2101 Wilson Boulevard, Suite 610
Arlington, VA 22201
(571) 527-2601
September 23, 2016
Hon. Mel Watt
Director, Federal Housing Finance Agency
400 7th Street SW
Washington, DC 20024
CHLA Comment Letter – FHFA Single-Family Credit
Risk Transfer Request for Input (RFI) – June 2016
Dear Director Watt:
The Community Home Lenders Association (CHLA) is writing to provide our comments in response to FHFA’s June 2016 Request for Input (RFI) on the subject of Single Family Credit Risk Transfers.
CHLA has warned repeatedly over the last two years that upfront Credit Risk Transfers would likely make it harder for small and mid-sized mortgage lenders to competitively originate Enterprise loans, which would reduce competition and access to mortgage credit, and thereby hurt consumers. CHLA is therefore appreciative that the Request for Input identifies a “Level Playing Field” as one of FHFA’s principles for Credit Risk Transfers and includes a separate section with questions on this topic.
1. CHLA’s main recommendation is that FHFA should not permit any Up-Front Credit Risk Transfers by the Enterprises. CHLA believes this is needed to ensure competitive access for all mortgage originators in order to maximize consumer choices and competition and lower mortgage rates and fees.
2. If, however, the Enterprises are permitted to engage in upfront Credit Risk Transfers, this should be limited to loan level risk sharing through guarantors [e.g., private mortgage insurers (PMIs)] – which MUST include three market protections: (a) a prohibition against volume discounts, (b) a requirement that any guarantor/PMI must serve all approved seller-servicers, and (c) full pricing transparency.
3. Additionally, if the Enterprises are permitted to engage in any upfront Credit Risk Transfers, such use should be conditioned on periodic certifications by FHFA, based on objective evidence, both that:
(1) a fully competitive cash window exists to serve all origination market needs AND
(2) mortgage originators can securitize loans on a fully competitive basis with vertically integrated mega-banks and other aggregators that structure and execute up-front risk sharing securitizations. And there should be complete transparency by the Enterprises with respect to G Fees, buy-up and buy-down grids, LLPAs, or any other mechanism that provides for pricing variability.
These recommendations reflect bipartisan concerns raised during debate on the 2014 Senate GSE Reform bill that vertical integration poses the significant risk that the major Wall Street banks could use securities affiliates to monopolize or dominate Enterprise loan origination – which in turn led to provisions effectively barring such practices that use securitization-based Credit Risk Transfers.
On a related issue, CHLA appreciates the development by FHFA and the Enterprises of a Common Securitization Platform (CSP) – and we renew our opposition to piecemeal Congressional proposals that would take the CSP from the Enterprises and effectively hand over its control to the large Wall Street banks. We believe such a proposal would also reduce competition and harm consumers.
Problems with Up-Front Risk Sharing
CHLA commends the FHFA and the Enterprises for their actions in recent years to reduce if not eliminate G fee disparities among lenders, that previously existed based on lender size or loan volume. These actions have created a more democratized mortgage origination market for Enterprise loan origination, which is beneficial for consumers. This has been timely, because, coming of out the 2008 housing crisis, many banks abandoned mortgage lending or curtailed it through credit overlays, and it has been small and mid-sized non-bank mortgage bankers that have filled the access to credit gap left by banks cutting back.
Unfortunately, up-front risk sharing, particularly where the securitizer generates the risk sharing funds through an MBS offering, threatens to undo this progress and return to a more concentrated government guaranteed mortgage market – dominated by the same Wall Street bank/securities firms that precipitated the 2008 Housing Crisis through their MBS securitization of large amounts of subprime loans.
The concern is simple. Currently mortgage loan originators can originate qualified Fannie Mae and Freddie Mac mortgage loans – and can either competitively sell them directly to the Enterprises through the so-called “Cash Window” or can pool and securitize these loans through a broad market of broker-dealers that are not vertically integrated with a mortgage origination affiliate.
However, the essence of up-front risk sharing is to create a chokepoint in this process – essentially a condition that loans cannot be delivered without the necessary risk sharing component in place. In this environment, the firm that controls the securitization process can control the MBS proceeds for the purpose of mortgage origination.
This concern is exemplified by the J P Morgan up-front risk sharing deals done to date. Regardless of the details of these transactions, one simple fact stands out: Chase Bank, J P Morgan’s bank affiliate, appears to have controlled 100% of the MBS proceeds for mortgage origination purposes.
It was precisely this issue – large vertically integrated firms that both underwrite MBS through a securities affiliate and have a large mortgage origination platform through a bank affiliate – that was a major focus of bi-partisan concern during debate in the Senate on the Johnson/Crapo GSE Reform bill, and that ultimately led to restrictions in the bill on this very type of activity.
We would also note that some of the upfront securitization deals to date have been done by non-vertically integrated aggregators. While these do not result in the proceeds being used exclusively by affiliated mortgage originators, there are also potentially anti-competitive effects of this particular practice. First, if these aggregators were to dominate the market, they could effectively turn mortgage originators into correspondents, thus reducing competition in the mortgage origination market at a time when an increasing number of non-bank mortgage lenders are securitizing loans. Secondly, this could force mortgage originators to originate loans on a seller-released basis – thus increasing concentration in the servicing of these loans into a smaller group of large bank and non-bank mega-servicers.
Finally, the up-front risk sharing deals done to date are much less transparent and competitive than back-end transactions. Unlike back-end risk sharing, upfront risk sharing G fee reduction levels are not negotiated on a competitive basis, and there is almost no disclosure about the details of the deals.
Up-Front Risk Sharing Threatens a Level Playing Field
Risk sharing already appears to be the “reality on the ground” – with some 90% of Enterprise loans over the last few years ultimately carrying some form of risk sharing. FHFA’s challenge is to move from a so-called “pilot” phase, testing out different forms of risk sharing, to a more permanent, stabilized market.
Some have argued that FHFA should allow unfettered development of all forms of risk sharing, and should not impose market rules, such as prohibitions against volume discounts by risk sharing providers.
However, CHLA believes that such an approach would likely mean that certain forms of risk sharing will dominate – at the expense of consumers, smaller lenders, and the objective of a level playing field.
Therefore, FHFA should not simply let all forms of risk sharing go forward, without any rules or guidance, and without addressing the significant concerns that Congress, consumers, and smaller lenders may have about vertical integration and about market dynamics inherent in up-front risk sharing.
Following are reasons why the up-front risk sharing securitizers could dominate or monopolize the Enterprise government guaranteed loan market, if FHFA followed a laissez-faire approach:
- G fee reductions negotiated with the large Wall Street banks or other large aggregators could be more favorable than other forms of risk sharing, such as back-end risk sharing or up-front loan level guarantee (PMI) risk sharing.
- The scale of the Wall Street Banks and the economies of scale for their MBS issuance could give them a significant pricing advantage over other competitors. Moreover, broker dealers that now help small and mid-sized lenders securitize Enterprise loans may not have the volume of business to justify the long, complicated process of negotiating with the FHFA and Enterprises to enter this market, and do not have the financial incentives to execute these transactions that the large Wall Street banks with mortgage origination affiliates have.
- The large Wall Street banks could cherry pick the best loans (i.e., only the highest FICO scores or lower LTV levels) – a likely outcome, as their bank affiliates have significantly tightened their credit box in recent years. This would facilitate a two-tiered market, balkanized by borrower credit quality. In contrast back end risk sharing does not result in this type of loan cherry picking.
- Congress has already given preliminary approval for provisions which we believe favor the largest market players, including a requirement for more upfront risk sharing and a provision to turn over the Common Securitization Platform (CSP) to the private sector, which would then be likely to be dominated by the biggest Wall Street Banks. For example, statutory language, or even the threat of this, could easily result in policies which create a preference or competitive advantage for up-front securitization based risk sharing.
Up-Front Risk Sharing Is Detrimental to Consumers
There are many reasons why up-front risk sharing could hurt consumers:
- More Concentration Means Less Competition – and Higher Mortgage Rates and Fees for Borrowers. As some of the large banks have exited the mortgage market or increased credit overlays in recent years, non-banks are the major source of affordable loans. Thus, as explained in this letter, if upfront securitization risk sharing leads to more concentration among mortgage originators, this would reduce competition and hurt consumers’ access to and pricing of loans.
- Fewer Loans Affordable to Minorities and Borrowers with Lower FICO scores or down payment capabilities. Setting the credit screen up front incentivizes the vertically integrated Wall Street banks and other large aggregators to establish more credit overlays, in order to gain a pricing advantage over alternative forms of risk sharing, such as back-end risk sharing, where the Enterprises are serving the market more broadly. This could negatively impact the availability of affordable mortgage loans to minorities, and borrowers with lower FICO scores or lower down payment capabilities. Ultimately, if upfront securitization risk sharing dominates the market, it could be more difficult for the FHFA and the Enterprises to carry out specific affordable policy requirements, such as Duty to Serve and Housing Goals.
- More Concentration in the Servicing of Mortgage Loans. Upfront securitization risk sharing increases the market power of large aggregators and thereby decreases the ability of lenders without securities capabilities to competitively securitize loans. This is the method under which lenders retain servicing on loans they originate. Thus, these lenders might be forced to sell loans on a servicing released basis, which would increase the number of loans held by the largest mortgage servicers to the aggregators. This would mean less personalized services and more systemic risk related to the large mega-servicers.
- It Will be Harder to Reduce or Eliminate Risk Sharing Requirement in Turbulent Market Conditions. There is a growing consensus that to ensure a stable mortgage market, there must be flexibility to reduce or eliminate the level of risk sharing in a turbulent market, where risk sharing may not be available or is excessively expensive. It is easier to monitor and to make these adjustments under back-end risk sharing than under up-front risk sharing.
So-called “Recourse” Upfront Risk Sharing
The FHFA is seeking comment about so-called “collateralized recourse” up-front risk sharing. CHLA appreciates the interest of the FHFA and the Enterprises’ in having collateral set aside to cover losses. To that point, we would simply point out that the investor-based back-end risk sharing deals done to date can be just as effective in achieving this objective. Thus, this is not a justification for up-front risk sharing.
CHLA also appreciates the principle behind lender “recourse” – that the mortgage originator retains ongoing loss risk on loans it originates. Unfortunately, because the deals done to date are almost completely opaque, it is almost impossible to determine how these deals have been structured and priced.
However, from our limited understanding, the deals completed so far do not appear to be “recourse” in the true sense of the word, as the risk is apparently only a very limited piece, and possibly not even a first-loss risk position.
Moreover, it is our understanding that on many deals, the risk is not recourse to the actual mortgage originator, but instead to some other investors or entities. If true, then the commonly perceived benefits of recourse – where the actual loan originator has real skin in the game – do not exist.
We are aware that BASEL III concerns, as well as true sale considerations, may make it difficult for banks (as mortgage originators) to have recourse on such loans. This has led to structures that use approaches that rely on third party investors (for example, investors of the securities affiliates of the Wall Street banks underwriting the MBS or REIT aggregators that are distinct from loan originators).
More importantly, it would be inappropriate to use the term “recourse” to justify the utilization of a risk sharing structure that, were it to proliferate and monopolize or dominate the market, could completely undermine the objective of a level playing field for mortgage loan originators.
Moreover, in the case of the large Wall Street Banks, this is the essence of vertical integration – a large securities firm leveraging their distinct securities powers and securities economies of scale to exclusively generate MBS funds for mortgage origination exclusively by their bank affiliates.
Similarly, in the case of the few deals done to date by large aggregators (REITS), we assume that it is the REIT itself that funds investments in the risk pieces – as opposed to true recourse by the mortgage originator. And, similar to concerns about vertical integration, these aggregators could use their securitization capabilities and economies of scale to control the proceeds of funds available for Enterprise loans – thus potentially turning mortgage originators into loan correspondents.
Finally, there is the potential risk that we could experience a repeat of some of the practices of the subprime mortgage crisis, where complex, non-transparent transactions resulted in sale of risk sharing pieces to unwary investors, motivated by the profits from both the securities deals and the mortgage origination. At the extreme, without proper vigilance, we could experience a repeat of the bad incentives from the housing crisis – such as the Magnetar example, where extremely risky investments were made in the highest risk piece, because of offsetting benefits or investments elsewhere in the transaction.
Private Mortgage Insurer Loan Level Risk Sharing
Historically, CHLA has been supportive of allowing PMIs to offer up-front loan level risk sharing on Enterprise loans. This support was rooted in the simple fact that PMIs, unlike the large vertically integrated Wall Street banks, are not competitors in mortgage origination. Therefore, they cannot leverage the risk sharing activity to monopolize or dominate mortgage origination or turn mortgage lenders into mere correspondents.
However, CHLA’s support has been conditioned on protections to ensure that the benefits of such participation extend equally to all small and mid-sized mortgage loan originators – and is not used merely as a tool to allow large lenders to negotiate pricing discounts and gain a competitive advantage – so-called one-off deals. The three specific protections CHLA believes are essential are:
(a) A prohibition against volume discounts,
(b) A requirement that any participating PMI must serve all approved seller-servicers, and
(c) Full pricing transparency.
If FHFA (or Congress through reform) were to establish such market protections as legally binding protections, CHLA could support up-front PMI risk sharing. However, the 2014 GSE Reform bill did not include such protections and there is noting in FHFA’s June RFI referring to such protections.
Therefore, in the absence of these protections being put in place, CHLA cannot continue to support this particular form of up-front risk sharing, and is withdrawing its support for this approach.
FHFA and Enterprise Concerns about Counterparty Risk.
CHLA is mindful that FHFA and Enterprises may have concerns about counterparty risk with respect to the PMIs. CHLA is not in a position to evaluate such concerns in depth. However, we would make two points to put those concerns in perspective.
First, it is true that a first loss risk sharing of MBS pools obviously offers a higher chance that taxpayers will incur no losses on particular pools, while deep loan based PMI coverage will inevitably result in some losses on individual loans. However, we would point out that the overall taxpayer risk of the former may be higher than the latter, particularly in high stress environments like the 2008 housing crisis. That is because once the first loss MBS level is pierced, taxpayers would take on increasingly accelerated risks – while PMI loan based coverage offers much deeper private sector exposure on individual loans.
Secondly, having different sources of risk sharing, such as through PMIs – whether back-end or front-end – provides more diversification in terms of exposure to counterparty risk.
Elevating the Importance of a Level Playing Field
CHLA appreciates the other objectives FHFA identified in assessing how to do risk sharing – including protecting taxpayers, stability through economic cycles, having repeatable, scalable transactions, and addressing counterparty concerns. CHLA believes that all of these objectives can be achieved through back end risk sharing, and the following section addresses these factors.
If, however, FHFA elects to permit some forms of up-front risk sharing, CHLA strongly recommends that the importance of a level playing field – promoting full competition to protect consumers – should be one of the highest priorities. Moreover, this should not just be rhetorical – the proof is in what takes place in the market. In order to address these concerns, CHLA recommends that if the Enterprises are permitted to engaging in any form of upfront Credit Risk Transfers, such use should be conditioned on periodic certifications by FHFA, based on objective evidence, both that:
(1) a fully competitive cash window will exist to serve all origination market needs AND
(2) mortgage originators can securitize loans on a fully competitive basis with vertically integrated mega-banks and other aggregators that structure and execute up-front risk sharing securitizations.
Achieving FHFA’s Other Goals and Objectives
- Reducing Taxpayer risk in an Economically Sensible Manner
FHFA states that up-front credit risk transfers should transfer a meaningful amount of credit risk to private investors, ensure that private investors successfully retain purchased credit risk, and avoid situations where the risk is returned back to taxpayers.
CHLA’s recommendation that the Enterprises should not engage in up-front risk sharing is predicated on the assumption that any investor or guarantor sources that could be accessed on the front end could be accessed on the back end – particularly with respect to securitization structures in which risk sharing strips are the basis for the generation of risk sharing dollars.
- Transparency
FHFA states that parties to a transaction should provide public disclosure of transaction information.
CHLA believes this should be a top priority – that there should be transparency in all risk-sharing deals. In practice, though, it is back-end risk sharing transactions that have had more transparency – with almost no transparency on the up-front risk sharing transactions. Put simply, there is no way to know such basic facts as how much G fees were reduced on up-front risk sharing transactions done to date – much less understand the competitive implications of these types of transactions.
Therefore, the objective of transparency is best maintained through back-end risk sharing.
- Stability through Economic and Housing Cycles
FHFA states that the transfers should be structured to ensure that the majority of investors will remain in the market through stressful phases of the housing price cycle, including during economic downturns.
CHLA’s conclusion that the Enterprises should not engage in up-front risk sharing is consistent with FHFA’s goal of ensuring stability through economic and housing cycles. Put simply, it is easier to turn off or adjust the back-end risk sharing process in periods of turbulence than it is to make adjustments to up-front risk sharing, which is more tied in to the generation of funds to originate loans. Moreover, back-end risk sharing transactions are inherently more transparent, with clearer after-market valuations – thereby making it easier for FHFA and the Enterprises to make adjustments to risk sharing level requirements in turbulent markets.
- Repeatable, Scalable Transactions with Broad Investor Base
FHFA states that the program should include different transaction structures to attract a diversified and broad investor base, with the objective of improving pricing, increasing secondary market liquidity, and promoting market stability.
CHLA believes that up-front credit risk transfers by nature have a limited investor base, particularly if it is controlled by a small number of large Wall Street Banks or other large aggregators. In contrast, back-end risk sharing potentially enables a more diverse investor base for the risk sharing.
- Counterparty strength
FHFA states that in transactions where the credit risk being transferred is not fully collateralized, credit risk transfer counterparties to the GSEs should be financially strong and stable companies that are consistently able to fulfill their financial commitments in the transaction even in adverse markets.
CHLA agrees that all parties to credit risk transfers should be financially strong and stable. However, we caution that counterparty strength is a two-edged sword – that the potential for re-aggregation of risk, in the hands of a small number of Wall Street banks or other large counterparties might put risk in the hands of large Too-Big-To-Fail entities, which may appear to reduce risk to the Enterprises, but which may in the process increase Systemic Risk.
In closing, we appreciate your consideration of these comments and recommendations.
Sincerely Yours,
COMMUNITY HOME LENDERS ASSOCIATION