cally would you suggest the GSE's should do to promote affordable multifamily housing? Please elaborate.

A.1.d. The Congress should evaluate the GSEs' effectiveness in promoting multifamily housing and other affordable housing initiatives. The Federal Reserve's primary interest is that GSE activities are conducted in a safe and sound manner that does not increase or promote systemic risk in the financial system.

Q.1.e. In the context of GSE regulatory reform, the Administration has suggested creating new affordable housing goals and subgoals. Do you believe that these proposed goals and subgoals are as rigorous as should be, in light of recent GSE financial performance and the implied GSE benefits? Why or why not?

A.1.e. If the Congress desires the GSE's to pursue affordable housing goals and to lower mortgage rates, it should put in place rigorous systems to measure and evaluate the extent that these goals are being accomplished. As I stated in my testimony, a more efficient approach, and one that greatly reduces the potential for systemic risk, would be to encourage the GSE's to pursue mortgage securitization without the accumulation of enormous GSE mortgage portfolios and for the Congress to pursue affordable housing goals through on-budget programs targeted toward homeownership. Q.2.a. In GAO's 1997 Report, "Advantages and Disadvantages of Creating a Single Housing GSE Regulator (GAO/GGD-97-139)," GAO argues, "Having an independent board would allow it to be structured to provide equal links to HUD, due to its role in housing policy and Treasury, due to its roles in finance and financial institution oversight. Having a single director, rather than a board, as head of the regulatory agency might provide for management efficiencies and clearer accountability. However, such an arrangement would sacrifice the advantages of having the different perspectives, expertise, prestige, and stability a board could provide." Do you concur with this preference for a board over a director agency structure? Why or why not?

A.2.a. As I stated in my testimony, world-class regulation, by itself, may not be sufficient and indeed, may even worsen the situation if market participants infer from such regulation that the Government is all the more likely to back GSE debt during a financial crisis. This is the heart of a dilemma in designing regulation for the GSE's. On the one hand, if the regulation of the GSE's is strengthened, the market may view them even more as extensions of the Government and view their debt as Government-backed. The result, short of a marked increase in capital, would be to expand the implicit subsidy and allow the GSE's to play an even larger unconstrained role in the financial markets. On the other hand, if we fail to strengthen GSE regulation, the possibility of an actual crisis or insolvency is increased. Regardless of whether the Congress creates an independent board or an agency director, the key issue is whether the Congress intends to bailout a GSE in the event of default. Any structure created by the Congress should incorporate clear and detailed provisions that spell out how the regulator is to deal with a GSE that has failed.

Q.2.b. In your testimony, you argued that the Chairman of the Board of Governors of the Federal Reserve System would have a substantial conflict of interest if placed on a board having oversight over the housing GSE's, as proposed by Chairman Shelby in a question to Comptroller-General Walker on February 10, 2004. You recommended against including the Chairman of the Board of Governors of the Federal Reserve System on the proposed board. Chairman Shelby also suggested that the proposed board also include the Chairman of the Securities and Exchange Commission. Do you think there is a similar conflict of interest with placing the Chairman of the Securities and Exchange Commission on the proposed board?

A.2.b. Other financial market regulators may or may not have conflicts of interest that might lead them to not want to be involved directly in the affairs of the GSE's. In general, the Congress may decide that it is best not to place regulators on a GSE board who, in the course of their primary duty, may be involved in situations that require judging the actions of a GSE and, perhaps, coming into conflict with its regulator. The Congress should assess whether or not such a conflict would exist with the Chairman of the Securities and Exchange Commission.

Q.3.a. In the same report, GAO argues that, "Our ongoing work has strengthened our belief that the housing GSE regulators would be more effective if combined and authorized to oversee both safety and soundness and mission compliance. Nothing we have observed has caused us to modify our criteria for an appropriate regulatory structure." For the record, do you support having a single regulator over all housing GSE's, Fannie, Freddie, and the Federal Home Loan Banks? Please elaborate.

A.3.a. Yes. These GSE's play essentially a similar role in the mortgage finance system and should be governed by the same regulatory framework.

Q.3.b.1. In your testimony, you appear to concur with this position. You said, "In my remarks, I will not focus on the Federal Home Loan Banks, although much of this analysis applies to them. In fact, because the Home Loan Banks can design their advances to encompass almost any type of risk, they are more complex to analyze than other GSE's and, hence, raise additional issues." Please elaborate on what you believe are the complexities involved with the risk profiles of the Federal Home Loan Banks.

A.3.b.1. The Federal Home Loan Banks (FHLB's), until recently, did not buy mortgage loans but instead provided loans (advances) to mortgage lenders who used their mortgage loans as collateral for the FHLB advances. In the past, advances simply passed on to mortgage lenders the borrowings of the FHLB's. Nowadays, however, FHLB advances can be custom designed for lenders, can include many financial options, and thus can pose complicated risks. Moreover, the mortgage loan portfolio of the FHLB's has increased significantly in recent years. The Congress should evaluate the growth in both the size and complexity of the FHLBs' portfolios and evaluate whether this growth is consistent with Congressional goals for the FHLB System and the safety and soundness of the financial system.

Q.3.b.2. Similarly, please elaborate on what you believe are the additional issues raised by such complexities.

A.3.b.2. As I stated above, traditionally the FHLB's have been portfolio lenders who acted mainly as intermediaries between capital markets and mortgage lenders. The FHLB's essentially passedthrough subsidized capital market funding to mortgage lenders, taking a passive role in the process. Over the past decade, however, the FHLB's have evolved into active portfolio managers who offer financial options to their members. With their large portfolios, the risks involved in hedging these options can be large and, as with Fannie Mae and Freddie Mac, difficult to evaluate and worrisome given the size and scope of the portfolio.

Q.3.b.3. What important concepts do you recommend that Congress consider if and when designing legislation to create a single regulator for all housing GSE's? Please elaborate.

A.3.b.3. As I explained in my testimony, the regulator must be structured in a way that does not reinforce investors' perceptions that the Government implicitly backs GSE debt.

Q.4. In his testimony to the Committee on February 25, 2004, Chairman Raines submitted a chart detailing the differences between 30-year conforming and jumbo mortgage rates in 2003 to demonstrate a 21 basis point difference in Fannie's mortgagebacked securities yields. In addition, after including guaranty fees, servicing costs, "carry" costs, and other miscellaneous mortgage rate expenses, he argued that there was a 26 basis point difference between the conforming and jumbo rates that are offered to homebuyers. How do you respond to such evidence? Don't these data demonstrate the difference in mortgage transaction costs, as well as the mortgage transaction prices? Why or why not?

A.4. There is clearly a difference between GSE and non-GSE yields in the MBS market, as well as a difference between conforming and jumbo mortgage rates. As shown in Wayne Passmore's study, jumbo mortgage rates are usually 15 to 18 basis points higher than conforming mortgage rates. The question is, how much of the difference can be attributed to the GSE status of Fannie Mae and Freddie Mac? Passmore's study indicates that most of these differences are due to the economies of scale and other pricing efficiencies of the conforming loan and MBS markets. These market attributes would likely continue to exist even if Fannie Mae and Freddie Mac were not GSE's. The implicit subsidy is the element of the difference that would disappear should investors become convinced the GSE's were not Government-backed. According to Passmore's work, about 2 to 4 basis points of the conforming guarantee fee (embedded in the total guarantee fee, which is shown as 18 basis points in the chart you mention) is due to the subsidy. Indeed, if Fannie Mae and Freddie Mac were passing through some of their implied Government subsidies, one might generally expect the guarantee fee for conforming MBS to be less than the equivalent fee in the jumbo market, both because conforming mortgages are generally very safe assets and because Fannie Mae and Freddie Mac do not need to provide the same degree of credit enhancements for their MBS because of their implicit Government backing. In the Fannie Mae example, the guarantee fees are assumed equal.

Q.5. In the context of reviewing the regulation of the housing GSE's, do you believe that the current minimum capital standard of 2.5 percent for Fannie and Freddie is too low or too high? A.5. Determining the appropriate minimum level of capital for the housing finance GSE's is a difficult and technical process that is best accomplished by the housing finance GSE regulator, which should have full access to information concerning the GSEs' activities, risks, risk management, and controls. The Federal Reserve generally does not have access to nonpublic information concerning the housing finance GSE's and has not conducted the analysis that would be required to express an opinion on the adequacy of the current minimum capital standard for the housing finance GSE's. Q.5.a. On what basis do you believe the minimum capital standard should be set?

A.5.a. The housing finance GSE regulator should be encouraged to consider the three pillars of the proposed Basel Capital Accord in setting a minimum capital standard-the minimum capital requirements, the supervisory review process, and the market discipline or disclosure pillars.

Minimum capital standards should be established by the housing finance GSE regulator at a level that ensures that the major risks to which the housing finance GSE's are exposed-credit, interest rate, liquidity, and operational risks—are adequately covered, taking into account their ability to manage these risks. Ideally, one could rely primarily on a risk-based capital standard, but riskbased capital standards are not perfect and are still under development. Indeed, it likely will be many years before regulatory capital measures will be able to fully quantify and appropriately reflect all of the risks to which an entity may be subject over time. A leverage ratio places an overall constraint on the degree to which an institution can leverage its capital with debt that is implicitly or explicitly subsidized by the Government and works to limit the extent to which an institution can arbitrage the capital standard. It has been the Federal Reserve's experience that it is difficult for a banking organization to arbitrage simultaneously both the risk-based and leverage capital standards, which helps to ensure that the overall level of capital in the institution remains adequate in relation to its risk exposure.

In addition and importantly, the housing finance GSE regulator should have broad statutory authority to adopt regulations or take other actions that are necessary or appropriate to ensure the safety and soundness of the institutions it supervises. This authority, which is an important complement to the capital authority, would allow the housing finance GSE regulator to address financial and operational weaknesses at an institution.

Q.5.b. Do you believe that allowing Fannie and Freddie's regulator to have discretion only over risk-based capital is insufficient to maintain the safety and soundness of the GSE's? Why or why not? A.5.b. As noted in my response to the prior question, a leverage ratio provides an important complement to a risk-based capital standard because risk-based standards are still relatively new and require significant ongoing development to encompass the many risks and arbitrage possibilities that exist in today's financial mar

kets. Both types of standards are needed and, in my view, it is important for the Congress to provide the housing finance GSE regulator with clear authority to adjust standards for the entities under its jurisdiction.

History indicates that supervisors need the ability to revise and modify both the risk-based and leverage capital standards to ensure that these capital measures appropriately reflect changes in the financial markets, capital instruments, and the structure, operations, and risks of supervised institutions. In the banking area, the Federal banking supervisors have modified the leverage capital requirement in several significant ways since it was first adopted in the early 1980's. Most of these revisions altered the definition of capital as used under the leverage ratio requirement in response to accounting changes, financial product innovations, and market developments. The Congress should provide the housing finance GSE regulator with similar flexibility to update the minimum capital requirements for the GSE's as needed to ensure that they adequately capture risks and reflect the rapidly changing housing finance industry.

Q.5.c. Do you believe that it would harm the ability of Fannie and Freddie's regulator to perform its oversight functions if Congress placed restrictions on its ability to adjust the minimum capital standards? Why or why not?

A.5.c. Statutory restrictions on the ability of the housing finance GSE regulator to adjust the minimum capital standards would impede the ability of the regulator to modify those standards to respond in a timely way to market innovations in capital instruments, revisions in accounting rules, advances in techniques for measuring exposure to risk, and the natural tendency of institutions to arbitrage capital rules. Accordingly, such restrictions have the potential, over time, to cause the minimum capital standards to become increasingly disconnected from the supervised entities' risk profiles, and to significantly diminish their usefulness in ensuring their safety and soundness.

Q.5.d. At a staff briefing given by Fannie Mae on the issue of minimum capital, a Fannie representative said that raising minimum capital would require the GSE's to raise mortgage rates in order to keep earnings per share at current levels. In the context of capital standards, is there a way to ensure that more of the implied GSE benefits go toward the GSEs' housing mission, and less to the shareholders?

A.5.d. In my view, capital standards should be used solely as a mechanism to ensure the safety and soundness of supervised entities, rather than as a mechanism to promote the GSES' housing mission. As I mentioned in my testimony, channeling the activities of the housing finance GSE's toward mortgage securitization could help to ensure that more of the implied GSE benefits accrue to the housing mission rather than investors. Refocusing the housing finance GSE's in this manner also would be consistent with the original Congressional intent that the GSE's provide stability in the market for residential mortgages and provide liquidity for mortgage investors.