson of the S&Ls
The S&L system was established in 1932 to provide housing finance. At the time, this was only the latest in many government efforts during the 20th century to increase home ownership in the United States. For good reason, Americans believe in home ownership. It has many indirect benefits--better housing, better neighborhoods, better family conditions, less delinquency, and others--that are worthy of government support if they do not occur through operation of the market alone. But recalling the history of the S&Ls and their collapse as an industry is important for understanding what we should do with Fannie Mae and Freddie Mac today.
This is true for two major reasons. First, the S&Ls were an attempt by Congress to use a financial mechanism--a depository institution--to achieve a social purpose, an increase in home ownership. S&Ls were limited to making housing loans, which meant that they were locked into a structure in which they were compelled to carry long-term assets with short-term liabilities--a prescription for depository institution failure if there ever was one. Commercial banks also carry long-term assets with short-term deposits, but they have many more short-term assets they can acquire. Fannie and Freddie were initially developed simply to make the secondary mortgage market function more effectively by purchasing mortgages from banks and S&Ls, thus making these long-term assets more liquid. However, after the S&L industry failed, Congress adopted another way to increase home ownership. In this case, Congress gave Fannie and Freddie an additional responsibility--an affordable housing "mission"--intended to direct housing finance resources to certain groups that were thought to be underserved in the normal credit markets. This effort worked too well. The loans made to the borrowers Congress designated, borrowers who did not have the financial resources or the credit standing necessary to meet their obligations, were largely responsible for Fannie and Freddie's insolvency. Accordingly, the failure of both the S&Ls and the GSEs should tell us that attempts to manipulate financial institutions in order to achieve a particular social purpose are likely to end badly.
Second, the S&Ls failed because the system was not flexible enough to survive in a market where interest rates were set by supply and demand. This is important, and bears on many of the issues raised by the financial crisis. The S&Ls remained a stable source of housing finance only during a unique time--when the deposit interest rates at banks and S&Ls were controlled by regulation and depositors had nowhere else to go. Under what was known as the Fed's Regulation Q, banks could not pay more than 5 percent interest on deposits and S&Ls could not pay more than 5 and 1/4 percent.
This created a very stable banking system for many years, and many of the advocates of greater regulation today point to this period--roughly from the end of World War II through the end of the 1970s--as a period of "great moderation" when we didn't have many banking crises. The implication is that we should have more regulation now. What the proponents of regulation don't mention is that during this period we had many very serious recessions and housing finance crises when there was insufficient liquidity in the economy because the banks and S&Ls could not raise sufficient funds for lending when money market rates exceeded what Regulation Q permitted them to pay for deposits. After the deregulation of deposit rates in the early 1980s, we really did have a great moderation, with only two mild recessions until the early 2000s, when the Dot-Com bubble deflated. Even that recession did not implicate the banking system, which remained reasonably strong from the time of the S&L collapse (when almost 1600 commercial banks also failed) until the current crisis.
The lesson of the S&L collapse is that it was a serious policy error to impose a rigid regulatory structure on an institution that is supposed to be operating in a market where the cost of its principal raw material--i.e., money--is subject to the law of supply and demand. The policy worked for a while, as long as the general public had no other choices, but with the advent of ordinary money market mutual funds, people could get access to higher rates and safe short-term investments in the money markets and withdrew their funds from banks and S&Ls. These institutions were forced to replace these fleeing depositors with funding sources that required them to pay higher rates, and the losses that resulted (paying more for funds than the assets that they were holding were yielding) caused the collapse of the whole industry. It is important to recognize that Regulation Q penalized the public and trapped them in low-paying deposit accounts. In effect, they were freed by technological changes--primarily the advent of computers--that made it possible for mutual funds to calculate their net asset value at the end of every day.
The S&L analogue at Fannie and Freddie is that no financial institution can serve two masters. Government-sponsored enterprises--to the extent that they are owned by shareholders but also have a government "mission"--are living contradictions. They were set up to achieve two government purposes--creating a more liquid and efficient secondary mortgage market and reducing the interest rates on mortgage loans. But they are also private, shareholder-owned companies, and their managements have a fiduciary duty to maximize value for the shareholders. Just as the S&Ls' rigid structure could not survive in a market where their depositors had alternative investments, Fannie and Freddie could not serve both their government purpose and their shareholders at the same time. In the end, the shareholders come first--if only because in serving the shareholders the managements could assure themselves of rich rewards by exploiting the government franchise they had been given. This is part of the story of Fannie and Freddie, and why they did not actually reduce mortgage interest rates for the great middle class of the United States. As the GAO points out, Fed studies showed that the interest rate reductions attributable to their operations actually amounted to only 7 basis points.
One of the reasons that they achieved only this paltry sum for home buyers is their affordable housing mission, which was adopted by Congress in 1992. This created another inherent conflict of interest in their charter. In this case, Fannie and Freddie were required to devote a substantial portion of their resources to purchasing loans made to home buyers at or below the median income. When HUD first began to implement this mandate, the requirement was 30 percent, but it was ratcheted up over time, and by 2005 HUD's affordable housing regulations required that 55 percent of the loans Fannie and Freddie purchased had to be loans to home buyers at or below the median income, including 25 percent to low-income home buyers. Of course, the HUD regulations said that these loans were to be prudent, but Fannie and Freddie were also importuned to be "flexible" in their standards, and that resulted in their looking for and buying loans that had been made to people with blemished credit or limited ability to make downpayments. By the early 2000s, Fannie and Freddie were buying loans which involved no downpayment at all. The result is clear today. At the time Fannie and Freddie had to be taken over by the government, they held or had guaranteed 10 million subprime and Alt-A loans, with a total value of $1.6 trillion. These loans are defaulting at unprecedented rates, and when it is all said and done, cleaning up the mess at Fannie and Freddie will probably cost the American taxpayers $200 to $400 billion.
The losses that finally overwhelmed Fannie and Freddie were hidden for a long time in the huge profits that the two companies were able to earn from exploiting their government franchise. It looks today as though the allocation of those profits was pretty much as one would expect--first to the management, then to the losses incurred in their affordable housing mission, then to the shareholders in the form of dividends, and finally 7 basis points of benefit to home buyers. Of course, the embedded losses were reserved for the taxpayers, who never had an opportunity to reject the honor.
There are certainly good policy reasons for the U.S. government to encourage home ownership, but imposing the burden on companies that are supposed to be shareholder-owned and profit-making is not the way to do it. A government program that provides downpayments for people who can't afford them would make a lot more sense. Then the losses, if any, would be visible and could be balanced against the gains from increasing home ownership. But requiring Fannie and Freddie to perform this mission--to find an increasing number of "prudent" loans that met HUD's requirements--was a mission impossible, and the result is the insolvency of the two companies and huge eventual losses for the taxpayers.
The Future of Fannie Mae and Freddie Mac
What do these two lessons say about the future of Fannie and Freddie? First, I think the GAO's conclusions about the options available to Congress are correct. The realistic options are only nationalization, privatization, or a return to GSE status, but there are some ideas that are refinements of these general categories. For example, the Mortgage Bankers Association has proposed a well thought-out proposal that falls somewhere between privatization and GSE status. In addition, the Treasury under Hank Paulson published a plan for a covered bonds structure that might get the government completely out of the mortgage market. An evaluation of these ideas as substitutes for Fannie and Freddie is beyond the scope of this testimony. Instead, I'd like to review each of the possibilities raised by the GAO, beginning with a return to GSE status.
Fannie and Freddie as GSEs
From what I have said above, a return to GSE status would the worst choice, especially if Fannie and Freddie were to continue to have an affordable housing mission. That mission seems unnecessary when FHA's activities could be expanded to achieve the same result, and if the objective is to increase home ownership, a program that provides downpayments for prospective homebuyers with otherwise good credit records is likely to be more effective. At least such a downpayment subsidy program would be transparent, which Fannie and Freddie's affordable housing losses certainly were not.
But even if Fannie and Freddie are no longer required to support affordable housing, and even if their activities are limited to securitizing mortgages (so that they are prohibited from holding portfolios of mortgages and mortgage-backed securities), it would be a mistake for them to be set up again as GSEs. The GSE form is a prescription for moral hazard. If there had ever been any doubt that GSEs are backed by the federal government, the federal takeover of Fannie and Freddie in 2008 removed it. If they are again set up as GSEs, creditors will assume that the government will rescue them again if they get into trouble. There will be no market discipline, no market-based restraint on their risk-taking. They and their supporters will argue that strong regulation will prevent substantial risk-taking, but this is an error. Since FDICIA was adopted in 1991, in the wake of the S&L crisis, we have relied on the strongest regulation we could think of to make sure that insured banks were safe and sound. Yet, today, we have the worst banking crisis since the Great Depression. Accordingly, it seems clear that strong regulation cannot overcome the incentives of management--indeed their fiduciary obligations as managers of shareholder-owned companies--to exploit the GSE franchise to the maximum possible degree. No regulator will be able to tease out the myriad ways in which the management of a future GSE will be able to take risks in order to enhance the returns with which they and the shareholders will be rewarded. Risk-taking is appropriate for private companies--they should take risks for profit--but not when companies are operating with the taxpayers' credit card. Yet that is exactly what we will be doing if Congress accepts the facile argument that strong regulation will prevent serious risk-taking and losses.
I should add here that if Fannie and Freddie return as GSEs, and creditors assume that their liabilities are backed by the federal government, the potential losses on their activities will be greater than the potential losses to the government arising out of the FDIC's insurance on bank deposits. Bank deposits are only insured up to $250,000, but all of Fannie and Freddie's debts will be covered in the event of another failure in the future. So the stakes will be high for the taxpayers if Fannie and Freddie are returned to GSE status.
Nationalization of Fannie and Freddie
The next question, then, would be whether nationalization would solve this problem. In other words, if Fannie and Freddie were combined into a single government institution, could they more effectively perform their secondary market role without a danger of excessive risks to the taxpayers? In effect, the new entity would be doing what Ginnie Mae is doing, but for a broader range of mortgages. Because a government agency would have no profit motive and no capital requirement, it could, in theory, offer less expensive mortgages.
Again, the question arises whether the new entity would have an affordable housing mission, and whether that obligation would require them to take on the excessive risks that Fannie and Freddie seem to have taken on in pursuit of that mandate. In addition, while there is little incentive for a government entity to take risks, there is also little incentive to be careful about the credit risks they might be taking on inadvertently. Unless the agency can pay the salaries necessary to attract high-quality employees, its staff may not be able to understand the complexity of the mortgages that might be created in the future. As we saw with Fannie and Freddie, these risks can build up over a long period and not come to light. When they do, the losses can be substantial, in this case for the account of the government. In this connection, the government entity could be securitizing trillions of dollars in mortgages, and only small errors in risk management could be very costly over time.
The GAO report does not consider the budgetary impact of nationalizing Fannie and Freddie. Fannie was originally turned into a private company in order to take it out of the budget process. As the housing market grew, Fannie's purchases of mortgages were larger than the revenues it received on the sale or refinancing of the loans, and this added to the budget deficit. The same problem would appear to arise if Fannie and Freddie were now to be nationalized. Even if they are no longer permitted to accumulate portfolios of mortgages, their purchases of mortgages will precede the sale of these loans to trusts or other special purpose entities in the securitization process, and this will add to the deficit. This phenomenon will be more pronounced in a growing housing market, when the size of the GSEs purchases will precede the proceeds of sale in a securitization.
The Nexus between Fannie and Freddie and the Administration's Reform Proposals
At this point, it is worthwhile to consider the nexus between the issues that concern the future of the GSEs and the issues that arise in connection with the Committee's consideration of various proposals to prevent a repeat of the financial crisis. The GAO did not address these issues, but they should be of concern. As part of its effort to prevent another financial crisis, the administration has proposed that in the future banks hold more capital and the securitization process be revised so that both loan originators and securitization sponsors retain some portion of the credit risk associated with the securitized mortgages. These proposals have important implications for the restructuring of Fannie and Freddie--whether they are re-established as GSEs or merged into a single government entity. The new capital requirements and securitization rules, if they go into effect, will increase the costs of securitization-based credit. This seems to be acceptable to the administration, apparently because it believes it will reduce or eliminate the risk of another financial crisis.
But these new capital requirements and securitization rules will have major implications for Fannie and Freddie as GSEs or as a consolidated nationalized entity performing a secondary market function. In both cases, explicit government backing--or its equivalent in the case of the GSEs--would have the potential to substantially reduce the cost of the mortgages that go through a securitization process run by a GSE or a government entity. At the same time, the new capital and securitization requirements for private sector operators would substantially increase their costs. The gap between the cost of mortgages in the two systems--government and private--could be very wide. Mortgages that fall within the conforming range for government or GSE securitization would have major advantages over those that do not, and this could distort investment in the housing market. This will substantially increase pressure for the government or the GSEs to take over all secondary market securitization. The usual groups--homebuilders, realtors, and others in the business of constructing or selling homes--will press Congress to cover all mortgages, not just those that are at or below some maximum permitted size. If Congress accedes to this pressure, it will significantly increase the amount of mortgage debt that becomes a government risk. For this reason, when the Committee gets to a consideration of the administration's proposals for reform of bank capital and the securitization process, it should weigh these in light of their effects on the future role of Fannie and Freddie in the housing finance system.
In addition, if Fannie and Freddie survive as GSEs, there is a question whether they will be subject to the both the new capitalization and securitization requirements. If we assume that they will be, then we are starting down the track of allowing Fannie and Freddie again to accumulate a portfolio of interests in mortgages. To carry these interests, they will be required to borrow, and if they borrow they will be required to hold more capital in order to protect the government against losses. However, the government's potential exposure would grow over time, and could get quite large if the GSEs take on growing numbers of mortgages for securitization.
The previous discussion suggests that there are serious flaws and taxpayer risks associated with both the GSE and the government agency structures. Ideally, Congress and the administration should be considering new and better ways to finance residential housing in the United States, but there is no indication that any effort is being made to address this issue. Under these circumstances, the Committee should consider the privatization of Fannie and Freddie as a better policy than the two flawed approaches we have previously discussed.
There is no reason in principle why mortgages cannot be securitized through solely private sector activity, like any other asset that creates a cash flow. Car loans, boat loans, insurance premiums, credit cards, and many other assets have been securitized without problems. The difficulties in the mortgage market come primarily from government interventions to promote home ownership in ways described above. Indeed, the current freeze-up in the asset-backed securities market was caused by investors' loss of confidence in mortgages and rating agencies after unexpected losses appeared in pools of subprime mortgages that had been rated AAA. The resulting losses to investors caused the entire asset-backed market to shut down in 2007, and it has remained largely closed since then. The right kind of reforms--simple requirements, such as downpayments for mortgages and transparency for the underlying rationale of the rating it received--will encourage a return of investor confidence, although it will take time.
There are many advantages to a fully private housing finance market--some of which were clear in the lessons of both the S&L collapse and the failures of Fannie and Freddie. Principal among these advantages is the fact that the taxpayers are unlikely to suffer any losses on a fully privatized mortgage finance system. Failures in today's mortgage financing system are increased, not reduced, by government backing. Government support creates moral hazard; creditors don't pay attention to risk-taking because they believe the government will ultimately bail them out, and regulators regularly fail to prevent excessive risk-taking. All these factors increase the risk of failures. In a fully private system, however, creditors will not lend to a Fannie or Freddie if they believe the company is undercapitalized or taking excessive risks. If the mortgages are securitized through structured arrangements, investors will insist on full disclosure concerning the nature and risks of the securities they purchase, and, given the recent track record of rating agencies, will want to know how a rating on a particular tranche in the structure was established. This will mean that the taxpayers will not again be burdened with hundreds of billions of dollars in losses by government-backed vehicles that were able to take unreasonable risks because of their government support.
In addition, a private system will encourage more innovation, efficiency, and competition; with many other players joining the secondary mortgage market, competition should bring down mortgage rates. Privatized entities would have the flexibility to react to changes in the economy and the financial markets, and the incentives to do so. Finally, privatized companies are not likely have an obligation to provide affordable housing financing to targeted groups--a mission that was responsible for Fannie and Freddie's overwhelming losses. This mission would be assigned to government agencies such as FHA, so the losses--if they occur--will be transparent.
Privatization can be achieved relatively easily after investor confidence in securitization returns. The simplest way would be to gradually reduce the size of the loans that Fannie or Freddie are permitted to buy. This will gradually move them out of the market and make room for new private sector entrants. It will also probably stimulate the development of new ways of financing mortgages, such as covered bonds, or the MBA's recent proposal, which--although designed to use a government guarantee--could work as a fully private sector structure. If at any time the reduction in the GSEs' role is interfering with the orderly financing of mortgages, the process can be stopped. In current market conditions, it would not be good policy to reduce conforming loan levels--investors are still too nervous about the private securitization process--but as investor confidence returns, and in the absence of any new thinking on how to finance housing in this country, this approach would be the best way to prevent future taxpayer losses while creating a viable housing finance system.
To conclude, the choices available to the Committee are rather limited. Both the GSE and nationalization option have serious flaws that probably make them unworkable. That leaves some form of privatization. There are many good reasons to adopt a privatization strategy as the future of Fannie and Freddie, but the best is that as private entities without an affordable housing mission, they will not create losses for the taxpayers. Ultimately, however, we must develop a better system of financing housing in the United States, and it is an unpleasant fact that no serious thinking along these lines appears to be going forward in Congress or the administration.
Peter J. Wallison is the Arthur F. Burns Fellow in Financial Policy at AEI.