In an exclusive interview, Federal Housing Finance Agency Acting Director DeMarco talks about the ongoing policy discussions to restructure the nation’s housing finance system. He also cites some of the activities that led to billions of dollars of losses at Fannie and Freddie.
Mortgage Banking
May 2010
By Robert Stowe England
Last August, President Obama designated Edward J. DeMarco acting director of the Federal Housing Finance Agency (FHFA)--the regulator of Fannie Mae, Freddie Mac and the 12 Federal Home Loan Banks.
Prior to August, DeMarco served as chief operating officer and senior deputy director for housing mission and goals for FHFA shortly after its creation as a government agency. DeMarco joined the predecessor agency, the Office of Federal Housing Enterprise Oversight (OFHEO) in 2006, where he served as chief operating officer and deputy director.
Before coming to OFHEO, DeMarco worked at the Social Security Administration (SSA), where, as assistant deputy commissioner for policy, he led SSA’s policy, research and statistics functions. DeMarco previously served as director of the Office of Financial Institutions Policy at the Treasury Department, where he oversaw analyses of public policy issues involving government-sponsored enterprises (GSEs) and other financial institutions. Prior to his 10-year tenure at Treasury, he worked at the U.S. General Accounting Office for seven years.
DeMarco received a doctorate in economics from the University of Maryland and a bachelor of arts degree in economics from the University of Notre Dame, Notre Dame, Indiana. (ital) Mortgage Banking (end) caught up with DeMarco at his FHFA offices in March.
Q: It’s interesting that in the new proposal for enterprise affordable housing goals published in the Federal Register in February, the overall affordability goals for Fannie Mae and Freddie Mac lending are gone and now FHFA is proposing what are called more-specific goals as a share of designated identified targets within the population.
A: Those changes were brought about by the Housing and Economic Recovery Act of 2008, or HERA.
Q: Does HERA require specific goals or did it just give you the authority to change the goals?
A: HERA requires FHFA to set a number of goals. Within the single-family market there are three purchase market goals and one refinancing mortgage goal. In addition, FHFA has authority to set both an affordable-housing goal and a sub-goal for the multifamily market.
FHFA is proposing benchmarks goals for 2010 and 2011; however, the enterprises will also be measured against actual market share each category represents in originations for those years. For a given enterprise, if originations in each category are at least as high as the actual shares of the origination market represented by each category, then this will be judged as having met the enterprise’s goals for each category.
Q: As I understand it, for the proposed single-family affordable-housing category, the purchase money benchmarks are as follows: 27 percent of acquisitions from low-income families; 8 percent of acquisitions from very-low-income families; and 13 percent of acquisitions from low-income areas. For single-family refinance, the low-income goal is 25 percent. Plus, there are the multifamily goals. What is the advantage of going from a general goal to specific goals?
A: It allows for closer measurement, separately, of the enterprise’s activity in the single-family purchase money market. It allows for separate measurement of the enterprise’s activity in the single-family refinance market. And it allows for individual attention to the enterprise’s activity in the multifamily market.
Breaking it out that way--it allows the goals and measurements and assessment of enterprises’ performance to be broken out into three areas to measure how an enterprise is doing with respect to supporting that segment of the market. Previously, HUD [the Department of Housing and Urban Development] was adding both single-family and multifamily lending activity [for an overall lending goal]. So, it’s a different approach to it.
The other change that was made with respect to defining low income and very low income was to align those definitions [for measuring affordable-housing lending performance by Fannie Mae and Freddie Mac] with the definitions that are used in the Community Reinvestment Act, or CRA, to which many mortgage originators are subject. So, the changes there provide greater alignment between CRA requirements and the housing goals of the enterprises.
Q: And the data for measuring the performance of the enterprises--is that the data collected under the Home Mortgage Disclosure Act of 1975, the so-called HMDA data?
A: Right. The market is typically measured using HMDA data.
Q: Now, I noticed that the specific goal of 27 percent for the category of low-income single-family home purchase mortgages is slightly higher than the range from 2004 to 2008, which was between 23.5 percent and 27.1 percent. Was that the idea?
A: The idea in setting the benchmark goals, or the prospective goals, was to come up with a range of what the expectations for that market looks like in the coming two years and then take a midpoint or median of range. There’s a detailed technical appendix that walks through the econometric methodology.
Q: This appendix is in the notice that was published in the (ital) Federal Register? (end)
A: It was not included in the (ital) Federal Register, (end) but it is available on FHFA’s Web site. The appendix goes through the methodology. But the idea, in simple terms, is to try to get a reasonably strong estimate of the range of what the low-income single-family purchase money market is going to look like as a share of mortgage activity and to set the goal in that range.
Q: So, the target for low-income single-family purchase money mortgages is slightly higher because the expected range is higher than recent experience--and this higher target is in the median range of expected originations?
A: I would say that one of the difficulties--and this is about looking at that over time--is that the share of mortgage activity is going to directly reflect those ratios. If you look at 2004, where there was a lot of refinance activity, [you’ll see that] the share of purchase money market activity is going to be relatively less. So, part of the estimating process that went into giving out the benchmark goal was to incorporate estimates of that kind of activity. But HERA, recognizing this, as I said, directed us to come up with a regime in which purchase money mortgages and refinance mortgages are treated separately.
Q: So, the very low income purchase money goal is 8 percent--which, again, is higher than the range from 2004 to 2008, which runs from 5.8 percent to 7.1 percent?
A: The goals that were set are not necessarily pushing the GSEs to do significantly more in this area, but reflect what they have been doing, with the assumption that this is good to operate within that range or close to it. It is to take the middle of a range of estimates of where the market might be in 2010 and 2011 and go to the middle of that range.
Q: Could you briefly describe the methodology for setting the prospective benchmarks?
A: FHFA has developed an econometric model for forecasting the shares of mortgages in the primary mortgage market that would qualify for the single-family housing goals. Specifically, for the two income-based home purchase goals, the model is based on housing affordability (as measured by the National Association of Realtors®), the unemployment rate and the steepness of the yield curve.
For the low-income areas’ home purchase goal, the model is based on changes in mortgage interest rates, FHA’s [the Federal Housing Administration’s] share of the mortgage market and the ratio of housing starts to home sales. For the low-income refinance goal, the model is based on housing affordability, short-term interest rates, the steepness of the yield curve and the projected refinance share of all originations.
Q: One of the reasons I ask about how these benchmarks are chosen is that in all of the hand-wringing of where the GSEs went wrong, a common theme is that they took on a lot of bad mortgages trying to achieve high affordable-lending goals, which were set above 50 percent of mortgage activity for many years. Also, the GSEs were also purchasing subprime private-label mortgage securities and alternative-A loans, and all that was thrown into the mix of the overall percentage.
A: It’s a complicated story, and you need to be a little bit careful with what you include in affordable lending. Not all of the examples you gave might have qualified for full goals credit. You take an alt-A loan or low-doc or no-doc loan, and there is no documentation of the borrower’s income--then you can’t necessarily get full goals credit for any of the goals that are income-based. HUD, and now FHFA regulations, do allow for an estimation of income-based goals credit based on HMDA data when there is missing income documentation. There are caps to this type of credit to encourage the collection of income data.
Q: I see.
A: Because you don’t know what the borrower’s income is, an estimation process occurs. So, they certainly acquired alt-A, low-doc, no-doc loans, but there were other motivating factors for that. You might also want to take a look at a recent mortgage market note 10-2 we released [on] Feb. 1. It reviews the housing goals of Fannie Mae and Freddie Mac in the context of the mortgage market from 1996 to 2009.
That will give you data on what the past goals were, what the enterprises’ actual performance has been and what the market shares actually looked like over a period of time. That was meant to be background information for analyzing the proposed specific goals.
The one thing you have to keep in mind when you look at it is that you can’t compare it to the new goals regime because the definition of low and moderate income was different from the definition in the proposed rule, and the old measurement was counting both single-family and multifamily.
Q: It differs in that the definition of categories in the old regime were broader and included more people?
A: Yes, that’s right. In the past, the overall affordable goal in single-family was for incomes up to 100 percent of the area median, while in the proposed rule for low-income purchase money mortgages, it includes families with incomes no greater than 80 percent of the area median.
Q: So now, with these goals, you believe that the FHFA is working to ensure the safety and soundness of the enterprises as well as meeting affordable-housing goals, and doing so in a way that serves both purposes?
A: That’s right.
Q: And the idea of ensuring the safety and soundness of the enterprises went into calculating these numbers, even though the benchmarks are higher than recent experience?
A: The benchmarks are based on our projection on what the market will likely produce in 2010 and 2011. So, what the market has been doing is an input into those projections. But the benchmark goals of 27 percent in 2010, for example, or the 8 percent for very low income, result from a projection of what the market share is likely to be. Now, there are two ways to meet the goals. That’s another key innovation.
Because we’re basing this off a projection of what the market will look like, but that projection could be off. It could be higher or lower. And so the other way you reach the goal is to have purchase activity that actually equals or exceeds what the market actually produced. So, if purchase money mortgages to low-income families ends up being 24 percent of the market in 2010, and the share of an enterprise’s purchase money mortgages is 24 percent, then [it satisfies] the goal established in the proposed rule.
Q: OK. So, that gives them flexibility. It allows them to hew the line on safety and soundness by not over-representing that segment?
A: A key part of it, since none of us has perfect insight on what the market will actually be, is that it allows that kind of flexibility. I would note that when Congress made changes to the housing goals regime in HERA, Congress itself recognized this conundrum, if you will, with the way it had been done in the past. Congress specifically authorized the FHFA to make changes to the goals if it looked like the market was producing something other than expected. And so we basically took that concept Congress had in the law and said here’s how we can do it, so we basically incorporate it from the start.
Q: Going to another topic, the Treasury announced on Christmas Eve it was removing the caps on the amount of money it would commit to keeping the enterprises solvent. At this point the potential funding of the enterprises is theoretically unlimited. Was the FHFA’s input sought into that decision? If so, what was the input?
A: Since the establishment of the conservatorships, what was put into place was a senior preferred stock purchase agreement with the Treasury Department as the vehicle for providing direct financial support to each enterprise to make sure that its net worth remains at or above zero. With the establishment of that agreement, ever since [then] there has been ongoing discussion between FHFA and Treasury. We have to notify Treasury every quarter about requesting draws on that agreement. And so we certainly have regular communication with Treasury. You would have to go to Treasury to find out why exactly this approach [was chosen] and what the goal was.
I can tell you that we were in communication with Treasury about the prospects for, and possible timing of, future losses and what potential draws might be. But, as my letter to Congress [in early February] also points out, from the beginning, we have had a key goal of providing assurance to investors as to soundness of an investment in enterprise debt and mortgage-backed securities [MBS].
And so this senior preferred agreement has had the dollar amount of the backstop adjusted twice since it was first put in place. And each time, it’s been about providing added assurance and confidence to capital market investors so that they continue to see enterprise MBS and debt as a sound investment. In that way, we’re making sure that the enterprises continue to be a stable source of funding for housing in the future.
Q: The next question is broader, and it’s a question a lot of people are asking. What is the exit strategy? FHFA has been focused--and rightly so--on keeping the housing finance market alive and functioning so that the housing market could recover. We’ve obviously made a lot of progress it would seem, in that direction. Where are we going with the enterprises? Do we want to restore them back to health and make them GSEs or is it still just an open question?
A: It’s an open question. I would point to a couple of things. When Congress created FHFA and gave us authority to appoint a conservator or receiver for the enterprises, in the statutory authority given to us, there was only one end game that was made available to FHFA--and that is to reconstitute a failed enterprise under the current charter that they have. So, while there are lots of options that are being discussed and there are lots of options available to Congress, the only one that is authorized in current law for FHFA to do would be to reconstitute Fannie or Freddie using the Fannie or Freddie charter as it exists today.
I think it’s pretty clear everyone wants to take a broader look at the options. I talked about some of the ways we might look at the future of the enterprises in my October testimony to Congress. But I would point you to others as well. The Government Accountability Office [GAO] has put out a report outlining an array of possible structures and options. [The report is titled (ital) Fannie Mae and Freddie Mac: An Analysis of Options for Revising the Housing Enterprises’ Long-Term Structures, (end) published Oct. 8, 2009.]
Various industry groups have come forward with specific proposals. The administration has said it is working on its thoughts here, too. And [House Financial Services Committee] Chairman [Barney] Frank has said that he is looking for a rethink and restructuring. So, I think a lot of participants in the policy discussions are seeing that this is more than how do we reconstitute Fannie and Freddie. It is an opportunity to reconsider the structure and public policy goals of the nation’s housing finance system. And I believe that a lot of the discussion about what happens to Fannie and Freddie post-conservatorship will take place in that larger context.
Q: Of course, you mentioned Chairman Barney Frank, and he has famously said that Fannie and Freddie should be eliminated.
A: I don’t have the transcript in front of me. He also said something to the effect that they should be eliminated in their current form.
Q: Yes, I remember that.
A: Which is consistent with the notion that there are a variety of forms and structures that are out there. And the business platforms and human capital resources of Fannie and Freddie could be transformed into one of these other forms, structures or charters in the future.
Q: The common view today is that the vast majority of the loans that go through loan modification are going to end up in foreclosure anyhow. But, of course, I guess the effort has to be made. What are your thoughts about what the GSEs are doing or should be doing?
A: The analysis of the re-default risk on modified mortgages is frequently made by looking at loan mods that were done in 2007 and 2008, where the loan modification actually resulted in an increase in the monthly payment [for] the borrower. One of the key characteristics of the HAMP [Home Affordable Modification Program] program is that these are loan modifications that are resulting in decreases in the monthly mortgage payment. And so I think it’s questionable to apply the re-default data from an earlier loan-modification regime to this new regime in which the new monthly payment is targeted at a certain affordability target and in which the monthly mortgage payment for a typical family is going down, not up.
Q: Treasury is reporting that by the end of February, about 1 million households had benefited from lower mortgage payments under HAMP. Of these 1 million, only 168,703 had permanent loan modifications while 835,194 were still in the trial stage. The goal in the modifications was to get the payment down to a 31 percent debt-to-income ratio, which is a more sustainable arrangement.
A: That is certainly the hope and the goal. How it actually works still remains to be seen. But we are certainly pursuing this as a lower-cost alternative to foreclosure, as a fair opportunity to a family that wants to stay in its house but does not currently have a mortgage that they can sustain, to try to recast that mortgage to a sustainable payment. So, I think it’s very fair to offer loan mods to borrowers that are in troubled mortgages. It represents a real opportunity to reduce foreclosures and help families keep their homes. And, very important, from my standpoint, it is an opportunity for the investors in these mortgages to reduce their losses relative to a situation where the delinquent mortgage went through foreclosure.
A critical goal of conservatorship is to minimize the credit losses on all these delinquent mortgages that Fannie and Freddie own. So, we are starting with the HAMP program. This loan-modification effort is a first step, but that is not the only tool in the toolbox. And if the HAMP modification is not going to work for a particular troubled borrower, we’re looking at a range of other loss-modification activities short of foreclosure such that the goal is to reduce the loss to Fannie or Freddie relative to the losses that result from foreclosure.
Q: Some of these, could I assume, could include principal reduction as well as short sales, which is also a principal reduction, which can reduce the amount of loss?
A: That can be the case. I would point you to our monthly foreclosure and refinance report, in which we provide data on a range of loss-mitigation activities for Fannie and Freddie and that provides data on the volume of short sales. That’s on our Web site under agency reports.
Q: So, principal reduction is already indeed a part of the effort?
A: The HAMP program currently uses principal forbearance as one of the tools to get to the 31 percent debt to income What that means is that the principal is not written off, but there is basically zero interest charged against the principal.
Q: Critics have said that the GSEs are becoming more subject to political pressure than ever. In your view, have the GSEs become more subject to political pressure since FHFA has been established? Or is this just merely an issue of perception? Are the new affordable-housing goals the result of political pressure?
A: I would say for this question, I think one of the key goals of my letter to the Hill on Feb. 2 was to report on FHFA’s activities and to provide some context around how we were managing the conservatorship, given the state of Treasury’s support to the senior preferred and to provide this report to Congress in a way that I was very specifically responding to this sort of question. I would take the bulk of that letter as being my response to that question that some critics may make that the enterprises are being operated not for safety and soundness but to achieve political ends and not to achieve other ends. I’m trying to provide a clear statement of the goals of conservatorship and the activities that we are ensuring the enterprises undertake in conservatorship to satisfy those goals.
I really think the bulk of the letter gets at those things. It talks about the change in the amount of coverage under the senior preferred. It talks about loss mitigation. It talks about, for instance, where I say that I will not proceed with entertaining ideas to get into new products and new lines of business, and that is consistent with the goals of conservatorship. Again, with the affordable-housing goals, I state one of the goals of conservatorship is to make sure that the companies continue to operate to serve their core statutory purposes. In fact, the housing goals serve a useful benchmark to see whether that goal of conservatorship is being satisfied or not.
Q: I noticed that in their efforts to meet your affordable-housing goals, the enterprises can not buy private-label mortgage-backed securities.
A: Remember, this is a proposed rule. We have proposed that investment acquisition of private-label securities not be counted toward meeting the goals.
Q: The purchase of private-label MBS is seen by many as part of the cause of the problems that brought down Fannie and Freddie and required them to be rescued.
A: They certainly suffered a good deal of losses from [private-label securities they purchased].
Q: Losses at Fannie and Freddie are of great public concern. Where are we now in terms of total losses to date, and how high do you expect these losses to get?
A: Let me tell you where we are now. Through the fourth quarter of 2009, Fannie and Freddie together have reported net losses of $93.5 billion for the 12 months ending Dec. 31, 2009.
Q: Do you have any idea what the total losses will be before Fannie and Freddie return to profitability?
A: As a regulator, these are the sort of analyses we regularly do, but that is not something we are going to make public.
Q: You raised several questions about the future of housing finance in your October testimony to Congress. For example, you asked what role in the future should the federal government have in ensuring the mortgage market has adequate sources of liquidity; in absorbing credit risk; and in promoting the availability of credit? I was wondering what response you have received to the questions you raised?
A: I would like to think a few of the questions are part of the policy discussion that is going on now in Congress and within the administration. It was meant to provide some thoughts about how the overall public policy goals of our housing finance system might help guide the discussion about the actual mortgage structures and regulations. I do think both Congress and the administration are thinking broadly about the housing finance system and the institutional arrangements that serve it, and I’m looking forward to a thoughtful public discussion of this in the coming months. MB
bio: Robert Stowe England is a freelance writer based in Arlington, Virginia. He can be reached at
