Blog

Senate Draft's Cross Subsidy Approach Misses the Mark

March 01, 2018

A wholesale reconstruction of the country’s mortgage finance system is a rare opportunity to take all of the federal government’s direct and indirect supports for home finance and use them to create a sensible system that protects taxpayers, assures full access to credit, and supports liquidity for the long term, fixed rate mortgages that consumers prefer.  But the draft GSE reform bill that was widely circulated in late January instead perpetuates the current separation of the so-called “conventional” market from the government market supported by FHA, VA and RHS.  The draft could use the government’s various existing forms of mortgage credit insurance as a conscious part of a new secondary market model in order to assure a broad and responsible spectrum of credit risk at the most affordable cost to consumers.  Instead, the system would divert the proceeds of a 10 basis point fee on guaranteed securities from supporting rental housing and community development efforts for very low income households to a new form of cross subsidy that effectively will subsidize private Mis and the new guarantors.

Even with the draft’s proposed subsidy, the draft will leave a significant portion of the market to FHA.  Borrowers with very little cash to put down and moderate credit scores will still find FHA their most economical choice.  The subsidy will pull some share of borrowers into the private system, but it will not do so for them all.  Even with its problems, FHA will remain a major part of the federal system of support for home ownership finance.  The draft’s failure to incorporate it more intentionally into its proposed new structure is a missed opportunity.

There are other concerns with the draft, including the lack of enforceable standards of service to under served communities and borrowers for the new guarantors.  Subsequent versions may change its outlines significantly.  But Congress faces what will hopefully be a once in a generation opportunity to redefine federal support for mortgage finance.  Its goal should be a durable, comprehensive system.  Trying to solve broad challenges of access in the market by isolating a new privately funded system from the government’s other supports will only perpetuate the uncoordinated and unsatisfactory system we have today.  We can and should do better than that.

Private Guarantors and Risk Based Pricing

The draft would rely on newly chartered private guarantors operating with market-priced capital.  It also would require these guarantors to off load much of the credit risk to others, like MIs. In turn, this would drive risk-based fees for their credit insurance and that of their risk sharing partners.  Borrowers with higher risk profiles – typically those with lower down payments and lower credit scores – would be charged progressively higher fees.   The bill would use the 10-basis point fee to offset these risk-based costs for certain borrowers. The draft defines these as borrowers with incomes below 80 percent AMI (or first-time homebuyers below 100 percent AMI).  A recent Urban Institute analysis suggests that this could range from between $929 and $236 in yearly mortgage cost savings, depending on a family’s income.   This effectively would shift the fee’s proceeds from supporting affordable rental housing for families at or below 30 percent of AMI, and community development efforts that support LMI communities, to helping some unknown number of aspiring homebuyers with a modest subsidy.

Such a subsidy might be justified if many consumers would be denied mortgage credit without it, or a ready means to provide lower cost insurance to the same borrowers was not readily available.  But, in reality, FHA is and will remain consumers’ alternative when it can offer a better “sticker price” than privately capitalized credit insurance.  (The same is true of insurance programs offered by the VA and USDA’s Rural Housing Services, but FHA will be the likely alternative for most consumers.)

An expansive and inclusive approach that consciously incorporates existing mission driven government supported mortgage credit insurance alongside that offered by the privately capitalized guarantors and their risk sharing partners like MIs would assure a broad range of credit pricing and availability through the proposed new securitization platform without needing to use the proposed new fee to subsidize market-priced risk.  This might even spur competition from these private insurers, which could expand private capital’s role in serving more borrowers without requiring the proposed subsidy. It also could help strengthen FHA's book by adding borrowers with relatively stronger credit and countering some of the adverse selection it otherwise will suffer.

Instead, the draft would use most of the 10-bps fee to reduce the impact of market driven pricing for some borrowers, and enable Mis and the guarantors to employ full risk based pricing without losing some borrowers to a cheaper FHA execution.  Relying on FHA to provide this cross subsidy in the overall system also would reduce the impact of market pricing without needing to divert the fee.

The 10-basis point fee’s original purpose was to generate a dependable source of funding for desperately needed affordable rental housing production and preservation, and support for community development investments, along with a modest set aside for supporting market expanding activities through a newly structured mortgage finance system.  It was meant to replace a far more modest annual assessment under the 2008 HERA legislation governing Fannie Mae and Freddie Mac, which has generated several hundred million dollars in support for the Housing Trust Fund and Capital Magnet Fund in each of the last few years.  This would still be the highest and best use of the fee.  Diverting it might move the proposed system’s reach a little further in FHA’s direction.  But it would not actually add any new mortgage borrowers to the federal government’s overall coverage.  Drafters could restore some of this funding by increasing the fee.  But this would only add more cost for all consumers in the system while bypassing FHA’s ability to support a cross subsidy for higher risk borrowers.

Alternative Approaches

There are multiple ways the bill’s higher market-driven costs to consumers could be mitigated.  The Mortgage Bankers Association 2017 proposal for multiple private guarantors, which is a partial blueprint for the draft, included utility-like regulation of new guarantors’ returns, controlling  their costs and potentially enabling lower overall charges to consumers.  Private mortgage insurers and guarantors could be regulated to use less loan level risk-based pricing and more broad pool pricing of risk. The new guarantors could be required to accept lower returns for loans serving a defined group of borrowers, as Fannie and Freddie are required today, forcing the cross subsidy to be borne by the guarantors benefitting from the securities guarantee. The drafters could reduce the amount of insurance in front of the government and use the government securities guarantee, which would not rely on loan level risk-based pricing, to cover more of the risk.  They could encourage the development of risk sharing pilots combining FHA and private mortgage insurance as has been suggested by others for years.  It also could include participation by other mission-driven market participants such as state housing finance agencies, Federal Home Loan Banks, and CDFIs like Self Help Venture Fund, which already operates a competitive and successful credit enhancement program with Freddie Mac, to diversify credit insurance offerings and deliver more value to consumers. 

Even if the entire 10 bps in the Senate draft, estimated to be an annual flow of about $5 billion once the new system is up and running, is diverted to cross-subsidizing the private credit insurers, there will still be borrowers for whom the price of a loan under the new system is higher than the price of a loan with FHA insurance.  Indeed, a recent Urban Institute analysis concludes that the draft’s proposed use of the fee is unlikely to reduce FHA’s market share from what it has today.  But it will cost billions in explicit subsidies to produce this result.  Relying more on FHA would expand its market share. But it would help create a broader range of credit pricing across the system, and probably strengthen rather than weaken FHA's own insurance book.  

Government programs like FHA and VA offer alternate, less costly mortgage insurance featuring broad risk pooling and attendant average risk pricing.  They charge lower rates for weaker credit borrowers than privately capitalized insurers will offer because of their explicit government sponsorship and support, and their mission-based approach.  They have a track record over more than 80 years.  The sensible course would be to exploit this valuable resource and rely upon it as an affordable alternative form of primary credit insurance alongside the draft’s totally free market pricing model.  Integrating these approaches would produce a broad policy approach to mortgage credit access that would mix public and private support to reach the best solution for taxpayers and borrowers.  In contrast, the net effect of the proposed cross subsidy in the Senate draft really will be to use the 10-basis point fee to shift market share from affordable government-supported insurance that is already available at a competitive price to private insurance. 

The draft also would give FHFA or its successor new responsibilities to design, execute and monitor other more direct forms of subsidy contemplated in the draft, such as down payment assistance, or consumer counseling support.  But HUD, the VA and the Department of Agriculture’s Rural Development Administration, and state and local governments, already have the capacity and expertise to do this.  If this is such an important outcome, the Housing Trust Fund’s purposes could be enlarged to include specific attention it.  Why add these tasks to yet another federal department that should be laser-focused on the government’s total exposure to mortgage credit risk and the larger mortgage finance system’s stability and liquidity?

FHA Also Has Issues

New legislation need not tackle the myriad administrative problems plaguing FHA for the FHA (as well as VA and RHS) to continue to play a principal role in making mortgage credit affordable.  But there is no doubt that much should be done to tackle FHA’s outdated technology, inadequate quality control and counterparty risk management.  Lenders remain wary of FHA’s reliability.  The loan level certifications lenders must sign, and the risk of False Claim Act prosecutions for errors in loan manufacturing are handicapping FHA’s market role.  HUD’s leadership, lenders, consumer advocates and congressional leaders are pushing for sensible reforms to FHA and they need to continue to do so.

But FHA in the meantime will continue to play a major role in the mortgage finance system.  Not including its ability to offer lower cost mortgage insurance that is average priced and provides access to lower wealth borrowers and those with weaker credit wastes a valuable resource and perpetuates a separation among the government's various direct and indirect supports for mortgage credit.  

The leaked draft's design may not move forward.  I joined with other co-authors to recommend a different approach that did incorporate FHA more fully. But any system that relies on private capital to absorb losses ahead of a government securities guarantee will have to confront the challenge of how to insure broad access for responsible borrowers in the face of private sector return expectations.  Failing to take advantage of FHA in a comprehensive system would miss an important chance to use an existing tool and to more closely integrate the government's approaches to supporting mortgage finance.

Mortgage Finance Reform Panel Dicsussion, CFA Financial Services Conference

December 02, 2017

CFA's 2017 Financial Services Conference featured a panel discussion on Dec 1 with Michael Stegman, Milken Institute; Mark Zandi, Moody's Analytics; Steve O'Connor, Mortgage Bankers Association; and Robert Henson, Credit Union National Association.

Credit Access and Mortgage Finance Reform

May 08, 2017

In late April, 2017 I was invited to participate in a panel on affordability issues in mortgage finance reform at a New York workshop sponsored by the Federal Reserve Banks of New York and Atlanta, the Federal Reserve Board, the Wharton School at University of Pennsylvania, and UCLA's Anderson School of Management. I summarized what I think are the critical pillars of a mortgage finance system, which I’ve edited into the bullets below.  Four colleagues and I expanded on the need for and an approach to affordability in mortgage finance reform in January as part of our series of essays describing A More Promising Road to GSE Reform, which can be downloaded through the link.

Assure that entities that use government support do not “cream” the market

This first point seems obvious and hard to oppose – a mortgage finance regime supported by the government should not be allowed to exclude credit worthy borrowers or reduce liquidity for primary market lenders serving these communities without good cause.  At a minimum, the book of business at these entities should reflect the distribution of loans being originated across income, race and other characteristics.  This is the central, original purpose of the housing goals.  Yes, they are imperfect and sometimes have been misused.  But you cannot assess what you cannot measure.  Some form of bench-marking, whether imposed by the government regulator based on market research and projection, or by the entities themselves using market data and economic projections approved by a regulator as some have suggested, is needed to monitor guarantors’ performance.  Fannie and Freddie have been under a revised housing goals regime since coming under conservatorship in 2008 and the HERA revisions and a de-politicized goals process have improved them.  In fact, GSE credit terms are significantly tighter than in the early 2000's under this regime, a continuing issue that still needs to be addressed.

Create a mandate for “leading the market” into new areas through research, pilots, and partnerships that extend secondary market liquidity responsibly

But matching the market is not enough.  The government’s support for these entities should include an obligation to use the secondary market’s central role in the marketplace to responsibly expand credit.  This is what the often-abused term “lead the market” means to me.  This “duty to serve” requirement was adopted in the 2008 HERA GSE amendments to focus attention on three specific market sectors – rural housing, manufactured housing and affordable housing preservation. Fannie and Freddie published their first proposed DTS plans under this provision on May 8.  What was speculative in 2013 has become an operating reality.  It should not be cast aside as reform discussions move forward.

Charge a fee on government backed MBS to support market expanding activities and contribute to support for low income rental assistance programs

Finally, reform should include a fee on all government backed MBS in the form of a strip collected every year on outstanding securities.  These proceeds should fund the market expanding efforts in the duty to serve, through risk sharing on new, untested products, on helping to create additional effective demand from consumers, and so on, as well as housing and community development funding outside of the mortgage finance system.  Fannie and Freddie's DTS plans should show further examples of how the GSEs can help develop and prove new markets. The GSEs today already pay an assessment on each year’s business to fund the Housing Trust Fund and Capital Magnet Fund, this year amounting to $455 million.  Shifting this volume-based one-time fee to a yearly ongoing strip will provide a significantly larger and more reliable source to continue that funding in addition to the market expanding efforts in the duty to serve. 

The Promising Road proposal I co-authored and the MBA’s new effort both envision some form of this combination of requirements in a future system.  The recent Milken Institute proposal includes the fee in the form of a 10 basis point strip -- as did the ill-fated Johnson-Crapo legislation in 2014 --  and suggests further elaboration in a later paper. These features are already required under conservatorship.  Reform should maintain and improve them.  Without these or similar requirements that blend quantitative measures of broad service to the market, an expectation that enterprises benefiting from a federal guarantee will help lead the market into new opportunities, and funding to support such work through a fee, reform could easily wind up sponsoring mortgage credit in ways that exclude households and communities from its benefits, an unacceptable outcome.

Fannie Mae’s Game Changing Announcement

August 31, 2015

Fannie Mae last week announced a significant new product, HomeReady^TM, designed to expand the company’s ability to provide financing for low and moderate income borrowers.  HomeReady builds on the company’s historical family of low-down payment mortgage products with flexible underwriting features, starting with Community Home Buyers^TM in 1993 and continuing through the My Community Mortgage® product. It is available to borrowers with incomes at or below 80 percent of their Area Median Income (AMI) or borrowers of any income living in designated Census tracts.

The product adds underwriting flexibilities and caps risk-based fees for loans above 80 percent LTV with borrower credit scores of 680 at 150 basis points, less than for similar non-HomeReady loans.  It also requires only 25 percent MI cover on HomeReady loans with LTV’s above 90 percent. These should make Fannie Mae mortgages more competitive for the targeted low wealth and LMI borrowers and enable lenders to make more loans to these borrowers with a secondary market execution.

But the really game-changing features of HomeReady are its adoption of a standardized consumer education requirement and the incorporation of the product into Fannie Mae’s Desktop Underwriter® (DU) automated underwriting system while offering unlimited lender access to the product.  With the first Fannie Mae has established a new standard that will affect the entire housing education and counseling industry.  With the second it has broken a long-standing tradition of offering HomeReady’s antecedents in only limited quantities as part of individual lender contracts. By promising that DU® will flag loans that qualify for HomeReady even if the lender does not specify them as such, and by offering them without volume limits, Fannie Mae has brought a flexible low down payment loan into its mainstream business in an unprecedented way.

Homebuyer Education

Fannie has been a supporter of homeownership education and counseling since the beginning of its community lending business in the early 1990’s.  Both Community Home Buyer and My Community Mortgage required borrowers to certify they had received housing counseling (although this requirement was briefly suspended for MCM in the mid-2000’s).  While at first this requirement could only be fulfilled through a HUD-certified counseling agency, years of pressure from lenders and mortgage insurers led Fannie to broaden the requirement to permit these other actors and other methods in the mortgage process to provide it.  Given the self-interested nature of these players, and the increasing pressure to increase the velocity and efficiency of mortgage originations, these products’ counseling requirement became less standardized and, according to some, less comprehensive and meaningful at the products aged.

HomeReady requires that the borrower complete an on-line homeownership education course offered through a platform called Framework®.  It requires borrowers to complete the course and to foot a $75 fee for it.  While Fannie is encouraging consumers to also consult certified housing counselors (who would receive 20 percent of the fee if they refer a borrower to Framework) this will not be required.  Framework is sponsored by the Housing Partnership Network (HPN) and the Minnesota Homeownership Center and meets the requirements of the HUD Housing Counseling Program and the National Industry Standards for Homeownership Education and Counseling.

By picking one platform that is accessible to any consumer with online access in order to obtain HomeReady’s flexibilities and pricing, Fannie has reset the table for the future of homeownership education and counseling. Fannie has neatly cut through a decades-long debate about how to incorporate education into the loan process by adopting it.  The education and counseling industry will have to adjust to this standard now.  Efforts to further refine or expand the use of classroom or face to face counseling will have to start with the premise that targeted consumers will use the Framework curriculum and certification.

DU and Community Lending

The announced availability of HomeReady through DU without contract limits is a huge paradigm shift.  If the product actually succeeds in qualifying more LMI borrowers it means its growth will be hindered only by effective demand, while MCM and Community Home Buyer were constrained by budgeted supply.

If the biggest story in mortgage underwriting in the late 1990’s and early 2000’s was the introduction and widespread adoption of automated underwriting, one of the biggest in 2015 will be Fannie’s decision to make their new HomeReady product available on the platform.  MCM was offered only through manual underwriting until 2006.  This required lenders to pull these loans out of their newly developed default business processes, which slowed its adoption.  Moreover,  lenders only could access it through individual contract provisions and for fixed amounts specified in them.  As described in last week’s announcement, this is not the case with HomeReady.

Enabling DU to flag loans that are eligible for HomeReady is another accelerant that should increase its market penetration.  Rather than manually assess a loan’s eligibility for the increased flexibilities, lenders should be able to run all loans through DU and get a recommendation from the system if a loan qualifies.  This removes another process barrier for lenders and should increase consumer access to the product compared to a non-DU execution, or one that requires the lender to flag the loan as HomeReady eligible.

Lenders will be able to commingle HomeReady and other loans in MBS pools and whole loan commitments, another feature that should make using the product easier and more seamless for lenders.

New Product Features

The HomeReady product adopts a number of underwriting flexibilities that are designed to increase credit access for LMI borrowers.

• Non borrower income will be included in calculating debt to income ratios up from the default limit of 45 percent to as high as 50 percent, which should increase such households’ buying power.  Fannie Mae states in its product materials that its research shows such income is “sticky” and persistent, and as stable as income that does not include such sources.

• Allows non occupant borrowers, and rental payments, such as from a basement apartment, and boarder income can augment the borrower’s qualifying income

• Lower mortgage insurance requirements, though these weren’t specified in the announcement

• Use of nontraditional credit

• Allowing gifts, grants, Community Seconds® and cash-on-hand to be used for down payments

• Allowing manufactured homes and HomeStyle® renovation loans up to 95 percent LTV

Freddie Mac has not announced a comparable product yet.  But historically, Freddie has followed Fannie Mae in such initiatives after some period of time.  Fannie’s announcement promises further details through the rest of the year and expects to go live with the product on DU and accept HomeReady loan deliveries late in 2015.

Progress on Mortgage Finance Reform?

March 12, 2014

More than five years after the housing bust pushed Fannie Mae and Freddie Mac into conservatorship, housing finance reform seems a step closer to reality with the announcement  on March 11, 2014 that the Chair of the Senate Banking Committee and the Ranking Minority Member have reached agreement on a comprehensive bill.

In announcing the agreement yesterday, staffs from both Chairman Tim Johnson (D-SD) and Ranking Member Mike Crapo (R-ID) provided a high-level summary of the bill they have been drafting for months, following an extensive series of hearings and meetings with stakeholders.  Staff was only able to provide a brief, one-page summary of the developing legislation’s details yesterday at a hastily convened briefing for stakeholders.  They said that actual legislative language will be circulated to committee members “shortly,” with the expectation that a full draft bill will be available within weeks and a mark up in the full  Senate Committee scheduled within months.

Staff said that the bill is “built” on S. 1217, the bill introduced last year by Sens. Bob Corker (R-TN) and Mark Warner (D-VA) and cosponsored by a bipartisan group of 10 Senators.  That bill built on recommendations from, among others, the Bipartisan Policy Center’s housing commission, and would wind down Fannie and Freddie and replace them with a full faith and credit guarantee on mortgage backed securities, offered through a new public regulator to issuers of securities that first obtained significant private guarantees that would protect investors ahead of any government-funded guarantee.  The bill attracted significant attention throughout the year, and has become the “chassis” for other approaches.

Staff yesterday said that their bill will change some important aspects of S. 1217 in response to all the hearings and stakeholder feedback.  These changes include the following important features:

• All guaranteed securities would be issued by a single entity, created by the public guarantor (called the Federal Mortgage Insurance Corporation, FMIC, in S. 1217) and operated as a utility mutually owned by the private entities making use of it.  Other, non guaranteed securities could use the same platform, but all guaranteed securities would be required to do so.  It was unclear if such PLS customers would have to join the cooperative operating the platform or not.  This single securitization platform would build on one that the Federal Housing Finance Agency (FHFA) has sponsored through a new corporation jointly owned by Fannie and Freddie. (Progress on this new platform has been slow, with no Chairman or CEO yet chosen to run the company, although it has leased space in suburban Washington, D.C.)  This platform would establish consistent underwriting, servicing and other requirements for insured bonds.  The servicing rules would build on those issued by the Consumer Financial Protection Bureau (CFPB) last year as part of the Dodd-Frank implementation; how they would elaborate on that rule is not clear. Staff said the underwriting requirements would “mirror” the so-called “Qualified Mortgage” definition issued by the CFPB that became effective earlier this year, but no details were provided.  The current maximum single family mortgage amounts would be retained.    
• The bill would require a minimum down payment of at least 5 percent, phased in over a short period.  First time homebuyers could put down a minimum of 3.5 percent. 
• FMIC is modeled in part on the FDIC.  It will collect guarantee fees to finance its work and to fund a reserve that would be used to cover timely payment of principal and interest to investors in the event the private capital required is exhausted.  FMIC would require aggregators to obtain private credit guarantees backed by capital equal to at least 10 percent of the exposure, in a form determined by the FMIC.  This could include straight equity as well as back-end risk sharing arrangements through capital markets structures.  Private guarantees provided by capital market structures alone would have to include a full 10 percent first loss.  
• Mortgage assets would be aggregated by private firms for issuance by the new utility.  To facilitate participation in the system by smaller lenders, the bill would establish a new, mutually owned lender cooperative open to any lender with less than $500 billion in assets which would operate a cash window and keep servicing rights for lenders who wish to use that execution, a holdover from S. 1217.  In addition, FMIC would require all aggregators to offer a “level playing field” for lenders by barring variable pricing based on size.  
• Fannie and Freddie would be wound down over a 5 year transition period, which could be extended if certain benchmarks in developing the new system are not met.  Their existing MBS would receive a full faith and credit guarantee to assure continued liquidity after the new system is launched.  
• Staff said that their bill will include a mandate that the system facilitate the broad availability of credit, and monitor consumer and market access to credit.  S. 1217 lacked both a strong mandate and a viable mechanism for monitoring the performance of both the system and its users.  Whether the new bill accomplishes this important objective successfully will depend on the actual language.  The bill would terminate the current housing goals regime for Fannie and Freddie.   
• Like S. 1217, the bill would impose a new fee on all insured mortgage backed securities in the form of a 10 basis point strip on the bonds’ outstanding UPB.  This fee would directly fund affordable housing and community development activities through the Housing Trust Fund at HUD and the Capital Magnet Fund for CDFIs at Treasury.  Whether and how the fee would finance a so-called “Market Access Fund” within the FMIC, a key goal for many consumer and progressive groups, is unclear. Staff also said that the fee could vary depending on issuers and guarantors success at meeting certain objectives to fully serve the market as determined by the FMIC. But they said while fees could be lower or higher for individual issuers, the average charged fee on each year’s production would have to be 10 percent,  
• The bill would extend a federal guarantee to multifamily mortgage bonds.  It would retain the current risk-sharing models at Fannie and Freddie and spin out their current multifamily businesses.  Multifamily issuers and guarantors could be the same entities, which how the current GSE risk-sharing models are run.  This this is not the case for single family guarantors, who have to be separate and, if part of a larger institutions, separately capitalized.  Staff described but did not detail affordability requirements that would apply ao all insured multifamily securities.

As in all things congressional, the devil of this new attempt to reform the mortgage finance system will be in its details.  These will become clear once actual legislative language is available for review.  But the agreement on these major “architectural” features, and the commitment to move forward, are very important steps.  And based on the staff briefing, the new draft has incorporated a number of important new features sought by progressive and consumer groups, including the preservation of the fee at the maximum level of 10 bps included in the early version of S. 1217; the addition of what staff described as a strong, clear mandate to make sure the system supports widespread access for the widest range of credit-worthy borrowers and communities; a clear expectation that the FMIC judge the performance of both the system and its participants; and a significant approach to ensuring federal support for rental housing finance, including a requirement that it primariliy serve tenants at affordable rents.