SPEAKERS       CONTENTS       INSERTS    
 Page 1       TOP OF DOC

OFHEO's PROPOSED RISK-BASED CAPITAL REGULATION

WEDNESDAY, MAY 12, 1999
U.S. House of Representatives,
Subcommittee on Capital Markets, Securities and Government Sponsored Enterprises,
Committee on Banking and Financial Services,
Washington, DC.

    The subcommittee met, pursuant to call, at 10:07 a.m., in room 2128, Rayburn House Office Building, Hon. Richard H. Baker, [chairman of the subcommittee], presiding.

    Present: Chairman Baker; Representatives Ryan, Sweeney, Biggert, Terry, Royce, Kanjorski, Bentsen, C. Maloney of New York, J. Maloney of Connecticut, Hooley, S. Jones of Ohio, Capuano and Vento.

    Also present were Representatives Metcalf and Vento.

    Chairman BAKER. I would like to call this meeting of the Capital Markets Subcommittee to order. As has been established by the notice for the hearing, the purpose of the hearing today is to review OFHEO's proposed risk-based capital regulation for housing-related Government Sponsored Enterprises.

    The Office of Federal Housing Enterprise Oversight was established as an independent entity within the Department of Housing and Urban Development by the Federal Housing Enterprise Financial Safety and Soundness Act of 1992 and charged it with the responsibility of implementing a risk-based capital standard, or ''stress test'' as it is known, to apply to housing GSEs, and determine the amount of capital necessary for the enterprises to maintain a positive net worth during periods of economic stress. OFHEO's proposed risk-based capital standard is vital to the secondary mortgage market because it provides an early warning signal of hidden weaknesses in the system that appears to be perfectly safe and sound under normal operating conditions.
 Page 2       PREV PAGE       TOP OF DOC

    I would like to point out that all of our discussion here today about the risk-based capital standard does not in any way reflect a concern that either of the housing GSEs are improperly capitalized today or that they are not profitable, or that they in any way have inappropriate management. In fact, a current market assessment of their capabilities would indicate both GSEs are very well-funded and are operating at very successful levels, providing needed liquidity in the housing markets.

    But our job does not look at these entities just in times of economic success, but to prepare for the unfortunate events that occurred during the 1980's, particularly in Louisiana, Texas, and Oklahoma where we saw housing values plummet and loan demand evaporate. Ultimately Congress must protect the taxpayer from the burdening costs of financial bailouts while protecting capital markets from significant disruptions should investors lose confidence in the marketplace.

    That is why it is imperative that OFHEO run an accurate stress test on each enterprise's capital adequacy and publish the results quarterly so that weaknesses can be properly addressed before facing significant capital losses.

    Although this rule has taken much longer than anyone initially expected to be promulgated, I certainly want to compliment OFHEO's effort and say that despite any criticisms that may be leveled against the particular formula that has been constructed, in principle it moves in the direction that Congress has asked, and I believe that with modest adjustments that are found appropriate by the agency, that we can in fact put into effect a regulatory standard that protects the public interest, and that is what we should be concerned with. And I thank you very much for your participation here this morning.
 Page 3       PREV PAGE       TOP OF DOC

    Chairman BAKER. Mr. Kanjorski.

    Mr. KANJORSKI. Thank you very much. Mr. Chairman, I want to congratulate you for calling these timely hearings on this sexy and attractive issue. I know all of America is just waiting with bated breath to hear the testimony today and our analysis of it.

    But quite seriously, Mr. Chairman, I supported the 1992 legislation from which this rule grows. It has been a long time in coming, not due to any failure on the part of the organization, OFHEO, but as a result of its desire to do the correct and proper thing.

    This issue is very complicated. As a matter of fact, when we passed this legislation back in 1992, the potential of risk to the system was about $1 trillion. In just seven years it is now potentially $2 trillion. These two GSEs have really grown 100 percent in seven years. It is probably more important today that we insulate them from unusual stress in the economic system and protect against a high risk to the taxpayers.

    First and foremost, all of us favor safety and soundness. That is what this rule is all about. Second, it is about making sure that we have good public policy and are carrying out the mission that Congress intended when it originally established these entities. That mission is to provide affordable housing and to keep the rate of mortgages at their very lowest level so that really the subsidy offered to these organizations by the Federal Government does eventually go through to the individual homeowner and buyer. That, in turn, helps stimulate the economy and improve the quality of life of all Americans who participate.

 Page 4       PREV PAGE       TOP OF DOC
    As I said, it is a highly technical area. A lot of people will not recognize some of the problems that the organizations face, or the names of the two entities being regulated.

    I would only offer the following suggestion, having spent a little time yesterday, but a little more than we had originally allotted, in discussing this issue. I looked at my watch, it was something like 6 o'clock when we ended our discussion. I certainly gained a great deal of information informally from the leadership and our witness today, and it brought me to the conclusion that because this is so technical, we should potentially hold a roundtable-type of discussion, and invite the regulator that is going to be testifying today and the two major entities that will be regulated, Fannie Mae and Freddie Mac. We should let them discuss with the regulator openly, and with the Congress, how this rule impacts them; whether this rule is acceptable in the state it is in, or how it can be modified. Then, maybe we could reinvent legislative Government in the way the Administration is reinventing Government.

    I think it works particularly well in this instance because of the technical nature of the rule, the limited number of parties that are immediately impacted, and the importance to the economic system. We should get this resolved as early as possible so that we can continue to have the third proposition for which these entities are constructed. I also want to protect safety and soundness and prevent stressful economic conditions.

    I want to encourage their innovative nature that is a hallmark for these two institutions and the credit system as it impacts on mortgage making for residential housing in this country. In fact, we are the model of the world. And so that we are inventing and creating something here that will be duplicated, I think, not only in this country but many times over around the world. We should try and get it as right as we can.
 Page 5       PREV PAGE       TOP OF DOC

    So I urge having this type of a thing and I want to thank you very much, Mr. Chairman, for taking the time and effort and giving us the opportunity to have this informative hearing today.

    Chairman BAKER. Thank you, Mr. Kanjorski. The topic is so exciting to everyone, it was not a difficult decision.

    Does anyone else have an opening statement? If not, I would at this time introduce our principal witness this morning, Mr. Mark Kinsey, who is the Acting Director of OFHEO, and who has had the principal responsibility for completion of the preparation of the stress test that we are reviewing this morning. Welcome, Mr. Kinsey.

STATEMENT OF MARK KINSEY, ACTING DIRECTOR, OFFICE OF FEDERAL HOUSING ENTERPRISE OVERSIGHT; ACCOMPANIED BY PATRICK LAWLER, CHIEF ECONOMIST

    Mr. KINSEY. Thank you. At this time I would like to introduce OFHEO's Chief Economist, Patrick Lawler. I asked him to join me in the hot seat today to make sure I can answer any and all questions you and the subcommittee might have.

    Mr. Chairman and Members of the subcommittee, I appreciate the opportunity to appear today to discuss OFHEO's proposed risk-based capital regulation for Freddie Mac and Fannie Mae. I especially want to thank you personally, Mr. Chairman, for your ongoing support of the regulatory activities of OFHEO. As always, this testimony represents the views of the Office of the Federal Housing Enterprise Oversight, which are not necessarily those of the President or the Secretary of Housing and Urban Development.
 Page 6       PREV PAGE       TOP OF DOC

    Regulating the amount of capital that Freddie Mac and Fannie Mae are required to hold is important because the failure of an enterprise could have serious consequences. The enterprises are critical for our housing finance system. When Congress created OFHEO back in the fall of 1992, the total obligations of the enterprises—and I am defining this as debt- and mortgage-backed security guarantees of the enterprises—stood at about $1 trillion. Today, in a little over six years, the obligations of the enterprises have doubled to nearly $2 trillion. The fastest growing part of their business is their retained portfolio of mortgages and, increasingly, of mortgage-backed securities. These assets have quadrupled. At over $725 billion, the combined retained portfolio of mortgage assets at Freddie Mac and Fannie Mae now exceeds, by well over $100 billion, the combined holdings of mortgage assets by the entire thrift industry.

    I am happy to report that currently the enterprises are very healthy and well-managed institutions. While we do not see any problems today, we cannot take it for granted that current circumstances will prevail indefinitely. Government-sponsored enterprises are not immune to problems. The Farm Credit System experienced severe problems in the late 1980's and required Federal assistance. And while Fannie Mae did not receive direct Federal assistance, at one point in the early 1980's when it was a much smaller company, it was losing nearly $1 million a day and on a mark-to-market basis the value of its liabilities far exceeded the value of its assets.

    Another important matter to OFHEO is that market forces cannot be relied on to regulate capital because the enterprises are not subject to the same normal market discipline that fully private firms face. What I am referring to here is the disabling of the normal market forces resulting from their GSE status. Freddie Mac and Fannie Mae are able to borrow money from the credit markets more cheaply than fully private AAA-rated firms, and at times nearly as cheaply as the U.S. Treasury. But investors' willingness to lend money to the enterprises is based primarily on the perception of an implicit Government guarantee and not on an evaluation of capital adequacy. Therefore, the enterprises have the ability to increase their risk-taking, possibly significantly, without much effect on their cost of funds.
 Page 7       PREV PAGE       TOP OF DOC

    Even in the extreme case when Fannie Mae had negative economic net worth in the early 1980's, Fannie Mae was still able to borrow money from the capital markets. For a fully private firm in a similar position, the flow of money from the capital markets would have dried up well before such a condition even developed. The sheer size and importance of the enterprises to the housing markets, together with the distortion of the market forces that their GSE status creates makes it critical that a strong risk-based capital standard be put in place. A strong risk-based capital standard will help ensure that the enterprises remain financially strong even under severe economic stresses.

    OFHEO's proposed risk-based capital standard represents a new and innovative approach for determining the capital adequacy of Freddie Mac and Fannie Mae. As such, we expect, and in fact welcome, constructive suggestions from a wide range of interested parties. We believe that the basics of the rule are on target. And we are encouraging everybody to take the time to study the proposal and provide us with their insights.

    Currently the public comment period ends on August 11. However, several parties have informally expressed a desire to have OFHEO extend the comment period. It may be necessary to provide additional time for all interested parties to fully comment on the proposal. We are evaluating the need for extending the comment period and expect to make a decision on this matter within the next couple of weeks.

    A stress test is the best way to determine the capital adequacy of Freddie Mac and Fannie Mae. It closely links capital to risk by calculating the risk exposure ''holistically.'' Now, that is my favorite one-word description of how the stress test works. It allows OFHEO to capture the bottom-line exposure to the enterprises from both credit and interest rate risk simultaneously. It not only captures the risks inherent in the enterprises' assets and off-balance sheet obligations, it also captures the effectiveness of hedges and credit enhancements that the enterprises put in place. No other regulatory capital standard explicitly gives credit for risk management activities.
 Page 8       PREV PAGE       TOP OF DOC

    The beauty of a stress test is that it is an early warning indicator of financial distress. A stress test can expose hidden weaknesses in a system that seems perfectly healthy and sound under normal conditions. This is the same concept used by doctors when they ask individuals to run on a treadmill in order to stress the heart so they can observe how well it functions. This enables the doctor to treat any weaknesses today, while the patient is relatively healthy, rather than after the patient suffers a heart attack when the treatment may be more difficult or too late. The same can be said for OFHEO's stress test. It assists in the discovery of otherwise hidden financial weaknesses, allowing an enterprise to fix them today before it becomes too weak financially to do so.

    While the stress test-based capital standard is a new and innovative approach for determining regulatory capital requirements, the use of a stress test is not new to the enterprises or to other large financial institutions. In fact, Fannie Mae and Freddie Mac each has been using its own stress test for assessing its capital needs for many years. As you may recall, Mr. Chairman, both enterprises played a major role in drafting the specifics of the legislation that created OFHEO and the stress test.

    In designing the proposed regulation, we had four goals in mind. First, we sought to create a model that measures risk consistently across the two companies. For example, if both enterprises purchased mortgages with the same risk characteristics and funded them in exactly the same way, they should have the same capital requirement. The use of a single model is important for this goal. Prior to the public release of our proposal, both enterprises had stated publicly that they had developed their own versions of a stress test that they believed were consistent with the statutory risk-based standard. Not surprisingly, both enterprises passed their own versions of the stress test. When using OFHEO's proposed standard, it became clear that using a common yardstick to measure risk resulted in quite the different picture.
 Page 9       PREV PAGE       TOP OF DOC

    Second, we needed to meet the statutory requirements for transparency and replicability. The model cannot be a black box. It has to be something that the enterprises can use to anticipate what their capital requirements will be. It also has to be transparent so that everybody else can evaluate it, including investors. So when we say the enterprises are adequately capitalized, investors will know exactly on what basis we make such decisions.

    Our proposed rule is, of necessity, lengthy. Every single equation and parameter in our model that are needed to reproduce our proposed capital standard are detailed in the proposal. In addition, we provided an extensive preamble to explain how the capital standard works and OFHEO's rationale in the choices made in developing the proposal. Actually the regulation itself is less than a third of the proposal. However, it will take time to understand completely how the proposal works, and OFHEO is committed to working closely with the enterprises and other interested parties so they can appropriately comment on the proposal.

    Third, we sought to develop a stress test that provided both OFHEO and the enterprises with flexibility. The stress test allows the enterprises to manage their own businesses according to their own individual strategies. It does not dictate the kinds of activities they should be involved in or how they should manage the risks. It is only a tool to measure overall risk and set capital requirements for that risk.

    The stress test also needs to be flexible enough to address innovation. The proposal must be able to capture the risk of new products when they are introduced and, just as importantly, capture the reduction in risk from the effective use of new financial instruments. This type of flexibility keeps the stress test from becoming obsolete over time.
 Page 10       PREV PAGE       TOP OF DOC

    The enterprises' activities are restricted by their charters to those associated with providing a secondary mortgage market. Before an enterprise can introduce a new program, the company must seek approval from the Secretary of HUD. During the new program review, HUD will have up to 45 days to make its decision. During this timeframe, OFHEO will be able to analyze the new program for capital treatment under the risk-based standard. Any new activity that is not a new program will generally be a variation of an existing activity. In this case, the existing techniques for measuring risk already in the proposal can easily be adjusted on a timely basis to effectively capture the risk impact of the new activity.

    The proposal allows for OFHEO to use our judgment on how to initially treat any new activity for purposes of the risk-based capital standard. We will work with the enterprises to understand the new activity and our decision will be informed by how they view the relative risk of that activity.

    We will also encourage the enterprises to come to us as soon as practical to engage in these discussions. Once we make a decision, the enterprises will be informed of the specifics of how it will treat the new activity for capital purposes so that they can plan accordingly.

    Fourth and finally, the capital standard needs to recognize and reward appropriate management of risk. The proposed rule would encourage the enterprises to manage risk effectively because higher risk is associated with higher capital requirements, and vice versa. It also is intended to discourage regulatory capital arbitrage. By tying capital very closely to risk, our proposal should discourage efforts to restructure transactions simply to receive a lower capital requirement even though the risk has not changed.
 Page 11       PREV PAGE       TOP OF DOC

    To demonstrate how the risk-based capital standard works, the proposed regulation includes calculations of capital requirements for Freddie Mac and Fannie Mae for two dates: September 30, 1996, and June 30, 1997. Freddie Mac would have comfortably met its risk-based capital requirement on both these dates with a surplus of about $1.5 billion. Fannie Mae, on the other hand, would have had a capital shortfall of around $3.5 billion on both dates. The main reason Freddie Mac met its requirement and Fannie Mae did not is that Fannie Mae had more interest rate risk than its capital base could support. Both enterprises have relatively low credit risk. Freddie Mac simply reduced its interest rate risk exposure to a greater extent than Fannie Mae had.

    These results should be interpreted with caution. Given that underlying economic conditions and the enterprises' risk profiles constantly change, these results do not necessarily reflect what an enterprise's current or future risk-based capital requirement might be.

    Furthermore, a projected capital shortfall, even a large one, does not imply that an enterprise actually has to raise that amount of capital. For example, let's focus on Fannie Mae's $3.68 billion capital shortfall on June 30, 1997. The number sounds very large. But because the proposal is so flexible, Fannie Mae could have met its capital requirement by reducing its interest rate risk in a variety of ways at an annual cost of about $70 million. Now, this estimate is not mine, by the way. It is from Fannie Mae's chief financial officer. Also, Fannie Mae's CFO stated that investors would see no perceptible change in the company's future performance as a result of this proposed capital standard, even if it is unchanged.

 Page 12       PREV PAGE       TOP OF DOC
    Let me conclude my remarks this morning by talking a little bit about the impact of our proposal on the housing market. Our proposed capital standard is unlikely to cause any change in mortgage rates nor will it curtail the enterprises' ability to finance affordable housing.

    The reasons why we believe our rule will have no negative impact on the housing market are straightforward. The first reason is competition between Fannie Mae and Freddie Mac. Since Freddie Mac comfortably met the capital standard, there would be no incentive for the company to raise the prices it charges lenders. If Fannie Mae attempted to raise its prices to recoup the costs associated with complying with the risk-based capital rule, Fannie Mae would lose profitable business to Freddie Mac. Competition for market share has heightened in recent weeks following the announcement of the exclusive Freddie Mac/Norwest deal.

    Second, the capital costs for credit risk for the two stress periods we looked at are, on average, about the same as they are in the minimum capital requirement at about 45 basis points. Since both companies currently meet their minimum capital requirement—you recall the minimum capital requirement is 2.5 percent of balance sheet assets and .45 percent of balance sheet obligations—there would be no incentive for either enterprise to change their guarantee fees that it charges lenders for accepting credit risk.

    As I mentioned previously, credit risk on affordable housing loans was not the reason Fannie Mae did not pass its risk-based capital requirement. We did discover that the credit risk on Fannie Mae loans was on average about 12 percent higher on a per-dollar basis than on Freddie Mac loans. However, if we imposed the higher Fannie Mae credit losses on Freddie Mac's loans, Freddie Mac still would have easily met its risk-based capital requirement on both dates.
 Page 13       PREV PAGE       TOP OF DOC

    The reason that Fannie Mae failed its risk-based capital requirement and Freddie Mac passed, even with the higher Fannie Mae loss rates, is because Fannie Mae had more interest rate risk. Freddie Mac chose to hedge its interest rate risk more extensively than Fannie Mae did on those selective dates.

    Affordable housing loans have been quite profitable for both enterprises. The credit risk associated with these loans remains relatively low with loan-to-value ratios that mirror those of their entire portfolios. The far more important variable is the way an enterprise chooses to fund all the mortgages and mortgage-backed securities it holds in portfolio. It is these funding choices that comprise the interest rate risk that caused Fannie Mae to fall short of its capital requirement.

    We believe our proposed rule will have a long-term positive impact on the housing market. Our proposal will help to ensure that the enterprises will remain financially strong and able to perform their important public purposes in good as well as bad times when they will be needed the most. Also, our proposal will provide market participants with additional information concerning the exposure of the enterprises to credit and interest rate risk.

    Mr. Chairman, I appreciate the opportunity to testify before this subcommittee and I will be pleased to respond to any questions.

    Chairman BAKER. Thank you very much, sir. Maybe we can move this fellow off into the secondary market who's beating around upstairs. I first want to say I appreciate the work that has been done.
 Page 14       PREV PAGE       TOP OF DOC

    Mr. KINSEY. Thank you.

    Chairman BAKER. I will try to stick to the five-minute rule on my own time as well, and so I have a series of things that I want to cover with you this morning and probably necessitate coming back for a second round.

    In order to put the importance of this in context for people not following it very closely—not that there is a parallel between LTCM and GSEs, just asset size—at the time of LTCM's difficulty, they were about a $125 billion asset institution. Ballpark combined assets of Fannie and Freddie would be in the range of $750 billion, some 5 to 6 times larger.

    If LTCM presented a potential for systemic risk because they were using their own internal risk models to govern their risk-taking investments, it would seem very clear to me that it is very prudent for Congress to look for the highest and best stress test to ensure that the public interest is not put at risk by an unexpected downturn in our economy.

    Given that, I want to try to understand, and I am not sure that I have, the manner in which this test was constructed. It would appear that the ten-year period used in order to facilitate the worst economic conditions was 1984 to 1993, where prices of housing during that ten-year period dipped in years four and five, recovered in years six through nine, and in year ten the housing prices were actually higher than at the initiation of the stress test period. The total drop in pricing in average houses from the top of the peak to the bottom of the trough was 11 percent.

 Page 15       PREV PAGE       TOP OF DOC
    I have heard it explained that this model, if put into effect in its current form, would represent the newest, highest, most stringent requirement for safety and soundness in business regulation. I then learned in Moody's, rating of a AAA-rated corporation in the Investors Service Great Depression scenario, which is an essential component of being found a AAA rating: that it mandates a 27 percent drop over a ten-year period, the same period of OFHEO's test, and that the Moody requirement is that the business enterprise has enough capital to maintain a positive business stature throughout the ten-year period to receive a AAA rating. And we can all assume that there are a number of AAA-rated corporations in America.

    Additionally, I went to the Standard & Poor's rating to understand how they come to achieve their highest rating. And they have the Texas scenario. Thank God it wasn't the Louisiana scenario. And it basically has a similar set of constraints over a ten-year period, frankly much worse than the OFHEO test.

    Why is it that you used a stress test evaluation period that only resulted in 11 percent reduction pricing, when I know what occurred in the 1980's in the South and the oil patch was far worse than what you have constructed your stress test with? Is there a reason for this?

    Mr. KINSEY. We were in fact constrained by the requirements of the legislation. Let me explain what the requirements were. The requirements were that we were to look at the worst regional credit experience, defined by mortgage losses and default rates, associated with a regional aspect of the country. Region was defined in law as a contiguous area containing at least 5 percent of the U.S. population.

 Page 16       PREV PAGE       TOP OF DOC
    So based on those parameters and based on the combined loan historic database that we have at OFHEO for both Fannie Mae and Freddie Mac, we constructed a giant computer model that searched for the worst regional experience. What we discovered was that experience was for loans originated in the 1983–1984 period for basically the eastern oil belt. The four States I believe are Mississippi, Arkansas, Louisiana, and Oklahoma.

    At that time in the data that we had, we looked at 30-year fixed rate mortgages. What we discovered was that on a gross basis, the loss rate associated with those mortgages made in that timeframe, in that region of the country, were 9.4 percent. That is a fairly high loss rate.

    I think you mentioned the Moody scenario. I believe the Moody scenario is a tougher scenario than the scenario I just described. That is really a question for Congress to decide as to what is the appropriate protection we should require for the enterprises, but we didn't really focus on any other type of scenario. We were just focused on——

    Chairman BAKER. You were basically constrained by the law, but we would both acknowledge that there were and had been economic conditions far more stringent than what you are applying in the——

    Mr. KINSEY. The economic conditions, possibly yes.

    Chairman BAKER. Thank you. My time has expired.

    Mr. Kanjorski.
 Page 17       PREV PAGE       TOP OF DOC

    Mr. KANJORSKI. In your opinion, should you come back to Congress and ask us to adopt a standard on the rule?

    Mr. KINSEY. I think the standard we have proposed is a reasonable one provided that the standard very precisely measures risk. We think our proposal does that. As long as it does that and as long as it is very sensitive to risk, then in combination with a floor capital requirement, which is the minimum capital requirement, we think that is probably sufficient protection.

    Now, once we put this in place and we have an opportunity to explore alternative scenarios, I might have a different opinion, but right now I think it is sufficient.

    Mr. KANJORSKI. Some of the opposition to the regulation, I am just starting to realize, is that the regulation for the first time forces the two institutions involved to put forth their internal, anticipated products to the regulator to be worked on the model. I think there is some fear out there that if they spend time and energy to develop a new product and then they have to turn it outside of the organization to the regulator to be worked on the model to gain approval, it will escape and the proprietary nature of their new product will go to their opponent, Freddie Mac to Fannie Mae, or Fannie Mae to Freddie Mac.

    In either case, the expenses attendant to the development of the new product would be an absolute loss to the innovator, and they think, therefore, it will stymie good economic business, internal business sense of being innovative, and creating new products. What would be your response?
 Page 18       PREV PAGE       TOP OF DOC

    Mr. KINSEY. I have two responses to that. First of all, our proposal does not require us to approve anything. They can do whatever they want to do. Our requirement is that once they do it, we are to measure the risk associated with that activity and to apply an appropriate capital treatment.

    Mr. KANJORSKI. But at that point would you require information from them to put into your model?

    Mr. KINSEY. We absolutely have to have that information, yes.

    Mr. KANJORSKI. How would you protect their confidentiality?

    Mr. KINSEY. The same way that we do right now. We have, for example, ongoing examinations at both companies. We already know what products are in development. This is really not——

    Mr. KANJORSKI. How do you keep that confidential from the competitor?

    Mr. KINSEY. That is part of the ethical requirements of our examiners. We don't share secrets.

    Mr. KANJORSKI. When you retreat to the word ethics in Washington, it sort of makes us all shudder.
 Page 19       PREV PAGE       TOP OF DOC

    But then there is some merit that they have fear that they may spend——

    Mr. KINSEY. I don't think there is any merit at all. We have been privy to their secrets for five years now, and to my knowledge, not a single one has leaked to the other. That is what we are in business to do, is to understand how the two companies operate and understand their strategies and to opine on that from a safety and soundness perspective. We absolutely do not share trade secrets between the two companies.

    Mr. KANJORSKI. Not that you share, but do you feel there is such confidentiality in your organization that there could be any leaks? We hear what happens inside the CIA every day. Are you better than they are? You have better maps than they do.

    Mr. KINSEY. All regulators have the same sets of issues as we do, and you know, I think we are no better, no worse, but that is what we must do: keep secrets.

    Mr. KANJORSKI. Let me ask you something that is important, because I said our interest is not only safety and soundness, first of all, and then the mission. I understand that HUD is planning in the not-too-distant future to issue new affordable housing goals for both Fannie Mae and Freddie Mac.

    Mr. KINSEY. I think that is correct.

    Mr. KANJORSKI. In your testimony you talk about the fact that affordable housing loans have been quite profitable for both enterprises. The risk associated with these loans remains relatively low. In light of that, would increasing the requirements for new GSEs to purchase additional low- and moderate-income mortgages of necessity result in any adverse impacts on the safety and soundness of the two companies?
 Page 20       PREV PAGE       TOP OF DOC

    Mr. KINSEY. We don't think so. We looked at the risk characteristics, as I mentioned in my testimony, of the affordable housing loans that they currently purchased to meet the requirements, and we discovered that there is really not a significant difference between those and other loans. And so if they were to increase those types of loans, assuming they have the same types of risk characteristics, no, we wouldn't think there would be a problem at all.

    Mr. KANJORSKI. So the increased problems or burdens that the Secretary would place on these new structures would not have an effect on the capital requirements of the two enterprises?

    Mr. KINSEY. Any type of increased activity will have an effect on the capital requirements. If the question is would it have an adverse effect on their ability to do affordable housing financing, the answer to that question is no, it will not.

    Chairman BAKER. Thank you, Mr. Kanjorski.

    Mrs. Biggert.

    Mrs. BIGGERT. Thank you, Mr. Chairman.

    Mr. Kinsey, you mentioned in your testimony that there would be a public comment period for approximately four months. And then you mentioned something about possibly extending that. Do you think that Fannie Mae or Freddie Mac and the public can analyze duplications of these regulations within that four months?
 Page 21       PREV PAGE       TOP OF DOC

    Mr. KINSEY. That is what I am in the process of determining right now. I have had discussions with both enterprises, but I also need to have discussions with other interested parties. And once I have the full range of those discussions, I will make a decision.

    Mrs. BIGGERT. What do they have to do within that four months?

    Mr. KINSEY. I believe that our rule, as proposed, gives the enterprises sufficient information to appropriately comment on the structure of the capital requirement. They need to analyze how we treat the risks, how we treat the risks relative to one another, to make sure that we are giving the right incentives to the two companies. And I think that is a very important role that they play, and I am sure they are busy working on that right now.

    Mrs. BIGGERT. Do you, as far as public comment when you—with regulations like this, is there really public comment? People come forward to——

    Mr. KINSEY. I am expecting to get a lot of comments on this. This has generated, frankly, more interest than I anticipated in terms of people wanting to understand how this rule works. I think as the Chairman pointed out, this rule may have, in fact, implications for financial institutions even outside of the GSE world, because it is really the first time that anybody has systematically looked at the specific risks, for example, associated with holding mortgages. And I think people are interested in that.
 Page 22       PREV PAGE       TOP OF DOC

    Mrs. BIGGERT. How long has it been since there was any change in regulation? Was that six years?

    Mr. KINSEY. For Fannie Mae and Freddie Mac?

    Mrs. BIGGERT. Mm-hmm.

    Mr. KINSEY. This is our first proposal on risk-based capital. The 1992 legislation required us to develop risk-based capital regulations. We have, as people have noted, taken quite a bit of time to do this, but we think we have done it right, and so this is our first attempt at showing the world what our proposal is.

    Mrs. BIGGERT. Well, what was the emphasis for doing that? Forgive me if I——

    Mr. KINSEY. Required by the law.

    Mrs. BIGGERT. Were there standards then set in 1992 that had to be implemented?

    Mr. KINSEY. There were guidelines given to us on how to develop the risk-based capital requirement. Specifically it told us that we needed to use a stress test to develop the risk-based capital requirement and it gave us lots of guidance on how to determine what the interest rate stress was and how to determine what the credit stress was and what variables to use.
 Page 23       PREV PAGE       TOP OF DOC

    Mrs. BIGGERT. Between 1992 and 1998, what has been the standard?

    Mr. KINSEY. The standard has been the minimum capital standard, which is just a simple leverage ratio, 2.5 percent of on-balance assets and.45 percent of off-balance sheet guarantees, which are basically their mortgage-backed securities. That was intended to be a floor capital level. The risk-based standard was intended to complement the minimum standard such that they would have to hold the higher of the risk-based requirements or the minimum requirements. The risk-based standard was designed to make sure that they could not suddenly dramatically increase their risk exposure without having to hold more capital. The minimum capital standard cannot ensure that because it is not risk-based.

    Mrs. BIGGERT. Thank you.

    Thank you, Mr. Chairman.

    Chairman BAKER. Thank you, Mrs. Biggert.

    Mr. Bentsen.

    Mr. BENTSEN. Thank you, Mr. Chairman.

    A couple of questions. Mr. Kinsey, on this new product, the 1992 Act set up the structure for HUD to review new products.
 Page 24       PREV PAGE       TOP OF DOC

    Mr. KINSEY. New programs.

    Mr. BENTSEN. New programs or products, I guess. But the credit review of a new product by OFHEO is—the genesis of that is in the rule itself.

    Mr. KINSEY. We have a statutory role in new program approval in the sense that we have to evaluate and report to the Secretary whether or not this new program would risk significant deterioration of their capital position, and that is a determination that I make to the Secretary that he uses in deliberating whether or not he would approve a new program or not. We have that statutory role until our risk-based capital standard is finalized and enforceable. At that point in time, our role is simply to, whatever they do, just measure the risk and require them to hold capital. So we have no approval authority at all.

    Mr. BENTSEN. Assuming this rule goes through and all and you set up this new standard, then you—I mean, I guess the way I heard your testimony, you would still have pre-credit approved——

    Mr. KINSEY. There is no preapproval of anything. We are required to make sure that they have adequate capital for the risks that they undertake. It is just the same way, anytime there is a new——

    Mr. BENTSEN. I don't mean to interrupt you, but in response to Mr. Kanjorski, because he raised a question of how would you keep confidential product line that is in the development stage while you are reviewing it, which I think you could. I have no qualms about that, but I guess my concern is more once you have a risk-based capital structure and a model that has been adopted, in effect the GSE should—you just look at the whole picture of the GSEs and if they have a product line that they come out with which erodes capital or has too much risk that erodes capital, then it would be pinpointed at that time.
 Page 25       PREV PAGE       TOP OF DOC

    But the way I understood your testimony was even after this rule went through, you would still put it through some credit approval process before the product could be offered, which Mr. Kanjorski raised the question of corporate secrets, I guess, or whatever, but the other concern—the only concern I would have is that you slow down product development.

    Mr. KINSEY. Let me see if I can clarify my answer. Here's how this would work. Let's assume we have the risk-based capital standard in place and it accounts for all the activities that they currently do. Now they come up with a new activity. We are going to need to figure out a way to treat that activity in the risk-based capital standard. We would expect to have a conversation with either of the enterprises at the earliest possible date, when they really have a good understanding of what they are going to do and how they are going to do it, to help them and help us decide what the capital treatment might be on that particular activity if they chose to engage in that activity. That is our responsibility.

    We will not release that information until they actually do the activity, until it is public. Once it is on their balance sheet or if they have an obligation, then we are required within the quarter to incorporate that in the risk-based capital standard because we have to incorporate all of their activities in the standard.

    Mr. BENTSEN. Obviously HUD, for whatever purposes, could hold up a product. But you wouldn't necessarily hold up a product while you are trying to figure out how you can model it later on?

    Mr. KINSEY. No. And if we couldn't figure out how, or there were a disagreement on how to treat it, we would treat it conservatively until we were able to figure out a more accurate way to do it; and once we did, we would tell the enterprises or the enterprise exactly how we would treat that so they would understand how to calculate their capital requirements. They could plan accordingly.
 Page 26       PREV PAGE       TOP OF DOC

    Mr. BENTSEN. Thank you.

    Thank you, Mr. Chairman.

    Chairman BAKER. Thank you, Mr. Bentsen.

    Mr. Terry.

    Mr. TERRY. Thank you, Mr. Chairman. Last night one of my staffers said, ''I need to brief you real quickly.'' I sarcastically retorted, ''Isn't that the same thing?'' What I am learning in Washington is that it necessarily isn't so.

    On September 30, 1993, in signing his Executive Order on regulatory reform, which is currently a new cause on Capitol Hill, one that I am glad to participate in—but President Clinton said, ''Today I am signing an Executive Order to create a fair, open, streamlined system of regulatory review for our Government to eliminate improper influence, delay, secrecy, and to set tough standards and time limits for regulation.''

    It goes on to reference, then, his September 11 Executive Order that he signed directing our agencies to eliminate 50 percent of their internal regulations.

    Now, one thing that I have learned today is this work in progress has been six years and is nearly 700 pages long. Now, I am confused at how this is a ''streamlined and limited regulation'' in light of what our goals are today and as President Clinton stated in his Executive Orders. Could you explain?
 Page 27       PREV PAGE       TOP OF DOC

    Mr. KINSEY. First of all, the actual regulation is only about a third of the 647 pages. It is actually much less than that in the Federal Register. But we are responding to very explicit directions from Congress on how to develop a stress test. It is quite specific in the variables that we are to use and how we are to go about doing that.

    I believe the desire of Congress in passing the law requiring us to do this was to create a capital standard that on the one hand would protect the taxpayer from unnecessary risk but on the other hand, and just as importantly, would not result in the GSEs holding excess capital, because Fannie Mae and Freddie Mac have very important public missions, and those missions are to assist housing, especially for the low- and moderate-income families.

    So the desire was to create a capital standard that could very precisely measure risk so that we could get the right amount of capital, not too much, not too little, but the right amount.

    The enterprises are engaged in a very complicated and complex business. It is true that they only do mortgages, but mortgages have lots of optionality in them which, from a risk measurement perspective, is very sophisticated and complicated to measure. That, I believe, was the goal of the legislation, and we followed, we think, the spirit of that in developing this rule.

    Mr. TERRY. One last question, and this was touched on in both the colloquy with the Chairman and the Ranking Member, but why do you think that your modeling is a better test of capital adequacy than that being used by the private enterprise sector today, especially in light of its high credit rating and stock value and its ability to serve the housing market?
 Page 28       PREV PAGE       TOP OF DOC

    Mr. KINSEY. Well, I touched upon that in my testimony. First of all, the enterprises aren't subject to normal market discipline, so we can't assume that their models will give them the proper incentives to manage risk under all economic scenarios that you can think of. Adequate discipline doesn't exist.

    Second, we only regulate two companies. As I mentioned also in my testimony, if we relied on their versions of internal models, we would probably have an unfair capital standard, because, as I mentioned, one company has a different view of risk than another company. We would have an incentive for the company that was more risk averse to take on more risk from a competitive perspective. That is not good public policy.

    Third, we think it is a very good idea, and it is consistent with the rest of the way in which we regulate Fannie Mae and Freddie Mac, to have the deliberations be in an open and public forum. For example, we are required to report the results and conclusions of our examinations in our annual report to Congress June 15 every year.

    No other regulator is required to do that. We regulate the enterprises in a very open environment. This is very consistent with that.

    Mr. TERRY. Thank you.

    Chairman BAKER. Mrs. Maloney.

    Mrs. MALONEY. Thank you, Mr. Chairman.
 Page 29       PREV PAGE       TOP OF DOC

    OFHEO's proposal has been criticized for conflicting with real business world risk-management incentives and for restricting innovation. What is your reaction to this?

    Mr. KINSEY. I don't think that it does conflict with the way the enterprises measure and manage risk. In fact, the rule was based, to a very large extent, on how Fannie Mae and Freddie Mac internally measure and model credit risk. Plus we went out to the industry, the private mortgage insurers, large thrifts, Wall Street firms, to understand how they measure risk.

    We think this is, in fact, close to state of the art in terms of the way that sophisticated companies measure risk associated with mortgage holdings.

    Now, on the issue of stifling innovation, I quite frankly don't understand that issue, and I would hope that somebody could be more specific with me on that issue. Our rule doesn't prescribe anything for the enterprises, other than the bottom-line risk that they hold, they have to have capital commensurate with that risk.

    That is the only thing it prescribes. It doesn't say ''You can buy this, you can't buy that.'' It doesn't say ''You have to issue this debt.'' It doesn't say ''You need to hedge the risk in a particular way.'' All of those choices are left up to the enterprises. They can manage as they see fit. They can reduce risk, they can raise capital, but the bottom line is they have to meet the standard. They have literally dozens and dozens of choices on how to do that.

 Page 30       PREV PAGE       TOP OF DOC
    Mrs. MALONEY. I understand that your test used a wind-down scenario for its stress test. Could you explain the wind-down method?

    Mr. KINSEY. What it does is say, at a particular point in time, what are the risks associated with the particular book of business that you hold on that particular day? That is what we are measuring. We are assuming that that book winds down, meaning that mortgages either default, prepay, or they mature based on a set of economic stresses.

    It does not assume new business occurs. It assumes that the current set of risk management, hedges, and activities they have in place, stay in place and don't change. It doesn't assume that the enterprises would adjust to anything. It doesn't assume they would be stupid and take more risk. It just assumes we are going to measure the risk inherent in the book of business at that particular point in time.

    Mrs. MALONEY. Is there a policy rationale for imposing lower minimum capital requirements on Fannie and Freddie than we impose on larger banks and thrifts?

    Mr. KINSEY. I think there is.

    Number one, we are talking about Government-sponsored enterprises. From the perspective, for example, of managing interest-rate risk associated with mortgage holdings, the enterprises have the ability to control the liability side of their balance sheet to a greater extent than fully private firms. They can issue debt at any time, at any maturity, virtually for any amount, just by doing it, and then people will buy it at favorable rates. They also have the ability to issue callable debt at virtually any amount and virtually at any time or any maturity.
 Page 31       PREV PAGE       TOP OF DOC

    Private firms don't have that ability. By being able to do that, they have the ability to manage interest rate risk to a much greater degree than a bank or a thrift would.

    On the credit side, they have nationally diversified portfolios. Their underwriting practices limit the types of mortgage assets that they purchase primarily to the A market, and home mortgages are one of the safer forms of assets to hold in terms of loans.

    So, yes, I think that there are good reasons why their requirements should be lower than those for banks or thrifts.

    Mrs. MALONEY. Thank you very much.

    Chairman BAKER. Thank you.

    Mr. Ryan.

    Mr. RYAN. Thank you for coming to us today, Mr. Kinsey. I am looking at some of the criticisms we have been reading about the proposed rule. I would like to ask you to address a couple of them. I just walked in the room, so you may have addressed this earlier, but I ask you to reconsider that.

    They have argued that OFHEO has built a model that replicates business enterprises in great detail, but that is going to lag. That is going to effectively dictate risk management choices and that Freddie and Fannie are going to look more like as time and this model is applied. Can you address those concerns?
 Page 32       PREV PAGE       TOP OF DOC

    Mr. KINSEY. I think that's a red herring. I don't quite understand that.

    What we have done is we have said you do business the way you want to do business. We have got a consistent, fairly applied measurement tool for measuring risk. It measures the risk exactly the same for both companies. It measures the bottom-line risk and says you have to hold capital commensurate with that risk. It is your decision as to how much risk you want to take. It is your decision as to how much capital you want to hold. That is really what this rule is all about. It is just a tool for measuring risk.

    If what one of the enterprises means is that they have to be as safe and sound as the other, well, then, yes, that is true.

    Mr. RYAN. In promulgating this rule, did you look at the instances with Freddie Mac and the Freddie model in applying this rule to affect Fannie, and do you see your standards should apply equally to the two, and do you see major distinguishing marks between the two of them, the two GSEs?

    Mr. KINSEY. We determined what models we would use based on our understanding of what Fannie Mae does, what Freddie Mac does, and what the rest of the mortgage industry does. Our models don't replicate those of the GSE's or anybody's, because we built the models using combined data from both Fannie Mae and Freddie Mac. So we, in fact, are going to have slightly different models from anybody else, because we have the largest data set in the world right now to develop our models.
 Page 33       PREV PAGE       TOP OF DOC

    But I might point out that I have heard a lot of people suggest that the reason Freddie Mac passes is that our model more closely resembles theirs. It is very interesting. Just a couple of years ago, we had some of our economists share our general modeling techniques with the academic community. We received a series of inquiries on the part of Fannie Mae complaining that we were giving away proprietary information. They were complaining that the models we were developing looked very, very similar to their models.

    So I think this is just a red herring. They measure risk in very much the same way that we measure risk.

    Mr. RYAN. It is my understanding that once the rule is finally published, that there is a year before it takes effect.

    Mr. KINSEY. That is correct.

    Mr. RYAN. What is your time line. What do you envision as your timeframe?

    Mr. KINSEY. I will make a decision in the next couple of weeks whether to extend the public comment period. But once that closes, it will be dependent on how many comments we get and the nature of the comments.

    We will have to take whatever time is necessary to analyze those comments. If we believe that we need to make significant changes to the rule, we will repropose the rule in a different form. If we believe that the changes are minor and we can go forward with the final rule, we will do that.
 Page 34       PREV PAGE       TOP OF DOC

    I am thinking right now that whether we do a reproposed or a final, it could be as early as next year.

    Mr. RYAN. As early as next year. We are looking at quite a great deal of time?

    Mr. KINSEY. At least another two years before it is enforceable.

    Mr. RYAN. Thank you, Mr. Chairman.

    Chairman BAKER. Thank you.

    Ms. Hooley.

    Ms. HOOLEY. Thank you, Mr. Chairman. Thank you for your testimony. I was not here to hear you, but I did read it this morning before I came in. I am glad that you finally finished this important work.

    Mr. KINSEY. Thank you.

    Ms. HOOLEY. I trust the conclusions that you and your staff came to, but I worry about the ability of outside parties to evaluate your work.

 Page 35       PREV PAGE       TOP OF DOC
    Mr. KINSEY. Yes.

    Ms. HOOLEY. As I read over your testimony, one of the thoughts that came to my mind was how complex this whole situation was and the extensive use you must have used for computer modeling.

    I am aware that you published in the rule a detailed account of the variables you used in coming up with the GSE stress test, many groups want to evaluate your work during the comment period. Would you outline the steps you are taking to allow outside interests to evaluate your methods and conclusions?

    Mr. KINSEY. This rule has been in the Federal Register now for about a month. We have quite literally almost on a daily basis been briefing somebody on this rule. We have been giving briefings at the policy level and at the detailed technical level. We are committed to working with everybody that asks our help to get them to understand how this rule works. That has become quite literally a full-time job for a considerable amount of my staff right now, and we expect that to continue.

    I think as time goes on, people will start to understand this more and more. This is going to be a function of time. I think a lot of the responses that people have heard were basically initial responses. As people start to get into this and work through how this works, people are starting to understand this better.

    We think this is just going to be a function of time before everybody gets up to speed. The evaluation that I am going to be doing the next couple of weeks is to try to figure out whether I need to give people more time.
 Page 36       PREV PAGE       TOP OF DOC

    Ms. HOOLEY. If I may, Mr. Chairman, just follow up, you talk about doing briefings. Obviously—you and your staff spent a lot of time putting this together.

    Mr. KINSEY. Yes.

    Ms. HOOLEY. And a lot of extensive computer work, I am assuming.

    Mr. KINSEY. Yes.

    Ms. HOOLEY. Can somebody with some briefings, even over a short period of time, can they understand it well enough so they can evaluate what you have actually done?

    Mr. KINSEY. You said over a short period of time. I don't know.

    It depends on the sophistication of the people that are talking to us. A lot of the larger financial institutions that are engaged in similar types of activities have the capacity that Fannie and Freddie have to understand this.

    Now, everybody other than Fannie and Freddie are at a slight disadvantage in the sense that they don't have the input data necessary to crank through the model to come up with the actual numbers that I referred to in my testimony. But that information is proprietary and only Fannie Mae has Fannie Mae's information and only Freddie Mac has Freddie Mac's information.
 Page 37       PREV PAGE       TOP OF DOC

    So they won't be able to replicate the numbers, but they will be able to understand, over time, as we work with them, how the model works, yes.

    Ms. HOOLEY. OK, because I am concerned about the housing groups. Clearly they have some interest in this and how they are going to help evaluate what you have done.

    Mr. KINSEY. We are working with the housing groups as well.

    Ms. HOOLEY. Thank you.

    Chairman BAKER. Thank you.

    Mr. Sweeney.

    Mr. SWEENEY. Thank you, Mr. Chairman.

    Thank you, Mr. Kinsey, for appearing here. I may be covering some territory that has been covered already, and I heard part of your responses to Mr. Ryan's questioning, and I am still curious and confused as to your rationale for using the model that you have, when it appears to conflict with all of or the vast majority of those models developed in the private financial marketplace.

    So in order to try to reconstruct that, I would like to ask a couple of questions. When you, over the past six years, as you devised this model, how much consideration did you give to the changing marketplace? Because it is my sense, and I think the common sense, that financial markets are rapidly changing and it is tough to predict from now until five years from now what they are going to look like.
 Page 38       PREV PAGE       TOP OF DOC

    Maybe you could walk me through that process in terms of this 600-some-odd page regulation, how that is going to provide ample flexibility?

    Mr. KINSEY. Let me approach it this way: We spent the first probably two-and-one-half to three years of OFHEO's existence trying to understand how mortgage risk was estimated and understood at the enterprises, by mortgage insurers, by the market in general.

    The general approaches to measuring, for example, default risk on mortgages, are the same today as they were five or six years ago. The general modeling techniques for measuring prepayment risk on mortgages are the same today as they were five years ago.

    What changes is you have more information about these processes. We fully expect, for example, to periodically update our rule based on new sets of historical information that might result in behavioral changes in the way people choose to default on a mortgage or choose to repay their mortgage.

    A good example, I think, is sort of the very low-cost refinance that can occur today versus five or six years ago when there was a significant cost associated with refinancing. We think that our models measure, attempt to measure, that change in behavior as well as the enterprises' models do.

    Mr. SWEENEY. What defines those changes? Is it not the private marketplace?
 Page 39       PREV PAGE       TOP OF DOC

    Mr. KINSEY. Yes, it is the activity as it finally occurs. We will monitor that and make sure our models are up-to-date.

    Mr. SWEENEY. What is the source of my confusion and some of my colleagues' is why wouldn't you more closely employ that private marketplace model?

    Mr. KINSEY. I think we are.

    You mentioned in your opening statement that our models conflict with what the market does. I don't think they do. That is what this public comment period is all about, for people to tell us whether they think we have done a good job. I haven't heard anybody tell us our models are not good models of risk.

    Mr. SWEENEY. I will send you the names and phone numbers of those that talked to my office, because I have heard nothing but that from them.

    Let me just close with this statement. It may better be a reflection of the fact that the period of time to develop this regulation, six years, seemed an extensive period of time, the size of the regulation, 600-plus pages, may have seemed extensive, and the period for public comment, which I understand in your statements you have said you are reexamining whether that will be extended.

    Let me just conclude by saying if that is the definition of reinventing Government, that is not a good definition. We need to work at being a little more responsive to the groups that we work with and regulate over.
 Page 40       PREV PAGE       TOP OF DOC

    Mr. KINSEY. I just have to make one more comment, and I don't really mean to sound defensive, but OFHEO was a brand-new organization, and we have only been in existence for six years.

    We quite literally started with one employee and worked our way up, designing computer systems, buying the computer systems, installing them, getting the data from the enterprises, going through a process of discovering how the enterprises—they have never been regulated before—discovering how the enterprises operate, how they measure and look at risk.

    All of this took a long time. It is somewhat unfair to suggest it took six years for us to develop this. It did not take six years. It took several years to do this. In fact, we basically had the model done a year ago, but it takes a while to write a regulation in sufficient detail over a complex subject as this to make sure everybody has the opportunity to try to understand this.

    Mr. SWEENEY. I can sympathize with that as a former regulator. Sixty pages, and it did take six years, is a little excessive.

    Mr. KANJORSKI. If the gentleman will yield, I have heard two or three Members on the other side refer to reinventing Government in six years.

    May I say, first of all, when this piece of legislation was proposed, it was the Bush Administration that was in office, and it was the Bush Administration that was in charge of HUD. It was a bipartisan action of this Congress that passed this bill.
 Page 41       PREV PAGE       TOP OF DOC

    This is the result of the request of the non-regulating Bush Administration and Republican Party, not the result of the Clinton Administration and the Democratic Party. I just wish we would not make it partisan, is what I am saying.

    Always referring to this reinventing Government, this is not reinventing Government. I happened to vote for it; Mr. Bush was correct because he had just come off the S&L disaster. We wanted to make sure we did not have another $400 or $500 billion disaster with Fannie Mae and Freddie Mac. We said look at this and see if we have safety and soundness out there sufficiently so we do not have to revisit this problem again.

    Mr. SWEENEY. My distinguished colleague is absolutely correct. I think you further reinforce the point that the intention was not to develop substantially large and unworkable regulations. The desire to insure that that doesn't happen isn't a partisan issue in my mind.

    Chairman BAKER. I thank the gentleman.

    Mrs. Jones.

    Mrs. JONES. Thank you, Mr. Chairman.

    A couple of questions, three or four questions. Would there have been any way for you to issue the subject matter of these regulations in less pages than you used?

 Page 42       PREV PAGE       TOP OF DOC
    Mr. KINSEY. Boy, we sure tried.

    Mrs. JONES. Had you done so, you would have been beat up saying you weren't clear enough?

    Mr. KINSEY. I suspect that would have been the case.

    Mrs. JONES. There is no indication, is there, based upon all that is—the history of either of these two GSEs, that the managers or directors of them are not doing a good job; is that correct?

    Mr. KINSEY. No, there is not.

    Mrs. JONES. That may be part of the dilemma here, they are doing such a good job they are competing out here in the world with everyone else? Is that a likely statement?

    Mr. KINSEY. Yes. I mean, I think they are doing a good job, and they are moving into areas that other folks think they shouldn't be moving into, yes. I think that is a correct statement.

    Mrs. JONES. Can you suggest what impact there would be on the affordable housing market if these two agencies did not exist?

    Mr. KINSEY. I think there would be quite a substantial impact.
 Page 43       PREV PAGE       TOP OF DOC

    Mrs. JONES. In fact, currently in our state of affairs in this country, there is still—even with Freddie Mac and Fannie Mae—a dearth of affordable housing available to members or to citizens of this United States?

    Mr. KINSEY. Yes. I want to make it absolutely clear, I think the enterprises perform very important public missions, and I am absolutely in support of their missions.

    I just want to make sure they are able to do it whenever they are needed the most.

    Mrs. JONES. OK. There is no indication they are not able to do it. They are needed the most right now. There are more people in the United States than ever before, more people need housing. Our communities are growing by leaps and bounds.

    Mr. KINSEY. Yes.

    Mrs. JONES. Take me back to pre–1992, since I wasn't a Member of Congress or paying any attention to these issues at the time. Maybe Mr. Kanjorski has said this. The reason you were created is because there were no regulatory agencies for the GSEs?

    Mr. KINSEY. There was no effective oversight of the activities of Fannie Mae and Freddie Mac.

 Page 44       PREV PAGE       TOP OF DOC
    Mrs. JONES. So at the time Fannie Mae and Freddie Mac were operating, there was no reason necessarily to believe, but this was just in response to the savings and loan crisis?

    Mr. KINSEY. It certainly was in response to the savings and loan crisis. For example, there was a capital requirement for Fannie Mae and Freddie Mac that dated way back when Fannie Mae was just a portfolio lender.

    The requirement was just basically a simple leverage ratio for on-balance sheet assets. What happened was they invented the mortgage-backed security and suddenly billions and billions and billions of dollars of contingent liabilities were being racked up by both Fannie Mae and Freddie Mac, but there was no capital requirement that was being associated with those contingent liabilities. The capital regulation of Fannie and Freddie was outdated.

    Mrs. JONES. Everybody agreed that it was outdated.

    Mr. KINSEY. Yes.

    Mrs. JONES. There were no sides on that. Everybody said there had to be a new system.

    Mr. KINSEY. Absolutely. Including Fannie and Freddie.

    Mrs. JONES. I am scared to ask this question because I don't think I have enough time to get a real answer.
 Page 45       PREV PAGE       TOP OF DOC

    It is not that you are not capable of giving it to me. But how was the regulation risk-based standard first developed?

    Mr. KINSEY. This is the first development of it.

    Mrs. JONES. I understand that.

    Mr. KINSEY. How was the law crafted?

    Mrs. JONES. Yes. Not drafted, developed.

    Mr. KINSEY. Through very intensive negotiation. I was there. At the time, I worked for Treasury. I was in the Bush Administration at the time. It was basically a bipartisan effort, as Mr. Kanjorski said, where both sides worked out the details and Fannie Mae and Freddie Mac had significant input into the specifics of the stress test.

    I believe they crafted a stress test, I know they did, that very closely resembled the way in which they themselves determined their capital needs. So they should be very comfortable with this approach, and I think that they are.

    Mrs. JONES. That wasn't the import of my question, but, OK. Even with trying to apply these standards—I am almost done, Mr. Chairman—to Fannie Mae and Freddie Mac, there is no reason to suggest that they are not well-run companies, even if they are GSEs?
 Page 46       PREV PAGE       TOP OF DOC

    Mr. KINSEY. Yes.

    Mrs. JONES. Even if you applied a standard of a well-run company to Fannie Mae or Freddie Mac, it is likely they would pass that test; is that true?

    Mr. KINSEY. Well, with respect to our capital standard, Freddie Mac passes, and as I pointed it out, it would be relatively easy for Fannie Mae to pass our test as well.

    I might emphasize, you know, we do comprehensive examinations of the two enterprises, and, yes, they are well-run, well-managed companies.

    Mrs. JONES. I am done. I am out of time.

    Thank you, Mr. Chairman.

    Chairman BAKER. Mr. Vento.

    Mr. VENTO. Thanks, Mr. Chairman. I wish Mr. Sweeney was still here, because most of the rule and regulation review issues that we have acted on we have exempted regulators from actually complying with because of safety and soundness standards.

    So I think that the emphasis on reinventing Government, I don't know how that fits with this, because the reinvention here was that we were going to make certain that we didn't run down the path that we had with other chartered institutions.
 Page 47       PREV PAGE       TOP OF DOC

    I was struck by the statement that the total combined or retained portfolio of the two GSEs, $725 billion, is $100 billion more than the entire thrift industry, which I think, Mr. Chairman, underlines the importance of having a risk-based standard which I was involved, along with others, in terms of writing into law.

    I take it, Mr. Kinsey, that you don't object to or are looking for additional risk based change in terms of the formula that is in the law with regards to 5 percent area region, the 9.4 percent lost?

    Mr. KINSEY. I think the entirety of the proposal as it exists is a reasonable capital requirement.

    Mr. VENTO. How would you relate this to the relationship between capital standards that exist in terms of minimum capital standards versus the stress test? In other words, if, in fact, we were to or the law was to increase the minimum capital standards, it may translate into a less rigorous stress test.

    Mr. KINSEY. Well, I think the relationship is close in the sense that the minimum only makes sense, as I pointed out earlier, if we have a strong risk-based capital standard. By strong, my definition of that is it must very precisely measure risk. That is part of why it is complex. When you start taking away some of that complexity, you start to lose some of your ability to precisely measure risk. If you do that, then I can't be as comfortable in saying they are adequately capitalized.

 Page 48       PREV PAGE       TOP OF DOC
    Mr. VENTO. I just want to point out there is a relationship between these, in other words, the exchange is to have in essence a minimum capital test that is less rigorous if you have a stress test.

    Mr. KINSEY. That is correct.

    Mr. VENTO. That is all I am trying to point out. I think obviously other regulators have opted for that. Have the other regulators taken an interest now that you have the rule written?

    Mr. KINSEY. They are starting to.

    Mr. VENTO. Mr. Chairman, I guess I am not surprised, and I know you would not be because they clearly are going to be under the same types of pressures.

    So you are out there really where no one has dared to venture, I guess, in the words of space exploration, in terms of trying to get this. Of course, as you know, I have defended the ability or tried to build OFHEO from the ground up, and we think you have done good work. Obviously in this case, I think there are some questions that reasonably need to be answered.

    You suggest, for instance, the two differences between Fannie and Freddie here. Are you are suggesting that Fannie could resolve this matter by doing more hedging against interest rate risk, as Freddie does?

 Page 49       PREV PAGE       TOP OF DOC
    And second, you are saying OFHEO states nearly a fourth of the guarantee fees charged by Fannie and Freddie are profits above what is normal shareholder return. What are you suggesting, that they trim back their profits a little bit?

    Mr. KINSEY. I am suggesting they have the ability to comply.

    Mr. VENTO. By trimming back the profits and retaining more assets?

    Mr. KINSEY. There is a cost associated with complying with the rule. They have the ability to do that without affecting their ability to do their business.

    Mr. VENTO. The hedging issue, are you comfortable that would be another scenario? These are obviously your points. You are not saying how they do this?

    Mr. KINSEY. I am suggesting they need to lower their interest-rate risk to comply with the risk-based standard.

    Mr. VENTO. There was an exchange earlier with regard to HUD and programs with regard to OFHEO and activities.

    In fact, of course, we wrote those into law. Do you find that those types—that role, in terms of providing information to HUD, providing information to the GSEs, is troublesome in terms of trying to do the job as an objective regulator?

 Page 50       PREV PAGE       TOP OF DOC
    Mr. KINSEY. No.

    Mr. VENTO. Mr. Chairman, it is very unusual, because this collaboration—in other words, being able to shoot proposals over, I guess which could be done informally if it wasn't formally recognized in law, may be interpreted by some to co-opt the regulator in terms of this. So that is the concern that down the road obviously we would like to see collaboration, but very often regulators are stung with the thinking that they are compliant to those that they regulate.

    Obviously you have been able to avert that, at least with regards to one of your constituents in this case, Mr. Kinsey.

    Mr. Chairman, I appreciate the opportunity to raise these questions. There are many more that you have indicated. I commend the staff for their work in terms of providing us with some questions. We sound like we know something.

    Thank you, Mr. Chairman.

    Chairman BAKER. Thank you, Mr. Vento.

    I have a fairly long set of questions to go through with you, Mr. Kinsey, and I think Mrs. Jones would like to be recognized. I am going to suggest that if you don't mind, we have a vote on the floor, I would recognize Mrs. Jones at this time.

    Mrs. JONES. It is real short. Just on the point that my colleague, Mr. Vento, was asking, you are suggesting that Fannie Mae should consider more hedging. What would be the advantage of more hedging by them?
 Page 51       PREV PAGE       TOP OF DOC

    Mr. KINSEY. It would reduce their interest rate risk exposure. Basically what the stress test does is require Fannie Mae and Freddie Mac to insure for catastrophic losses.

    Right now we believe Fannie Mae is insuring for expected losses, as any normal business would, but the stress test is imposing a set of conditions that are unexpected. This is the purpose of a stress test. Normally when a financial institution goes belly up, it is because of some unexpected set of circumstances. It is not to say they are not smart management; it is just they don't expect it to happen.

    Mrs. JONES. Then the smart managers, and I don't know whether they have or have not suggested they should do more hedging, but what would be the disadvantage as a manager if I decided not to do more hedging?

    Mr. KINSEY. It will cost you.

    Mr. VENTO. One other one, there is the sort of argument over good and bad hedges. We had Mr. Soros explaining that to us.

    Mr. KINSEY. There are a variety of different ways to reduce interest rate risk.

    Mrs. JONES. What would be the disadvantage if I decide not to?

 Page 52       PREV PAGE       TOP OF DOC
    Mr. KINSEY. It would cost them money to do so, which has to come from someplace.

    Mrs. JONES. Thank you, Mr. Chairman.

    Chairman BAKER. Mr. Kanjorski.

    Mr. KANJORSKI. Going to that point, when you examined Freddie Mac as opposed to Fannie Mae, Fannie Mae had two options: They could increase their internal equity by selling shares, or they could hedge.

    Mr. KINSEY. Or reduce risk, correct.

    Mr. KANJORSKI. Hedge funds, as we know, are hedge funds. Some of them are not too secure even in the best of times. If they had hedged with Long Term Capital Management, they would have passed your stress test, but as we have discovered, Long Term Capital Management would have bellied up before Fannie Mae did. So it would have been a waste of $70 million of buying that premium.

    Mr. KINSEY. If they would have hedged with Long-Term Capital, they would have had as much as an 80 percent haircut, because there was no credit rating on Long-Term Capital, and we would have haircutted the value of that hedge.

    Mr. KANJORSKI. That is really going to the substance of my question. If they do use hedge funds, what authority do you have to examine the stability and the reasonable survival necessary of that hedge fund under the same stress test?
 Page 53       PREV PAGE       TOP OF DOC

    Mr. KINSEY. We don't have that authority specifically. What we do is take, as a proxy, the public credit rating associated with the party.

    Mr. KANJORSKI. We have to recognize, if we put these people under stress, all these institutions, particularly the hedge funds, will be under the same type of stress, and they will likely fail because they are much more narrowly funded than these institutions.

    So is it not sort of a false security?

    Mr. KINSEY. Not really. I mean, even a AAA gets a haircut. So we take a very conservative approach and say even if they are AAA, I agree with you, under stressful conditions, some AAA's are going to fail. So you are going to get a haircut even on a AAA. But the haircuts increase dramatically as the credit rating goes down. So we are encouraging the enterprises to use top-rated counterparties.

    Mr. KANJORSKI. Top rated. Lloyd's of London was top rated at one time. It is gone.

    Insurance companies, some of them are top rated. I have seen them disappear. Likely in a stressful economic situation a lot of these institutions that we think are AAA-rated will be gone, a lot sooner than Fannie Mae and Freddie Mac may be gone.

    So perhaps should we consider as the regulator, the Congress, not to look at the potential of hedging, but to look at firmer capital, more substantial capital?
 Page 54       PREV PAGE       TOP OF DOC

    Mr. KINSEY. What you are suggesting is a possibility.

    Mr. KANJORSKI. One other question: This is the first time you are being armed with the ability to enforce against these enterprises if they do not comply——

    Mr. KINSEY. Not yet. After it is final.

    Mr. KANJORSKI. When it is in place. If this standard had been in place when Freddie Mac was in difficulty several years ago and could not raise capital——

    Mr. KINSEY. Freddie Mac has always been able to raise capital.

    Mr. KANJORSKI. Would they have survived or would you have forced them to raise capital, and if so, how much?

    Mr. KINSEY. If you are referring to the situation they had trouble with their multifamily program?

    Mr. KINSEY. When Johnson took over.

    Mr. KINSEY. You mean Fannie Mae.

 Page 55       PREV PAGE       TOP OF DOC
    Mr. KANJORSKI. Fannie Mae, yes. Did I say Freddie Mac? Yes, I am sorry.

    Mr. KINSEY. We think, yes, in the mid-1980's, they would have been in serious problems. We probably would have put them in conservatorship.

    Mr. KANJORSKI. That may be an argument to the benefit of internal management deciding. You know, what you are trying to do—and I am sorry, Mr. Chairman, I will only take a second—in any way you can is to insure against everything. If you insure against everything you are apt not to have a profit, because profit comes from risk.

    That is the intelligence of successful businessmen as opposed to unsuccessful businessmen. So some people say, ''Well, I don't think I am going to have a flood in 400 years, so I am not going to spend $1 million to insure against that, and if I do not have a flood in my lifetime, I have saved $1 million a year, and I am that much more wealthier. On the other hand, if a flood comes along next year, I look like the dumbest fool that ever lived.''

    Mr. VENTO. If the gentleman will yield.

    The issue, of course, is hedge funds look OK today. This is not the last rule. Obviously as you go down the road, if there is an indication that the risk-based capital measurement that you have is not correct, you will be able to modify that.

    Mr. KINSEY. I think that is correct.

 Page 56       PREV PAGE       TOP OF DOC
    Mr. VENTO. I was trying to find out is it dynamic in terms of where we are going. Besides being the first risk-based capital regulation under the sun.

    Mr. KINSEY. Your question was if the requirement was in place at the time they were in trouble, and my answer was yes, they would probably be in conservatorship, because we have a set of FDICA type prompt corrective action authorities that are very similar to banks.

    But if the requirement were in place before they had this problem, we think that that probably would have forced them not to get in the problem to begin with, because as the problem started to get worse, we would have required them to start holding more capital or start reducing risk. So it is very likely they would have never been in that problem.

    Chairman BAKER. If I can jump in here, we are down to about six minutes.

    I want to come back. Just for the record, I want to establish I was a Member of the Banking Committee at the time this proposal was adopted, and I voted against it. I have some questions that will make that imminently clear upon my return.

    [Recess.]

    Chairman BAKER. If I could ask, we would reconvene our hearing at this time.

 Page 57       PREV PAGE       TOP OF DOC
    Mr. Kinsey, I suspect that I may be the only Member to return, so I will take my time in trying to get thorough explanations of concerns that have been raised by the regulated entities, as well as some of my own particular concerns.

    There has been some suggestion the 1992 Act requires, in addition to the risk-based capital that is determined by the stress test, that an additional 30 percent add-on attributable to managerial and operations risk should be added on to the risk-based standard, meaning that once the stress test determines the minimum capital required, that is multiplied by 1.3 to get the real standard.

    Mr. KINSEY. That is correct.

    Chairman BAKER. Some have suggested that this add-on, coupled with haircut requirements, particularly for BBB-rated counterparties, results in a 104 percent haircut, because they take the 80 percent, multiply it by the 1.3, and come up with this 104 percent number.

    In looking at the way this particular aspect of the formula is operating, you don't only do it on the risk side, but you do it on the credit enhancement side as well.

    Mr. KINSEY. Yes.

    Chairman BAKER. Taking that explanation, the haircuts start at a small level and escalate up. Under the presumption for a lower-rated Counterpart I, the longer you do business with them, the more likely there is to be an adverse consequence.
 Page 58       PREV PAGE       TOP OF DOC

    Mr. KINSEY. That is correct.

    Chairman BAKER. So that in the early years of this Counterpart I relationship, where the haircut is first leveled, is about 6.7 percent at year one, 40 percent at year five, 73 percent at year ten, and we don't ever get to this mythical 104 percent because they were missing the fact that the credit side of the formula also gets the benefit of the 1.3 multiplier.

    Mr. KINSEY. You are correct in your understanding of how the 30 percent add-on works. We basically determine the amount of capital that is required to pass the stress test, and then we multiply that by 30 percent.

    I view this 30 percent as basically capital for the unquantifiable risks. I know the legislation says management and operations risk. Well, that is just I think a descriptor of the nonquantifiable types of risk.

    Chairman BAKER. As Members earlier were saying in characterizing hedge funds, it is not necessarily valid guarantees to prevent loss.

    Mr. KINSEY. Yes.

    Chairman BAKER. This 30 percent I would call managerial risk, where you invest in the hedge fund which later does turn out not to be able to perform.

 Page 59       PREV PAGE       TOP OF DOC
    Mr. KINSEY. Yes, it is a potential cushion for those types of things that you do not, or are unable to measure, or can't measure very well. Yes, it is a cushion.

    Is it correct that it be proportional to the credit and interest rate risk? That was a determination that Congress made. As of right now, I think that is probably reasonable, given the current circumstances that the enterprises are in. I am not so certain that that makes good sense in the long run, that it be necessarily proportional. I think we could look at alternative ways of determining what that add-on might be.

    Chairman BAKER. Before the break, I am sorry—did you have a follow-up?

    Mr. LAWLER. If I could say a word about the 104 percent. The 104 percent conclusion suggests that by having a credit enhancement with a BBB-rating, you actually worsen your capital condition and require more capital because of that. That is simply not correct. You can't multiply any number that comes up by 1.3. We, in fact, give credit for all credit enhancements. The lower the credit rating, the less credit. But——

    Chairman BAKER. There is some value to it.

    Mr. LAWLER. There is value to it. Even with the 80 percent haircut, you never even get to 80 percent until the last month of the stress test. Even then, you would still be getting 20 percent of the complete credit. On average, in the fifth year, as you said, you would be getting 60 percent credit, and with higher rated credit enhancements, you would get that much more.
 Page 60       PREV PAGE       TOP OF DOC

    Chairman BAKER. Sure. Before the break, I indicated I had concerns about the stress test when it was originally debated in the Congress.

    One, I didn't think anybody could make it work. Congratulations.

    But, two, I frankly felt there were areas of the stress test—having been in the real estate business during the period of time in difficulty where the measurements weren't severe enough, and what we would create would be a false sense of capital adequacy by applying a formula which really didn't go quite far enough.

    Now, that was not a popular perspective in the arena in which we were debating this particular stress test.

    For example, here is an element. The haircut plan outlined in the rule gives more credit to highly rated mortgage insurers, 555 versus 55. It also gives more credit to the use of derivatives contracts over mortgage insurance as a means of credit enhancement.

    OFHEO proposes to give more credit to derivatives counterparties. As one individual said to me, given this principle, you would rather have us hold a derivative than a 555 rated security of a Wall Street firm.

    I then found out that the only case in which that is applicable is where the derivative has earmarked collateral to support the risk taking position.

 Page 61       PREV PAGE       TOP OF DOC
    Mr. KINSEY. That is correct.

    Chairman BAKER. So that we are really taking a fully collateralized position as a superior element of risk hedging to the 555 rated security.

    Mr. KINSEY. That is correct.

    Chairman BAKER. Tell me why that is wrong?

    Mr. KINSEY. It is not wrong.

    Chairman BAKER. Some have expressed that that view is skewed somehow. I was just trying to make sure I was understanding it properly.

    Mr. KINSEY. You understand it correctly. I do want to point out, however, that your statement that we give incentives to use credit derivatives over mortgage insurance I think is misleading. The law requires that Fannie Mae or Freddie Mac obtain one of three forms of credit enhancements for mortgages that have loan to value ratios greater than 80 percent. One is mortgage insurance, another is recourse, the third is some sort of participation.

    We will only give credit for purposes of the stress test for credit enhancements on high LTV mortgages if they have one of the specified types, which, in the predominant form, is mortgage insurance. They cannot use things like spread accounts or other forms of credit derivatives as credit enhancements for high LTV mortgages as the primary means for insuring that risk. If they have mortgage insurance on the high LTV mortgages and decide to layer additional credit enhancements on, we will give credit for that, but they cannot be used as the primary means, because that is against the law.
 Page 62       PREV PAGE       TOP OF DOC

    Chairman BAKER. And on the point of high loan to value loans, really my next question, there is the allegation that the capital requirements on these would range between 3 and 6 percent capital.

    That view comes from a simulation of the rule that shows a 3.1 percent capital requirement in the up rate, a 6.0 percent capital rate requirement on the down rate side. These rates are for newly originated loans in the month of origination. As the loans season, the capital requirements decline.

    Mr. KINSEY. That is correct.

    Chairman BAKER. So that if you were to take the 95 percent loan to value loan and look at the risk-based capital over the seasoned life of the loan, it would end up at about 1.3 on the upside and 2.1 on the downside.

    Mr. KINSEY. About a third probably, yes.

    Chairman BAKER. If the expected life of a loan is about seven years, which I understand is about a typical market perspective about loan life these days, and that may be really truly lengthy, the risk-based capital used for a guarantee fee calculation on a 95 percent loan to value loan would be around 1 percent; is that correct?

    Mr. KINSEY. I am going to let Pat answer that question.

 Page 63       PREV PAGE       TOP OF DOC
    Before he does, I do want to point out that some of the sensitivity numbers that are in the proposal itself are very difficult to interpret as marginal capital requirements for a couple of reasons that Pat will mention.

    But one important reason is that we used an average guarantee fee in these calculations. The fact of the matter is that for higher risk mortgages, like 95 LTV's, the enterprises would apply either explicitly or implicitly on higher G fee, which would affect the capital requirement associated with that activity. We did not use the more realistic G fee numbers, because we felt that would have been divulging proprietary information about how they priced their products.

    Chairman BAKER. Further, you don't look at individual loans. You look at the portfolio and require capital sufficient to offset the average of the portfolio?

    Mr. KINSEY. Yes. Pat will give you a more thorough answer.

    Mr. LAWLER. You are quite right. While a 6 percent capital charge for a loan is by itself a very high number, the perspective that a company properly would be concerned with if it were actually purchasing such a loan is not only what will the capital requirement be immediately, but what will the capital requirement be over the lifetime of that loan, or the time they are likely to be holding that loan.

    The results for four-year-old loans are much more representative of that. For Fannie Mae, where the rate scenario was the one that was the really relevant scenario, the binding scenario at the two dates we looked at, a four-year-old loan with a 95 percent LTV would have had about a 1.3 percent capital charge.
 Page 64       PREV PAGE       TOP OF DOC

    That is still a high number compared to requirements on other loans, but I think it is fair to point out that it is only a third of what a bank would have to have for the same loan.

    It is still high relative to other loans. Some loans are going to have relatively higher capital costs and some relatively lower.

    There are other important things to consider. One, as Mark mentioned, the guarantee fee behind that number was lower than a company would actually charge if it were just making high LTV loans, just buying that kind of loan.

    And second, the 1.3 percent result could be misleading in that one doesn't actually have to go out and raise that amount of capital. One can reduce other risks at a cost far lower than the cost of raising other capital. So while 1.3 percent seems higher than the 45 basis points they currently have to hold, the difference is a misleading representation of the increase in cost associated with capital on that loan.

    So there are all those considerations and on balance I think it is probably a very reasonable number. If it is similar to the kinds of results that the enterprises and others get for the relative risk of a 95 percent loan, then we have done our job.

    Chairman BAKER. Thank you.

    Mortgage banks particularly appreciate having a guaranteed outlet for their loan production because they are not really set up to fund loan portfolios for anything more than just warehousing purposes. There have been some concerns about the reduction or elimination of the use of loan commitment contracts would increase the risk of bankruptcy in the seller-servicing network.
 Page 65       PREV PAGE       TOP OF DOC

    It would seem to me that in order for that to continue, that a modest premium or delivery fee would cover the cost for the GSE to hold capital against these types of commitments and thereby basically hedge them. They do make—there is one point made, however, particularly with regard to Fannie Mae that I find valid or at least worth your consideration to review with regard to their delegated underwriting and services program.

    It would appear that as proposed, OFHEO tends, under the construction of the rule, to treat it rather harshly. I could argue that OFHEO should not treat the DUS co-insurance lenders simply as BBB-rated counterparties and would simply encourage you to take a look at the manner in which that system is being considered. I think there is a valid point that is being made there that the assessment is a little harsh.

    On the other hand, I have got a question that goes to something that is probably at the margins of what the stress test is designed to do. It hasn't been stated, but it is a capital adequacy requirement, not a mission analysis. At the same time, when the stress test determines that there is a capital inadequacy because interest rate risk has not been sufficiently hedged against, you then have to ask the larger question, what is it that is creating the interest rate exposure and is it those things held in portfolio. Of particular interest or concern is that of mortgage-backed securities.

    When you focus on that element of that investment portfolio, does that really put anybody else in a house?

    Mr. KINSEY. You raised a lot of questions.
 Page 66       PREV PAGE       TOP OF DOC

    I guess I will answer that one first. With respect to the mortgage-backed securities that are held in portfolio at Fannie Mae and Freddie Mac, if you look at year 1998 numbers, Fannie Mae's portfolio was a little over $415 billion. About 48 percent of their portfolio at the year-end was comprised of their own mortgage-backed securities.

    Chairman BAKER. What was that number say three years ago, five years ago?

    Mr. KINSEY. Far less. Far, far less.

    Freddie Mac at the end of 1998 had a mortgage portfolio of around $255 million. About 66 percent of that was their own mortgage-backed securities. I mentioned at the beginning of my testimony that that is the most rapidly growing area of the two companies. It continues to be.

    The flow tends to be around 75, 80 percent of new additions to both portfolios or their own mortgage-backed securities.

    Chairman BAKER. Explain what holding your own security does for the housing market?

    Mr. KINSEY. Less than providing a guarantee on a loan. The primary mission of Fannie Mae and Freddie Mac is to provide liquidity for lenders. They do that by buying mortgages from lenders and making them liquid either in the form of a mortgage-backed security which they swap back to the lender or they buy loans outright and package them as a mortgage-backed security and sell them to capital market investors.
 Page 67       PREV PAGE       TOP OF DOC

    In either case, the lender has a non-liquid investment changed to a very liquid investment. The act of purchasing back their own mortgage-backed securities does not change their credit risk exposure at all. They still own the credit risk exposure on those mortgages that back those securities whether they own them or whether somebody else owns them. From a capital perspective, what they are doing when they repurchase securities is they are making a decision to take on some interest rate risk. I believe they would probably argue that they are helping support the market for their securities by purchasing their securities thereby making the market more liquid and thereby having better execution for their securities.

    That is a very, very indirect way of providing assistance, I think, to families trying to buy homes.

    Chairman BAKER. You have not suggested as a result of your analysis that that practice be limited or in any way modified other than in the case of Fannie to perhaps engage in some hedging devices that might cost them $70 million?

    Mr. KINSEY. Our responsibility is to ensure that they have capital sufficient for the risk. We do not have the responsibility to look at whether that is an appropriate investment or not. That is up to you.

    Chairman BAKER. So I am just making the point very clear. Your analysis is capital adequacy, not mission related?

    Mr. KINSEY. That is correct.
 Page 68       PREV PAGE       TOP OF DOC

    Chairman BAKER. Once we determine that they are functioning in a safe and sound manner, the Congress needs to determine whether that manner is complicit with the charter that requires them to do certain things.

    Mr. KINSEY. If I might, I might address a couple of the other statements that you made.

    First on commitments. The fact of the matter is when an enterprise engages in a contractual commitment to purchase a mortgage from its seller/servicers, they are, in fact, taking on risk. Both enterprises acknowledge that, and, in fact, they talk about that in their annual reports. The law requires us to take into account the commitments outstanding at the time that this stress test is run, recognizing the fact that there is risk associated with these contracts.

    The way in which we treat commitments in the proposed rule is actually, I think, favorable to Fannie Mae and Freddie Mac. We assume that these things will be securitized and held outside of the corporation, so we are considering only the credit risk that would be associated with purchasing the mortgages under these commitments.

    We are not making the decision that a portion of these mortgages would be held in portfolio and, therefore, would be subject to interest rate risks. So from that perspective, we think we are treating them very fairly. And in terms of the up-rate or down-rate environment, in the down-rate environment we are assuming that all of the commitments are met and the mortgages come in rather rapidly. In the up-rate environment, we assume 75 percent of the commitments are met and that they come in more slowly.
 Page 69       PREV PAGE       TOP OF DOC

    So we think we are being very realistic in how the actual delivery of the mortgage product would be made to the enterprises given the economic circumstances with respect to the distribution of risk characteristics that would be associated with the mortgages delivered under the commitments, we basically say whatever you have purchased in the last six months, that is the same set of average characteristics that we will assume will get delivered to you. So we think that is also a very fair assumption.

    With respect to the delegated underwriting loans and the capital associated with that, to the extent that the seller-servicer in this program is not rated or does not have a public rating, we will treat them as a BBB entity in terms of a counterparty risk. There is real risk associated with relying on payments from these institutions as part of the delegated underwriting program.

    These institutions give Fannie Mae a form of credit enhancement on, particularly, the multifamily mortgages that they sell to Fannie Mae. Now, we think that even if one of these companies were to experience trouble in a very severe economic scenario, it would be very quite difficult for Fannie Mae to transfer, for example, the servicing rights on those types of mortgages to another company because part of that transfer involves the transfer of the credit enhancement that is associated with being part of that program.

    And under severe economic stresses, it may be very difficult to find somebody else to transfer the servicing rights to. But for the institutions that are rated, they do not get the BBB or up to 80 percent haircut. They get the haircut that would be associated with the rating. So for AAA, there is a maximum 10 percent haircut. For a AA, there is a maximum 20 percent haircut. For a single A there is a maximum 40 percent haircut.
 Page 70       PREV PAGE       TOP OF DOC

    Now the credit enhancement is used if the mortgage defaults. Lenders are to make good on the mortgages that they sell back to Fannie Mae. For the most part the defaults occur basically in the front half of the stress test. So in the ten-year stress period, the bulk of the making good on the credit enhancements for the DUS underwriter would occur basically in the first five years.

    As you pointed out earlier, our credit counterparty haircuts are phased in over time so even, for example, the BBB, the actual haircut of the cash flows would be far less than that because, in all likelihood, on average the credit enhancement would be made good in the first five years of the stress test, in which case the actual haircut would be far less than the 80 percent. So, yes, we will be looking into how we treat DUS, and these are the types of issues that I think are very appropriate for people to comment on.

    Chairman BAKER. No hearing would be complete without a Y2K question. I have been concerned not only as to GSEs but to the financial institution marketplace generally about the implementation of FASB's new method of accounting in addition to the complicity with all the Y2K problems.

    In the course of your analysis, have you come to any determination, looked at, or have any particular feeling about whether a delay in the implementation of FASB's rules would be of any consequence to the GSEs?

    Mr. KINSEY. Let me answer the question this way. I believe both Fannie Mae and Freddie Mac have committed sufficient resources to comply both with FASB 133 and to comply with Y2K, but I will say that this has really stretched the capacity of the enterprises. I would have concerns potentially with some of the other participants in the industry because not everybody has the resources that Fannie and Freddie have. If Fannie and Freddie are stretched, my guess is that other firms would be very hard pressed to do both.
 Page 71       PREV PAGE       TOP OF DOC

    Chairman BAKER. Let me get to sort of the end conclusion in my mind of all of this.

    On the one hand, we all want a liquid housing market and individuals who otherwise could not get access to a loan perhaps get access to loans as a result of the secondary market. We have an obligation, however, at the same time to ensure that the events of the 1980's do not, to the best of our ability, reoccur; and should they reoccur, at least not to the depth of difficulty which we found ourselves. That is on one side of the scale.

    On the other side is the question of if we impose reasonable regulatory standards and capital adequacy requirements, there will be some who say all you are going to do is limit access to some, increase mortgage rates for others, and overall reduce liquidity in the market.

    Do you really see that it is an either/or?

    Mr. KINSEY. No, I don't. What I have tried to point out in my testimony is that what is important here is that the rule precisely measure risk. And our proposal suggests that certainly Freddie Mac would meet the proposal and Fannie Mae could easily meet the proposal. Both companies could do this without basically changing the way that they currently do business. We think that as a result of that, it should have no impact on mortgage rates per se or no impact on their ability to do affordable housing lending.

    Now, if economic conditions get worse, clearly their capital requirements are going to go up and the costs of complying with the rule will go up; but it won't necessarily be the fault of the rule that the enterprises are having to hedge more, having to hold more capital. That would be the same decision that they would make even in the absence of this rule because they want to protect their stockholders as well.
 Page 72       PREV PAGE       TOP OF DOC

    They are well-managed institutions and they would be responding to the same types of circumstances as the stress test would be demonstrating. So I think this rule does not create any incentives other than what actually exists in the marketplace for mortgage rates increasing or for their ability to do affordable housing lending.

    Chairman BAKER. Thank you.

    Mr. Bentsen, you have been patient and quiet. Do you have some more questions?

    Mr. BENTSEN. Thank you, Mr. Chairman. I apologize for having to leave and come back. I had another meeting.

    I have a few questions. Let me go, for a second, to the mortgage-backed security issue based upon the questions that the Chairman raised.

    And you talked about—and this is a little bit off the topic but you talked about the level of mortgage-backed securities issued by the GSEs that they hold in their own portfolio. These are marketable mortgage-backed securities?

    Mr. KINSEY. Yes, they are.

    Mr. BENTSEN. What percentage—do you know what percentage they make of the original issue that they hold? Is it a large percentage? Is it a small percentage? And do you know—are they holding these for a long period? Short period?
 Page 73       PREV PAGE       TOP OF DOC

    Mr. KINSEY. Yes, they hold them for long periods. They are not in the market to buy and sell. When they buy them, it basically stays there until the mortgages underlying the security either default, prepay, or mature. In terms of percentages——

    Mr. LAWLER. They both hold about a fourth of their MBS securities.

    Mr. BENTSEN. And the vast majority you would say is not—not that they are holding while they are making a market in the issue itself or are they making a market?

    Mr. LAWLER. Freddie Mac does some trading to try and support the market for its securities. I don't think Fannie Mae has a similar operation, but the proportion of its securities that Freddie Mac holds, because of these trading activities, are a very tiny proportion of the total.

    Mr. BENTSEN. They are holding it because it is a good product——

    Mr. KINSEY. Normally. If you are in the business to help support the market, you are on both the buy and sell side. They are normally just on the buy side.

    Mr. BENTSEN. And in your risk analysis, you talked about interest rate risk as it relates to mortgages held—directly held in their portfolio. Do you also consider interest rate risk associated with the mortgage-backed securities?
 Page 74       PREV PAGE       TOP OF DOC

    Mr. KINSEY. Yes.

    Mr. BENTSEN. I was unclear on that. I assumed that was the case.

    Let me ask about the designated underwriting services because I don't know a lot about that. But that has—your concern there is the mortgage originator who is selling the mortgage, originating the mortgage and selling it to Freddie or Fannie and then they are retaining the servicing and passing through the guarantee fee and the PMI back to Fannie or Freddie.

    Mr. KINSEY. Correct. But there is also a form of credit enhancement associated with being in that program.

    Mr. BENTSEN. Which would be the mortgage insurance PMI or——

    Mr. KINSEY. It is basically recourse. If the loan goes bad, Fannie has the option to put it back. They get a coverage of the percentage of the loss associated with that. So it is a form of recourse.

    Mr. BENTSEN. So your concern with respect to the credit quality of the originator is——

 Page 75       PREV PAGE       TOP OF DOC
    Mr. KINSEY. It is a counterparty risk, just like that with a mortgage insurer is.

    Mr. BENTSEN. I understand that. On the originator it is a standpoint that even though Freddie or Fannie would have recourse to the loan or to the asset that is pledged to the loan that you would have—that you would have to go through an originator that is in trouble and so it just—it slows down the process.

    Mr. KINSEY. Well, they would have access to the actual collateral, but they also have an agreement with the actual lender that the lender would pick up some of the costs associated with the default.

    Mr. BENTSEN. In talking about the up-rate and the down-rate—and I am sorry I wasn't here, but I think I understand what you are saying about how you grade down-rate risk. Up-rate risk, you said 75 percent comes back in.

    Mr. KINSEY. The formula for purposes of the stress is basically that you start with a nine-month average of the ten-year treasury yield. That is the base rate, and then it either goes up 75 percent or down 50 percent with the 600 basis point cap. That is the interest rate stress.

    Then we have formulas that basically tie all other rates reasonably to this movement of the ten-year rate and that would be the interest rate stress, both the up-rate and the down-rate scenarios in the stress test. The actual capital requirement would be based on whatever situation is worse for them, whether it be the up-rate or the down-rate scenario.
 Page 76       PREV PAGE       TOP OF DOC

    Mr. BENTSEN. Based upon the——

    Mr. KINSEY. Based on the performance of their portfolio over the stress period.

    Mr. BENTSEN. And in structuring your models, how do your models comport with private sector models, Moody's, and S&P, and how they structure, not Fannie Mae, but separate mortgage pool?

    Mr. KINSEY. I think they generally look at same risk variables. The private sector is concerned more about mortgage termination. They don't really care whether the mortgage defaults or prepays because Fannie Mae is going to make good or Freddie Mac is going to make good on the payments associated with the security that——

    Mr. BENTSEN. I guess I am not saying—I mean, I think—I kind of agree that looking at a Freddie or Fannie mortgage pool is not—not the best thing to look at, because there are other issues at play that I think the rating agencies look at, but looking at non-conventional mortgage pools, do you use similar risk variables?

    Mr. KINSEY. Yes, like loan-to-value ratio is very important for measuring default. The size of the loan, whether it is owner-occupied or investor owned is important. The type of mortgage product is important. Is it a 30-year fixed rate? Is it an adjustable rate? Is it a balloon?

 Page 77       PREV PAGE       TOP OF DOC
    Mr. BENTSEN. Loss recovery.

    Mr. KINSEY. We have separate models for loss recovery. That is based on a number of variables, including interest rates for holding-period costs, what we expect them to be able to sell the property for, and the average time it takes to sell a property in a particular region.

    There are a lot of variables like that that are of the same or similar types of variables that the rating agencies look at.

    Mr. BENTSEN. Would you say that the model is more conservative than what Standard & Poor's or Moody's might use for a non-conventional mortgage pool?

    Mr. KINSEY. Well, they have different requirements for different levels of comfort. The strongest are for a AAA, something less for a AA, something less than that for a single A.

    Mr. BENTSEN. Right. In terms of what they would grade the pool or any security——

    Chairman BAKER. If I could jump in earlier on that point. Earlier in the hearing, I brought that up for consideration and asked Mr. Kinsey's opinion. As I understand it, that a Moody's rating for a catastrophic decade would require an average decrease in value of 27 percent as opposed to what the result of what the law dictated, not OFHEO's judgment, is about 11 percent decline in values in the stress test promulgated by OFHEO.
 Page 78       PREV PAGE       TOP OF DOC

    And the Moody's requirement says that you must have capital sufficient to withstand that real loss circumstance over the decade where OFHEO's average price of housing actually increases by year ten over where they started in year one.

    So that was one of the things that bothered me about the construction of this stress test. They really couldn't go out and get a Texas-Louisiana worst case. They were constrained by the way the model was drawn in the law and have really a significantly less stringent model requirement than Moody's or Dunn & Bradstreet.

    Mr. KINSEY. For AAA?

    Chairman BAKER. For AAA.

    Mr. LAWLER. I think there is an important thing to add to that if I might. In addition to the house-price scenario we have, which we use for convenience to make it similar to the stress period, we specifically calibrate the model to the results of the four States in the two origination years that form our benchmark loss experience.

    So after they go through this experience that is a severe house price decline, we have to increase the losses by about 15 percent over what our model projects so that we have calibrated to the benchmark experience, this worst regional experience, and after we have done that, we think we have a fairly tough and stressful period.

    When we compared the results of our proposal for selecting the benchmark time and place three years ago, we looked at very carefully what different rating agencies assume for different credit ratings; and we judged that the period that we had identified by the process that we were proposing looked to be about a AA-plus level of credit stress, looking at a number of rating agencies. Certainly not as tough as Moody's AAA scenario, but a tough scenario.
 Page 79       PREV PAGE       TOP OF DOC

    Chairman BAKER. I would also point out to the gentleman that the S&P standard is known as the ''Texas scenario.'' You might want to look at that.

    Mr. BENTSEN. I have. And the gentleman knows we have both—just for your benefit, we have both been sometimes skeptical of the rating agencies in our official capacity and, in some instances, in other parts of the country.

    In talking about the high loan-to-value mortgages, you—if I understand that correctly, you are looking at—you are looking at what the credit enhancement is, the mortgage insurance and the quality of that, and then you are signing a capital——

    Mr. KINSEY. It is a haircut. For example, let us assume the first 25 percent of the loss on a high LTV mortgage is covered by mortgage insurance.

If it is a AAA-rated mortgage insurer, we haircut that by up to 10 percent over the ten-year period. If it is AA, it is up to a 20 percent haircut; but again it is phased in over time.

    So the actual haircut would depend on precisely when the credit enhancement would be made good, whether the mortgage defaulted in period one versus period two.

    Mr. BENTSEN. On top of that do you assign the 3—where does the 3 to 6 percent capital assignment—it is not assignment, but it is——

    Mr. KINSEY. That was an exercise where we looked at what happens if we increase by some dollar amount the number of 95 percent LTV newly originated, 30-year fixed rate mortgages.
 Page 80       PREV PAGE       TOP OF DOC

    What would be the implication for capital requirements the model would produce. And that is where this number came up that the Chairman talked about; but as we pointed out, that is not really the real marginal capital requirement because of all these other factors that have to be taken into account. That is only the capital requirement that would be associated with the first year of that mortgage.

    As mortgages season, they become less risky, therefore, the capital requirement would be reduced over time. That calculation also is dependent on the level of G fees we assume they charge on these 95 LTVs.

    What I am telling you is we assumed a much lower G fee than what they would normally charge for such a mortgage. If they were charged a higher G fee, there would be less net losses on those types of mortgages, and they would therefore be less risky, so less capital would be required. So there are various other factors that would have to be taken into account if you wanted to actually know what it would cost to add on an X dollar amount of 95 LTV mortgages.

    Mr. BENTSEN. This may be oversimplistic. But if you used just a set guarantee fee in your model, at the end of the day aren't you still looking at total capital and total assets; and, ultimately, that is going to flow to the bottom line so you don't necessarily have to have an actual guarantee fee.

    Mr. KINSEY. You do when you are trying to do a marginal analysis like this because we are specifically saying what happens if we increase the type of business that is really at the high end of the risk spectrum and what would be the capital requirement associated with just an increase in that type of activity.
 Page 81       PREV PAGE       TOP OF DOC

    To get a more accurate measure of what the cost would be you need to associate with that what they would actually charge for that type of business. I am telling you that it would be much higher than what we assumed in the proposal.

    Mr. BENTSEN. Thank you, Mr. Chairman.

    Chairman BAKER. Thank you, Mr. Bentsen.

    You gave about one more follow-up, if I can put my hands on it. Mr. Bentsen got me sidetracked looking at his 3 to 6 percent question.

    The issue of how the stress test would affect potentially housing finance agencies, that the haircuts that you would provide for municipal-revenue bond portfolio investments may result in haircuts that are too steep and would impact potential for supporting affordable housing through housing finance agencies.

    As I understand it, housing finance agency bond issues are publicly rated. That rating is all too often based on not just the characteristics of the underlying portfolio, but you look at the strength of the finance agency itself. For example, some States put the full faith and credit of the Government behind that offering. If you were to deviate from the structure you have proposed and began looking at the risk characteristics of the underlying loans as opposed to the ratings of the agencies, we would likely get a much more adverse treatment than we get under the proposed rules.

 Page 82       PREV PAGE       TOP OF DOC
    Mr. KINSEY. I think that is correct.

    Chairman BAKER. I just want to make sure that we are not unintentionally doing something here that detracts from availability of public housing—from financing housing and I think that point needed to be made.

    Mr. KINSEY. We are trying to treat all counterparty risk consistently regardless of who the counterparty is. The tool with which we are using to try to get consistency is the public rating.

    Chairman BAKER. If something is not rated, then you treat it as a BBB.

    Mr. KINSEY. That is a BBB, yes.

    Chairman BAKER. Except in the case of the derivative where there might be specific collateral to offset the risk.

    Mr. BENTSEN. Would the Chairman yield on that?

    Chairman BAKER. Be happy to.

    Mr. BENTSEN. Is your point that Fannie Mae in particular—I don't know that Freddie Mac has done this where they have gone in and done a private placement with a local or State housing finance agency on the tax exempt side and you are saying that the—depending on the originators in the—who are part of the State or the local HFA's offering and/or the mortgage insurance that is used would raise the cost or you would require—you could—your model would require a larger haircut on that basis, which I guess your point is that might raise the cost of—the interest rate cost and the mortgage and make it less attractive.
 Page 83       PREV PAGE       TOP OF DOC

    Mr. KINSEY. This was an attempt to try to simplify our approach to counterparty risk. As the Chairman points out, we could have looked at the underlying collateral of the mortgages potentially. That would have made this a much more complex rule than it already is.

    Chairman BAKER. That is hard to believe.

    Mr. BENTSEN. Thank you.

    Chairman BAKER. Thank you, Mr. Bentsen.

    Just one closing comment. I know that this has taken some while to promulgate. It is technically confusing and may be difficult for the affected entities to adequately absorb considering the problem with the difficulty of recreating the model within their own agency that replicates your method of analysis.

    Having recognized that, I am certainly hopeful that you will find a mechanism to get closure on this at the earliest possible date. I recall meetings in prior years when I was perhaps a little upset with the agency itself and not having moved further in the time available to it to reach final promulgation.

    I would hate to think that the date now held out would be extended another year knowing that we have another year following that before the rule would become official. To that end, I will encourage you to stick as best you can given your regulatory responsibility and sense of appropriate fairness to the involved parties—that whatever extensions may be considered be minimal and to the best of our ability move forward in a timely basis.
 Page 84       PREV PAGE       TOP OF DOC

    I do want to express my appreciation to you. I think your appearance here today was very helpful to the subcommittee. I know it is a rather boring subject to most people, but this is a very vitally important financial issue, frankly, to the solvency of our financial system. And the work you are about, I think, is extremely important to the public's best interest.

    I want to thank you for that.

    Mr. KINSEY. Thank you, Mr. Chairman. I agree with you excessive delay is not in the public interest on this rule.

    Chairman BAKER. Thank you very much. Our hearing stands adjourned.

    [Whereupon, at 12:38 p.m., the hearing was adjourned.]