ABA's Letter to FHFA on the Uniform Mortgage-Backed Securities

The Honorable Mark Calabria
Director
Federal Housing Finance Agency
400 7th Street SW, 8th Floor
Washington, D.C., 20219

Director Calabria:

The undersigned associations are pleased to respond to the Federal Housing Finance Agency’s (FHFA) Request for Input (RFI) on Uniform Mortgage-Backed Security (UMBS) pooling. Together, we represent a wide range of the parties that would be affected by changes to mortgage-backed security (MBS) pooling practices, including loan originators, investors, market makers, and other businesses that rely on continued robust liquidity in the To-Be-Announced (TBA) market and the markets for specified pools and collateralized mortgage obligations (CMOs).

Several of the undersigned are also planning to submit individual comments, but we felt it was important to provide our shared views and concerns on a number of issues. While we appreciate the FHFA’s well-intentioned effort to address liquidity and other concerns in the UMBS market, the RFI includes proposals that could fundamentally change the MBS markets in ways that will have negative consequences for all of our members as well as for mortgage borrowers; those proposals should not be implemented.

While the initial operational transition from separate forms of MBS issuance by Fannie Mae and Freddie Mac (the Enterprises) to the UMBS was relatively smooth, we believe there is room for improvement in TBA market liquidity and that fundamental misalignments that existed prior to the creation of the UMBS have not yet been sufficiently addressed. Unfortunately, we do not believe that the proposals in the RFI will address these issues, and instead may worsen them.

I. The Proposed Approaches in the RFI Will Not Improve the TBA Market and Should be Reconsidered

FHFA’s stated purposes in issuing this RFI are to ensure more uniform cash flows for TBA investors to promote liquidity, while continuing to offer specified pools sought by investors on a limited basis, and to align the Enterprises policies around the actions to be taken when a seller/servicer exhibits prepayment behavior outside acceptable norms. To achieve these goals, the RFI would: (1) mandate that the majority of loans are placed into multi-lender pools; (2) significantly limit the production of specified pools; and (3) create a mechanism that would limit access to TBA-eligible pools for seller/servicers with anomalous prepayment performance.

While we share these overall goals, we believe that the approaches detailed in the RFI will not result in enhanced liquidity in the TBA market, will diminish the specified pool and CMO markets, and will cause harm to virtually every market participant, leading to higher costs or reduced access to credit that will ultimately impact mortgage borrowers. We strongly urge FHFA to reconsider this approach and offer the following comments.

II. Ginnie Mae is Not an Appropriate Model for Conventional MBS Markets

The RFI states that the primary intent is to “reap the advantages of forming much larger, multi-lender pools that stem from more consistency in prepayment behavior.” The RFI goes on to state that “FHFA believes this proposal would result in pooling practices similar to those under the Ginnie Mae II program.”3 Under that program, Ginnie Mae issues one large multi-issuer TBA-eligible pool per month, per coupon. As explained in separate comments filed by many of our associations, Ginnie Mae liquidity trails that of conventional markets by a variety of measures, and to move conventional MBS markets in that direction would be an unwise step backward. The market has clearly expressed its preference for the Enterprises’ MBS issuance model. Therefore, we oppose adopting the Ginnie Mae model for the conventional market.

A critical assumption of the RFI is that creating much larger pools, and limiting specified pool production, will increase homogeneity and enhance TBA liquidity. While homogeneity (of the type provided by large multi-lender pools) is needed to a point, we believe that market participants continue to prefer some degree of variability and that more numerous pools (even with somewhat less predictable performance) will foster greater liquidity. This is in part because singular or a very limited number of large pools with predictably bad performance will assuredly result in poorer TBA deliverables. In contrast, more numerous pools with limited variation will be more likely to lead to better liquidity given a broader range of strategies that can be employed even if the deliverable is somewhat less certain than market since the market is more attractive to investors. A limited level of variation allows investors to express their views on expectations of MBS performance, rewards seller/servicers that create more desirable MBS, permits a variety of trading strategies, and helps foster liquid and vibrant trading. In short, we believe investors prefer more options, not fewer, and that driving toward one or a very few large pools will reduce investor interest and liquidity, not increase it.

Additionally, the proposal eliminates incentives for lenders to create more desirable MBS, and it gives the Enterprises too much control over mortgage markets. Today, by choosing which lenders are eligible for which MBS programs, the Enterprises have too much influence in the market. The proposal extends, enhances, and further entrenches that power by removing the ability of market participants to individually price the risk of a variety of pool types. In a post-conservatorship environment, it is not clear what controls or limitations might be placed on this power, how any standards would be enforced, or how the Enterprises would be required to remain aligned as private-sector entities.

III. The Proposal Does Not Address the Underlying Problems

A fundamental challenge facing the UMBS market is that similar bonds issued by the Enterprises do not necessarily trade at the same prices. Market participants have identified important performance and alignment issues between the Enterprises that should be addressed, but these efforts remain ongoing, causing divergent pricing between some of the Enterprises’ MBS. Driving more loans into larger multilender pools does not remedy the misalignments of the MBS produced by each of the Enterprises; instead, it masks the problems without fixing them. It will also reduce the incentives for originators to produce MBS for which investors typically express greater demand, given that they will not be rewarded with increased prices for those bonds. Instead, originators of more desirable MBS will subsidize producers of less desirable MBS and, more problematically, any institutions that rapidly refinance loans in ways that do not provide tangible benefits for borrowers. Further, nothing in the RFI addresses the potential for inappropriate types of market share competition between the Enterprises, which could degrade MBS performance. Far from correcting the underlying problems that impact liquidity, the proposal would only mask such problems, and it has the potential to create ever-larger pools of reduced quality in a “race to the bottom.”

IV. Banning Originators with Fast Prepayment Rates Presents Challenges and Requires Greater Detail

The RFI proposes that the Enterprises would exclude certain originators from TBA pools based upon prepayment performance. While this attempt to address concerns over unjustifiably fast-paying seller/servicers—who in some instances could be churning loans or engaged in other abuses—is laudable, it lacks detail. Further, as already noted, requiring that more loans go into large multi-lender pools is likely to mask problems with particular originators. It is unclear from the RFI how the Enterprises would distinguish originators with unjustifiably fast prepayment rates from institutions with faster prepayment rates resulting from technological or other innovations that drive more efficient refinancing. A preferable approach would be to require the Enterprises to develop detailed, transparent standards for evaluating originator performance to ensure that bad actors are sorted out of larger, multi-lender pools. We note that in today’s market, originators can be rewarded for producing pools desired by MBS investors, and this incentive provides a check on originator behavior. The proposal in the RFI would, at best, weaken this market mechanism.

V. Specific Harms to Market Participants

As discussed above, we believe the proposals in the RFI could impair TBA market functioning and liquidity. Accordingly, we believe that the proposals will have negative consequences for nearly all market participants, including each of our respective memberships, as well as for borrowers. We discuss some of the specific harms we anticipate below.

A. Originators

Loan originators, whether they are community or large banks, credit unions, independent mortgage bankers, or other types of financial institutions, strive for best price execution. The level of originator activity varies with market conditions and may be best served by issuing single-issuer pools or other specified pools depending upon investor appetite. The RFI would greatly reduce flexibility by requiring that the bulk of originations are delivered into large, multi-lender pools. This will result in reduced profitability for many originators and reduced product availability (or increased costs) for borrowers as loans become more standardized and “cookie cutter” to fit into the larger pools. Significantly, by forcing most loans into larger pools, the proposal would be particularly disadvantageous to those institutions that originate loans most desired by investors.

B. Investors

Investors would have their investment choices and strategies substantially reduced by the proposal. Their options would be limited to the large, multi-lender pools, or fewer and more expensive specified pools (which may be further limited by policies set by the Enterprises). Overall, investment options will decline, which may result in investors seeking returns in other markets, reducing rather than expanding liquidity in the TBA market. We also note that investors will have fewer options in the CMO market, as the specified pools that are available to create desired cashflows in a CMO structure will either be unavailable or uneconomical. Finally, while FHFA attempts to make accommodations for real estate investment trusts (REITs), an important investor class, it is not clear that the accommodations would be sufficient or that they would be equitable to REITs or other types of investors.

C. Market Makers

Market makers stand between transaction parties to ensure a continuous flow of liquidity from buyers to sellers. TBA trading operations are commoditized, low-margin businesses. Specified pool and CMO trading, however, are vital sources of revenue to support the broader MBS trading operations. To the extent that specified pool and CMO markets become less vibrant, this will affect the TBA side of the business and result in decreased capital allocations to TBA desks. As a result, market makers would carry less inventory, take on less risk, and provide less liquidity to TBA markets than they do today.

D. Mortgage Borrowers

Reduced liquidity in the TBA market would likely lead to higher interest rates for borrowers. Additionally, the reduced number of specified pools may result in fewer loan product options for borrowers as loans become more standardized to fit into the larger, more generic multi-lender pools. The cost for more specialized loans, if they remain available, could increase. These increased costs could lead to a reduction in access to credit, putting homeownership out of reach for many Americans.

VI. Conclusion

For the reasons cited above, we strongly urge FHFA to abandon the proposed mandate for a high share of loans to be securitized in multi-issuer pools. Specifically, we urge FHFA to directly address the causes of misalignments among MBS between the Enterprises and to avoid taking actions that reduce the variety and optionality for investors and lenders that exists today in the TBA market. Efforts to address bad actors are well intended but need greater detail. FHFA could direct the GSEs to develop seller/servicer MBS performance standards and require that they be clear and transparent to all market participants and well aligned to prevent forum shopping by bad actors. We appreciate this opportunity to comment and look forward to working with FHFA to foster an efficient and liquid TBA market.

American Bankers Association

Housing Policy Council

Independent Community Bankers of America

Mortgage Bankers Association

National Association of Home Builders

Nareit

National Council of State Housing Agencies

SIFMA