Testimony

GSE Reform: Immediate Steps to Protect Taxpayers and End the Bailout

Subcommittee on Capital Markets & Government Sponsored Enterprises
Committee on Financial Services
United States House of Representatives

Chairman Garrett, Ranking Member Waters, and distinguished members of the Subcommittee, I thank you for the invitation to appear at today’s important hearing. I am Mark Calabria, Director of Financial Regulation Studies at the Cato Institute, a nonprofit, non-partisan public policy research institute located here in Washington, DC. Before I begin my testimony, I would like to make clear that my comments are solely my own and do not represent any official policy positions of the Cato Institute. In addition, outside of my interest as a citizen, homeowner and taxpayer, I have no direct financial interest in the subject matter before the Committee today, nor do I represent any entities that do.

Need for Reform
Given the central role of Fannie Mae and Freddie Mac in the financial crisis, the need for reform should be beyond dispute. What will be debated is the substance of such reform. While I believe a major overhaul of our federal mortgage policies should happen sooner, rather than later; reform should be done in a deliberate and thoughtful manner. The need for a deliberate and thoughtful process, however, does not preclude the necessity of taking immediate steps to protect the taxpayer and reduce the perverse incentives that permeate our financial system. My testimony will focus upon those steps which Congress and the Administration should take immediately.

Receivership, not Conservator
The most immediate and powerful step that can be taken to protect the taxpayer is to change the role of the Federal Housing Finance Agency (FHFA) from that of conservator to receiver. Section 1145 of the Housing and Economic Recovery Act (HERA) of 2008 establishes a resolution and/or reorganization process for the GSEs. Unlike the conservator powers in Section 1145, the receiver provisions allow losses to be imposed upon the GSEs’ debtholders, rather than the taxpayer.

It should also be noted that there is little, if anything, that a conservator can do that a receiver cannot. There is, however, a considerable amount that a receiver can do, which a conservator cannot. As mentioned, the most important difference is that a receiver can impose losses on creditors and other parties. This would also subject the remaining shareholders, subordinated debtholders and other creditors to potential losses.

Some might object to a receivership on the basis that it would “end” the GSEs. Such a position would be mistaken. Section 1145 specifically prohibits the receiver from revoking, annulling or terminating the charter of an enterprise. Quite simply, the charters of Fannie Mae and Freddie Mac would remain in place under a receivership. As a former staffer who worked on Section 1145, I very much recall that the purpose of this section is to “clean” a GSE and ready a new charter, not end the GSE model.

Another potential objection to receivership would be that it forces a solution before Congress has had sufficient time to deliberate. Such an objection would also be false. Again under Section 1145 of HERA, a limited-life regulated entity, essentially a bridge bank for GSEs, has an initial life of 2 years, which can be extended by FHFA for 3 additional 1-year periods. This would give Congress, the Administration, and FHFA five years to arrive at a suitable replacement for Fannie Mae and Freddie Mac. Again, as HERA prohibits FHFA, in its role as receiver, from “ending” the GSEs, a receivership still allows Congress the option of keeping the GSEs in their current form.

Another important feature of receivership is that it would help to lessen the perception that certain entities, including our largest bank holding companies, are “too big to fail”. The Dodd-Frank Act establishes a resolution process for both non-banks and bank holding companies. This resolution process mirrors in many ways the receivership provisions of HERA. Market participants have rightly questioned whether the resolution powers of Dodd-Frank would ever be used to impose losses on creditors. If we are unwilling to take Fannie Mae into a receivership, then most market participants will conclude that we would also be unwilling to take Citibank or Goldman Sachs into a receivership. Moving Fannie Mae and Freddie Mac into receivership will likely reduce the favorable funding advantage which “too big to fail” institutions currently enjoy (at the expense of the taxpayer).

Lastly, some might object to a receivership in that it would impose losses on creditors.

The concern being that as most of these creditors are other financial institutions, about 80 percent of Fannie and Freddie funding is provided by the remainder of the financial services industry, the imposition of losses could cause other financial institutions to fail or at minimum experience financial stress. I believe such a concern is overstated, particularly since we are past any “panic” in the financial markets. If Fannie and Freddie were to experience losses of another $100 billion, then it is likely that MBS holders would experience little loss and holders of unsecured debt would receive about 94 cents on the dollar. Subordinated debt would likely be wiped out. As insured depositories hold mainly MBS, additional resulting bank and thrift failures would be few. Money Market Mutual Funds would likely incur significant losses, with several funds “breaking the buck”. Foreign holders, particularly central banks, would experience losses, although these losses would be likely less than that already experienced due to exchange rate movements.

To summarize, I believe that shifting losses from the taxpayer to GSE creditors would have minimal disruptions on our financial markets in the current environment. More importantly, the taxpayer should no longer be on the hook for protecting the financial services industry from the consequences of its own mistakes.

Lower Loan Limits
In transitioning from a government-dominated to market-driven mortgage system, we face the choice of either a gradual transition or a sudden “big bang”. While I am comfortable with believing that the remainder of the financial services industry could quickly assume the functions of Fannie Mae and Freddie Mac, I recognize this is a minority viewpoint. Practical politics and concern as to the state of the housing market point toward a gradual transition. The question is then, what form should this transition take? One element of this transition should be a gradual, step-wise reduction in the maximum loan limits for the GSEs (and FHA).

If one assumes that higher income households are better able to bear increases in their mortgage costs, and that income and mortgage levels are positively correlated, then reducing the size of the GSEs’ footprint via loan limit reductions would allow those households best able to bear this increase to do so. As tax burden and income are also positively correlated, the reduction in potential tax liability from a reduction in loan limits should accrue to the very households impacted by such a reduction.

Moving beyond issues of “fairness” — in terms of who should be most impacted by a transition away from the GSEs — is the issue of capacity. According to the most recent HMDA data (2009), the size of the current jumbo (above $729k) is approximately $90 billion. Reducing the loan limit to $500,000 would increase the size of the jumbo market to around $180 billion. Since insured depositories have excess reserves of over $1 trillion, and an aggregate equity to asset ratio of over 11 percent, it would seem that insured depositories would have no trouble absorbing a major increase in the jumbo market.

Given that the Mortgage Banker Association projects total residential mortgage originations in 2011 to be just under $1 trillion, it would appear that insured depositories could support all new mortgages expected to be made in 2011 with just their current excess cash holdings. While such an expansion of lending would require capital of around $40 billion, if one is to believe the FDIC, then insured depositories already hold sufficient excess capital to meet all new mortgage lending in 2011.

Moving more of the mortgage sector to banks and thrifts would also insure that there is at least some capital behind our mortgage market. With Fannie, Freddie and FHA bearing most of the credit risk in our mortgage market, there is almost no capital standing between these entities and the taxpayer.

The bottom line is that reducing the conforming loan limit to no more than $500,000, if not going immediately back to $417,000, would represent a fair, equitable and feasible method for transitioning to a more private-sector driven mortgage system. Going forward, the loan limit should be set to fall by $50,000 each year. As this change could be easily reversed, it also represents a relatively safe choice.

GSE employees are Government Employees
The hallmark of a private corporation is that its owners (shareholders) bear the benefits and costs of its activities. This situation no longer holds for Fannie Mae and Freddie Mac. These entities will never be able to grow their way out of their current obligations to the American taxpayer. Any revenues going forward will help to reduce the size of the hole, while expenses dig it deeper. Given that the taxpayer is now the residual claimant to these entities, it should be clear that the employees of Fannie Mae and Freddie Mac are working not on behalf of the shareholders, but on behalf of the taxpayer. They should be paid like other government employees. I recommend that all GSE employees be transitioned to the GS pay scale as soon as possible. This would include the executive officers. Ever penny of the close to $7 million in total annual compensation paid to Fannie Mae’s President and CEO comes at the expense of the taxpayer. This is simply offensive. If FHA can adequately manage the mortgage risk in its business while paying its employees on the GS scale, then so can Fannie Mae and Freddie Mac.

Bank Buybacks
Credit losses suffered by Fannie Mae and Freddie Mac have in some instances been caused by the violation of representations and warranties by the originating lender. While the GSEs have made some efforts to recover losses from the originating lenders, there is simply not enough public information to gauge the aggressiveness of these efforts. Congress should examine in detail the agreements reached between the GSEs and the banks in regard to loan repurchases and recovery for losses on purchased privatelabel securities. I believe a GAO audit of these agreements, along with detailed information by lender, would help aid in the stemming of losses. Funds recovered should be used exclusively for off-setting previously provided taxpayer assistance to the GSEs.

“Pay it back”
Section 134 of the Emergency Economic Stabilization Act of 2008, better known as the TARP, directed the President to submit a plan to Congress for recoupment for any shortfalls experienced under the TARP. Unfortunately HERA lacked a similar requirement. Now is the time to rectify that oversight. Rather than waiting for a Presidential recommendation, Congress should establish a recoupment fee on all mortgages purchased by Fannie Mae and Freddie Mac. Such a fee would be used directly to reduce the deficit and be structured to recoup as much of the losses as possible. I would recommend that the recoupment period be no longer than 15 years and should begin immediately. A reasonable starting point would be 1 percentage point per unpaid principal balance of loans purchased. Such as sum should raise at least $5 billion annually and should be considered as only a floor for the recoupment fee. A recoupment fee would have the additional advantage of reducing the competitive position of Fannie Mae and Freddie Mac.

Bank Capital Standards
The structural flaws in our mortgage finance system were not limited to Fannie and Freddie, but also included the treatment of GSE debt within the bank capital standards. One of the rationales for the rescue of Fannie and Freddie was a concern as to the impact their failure would have on the rest of the financial system. According to the FDIC, holdings of GSE securities, bonds and mortgage-backed securities as well as preferred stock, constitute more than 150% of Tier 1 capital for insured depositories. This high level of concentration of GSE debt in our banking system was a direct result of the favorable treatment of GSE debt by bank capital standards. Whereas whole mortgage loans require a 50% risk-weighting under Basel II, GSE debt only requires a 20%. The result is that the overall system holds only about 40% of the equity behind the mortgage market as it would otherwise. Congress should direct bank regulators to remove the preferential treatment of Fannie and Freddie. This change would require the banking system to increase capital by approximately $24 billion; accordingly it can be implemented over a reasonable period of time.

Mortgage Credit Quality
The bulk of losses suffered by Fannie Mae and Freddie Mac were the direct result of declines in credit quality. In order to limit future losses, Fannie Mae and Freddie Mac should be restricted as to the quality of loans they can purchase. Under current law, Fannie Mae and Freddie Mac essentially set their own credit quality standards. This has allowed the GSEs to aggressive purchase poor quality mortgages. Going forward, the GSEs should be limited to purchasing only those mortgages that meet the definition of a “qualified residential mortgage” as will be determined by regulations promulgated under the authority of the Dodd-Frank Act.

As regulators are still crafting definitions for “qualified residential mortgage”; the following restrictions should be immediately placed by the GSEs: prohibit the purchase of mortgages for investment properties and second homes; require a minimum cash investment by the borrower of 10 percent of the purchase price or existing home value; and prohibit the purchase of mortgages which have a credit quality indicating a projected delinquency rates of 5% or higher. While there remains considerable debate as to the role of the GSE housing goals in driving their credit losses, I believe it is beyond debate that such were a contributor. Accordingly, the housing goals should be permanently suspended; their future should await the outcome of the broader reform process.

Reduction of Retained Portfolios
Although credit losses have so far constituted the majority, the retained portfolios of Fannie Mae and Freddie Mac continue to pose significant credit and interest risk to both the enterprises and the taxpayer. While the retained portfolios are projected for a gradual decline, that decline could and should be accelerated. The composition of their retained portfolios should also be restricted to mortgage-related investments only, wish some minor provision for cash and Treasuries.

Conclusions
Reform of our federal mortgage finance policies should be among Congress’ top priorities. While the complexity of reform demands a deliberate and thoughtful process, there are immediate steps that can be taken to protect both the taxpayer and our broader economy. Among these steps are: moving Fannie Mae and Freddie Mac into receivership; lower the current conforming loan limits; aligning GSE compensation standards with that of the Federal government; improving the credit quality of GSE loan purchases; and instituting a mechanism to recoup taxpayer assistance to the GSEs.