"Legislative Proposals On GSE Reform"

Introductory Comments

Good morning Chairman Frank, Ranking Member Bachus, and distinguished members of the Committee. I want to thank you for the opportunity to speak today about the Federal Housing Finance Reform Act of 2007 (H.R. 1427). This is important regulatory reform legislation that HUD believes is needed to strengthen the Federal government's oversight of the housing government sponsored enterprises: Fannie Mae, Freddie Mac, and the Federal Home Loan Banks. The legislation introduced last week improves the oversight of the GSEs by creating a regulator on par with the existing financial regulators. It is a regulator with the necessary tools to monitor safety and soundness, as well as the mission of the GSEs.

HUD endorses establishing a new regulator for all three of the housing GSEs that would combine safety and soundness authority with oversight of their respective housing missions, and we urge Congress to act now. This would include moving certain responsibilities, such as new program review, from HUD to the new regulator.

HUD and its predecessor agencies have played a key role in implementing Congress' long-standing housing policies, both through its Federal Housing Administration (FHA) and Government National Mortgage Association (Ginnie Mae) programs and as a regulator of Fannie Mae and Freddie Mac. Because of HUD's history in regulating Fannie Mae and Freddie Mac, as well as its role in ensuring that the Nation's affordable housing needs are addressed through both public and private initiatives, we are especially interested in ensuring that new legislation continues to promote important affordable housing objectives by building upon the progress that has been made over the past decade.

A Decade of Improvement in Affordable Housing Lending

The last ten years have been years of increased affordable lending for low-income and minority families in the conventional mortgage market. Lenders introduced and marketed special lending programs to low-income families and their neighborhoods, revamped their underwriting standards, to deal with the special circumstances of these families, and attempted to prevent mortgage defaults and keep new homeowners in their homes by offering homebuyers counseling programs and developing innovative foreclosure prevention programs. Others operating in the conventional markets, including Fannie Mae and Freddie Mac, also played important roles in what some have termed a revolution in affordable lending.

Home Mortgage Disclosure Act (HMDA) data show substantial growth in conventional lending to low-income and minority borrowers. This suggests that these new affordable lending initiatives have had a measurable impact. Most observers generally agree that in addition to low interest rates and economic expansion, enhanced regulation of depositories' Community Reinvestment Act (CRA) obligations and HUD's affordable housing goals also contributed to a renewed emphasis on low-income and minority lending in the conventional markets.

President Bush has made expanding homeownership opportunities, especially for first time homebuyers and minority homebuyers, a priority goal for his Administration. Fannie Mae and Freddie Mac are well positioned to have played a significant role in funding mortgages that serve these potential homeowners. Both are substantial market forces in the conventional markets, and originators often rely upon the GSEs' mortgage products and underwriting standards. The significant role of the GSEs in the conventional mortgage market means that they can, and should, continue to play a key role in funding mortgages for very low, low- and moderate-income homebuyers and those families which are historically underserved by the mortgage markets.

Today, we are talking about reforming the manner in which the GSEs will be regulated. An important part of that discussion concerns how the Government will continue to measure Fannie Mae's and Freddie Mac's performance in meeting the affordable housing objectives that will be set for them going forward. In a broader context, two other issues, calculating the conforming loan limits, especially in high cost areas, and managing an Affordable Housing Fund, also impact affordable housing, so I will also address those proposals today as well.

For HUD, the affordable housing goals have been a key focus of our regulatory oversight work, so I will begin with our experiences in this area.

HUD's Current Affordable Housing Goals

In 1992, Congress expressed concern about the GSEs' funding of affordable loans for low-income families, particularly those living in inner-city neighborhoods that had been "redlined" by primary market lenders. For this reason, Congress called for HUD to establish three annual housing goals that the GSEs must meet. These are the:

Low-and Moderate-Income Housing Goal, which targets families earning no more than area median income (AMI);

Special Affordable Housing Goal, which targets very-low income families (to 60% AMI) and low-income families (to 80% AMI) living in low-income areas; and the

Underserved Areas Goal, which targets low income and high minority neighborhoods.

The main objective of the housing goals is to encourage Fannie Mae and Freddie Mac to introduce new affordable lending programs and to make prudent adjustments in their mortgage purchase standards that take into account the special circumstances of low-income families and others who have found it difficult to access credit in the conventional mortgage market.

In carrying out its responsibilities to set, monitor, and enforce the housing goals, HUD established progressively higher goal levels by regulation in 1995, 2000, and again in 2004. These goal levels were based on HUD's estimates of how much the overall conventional conforming market, including single family and multifamily housing, and both purchase and refinance loans, would originate in qualifying mortgages over the time period covered by each regulation.

Current goal levels, which HUD established in its 2004 Rule, rise year-by-year beginning in 2005 and extending through 2008. HUD set the goals at levels that will move Fannie Mae and Freddie Mac up to market levels. For instance, HUD projected that, under normal circumstances, special affordable housing will account for about 27 percent of all single-family and multifamily housing financed in the conventional conforming market. Therefore, the 2008 goal is set at 27 percent so that the GSEs will "match the market" by 2008.

In its 2004 rule HUD also established three home purchase subgoals under each housing goal. These subgoals measure the GSEs' acquisition of qualifying owner-occupied home purchase mortgages in metropolitan areas. The subgoal levels were set about 2-3 percentage points above HUD's estimates of the market because it was HUD's intent that the GSEs should lead, rather than match, the market. HUD concluded that the GSEs have the expertise, resources, and ability to lead the single-family owner home purchase market, which is their core business.

Over the past few years, HUD's analysis of the GSEs' performance shows that Fannie Mae led the home purchase market from 2002 to 2005 on its special affordable housing business and from 2001 to 2005 on purchases serving low- and moderate-income families. In fact, on the two income-based goals, Fannie Mae sharply improved its performance beginning in 1999, thereby erasing its gap with the market. In the underserved areas category, Fannie Mae also sharply improved its performance beginning in 1999, exceeding market levels by 2002. However, Fannie Mae lagged the underserved areas market during 2004 and 2005.

Freddie Mac also improved its performance on the three home purchase subgoal categories, particularly between 2004 and 2005. The year 2005 was the first time that Freddie Mac led the market on the two income-based goals. However, like Fannie Mae, Freddie Mac lagged the market in 2005 in the underserved areas category.

The Proposed Affordable Housing Goals

With respect to affordable housing goal provisions of the H.R. 1427, HUD is pleased to see that Congress continues to be concerned with the role that Fannie Mae and Freddie Mac play in serving affordable housing markets and has retained the housing goal structure as a means of measuring their performance. We were also glad to see some improvements over the current statute. For example, establishing an 80 percent AMI ceiling for defining underserved census tracts is a positive enhancement as is providing monetary penalties for a GSE's failure to achieve a housing goal.

With respect to the proposed goals, however, we would like to offer some suggestions and "food for thought" based upon our own experience in analyzing affordable housing markets and the GSEs' role in serving these markets over a considerable period of time.

One of the first differences we noticed was the discontinuation of what are now the Low-and Moderate-Income Housing Goal and the Underserved Areas Goal. It seems that the new single-family home purchase goals are mostly derived from HUD's current Special Affordable Home Purchase Subgoal.

In HUD's regulations, the Special Affordable Home Purchase Subgoal is defined as home purchase mortgages financing very-low income borrowers (to 60-percent AMI) and low-income borrowers (to 80-percent AMI) living in low-income areas. Similarly, the new single-family goals are Very Low-Income (to 50-percent AMI), Low Income (to 80-percent AMI) and Low-Income Areas (to 80-percent AMI).

However, HUD also has a broader-based Special Affordable Housing Goal, defined in the same way as its home purchase subgoal, but including mortgages that finance both purchase and refinance transactions, and both single-family and multifamily mortgages. This broader definition ensures that affordable housing markets are served in the context of the types of housing the marketplace is financing in any given year. In other words, because goal levels are based on estimated markets that include all types of residential mortgages, the goal is not especially vulnerable to unforeseen trends or events in the marketplace. It is also large enough to warrant special programs, products and services from the GSEs that they would be less inclined to initiate for especially small goal levels.

In that regard, I would like to describe some concerns HUD has identified with the proposed single-family goals, including why we think some of them may not achieve the desired result.

Size of the Single-Family Home Purchase Markets

With respect to the Very Low-Income Goal, the conventional market for very low-income home purchase mortgages is so small, it would be unlikely to achieve the objective of parity with the Community Reinvestment Act (CRA). Very small markets also create a higher degree of uncertainty from year-to-year in setting the appropriate goal level.

To understand the small magnitudes of mortgage involved, consider these recent figures:

As shown, the percentages are quite low, in the single digits, and in 2005, both GSEs already exceeded the very-low income home purchase market.

HUD's Special Affordable Housing Goal overcomes this "small percentage" problem by raising the affordability category to 60-percent AMI. This extra 10 percent of coverage includes almost as many borrowers as the entire less-than-50-percent AMI group. If H.R. 1427 defined Very Low-Income as 60-percent AMI, then the home purchase goal would be in the double digits, ranging from about 12-13.5 percent. By creating an affordable goal of sufficient size relative to the market, there is greater likelihood that the GSEs could serve that market through targeted products and outreach.

HUD recommends that the Committee consider imposing either a single home purchase goal with incomes to 80 percent of area median income, which would increase market share to about 30 percent, OR retain the two income-based home purchase goals but define them somewhat more broadly at 0-60 percent AMI and 0-90 percent AMI. Goals that include a range of incomes offer sufficient market share for the GSEs to invest in the development of products targeted to lower-income borrowers and to maintain a consistent presence in those markets regardless of market trends over time. Goals set at very low ranges of affordability, such as to 50 percent of AMI, would encourage the GSEs to purchase private label securities, mortgage revenue bonds, and similar instruments that are already financed adequately by the capital markets.

FHA's Role

On another note, I think it is important for the Congress to consider FHA's role in these markets. Homebuyers at the very low-income level pose greater risks to private lenders and the capital markets. For this reason, FHA can provide the needed entrée too mortgage financing for these borrowers. For example, 14 percent of FHA-insured loans in metropolitan areas are for very-low income borrowers, compared with only 6 -7 percent of the conventional conforming market.

In fact, FHA's mission is to provide financing for those borrowers who are at the margins of homeownership. FHA serves an important function in the overall continuum of financing that is available in the marketplace. HUD is currently working with the Congress and will shortly submit legislation to modernize FHA so that it can continue to fulfill its mission of providing safe, accessible, and efficient financing to those who are most underserved by the mortgage markets. I would urge the Congress to consider the FHA Modernization proposal this spring.

New Goals Do Not Count Single-Family Rental Units

HUD is concerned that because H.R. 1427 does not include overall housing goals, which would include single-family rental mortgages, there would be less incentive for the GSEs to purchase "goals rich" rental mortgages.

The housing goals established in H.R. 1427 would not require the GSEs to play any particular role in the single-family rental mortgage market, a very important source of affordable housing. Specifically, American Housing Survey data indicate that in 2005, 18.4 million single-family rental units (i.e., rental units in 1-4 unit properties) accounted for 54 percent of all occupied rental units, and 45 percent of these units were occupied by households with incomes less than 50-percent AMI.

Thus, even though H.R. 1427 includes a Multifamily Goal, there would be no incentives for the GSEs to purchase single-family rental mortgages, an important market segment for lower income families and families in underserved areas. Final legislation should encourage the GSEs to continue to grow their single-family rental business. This is unlikely to occur if that market segment is removed from goals consideration.

Separate Multifamily Goals Would Be Difficult to Establish

With respect to multifamily markets, H.R. 1427 calls for at least three multifamily housing goals. The small market coverage of one, dealing with low-income housing tax credits, which HUD estimates to be about 7 percent of the multifamily dwelling units financed in 2003, makes it inappropriate for a separate goal. Again, it is a "small markets" issue, and the GSEs would likely meet any such goal by purchasing state-issued bonds that finance such units.

The other two multifamily goals (very low-income and low-income) have ample affordability coverage, but data are not readily available for these market segments. That is also true of the small 5-50 unit subgoal. In the past, HUD has had to piece together estimates of the multifamily market from different sources, recognizing that there was uncertainty with any estimate, particularly with respect to the number of dwelling units financed. Separate multifamily goals raise serious issues about the uncertainty of the data, making it less likely that meaningful and challenging goal targets will be possible.

Under the current goals structure, which is based on targets that include the overall market for both single-family and multifamily mortgages, HUD finds that the GSEs have to be active in the multifamily market in order to meet the targets that HUD has established. This is because the majority of multifamily dwelling units are affordable to families with incomes equal to or less than 60-percent of AMI. This compares with about 13-15 percent of single-family owner units. For this reason, the GSEs have every incentive to purchase multifamily mortgages in order to meet an overall goal target.

H.R. 1427 Eliminates Overall Performance Goals for the GSEs

H.R. 1427 does not include overall standards for evaluating the GSEs' performance in serving lower-income families and their neighborhoods. This would represent a significant reduction in legislative and regulatory scrutiny of the GSEs' affordable lending efforts, particularly since the housing goals have shown that they are effective tools for moving the GSEs from sub-par to market performance across all their books of business.

HUD's current goals include all mortgages financed by the GSEs. That is, each goal is based on market share that includes single-family and multifamily mortgage financing, as well as both home purchase and refinance mortgages. H.R. 1427 excludes GSE purchases of mortgages on single-family rental housing, a very important source of affordable housing, and fragments the other markets into separate categories. The reasons for doing this are unclear, especially in light of the substantial improvements that the GSEs have made in serving targeted markets:

For example,

• For the Low-and Moderate-Income Goal, which targets families with incomes below area median income, Fannie Mae's performance improved by 10 percentage points between 1996 and 2005. Freddie Mac's performance improved by nearly 13 percentage points.
  
  
• For the Underserved Areas Goal, which targets families in low-income and high-minority areas, Fannie Mae's performance improved by more than 8 percentage points between 1996 and 2005 while Freddie Mac's rose by more than 13 percentage points during the same period.
  
  
• For the Special Affordable Housing Goal, which targets very low-income (0-60-percent AMI) families and low-income families (to 80-percent AMI) in low-income areas, Fannie Mae's performance rose by 11 percentage points in the 1996-2005 period while Freddie Mac's rose by over 10 percentage points during the same period.

HUD's overall housing goals have been a major success in that they have led to a significant decrease in the gap between the GSEs' and the market's performance. For example, in 1996, the special affordable share of the primary market was 27.3 percent. This means that Fannie Mae's 15.4-percent performance on the special affordable goal in that year represented a market gap of 11.9 percentage points. By 2005, Fannie Mae had increased its special affordable performance to 26.3 percent, eliminating much of its gap with the market. And under its 2004 regulation, HUD established goals at levels that will cause the GSEs to match HUD's estimate of the market by 2008. If H.R. 1427 (without its overall housing goals) had been implemented in 1993, the above increases in overall goal performance would not have taken place and the GSEs would not have closed their gaps with the market. Given that the GSEs are now active in all major segments of the affordable market, it is important that their overall performance continue to be actively monitored and enforced.

Overall goals also allow for more challenging goal levels. As the table below indicates, the GSEs already meet or exceed most of the proposed single-family goals based on current originations as reported in HMDA:

Finally, overall goals are consistent with the GSEs' charters, which require them to, among other things, provide stability in the secondary markets for residential mortgages and to provide ongoing assistance to the residential mortgage market, including the affordable markets. This means they must be willing to purchase, within their charter limitations, all single family and multifamily mortgages, including purchase and refinance mortgages, that their selling lenders originate, including those that do not qualify as affordable mortgages. They are required to do this regardless of market trends. Goals that are too narrowly construed - and segmented - could lead to credit allocation at the expense of the other markets that the GSEs must also serve. It could also encourage a pull-back from the gains achieved to-date in closing market gaps across all of the GSEs' books of business.

HUD would like to see overall market based goals reinstated in the new legislation. Otherwise, it is likely that progress made in serving the housing needs of all income levels below area median income will not be sustained.

The "Duty to Serve" Provisions Require Clarification

HUD urges the Congress to, at a minimum, provide the Director with clarity on how the "duty to serve" provisions are to be implemented and managed, including their purposes, and how the Director should measure the GSEs' performance in either serving, or working with, these HUD programs.

Conforming Loan Limits Provisions in High Cost Areas

In addition to the housing goals, there are other proposals of concern to HUD, including the conforming loan limits provision.

H.R. 1427 would increase the conforming loan limit in high-cost areas to the lesser of the median house price in each area or 150 percent of the conforming limit. HUD questions the need for this provision, especially since the markets that are likely to qualify as "high "cost areas" are already well served by private lenders and the capital markets. In fact, we are not aware of any reported mortgage liquidity issues in those metropolitan statistical areas (MSAs) that might be classified as "high cost areas" under this provision.

With respect to affordable housing factors, HUD's preliminary analysis of 12 high-cost MSAs (as identified using the NAR median house price list) found only 1,207 low-income (to 80-percent AMI) loans out of the 153,950 current jumbo home purchase loans that would be added under this proposal. In this regard, the high cost provision seems inconsistent with H.R. 1427's emphasis on affordable housing markets. For example, at 50 percent over the current conforming limit of $417,000, the GSEs would be authorized to finance mortgage loans up to $625,500. HUD finds no reason to authorize this type of expansion to the GSEs' charter authorities, and for this reason, HUD opposes the "high cost areas" provision.

With regard to the conforming loan limit provisions in general, H.R. 1427 permits downward adjustments to the conforming loan limit when the housing price index falls from one year to the next. We think there also needs to be a provision that protects loans originated at the previous, higher limit but not delivered for sale until after the effective date of the new lower limits. This provision should also extend to loans refinancing in years when the current limit is less than the limit in effect at the time the original loan was closed.

The Affordable Housing Fund

While HUD does not advocate for the creation of an affordable housing fund, we believe any such fund should have a cap. HUD does, however, note that important improvements over the original proposals have been made in H.R. 1427. These include having the fund managed by the Director, rather than the GSEs, greater clarity regarding eligible grantee recipients, and a more precise sunset provision.

Conclusion

In conclusion, I would like to commend the Congress and especially its leadership, both present and past, for the extraordinary efforts it has made to the important and critical work of creating a new regulator. This is a difficult undertaking, but a worthwhile one that HUD fully supports. Because the housing government sponsored enterprises are so vitally important to the financial markets, anything less than a world-class regulator fully empowered to oversee their activities is simply not in the public interest.

HUD believes the time is right for a change, and we look forward to working with Congress in this endeavor.