Statement No. 164
For Information Contact:
Charles W. Calomiris
(212) 854-8748
Peter J. Wallison
(202) 862-5864
Congressman Richard Baker recently introduced a bill (H.R. 3703) to change the way Fannie Mae ("Fannie") and Freddie Mac ("Freddie") are regulated, and to eliminate their access to a "credit line" from the U.S. Treasury. Fannie and Freddie, together with the Federal Home Loan Banks (FHLBs), are the principal "housing GSEs"--government sponsored enterprises that channel implicit taxpayer subsidies to the home mortgage market through their activities in the secondary market for mortgages. The housing GSEs have received increasing attention because of their continuing rapid growth and ever-expanding range of activities, and because of their unique status as agents that mix public purposes and funding guarantees with private ownership and gains.
Congressman Baker's bill and his subcommittee's hearings have raised a number of important questions of public policy which point to the need to fundamentally reassess the missions and powers of the housing GSEs. The Baker bill contains several commendable features, but the Shadow Financial Regulatory Committee believes that the problem it addresses is broader in scope than the need to fine tune the regulation of Fannie and Freddie, and requires a more far-reaching set of reforms for the housing GSEs. For example, the FHLBs are largely omitted from consideration in the bill, yet their recent and likely future growth are at least as fraught with problems as that of Fannie and Freddie.
The implicit subsidies received by the housing GSEs consist of links with the federal government and unusual government-granted privileges that are not accorded to ordinary companies--all of which lead the capital markets to believe that these enterprises will be supported by the government if they encounter difficulties. The implicit subsidy--which is realized in the form of lower interest rates demanded by investors in GSEs' debts--can reach substantial proportions. In 1996, the Congressional Budget Office (CBO) estimated that Fannie and Freddie alone received a subsidy of $6.5 billion in 1995. A new study by the CBO is needed to determine its size in light of the substantial growth of Fannie and Freddie since 1995. As far as we are aware, there has been no study of the size of the FHLBs' subsidies, but there is every reason to believe that these subsidies are, or soon will be, of comparable or larger magnitude than those received by Fannie and Freddie.
It is not unusual for the government to provide subsidies to institutions that are carrying out a government mission. But in most cases the benefits of government support do not accrue directly to the persons or enterprises for whom the programs are purportedly designed. The housing GSEs are unusual in that a large portion of the benefits they receive from taxpayer support accrue to their owners--private shareholders in the case of Fannie and Freddie, and private financial institution members in the case of the FHLBs. The CBO estimated that almost a third of the subsidy received by Fannie and Freddie in 1995--$2.1 billion--went to their shareholders and managements rather than to the mortgage markets. In effect, according to the CBO, the taxpayers pay the full cost of $6.5 billion, but only receive about two-thirds of that cost back in the form of lower mortgage rates.
To be sure, the housing GSEs have done some good. They have helped to create a national mortgage market from what was once a collection of highly localized markets, and in the case of Fannie and Freddie the CBO study estimates that they reduce mortgage interest rates by about 30 basis points. In light of this, many commentators and members of Congress ask why anything should be done about the housing GSEs.
The fact that the housing GSEs do some good and create little visible harm, however, does not mean that they are as benign as their supporters contend. In the mid-1970s, anyone who pointed to the dangers inherent in the thrift industry would have confronted the same argument. Yet, when the economy and thrifts' fortunes changed direction a few years later, the thrift industry's collapse cost the taxpayers over $150 billion.
The housing GSEs represent threats of two different kinds. First, their implicit government support threatens taxpayers. At some point in the future, as with the thrift industry and the Farm Credit System, the taxpayers may be called upon to make good on the implicit support the government has been giving to these agencies. Furthermore, because there is no effective control on GSE growth, they represent a threat to nonsubsidized companies that compete with them. Without effective private competition, and with the weak existing regulation of their missions and financial health, the housing GSEs are literally out of control.
In the case of the FHLBs, recent growth in the use of advances as a source of financing mortgage expansion by thrifts, banks, credit unions, and insurance companies--all of which are eligible for membership in the FHLBs--has reflected the increasing reliance on public funding by FHLB members. From 1992 through 1999, the total membership in the FHLB system has grown from 3,624 to 7,378, and total advances have increased from $79 billion to $390 billion. The growth in the membership of community banks (any bank with less than $500 million in assets) in the FHLB system has been especially dramatic. From 1992 through 1999, the number of community bank members has risen from 1,554 to 5,089, and the advances received by these banks has grown over that period from $1.6 billion to $25 billion.
Borrowing from FHLBs--especially by community bank members--is likely to accelerate in the future because of changes of the past year in laws and regulations that expand the range of assets that can serve as collateral for FHLB advances (which now include community bank credit to small businesses, small farms, and small agri-businesses), and that could reduce the FHLB minimum capital requirement from 5% to as low as 4%. The expansion in the growth and the riskiness of the FHLB system, financed by taxpayer protection, is clearly a worrying trend.
In the case of Fannie and Freddie, the taxpayer subsidy takes the form of an expected bailout of these institutions if they became insolvent. In the case of the FHLBs, credit risk is subsidized through a different channel. The FHLBs face no significant risk of loss on their advances, just as the Fed faces little risk of loss on discount window lending, because advances and discount window lending are collateralized and enjoy a senior claim on bank resources in the event of bank insolvency. Instead, FHLB subsidies for risk taking by member institutions generate expected costs for the Bank Insurance Fund (BIF) and the Savings Association Insurance Fund (SAIF). In the wake of losses imposed on the FDIC by Fed lending in the 1980s, FDICIA in 1991 intended to restrict Fed lending to weak institutions. The growing FHLB subsidy contravenes the intent of FDICIA by expanding the means for exposing BIF and SAIF to the risk of loss from a risk-subsidizing protected senior lender.
Additionally, the Federal Housing Finance Board has proposed changes in ownership rules for the FHLB system that would allow one institution to control up to 20% of the voting power in any Federal Home Loan Bank. Thus, if this rule is adopted, three private institutions could effectively obtain control of an FHLB and use its protected status to further their private objectives, which might include anti-competitive behavior or the financing of extremely risky activities at low cost.
The benefits the housing GSEs offer are small relative to the size of the mortgage payments made by homeowners. The mortgage market is now mature, national in scope, and fully capable of securitizing mortgages to a global capital market without taxpayer subsidization. The lower mortgage interest rates attributable to the government subsidy also drive up home prices, further reducing the net value homeowners receive from lower mortgage rates. Furthermore, since the bulk of the mortgage subsidy is delivered to middle-income families who have the wherewithal to purchase homes, the housing GSEs do little to increase homeownership rates. Indeed, both HUD and researchers at the Federal Reserve Board have called attention to the fact that the GSEs channel very little of their subsidy to the poor, where it could potentially have a significant effect on homeownership rates. That distribution of benefits partly reflects profit maximization by the GSEs, and partly results from rules that mandate private mortgage insurance for high-loan-to-value mortgages.
The Shadow Committee believes that rather than tinker with the powers and regulation of the housing GSEs, the preferred policy response to the risks they pose is to fully privatize them. This would eliminate the risks now borne by the taxpayers, and create a more competitive mortgage market and financial system. Alongside privatization of the GSEs, it is worth considering ways to target assistance more effectively to low-income would be homeowners. For example, households could qualify for a one-time downpayment assistance grant that would match homeowners' contributions with a government subsidy. Such a government program would have the advantages of focusing assistance where it is likely to have the greatest effect, and making the cost of assistance transparent to taxpayers (as it would be accounted for in the budget). We mention down payment assistance only as one possibility; other options designed to target assistance where it would be more effective are also worth considering.
GSE privatization has been done before. In 1996, Congress privatized Sallie Mae by permitting the GSE to become a wholly owned subsidiary of an ordinary Delaware Corporation. The holding company was allowed 10 years to liquidate its portfolio of student loans and its government-backed liabilities, after which it will become a fully privatized company, able to engage in any line of business.
This could be done with Fannie, Freddie, and the FHLBs, as well, although in the case of the housing GSEs it would be desirable to divide these entities into a number of competing companies, to avoid excessive concentration of power in a handful of private firms, and the possibility of a too-big-to-fail bailout of the privatized successors to the housing GSEs. For example, dividing Fannie and Freddie each into two firms, and combining the portfolios of the FHLBs to create four additional firms would result in a highly competitive market structure. Given the homogeneity of the assets of the housing GSEs, such a restructuring would not be prohibitively difficult.
While the Shadow Committee believes that full privatization would be the best option, if Congress is unwilling to follow that course immediately, then a stronger and more effective regulatory framework for the GSEs would be a sensible second choice. Some of the provisions in Congressman Baker's bill offer an excellent first step.
But as an effort to control the GSEs, H.R. 3703 does not go far enough in a number of respects. First, in its current form, the bill mainly affects Fannie and Freddie, and does little to rein in the growing subsidy provided to the FHLBs. While it is true that Congressman Baker's proposals limit the seniority of FHLB claims on insolvent member institutions and to consolidate regulation of the housing GSEs, if implemented properly, might reduce somewhat the risks posed by FHLB expansion, the Shadow Committee believes that the risks from FHLB expansion would still remain largely unaddressed.
Second, Congressman Baker has proposed to eliminate the Treasury Secretary's blanket authority to purchase GSE securities. But this so-called line of credit at the Treasury is only one of a number of symbolic links to the U.S. government which have encouraged capital markets to believe that the housing GSEs are government-backed. Among other things, Fannie and Freddie are exempt from state and local taxes, the FHLBs are exempt from all taxes, and FHLB bondholders also enjoy substantial tax exemptions; many of the directors of these institutions are government appointees; none of the housing GSEs is required to register securities with the SEC; and the securities of all of them may be used as collateral for government funds held on deposit in commercial banks. All of these links should be eliminated.
Third, until the subsidy is credibly eliminated, we recommend on-balance sheet accounting for the value of the subsidy portion of the GSE debt issued each year. This will clarify for the public the extent to which government resources are being used in support of the housing GSEs and will force Congress to weigh the value of GSE debt issuance against other government costs.
Fourth, until full privatization occurs, the housing GSEs should be charged a guarantee fee to compensate taxpayers for the value of the subsidy they receive and the risk they create for taxpayers.
Fifth, the formal names of the housing GSEs should be changed to remove the words national and federal, and thus make clear that these entities are not linked to the U.S. government.
If full privatization is not immediately feasible, these reforms would substantially improve incentives and control the growth of the GSEs and the costs to taxpayers implied by that growth. Furthermore, such reform might cause the GSEs to consider pushing for full privatization. That is what occurred in the case of Sallie Mae, which was encouraged to seek privatization in part by the Congressional imposition of a 30 basis point guarantee fee in 1993. The same result for the housing GSEs would be highly desirable.
