agers of the credit risk of the mortgages they buy. Further, by maintaining exclusive focus on the residential mortgage markets, as required by law, the GSE's have developed extraordinary expertise in understanding the credit characteristics of borrowers. This has resulted in a steady lowering of downpayment requirements within the conventional market to the point at which the GSE's, with no explicit subsidy, are able to provide nearly the same benefit to borrowers as the Government provides through its on-budget FHA and VA mortgage programs.

Management of interest-rate risk also has been a notable success. Through the creation of mortgage-backed securities, the issuance of callable debt and the use of derivatives, the GSE's routinely and efficiently transfer interest-rate risk from individual households to global capital markets. Not only do the GSE's make it possible for originators to lend money to individual homeowners for long periods of time at better rates than many corporations can borrow, but they also permit borrowers to "put" the mortgages back whenever they desire to do so and at no penalty. This extremely valuable option makes the 30-year, fixed-rate mortgage the product of choice among U.S. homeowners; in 2003, 82 percent of all conforming purchasemoney originations were fixed-rate mortgages. Homeowners were able to profit from falling interest rates by refinancing into lower-cost loans, adding billions of dollars to our economy. Prepayable mortgages also help diminish friction in our economy by facilitating the mobility of the Nation's labor markets.

These innovations in mortgage financing made possible by the GSE's produce valuable benefits. Low-cost mortgage money is readily available. Families can get their loans approved in minutes. (In fact, during this hearing, Freddie Mac likely will have financed mortgages for about 2,000 families.) Today, more people own homesand higher quality homes-than at any time in our Nation's history and than in virtually any other part of the world.2 And wealth created through homeownership will help bear us into old age, taking some of the burden off Social Security and allowing us to pass something along to the next generation. Not a bad track record for Congressional inspired institutions that need no budget authority, pay significant Federal taxes, and employ thousands of people.

In United States, we tend to take these benefits for granted. However, very few countries can boast of such an efficient and effective mortgage delivery system.3 Despite the integration of world capital markets, the United States is still the only place where a long-term callable mortgage product is broadly available. Countries that want to provide long-term prepayable mortgages to their own citizens are considering creating GSE's. The European Union is currently considering the creation of a GSE-type agency to "enable lenders to provide their existing mortgage products at better prices and introduce long-term, fixed-rate mortgages without redemption penalties."4

Let us now consider U.S. housing finance without the GSE's. There are three key arguments I would like to address.

First is the view that Government sponsorship is no longer needed to attract capital to housing or to provide an abundant supply of 30-year, fixed-rate mortgages. This optimistic view contradicts the experience in other developed countries. That is, if homeowners in Northern New York or Washington State lived a few miles to the north in Canada, they would typically be restricted to a 7-year, fixed-rate mortgages, they would be locked into higher interest rates or have to pay heavy penalties if they wanted to prepay, and they would have to put 25 percent down.

This sanguine view of markets also reflects our collective amnesia about where we are in the credit cycle. History reveals that certain industries will slump, that certain regions will experience economic downturn, which, in turn, causes house values to fall and defaults to rise. We also know that with interest-rates at historic lows, the mortgages put on the books today, in all likelihood, will require financing for decades to come. In short, it is easy to dismiss the risks of mortgage lending when times are good.

GSE's were created precisely for those times when things are not going so well, however. GSE's absorbed significant losses during the oil bust in the 1980's and during the weakening of the economy in Northeast in the early 1990's. They also stabilized residential mortgage rates during the international financial crisis of 1998and again after-September 11-by continuing to provide liquidity to the secondary

2 "As a result of the very favorable conditions in the housing sector, the U.S. homeownership rate climbed to 68.2 percent in the third quarter of 2003-equal to its highest level on record," 2004 Economic Report of the President, p. 89.

3 Marsha J. Courchane and Judith A. Giles, "A Comparison of U.S. and Canadian Residential Mortgage Markets," March 2002.

4 Richard Adams, "Banks Back Cheaper Mortgage Plan," The Guardian, November 17, 2003.

market for conforming home loans. Their actions ensured that mortgage credit remained available and affordable.

A second argument concerns the allocation of capital to housing. The housing market has an enormous impact on the economy, directly accounting for more than one-third of the nominal growth in GDP over the past 3 years.5 And this does not begin to account for all the indirect support for consumption generated by record levels of refinancing in the past few years. Housing played an important countercyclical role in supporting the recent weak economy, as noted in the President's 2004 Economic Report:

Despite the similarities between the recent business cycle and previous ones, this most recent cycle was distinctive in important and instructive ways. One noteworthy difference is that real GDP fell much less in this recession than has been typical This relatively mild decline in output can be attributed to unusually resilient household spending. Consumer spending on goods and services held up well throughout the slowdown, and investment in housing increased at a fairly steady pace rather than declining as has been typical in past recessions.6

Finally, there are arguments about size and systemic risk. Residential mortgage debt outstanding grew at an annualized rate of 8.6 percent over the past decade. Not surprisingly, the GSE's also have experienced significant growth. But GSE size is not an accurate proxy for risk. On average, there is approximately 40 percent collateral in homeowner equity behind the loans Freddie Mac has guaranteed. Interestrate risk also is well-managed. Freddie Mac strives to maintain an extremely closely match between the duration of our assets and liabilities. Throughout 2003, for example, a period of extreme turbulence in financial markets, Freddie Mac's duration gap never exceeded 1 month.

Finally, there is no way that mortgage debt and the risks of investing in it would disappear by downsizing the GSE's or making other changes to the GSE charter. Rather, the burden of managing mortgage credit risk would shift from these institutions to those with explicit Government support, while interest-rate risk would shift onto individual households. Another likely outcome is that higher costs of conventional mortgage financing could cause borrowers to shift into the FHA market, thereby actually increasing Government subsidization of housing. For homeowners, restrictions on GSE growth likely would result in reduced availability of 30-year, fixed-year, prepayable mortgages and higher costs.

These uncertain benefits must be coupled with the potential risks of dismantling a highly efficient and successful housing finance system. We can get a glimpse of a world without GSE's by looking at the jumbo market. On any given day, it is possible to look in a newspaper and find that mortgage rates on conforming loans are regularly one-quarter of a percentage point lower than those in the higher-balance jumbo market. Borrowers in the jumbo market not only pay higher rates, but they are also more likely to have to settle for an adjustable-rate mortgage (ARM's).

ARM's have the obvious advantage of lowering monthly mortgage payments in the first few years of homeowning, but they require borrowers to bear the interest-rate risk on the loan-rather than the capital markets bearing this risk. This results in higher borrower defaults over the long-term. Jumbo borrowers also typically make larger average downpayments than conforming borrowers. Higher mortgage-interest rates and larger downpayments make it significantly harder for low- and moderateincome families to become homeowners.7

In summary, we are a Nation of homeowners-and from all I can tell, we want to keep it that way. While discussions of the optimal allocation of the Nation's capital have their place, I believe this Nation made the right decision 70 years ago to lend housing a helping hand. (You will have to excuse my passion on this subject, but homeownership was part of my Ph.D. dissertation 30 years ago.) Bi-partisan support for Federal housing policy has paid enormous dividends. Families build wealth. Kids do better in school. Neighborhoods are safer. And, in recent years, housing has been the backbone of our Nation's economy. Support for homeownership-whether explicit or implicit clearly has been good for this country.

But the task is not finished. There are millions of families still waiting to participate in the American Dream, and the homeownership gap between white families

5 These percentages are based on data published by the Bureau of Economic Analysis, U.S. Department of Commerce for 1996 through 2003 and data for the same years available upon request from Freddie Mac.

62004 Economic Report of the President, pages 30, 32.

7 Roberto Quercia, George McCarthy, and Susan Wachter, "The Impacts of Affordable Lending Efforts on Homeownership Rates," Journal of Housing Economics (Vol. 12, 2003), pp. 29–59.

and families of color is unacceptable. This is not the time to begin dismantling the world's finest housing finance system, or placing artificial limits on the GSE growth. The potential benefits of doing so are uncertain, and the risks are great.

Imperative for Regulatory Reform

Continued support for the GSE model of housing finance does not imply that improvements to the GSE regulatory oversight structure are not needed. They are. As a former regulator, I will be the first to say that world-class regulatory oversight is absolutely critical to the achievement of Freddie Mac's mission and to maintaining the confidence of the Congress, the public and financial markets. Freddie Mac strongly supports the enactment of legislation that provides strong, credible regulatory oversight. These enhancements are needed-even overdue.

I am sadly aware that Freddie Mac's accounting issues are the source of much of the current controversy regarding the role of the GSE's. However, as with any episode such as this, it is critical to get the ship back on course without overreacting at the wheel. Given the enormous benefits of the conforming mortgage market, which has proven its resiliency in all interest-rate and credit environments, zeal to improve this system must be tempered with an abundance of care. Borrowing a phrase from our friends at the Homebuilders, I urge the Committee to "measure twice and cut once."

To guard against potential negative unintended consequences, I would like to offer a set of principles, based on my experience as a former regulator. The new GSE regulatory structure must:

• Engender public confidence through world-class supervision and independence; • Ensure continued safety and soundness of the GSE's;

• Respond flexibly to mortgage market innovation; and

• Strengthen GSE market discipline through robust and timely disclosure.

With these principles in mind, today, I will comment briefly on key aspects of the regulatory structure under consideration in this Committee.

Structure and Independence

Freddie Mac would strongly support an independent board regulatory structure modeled on independent Federal agencies such as the Securities and Exchange Commission. Our preference would be for a three-member board, comprised of a Chair and two additional members. The President would appoint Board members, by and with the advice and consent of the Senate, subject to statutory criteria relating to qualifications of the nominees. For instance, we believe that at least one member of the Board should have significant housing industry experience. It would also be important to ensure that members have significant experience with complex financial transactions. As is typical with independent boards, we would suggest that not more than two of the Board members be members of the same political party. Notwithstanding the importance of housing and financial expertise, we would have some concern if the Board were to include representatives of cabinet departments such as the Department of Housing and Urban Development, the Department of the Treasury or other executive branch departments. The purpose of establishing an independent board is just that, independence. Inclusion of executive branch representatives on the GSE regulatory board could compromise this important component of world-class regulation.

Freddie Mac would have similar concerns should the Congress decide to locate the new regulatory office within the Department of the Treasury. To ensure independence, we would support applying the same operational controls as apply to the relationships between the Secretary of the Treasury and the Office of the Comptroller of the Currency and the Office of Thrift Supervision.8 Adequate firewalls are needed to avoid the politicization of the GSE mission and the critical role we play in the Nation's economy and global financial markets.

Funding of New Oversight Offices

Freddie Mac supports providing both the new regulator and the Secretary of HUD authority to assess Freddie Mac outside the annual appropriations process to pay for the costs and expenses of carrying out their respective responsibilities vis-a-vis the GSE's. However, we would suggest that the General Accounting Office regularly report to the Congress on the efficacy of the new regulatory structure and the reasonableness of the costs relative to other world-class financial regulators so that neither unnecessarily raise the cost of meeting our mission.

8 See 12 U.S.C. §§ 1, 250, 1462a(b)(2), (3), and (4) and 1464(d)(1)(A).

GSE Capital Requirements

Second to questions of GSE role and benefits, I have quickly learned that questions about GSE capital adequacy are highly contentious and can serve as "stalking horses" for other issues. There is no question these issues are of paramount importance. Capital adequacy is the touchstone of investor confidence and is key to our ability to attract low-cost mortgage funds. On that score, Freddie Mac consistently has exceeded both its minimum capital and risk-based capital standards.

However, from the perspective of a former regulator, I believe there are many difficult and sometimes confusing aspects about the direction of the debate on GSE regulatory oversight. The first is the view that the GSE's should be held to the same capital standard as for banks. Let me begin by stating the obvious: GSE's are not banks.

• There are nearly 10,000 banks and savings institutions in this country. There are two GSE's focused exclusively on housing.

• Banks are largely funded by deposits. GSE's must rely exclusively on the capital markets for their funding.

• Banks can (and do) invest in a wide range of higher-risk assets, ranging from unsecured loans, to commercial loans and loans to foreign countries. In contrast, GSE's are restricted to one line of business: Residential mortgages finance. We invest almost exclusively in conventional conforming mortgages, among the safest investment vehicles around.

Given these important distinctions, it is entirely appropriate that the GSE capital regime be distinct from the bank capital model. GSE capital requirements reflect the confinements of its GSE charter, such as the conforming loan limit and credit enhancement requirements for high loan-to-value mortgages. These charter limitations necessarily result in a lower GSE risk profile.

Since 1994, charge-off losses at the five largest banks have been, on average, 17 times larger each year than charge-offs at Freddie Mac. Even in these banks' best year, charge-offs were more than five times higher than Freddie Mac's worst year.9 Limiting the comparison to mortgage assets, the residential mortgages found in bank portfolios typically entail greater risk than those in Freddie Mac's portfolio. In 2002, FDIC-insured institutions had an average charge-off rate of 11 basis points on their mortgage portfolios, compared to 1 basis point for Freddie Mac. 10 Given this lower risk exposure relative to banks, we believe that the GSE minimum capital requirement is adequate and need not be changed.

The second troubling aspect of the current debate is the fixation on the GSE minimum capital ratio, when the risk-based capital standard is a far more effective regulatory tool. Leverage ratios are last year's capital "model." They have significant limitations and, depending on how they are enforced, can do more harm than good. I observed first-hand the problems with overzealous enforcement of simple leverage ratios during my tenure at the Federal Reserve Bank of Boston in the early 1990's. While many financial institutions in the Northeast were adequately capitalized on a risk-adjusted basis, the strict enforcement of simple leverage ratios required them to liquidate a substantial portion of their assets. This resulted in a drying up of commercial credit that greatly exacerbated the economic downturn. The infamous "credit crunch" had profound effects on small and mid-size businesses and employment in the Northeast. It turned a 2-year recession into a 5- to 6-year slump.11 I discuss these issues in two articles I wrote on this subject. 12

My experiences are consistent with leading international trends in capital management. Drawing from recent statements by the Basel Committee on Banking Supervision, risk-based capital regimes are preferable to the use of simple ratios to set capital standards. In its 1999 Basel Consultative Paper and the 2001 New Basel

9 Federal Financial Institutions Examination Council, Consolidated Reports of Condition and Income and Freddie Mac annual reports for 1994 to 2001. For 2002 Freddie Mac credit information, see http://www.freddiemac.com/news/archives/investors/2003/4qer02.html.

10 Federal Financial Institutions Examination Council, Consolidated Reports of Condition and Income and Freddie Mac. See http://www.freddiemac.com/news/archives/investors/2003/ 4qer02.html.

11 History of the Eighties, Lessons for the Future: An Examination of the Banking Crises of the 1980's and Early 1990's, vol. 1, part 2, Sectors and Regional Crises, Ch. 10, Banking Problems in the Northeast, Federal Deposit Insurance Corporation, 1997.

12 See Richard F. Syron, statement before the Subcommittee on Domestic Monetary Policy of the Committee on Banking, Finance, and Urban Affairs, U.S. House of Representatives, May 8, 1991, reprinted in "Are We Experiencing a Credit Crunch?," New England Economic Review (July/August 1991), pp. 3-10; and Richard F. Syron, "The New England Credit Crunch," Credit Markets in Transition: Proceedings of the 28th Annual Conference on Bank Structure and Competition, Federal Reserve Bank of Chicago (1992), pp.483-9.

Capital Accord, the Committee proposed a capital adequacy framework to replace the 1988 Capital Accord for U.S. bank capital standards, which relied heavily on simple ratios to set capital standards. The new framework, which is currently under consideration in this country, more accurately aligns capital requirements to the actual risks incurred by regulated institutions. 13

Notwithstanding my philosophic differences regarding the efficacy of leverage ratios, I can understand the need for regulator discretion to increase the leverage ratio in the event of a finding of an unsafe and unsound practice. We believe parameters should be put in place in statute that define the circumstances under which such an increase could be undertaken, as well as parameters for resetting the ratio to the statutory minimum once the unsafe and unsound practice has been satisfactorily addressed.

Discretion on Risk-Based Capital

In my view, greater discretion with regard to the GSE risk-based capital rule is the best way to avoid potential negative unintended consequences associated with strict enforcement of leverage ratios. Ten years in the making, the GSE risk-based standard is unique among financial services regulation. It requires Freddie Mac to hold capital sufficient to survive 10 years of severe economic conditions; under the risk-based test, both the credit and interest-rate risk of the GSE's mortgage holdings are stressed to historic proportions. Without a doubt, this rule is at the cutting edge of financial services regulation.14 It ties capital to the specific risks of an institution ensuring safety and soundness without raising costs unnecessarily or crippling the smooth flow of mortgage capital. It is the standard-bearer in capital regulation.

To ensure that the GSE capital standard remains at the forefront of capital regulation, the new regulator must have adequate discretion to keep pace with developments. Although the basic parameters of the risk-based capital stress test are set in law, our present regulator has significant discretion in adjusting the risk-based capital requirements. Additional discretion, such as provided to Federal banking agencies, could help ensure the GSE risk-based capital standard remains at the forefront of financial sophistication, while continuing to tie capital to risk.

Discretion must be balanced with continuity, however. Unnecessarily changing the risk-based capital standard harms those who made investment decisions based on a particular set of rules, only to find later that the rules were changed. This "regulatory risk" increases costs that are ultimately borne by mortgage borrowers. Therefore, until such time as an overhaul of the risk-based capital stress test appears warranted, the regulator should be encouraged to continue to apply the existing risk-based capital rule. The rule has been in effect for only 1 year and has yet to show signs of need for reform.

We also believe the new regulator should be encouraged to gather information over the entire business cycle before making changes. This could be accomplished by requiring that the current rule remain in place for a period of time and expressing Congressional intent to this effect. When a new rule appears warranted, policymakers should ensure that certain fundamental principles remain firmly intact. It would be our strong suggestion that any future capital standard must continue to tie capital levels to risk; be based on an analysis of historical mortgage market data; remain operationally workable and as transparent as possible; and accommodate innovation so the GSE's can carry out their missions.

Further, we would expect that any changes to the rule be accomplished through notice-and-comment rulemaking, with an adequate comment period for all interested parties to express their views, followed by an adequate transition period for the GSE's to make any necessary adjustments to comply with new requirements. In summary, Freddie Mac supports improvements to the GSE capital regime that reflect the unique role of the GSE's, while ensuring public trust in our financial strength. Based on my experience as a regulator, I fully support granting the regulator greater discretion to set risk-based capital levels that accurately reflect the risks we undertake. Discretion on risk-based capital greatly mitigates the need to

13 The New Basel Capital Accord, Consultative Document, Basel Committee on Banking Supervision (January 2001) (the 2001 Basel Accord).

14 According to an analysis prepared by L. William Seidman, former Chairman of the FDIC, the stringent risk-based capital standard applicable to Freddie Mac could be extremely challenging if applied to most other financial institutions. L. William Seidman, et al., Memorandum to Freddie Mac, March 29, 2000. More recently, the CapAnalysis Group, LLC, concluded that the risk-based capital stress test is "a much more stringent test for judging the safety and soundness of a financial institution than is a traditional capital-requirements test. CapAnalysis Group, LLC, OFHEO Risk-Based Capital Stress Test Applied to U.S. Thrift Industry (March 17, 2003), p.1.

The