Ghost of A Crisis Past

photo by Chandres

The Royal Bank of Scotland settled an investigation brought by New York Attorney General Schneiderman arising from mortgage-backed securities it issued in the run up to the financial crisis. RBS will pay a half a billion dollars. That’s a lot of money even in the context of the settlements that the federal government had wrangled from financial institutions in the aftermath to the financial crisis. The Settlement Agreement includes a Statement of Facts which RBS has acknowledged. Many settlement agreements do not include such a statement, leaving the dollar amount of the settlement to do all of the talking. We are lucky to see what facts exactly RBS is “acknowledging.”

The Statement of Facts found that assertions in the offering documents for the MBS were inaccurate and the securities have lost billions of dollars in collateral. These losses led to “shortfalls in principal and interest payments, as well as declines in the market value of their certificates.” (Appendix A at 2)

The Statement of Facts outlines just how RBS deviated from the statements it made in the offering documents:

RBS’s Representations to Investors

11. The Offering Documents for the Securitizations included, in varying forms, statements that the mortgage loans were “originated generally in accordance with” the originator’s underwriting guidelines, and that exceptions would be made on a “case-by-case basis…where compensating factors exist.” The Offering Documents further stated that such exceptions would be made “from time to time and in the ordinary course of business,” and disclosed that “[l]oans originated with exceptions may result in a higher number of delinquencies and loss severities than loans originated in strict compliance with the designated underwriting guidelines.”

12. The Offering Documents often contained statements, in varying forms, with respect to stated-income loans, that “the stated income is reasonable for the borrower’s employment and that the stated assets are consistent with the borrower’s income.”

13. The Offering Documents further contained statements, in varying forms, that each mortgage loan was originated “in compliance with applicable federal, state and local laws and regulations.”

14. The Offering Documents also included statements regarding the valuation of the mortgaged properties and the resulting loan-to-value (“LTV”) ratios, such as the weighted-average LTV and maximum LTV at origination of the securitized loans.

15. In addition, the Offering Documents typically stated that loans acquired by RBS for securitization were “subject to due diligence,” often described as including a “thorough credit and compliance review with loan level testing,” and stated that “the depositor will not include any loan in a trust fund if anything has come to the depositor’s attention that would cause it to believe that the representations and warranties of the related seller regarding that loan will not be accurate and complete in all material respects….”

The Actual Quality of the Mortgage Loans in the Securitizations

16. At times, RBS’s credit and compliance diligence vendors identified a number of loans as diligence exceptions because, in their view, they did not comply with underwriting guidelines and lacked adequate compensating factors or did not comply with applicable laws and regulations. Loans were also identified as diligence exceptions because of missing documents or other curable issues, or because of additional criteria specified by RBS for the review. In some instances, RBS disagreed with the vendor’s view. Certain of these loans were included in the Securitizations.

17. Additionally, some valuation diligence reports reflected variances between the appraised value of the mortgaged properties and the values obtained through other measures, such as automated valuation models (“AVMs”), broker-price opinions (“BPOs”), and drive-by reviews. In some instances, the LTVs calculated using AVM or BPO valuations exceeded the maximum LTV stated in the Offering Documents, which was calculated using the lower of the appraised value or the purchase price. Certain of these loans were included in the Securitizations.

18. RBS often purchased and securitized loans that were not part of the diligence sample without additional loan-file review. The Offering Documents did not include a description of the diligence reports prepared by RBS’s vendors, and did not state the size of the diligence sample or the number of loans with diligence exceptions or valuation variances identified during their reviews.

19. At times, RBS agreed with originators to limit the number of loan files it could review during its due diligence. Although RBS typically reserved the right to request additional loan-level diligence or not complete the loan purchase, in practice it rarely did so. These agreements with originators were not disclosed in the Offering Documents.

20. Finally, RBS performed post-securitization reviews of certain loans that defaulted shortly after securitization. These reviews identified a number of loans that appeared to breach the representations and warranties contained in the Offering Documents. Based on these reviews, RBS in some instances requested that the loan seller or loan originator repurchase certain loans. (Appendix A at 4-5)

Some of these inaccuracies are just straight-out misrepresentations, so they would not have been caught at the time by regulators, even if regulators had been looking. And that’s why, ten years later, we are still seeing financial crisis lawsuits being resolved.

It is not clear that these types of problems can be kept from infiltrating the capital market once greed overcomes fear over the course of the business cycle. That’s why it is important for individual actors to suffer consequences when they allow greed to take the driver’s seat. We still have not figured out how to effectively address tho individual actions that result in systemic harm.

Reps and Warranties Mean What They Say

Derek Jensen

The New York Appellate Division, 1st Department, issued a ruling in Bank of New York Mellon v. WMC Mortgage, 654464/12 (Dec. 1, 2015) that stands for the proposition that representations and warranties regarding mortgage-backed securities mean what they say and say what they mean. The opinion opens,

This breach of warranty action arises from a residential mortgage backed securitization called the J.P. Morgan Mortgage Acquisition Trust 2006-WMC4 (the Trust). The Trust was arranged and sponsored by defendant J.P. Morgan Mortgage Acquisition Corporation (JPMMAC), which made certain representations and warranties as to the quality of the mortgage loans in the Trust. We find that plaintiff’s interpretation of the language of the representations and warranty at issue is the only reasonable interpretation . . ..

The Pooling and Servicing Agreement represented and warranted that

“With respect to the period from [the] Whole Loan Sale Date to and including the Closing Date, [JPMMAC] hereby makes the representations and warranties contained in paragraph (a) . . . of Schedule 4 attached hereto . . . . [that] [t]he information set forth in the Mortgage Loan Schedule and the tape delivered by [WMC] to [JPMMAC] is true, correct and complete in all material respects.”

It also stated that if “JPMMAC breached a representation or warranty it made . . . it was to cure the breach within 90 days after notification; if it failed to do so, it was to repurchase the defective mortgage loan or substitute a qualifying loan for the defective one.” This is pretty standard stuff so far.

By 2012, it appeared that more than 40% of the mortgages remaining in the pool were delinquent and that the R&Ws had been violated. The certificate holders therefor demanded that JPMMAC repurchase the mortgages that were in breach of the R&Ws, which JPMMAC refused to do.

JPMorgan argued, against the plain language of the R&Ws, that it only covered defects that arose during a short period prior to the closing date of the securitization. The Court gave short shrift to this implausible reading of the R&Ws.

This opinion does not make new law, but one wonders what effect it will have on securitization business practices. R&Ws are driven by many things — concerns about credit risk, but also tax compliance with the REMIC rules, to name a couple.  I am curious as to how MBS R&Ws have changed since the early 2000s — and whether the parties to these transactions understand how R&Ws allocate risk among them.

Weaker Reps and Warranties on the Horizon

Inside Mortgage Finance highlighted a DBRS Presale Report for J.P. Morgan Mortgage Trust, Series 2014-IV3.  This securitization contains prime jumbo ARMS, some with interest only features. So, these are not plain vanilla mortgages.

The report raises some concerns about loosening standards in the residential mortgage-backed securities market, particularly relating to standards for the representations and warranties that securitizers make to investors in the securities:

Relatively Weak Representations and Warranties Framework. Compared with other post-crisis representations and warranties frameworks, this transaction employs a relatively weak standard, which includes materiality factors, the use of knowledge qualifiers, as well as sunset provisions that allow for certain representations to expire within three to six years after the closing date. The framework is perceived by DBRS to be weak and limiting as compared with the traditional lifetime representations and warranties standard in previous DBRS-rated securitizations. (4)

 DBRS noted, however, that there were various mitigating factors.  They included:
Representations and warranties for fraud involving multiple parties that collaborated in committing fraud with respect to multiple mortgage loans will not be allowed to sunset.

Underwriting and fraud (other than the above-described fraud) representations and warranties are only allowed to sunset if certain performance tests are satisfied. . . .

Third-party due diligence was conducted on 100% of the pool with satisfactory results, which mitigates the risk of future representations and warranties violations.

Automatic reviews on certain representations are triggered on any loan that becomes 120 days delinquent, any loan that has incurred a cumulative loss or any loan for which the servicers have stopped advancing funds.

Pentalpha Surveillance LLC (Pentalpha Surveillance) acts as breach reviewer (Reviewer) required to review any triggered loans for breaches of representations and warranties in accordance with predetermined procedures and processes. . . .

Notwithstanding the above,DBRS reduced the origination scores, assigned additional penalties and adjusted certain loan attributes based on third-party due diligence results in its analysis which resulted in higher loss expectations. (4-5)
All in all, this does not sound so terrible. But it is worth noting that the tight restrictions in the jumbo RMBS market appears to be loosening up. As the market cycles from fear to greed, as it always does, it is worth keeping track of each step that it takes toward greed. We can always hope to identify early on when it has taken one step too many.

Frannie Reps and Warranties Crisis Brewing

The Inspector General of the Federal Housing Finance Agency released an audit, FHFA’s Representation and Warranty Framework. Reps and warranties are a risk-shifting device that sophisticated commercial parties use in transactions. If a party makes a representation or warranty that turns out to be false, they other party may have some remedy — maybe the ability to return something or to get some kind of payment to make up for the failure to live up to the promise.

For instance, a lender may sell a bunch of mortgages to a securitizer and represent that all of the borrowers have FICO (credit) scores of 620 or higher. If it turns out that some of the borrowers had scores of less than 620, the securitizer may be able to make the lender buy back those mortgages pursuant to a rep and warranty clause. The IG undertook this audit because of recent changes to the reps and warranties framework for Fannie and Freddie. Before these changes were implemented,

the Enterprises’ risk management model primarily relied on reviewing loans for underwriting deficiencies after they defaulted as the representations and warranties were effective for the life of the loans. In contrast, the new framework transfers responsibility to the Enterprises to review loans upfront for eligible representation and warranty deficiencies that may trigger repurchase requests. If the Enterprises fail to do so within the applicable period, their ability to pursue a repurchase request expires if it is based upon a representation and warranty that qualifies for repurchase relief. (2)

The IG’s findings are disturbing. It “found that FHFA mandated a new framework despite significant unresolved operational risks to the Enterprises.” (3) It also found that

FHFA mandated a 36-month sunset period for representation and warranty relief without validating the Enterprises’ analysis or performing sufficient additional analysis to determine whether financial risks were appropriately balanced between the Enterprises and sellers. Freddie Mac, in contrast to Fannie Mae which provided analysis limited to a 36-month period, provided FHFA with the results of an internal analysis of loans that indicated loans with a 48-month clean payment history were significantly less likely to exhibit repurchaseable defects than loans with a 36-month clean payment history. Thus, losses to the Enterprise could be less with a longer sunset period. Therefore, FHFA cannot support that the sunset period selected does not unduly benefit sellers at the Enterprises’ expense. (3)

This is all very technical stuff, obviously. But it is of great significance. Basically, the IG is warning that the FHFA has not properly evaluated the credit risk posed by changes to the agreements that Fannie and Freddie enter into with the lenders who convey mortgages to them. It also implies that this new framework “unduly” benefits the lenders.

The FHFA’s response is unsettling — it effectively rejects the IG’s concern without providing a reasoned basis for doing so. Its express rationale for doing so is to avoid “adverse market effects” and because addressing the IG’s concern “may not align with the FHFA objective of increased lending to consumes . . ..” (32)

Whenever federal regulators place increased lending as a priority over safety and soundness, warning bells should start ringing. Crises at Fannie and Freddie (and the FHA, for that matter) begin with this kind of thinking. Increased lending may be important today, but it should not be done at the expense of safety and soundness tomorrow. We are too close to our last housing finance crisis to forget that lesson.

Consumer Protection in RMBS 3.0

The Structured Finance Industry Group has issued RMBS 3.0:  A Comprehensive Set of Proposed Industry Standards to Promote Growth in the Private Label Securities Market.  This “green paper,” frequently referred to as a First Edition, states that RMBS 3.0 is an initiative

established with the primary goal of re-invigorating the “private label” residential mortgage-backed securities (“RMBS”) market.

Initiated by members of SFIG, the project seeks to reduce substantive differences within current market practices through an open discussion among a broad cross-section of market participants. Where possible, participants seek to identify and agree upon best practices. RMBS 3.0 focuses on the following areas related to RMBS:

  • Representations and warranties, repurchase governance and other enforcement mechanisms;
  • Due diligence, disclosure and data issues; and
  • Roles and responsibilities of transaction parties and their communications with investors. (1 footnotes omitted)

RMBS 3.0 is expected to

1. Create standardization where possible, in a manner that reflects widely agreed upon best practices and procedures.
2. Clarify differences in alternative standards in a centralized and easily comprehendible manner to improve transparency across RMBS transactions.
3. Develop new solutions to the challenges that impede the emergence of a sustainable, scalable and fluid post-crisis RMBS market.
4. Draft or endorse model contractual provisions, or alternative “benchmark” structural approaches, where appropriate to reflect the foregoing.(2)

There is much of interest in this attempt at self-regulation by the now quiescent but formerly roaring private-label market. But I think that readers of this blog would be interested in its approach to consumer protection regulation. First, the green paper refers to it as “consumer compliance.” (See, e.g., 23) Unsurprisingly, the paper is only concerned with protecting industry participants from liability for violations of consumer protection/consumer compliance laws. It pays no lip service to the spirit of consumer protection — promoting sustainable credit on transparent terms. That’s fine given the constituents of the SFIG, but it only confirms the importance of active consumer protection regulators and enforcement agencies who will look beyond rote compliance with regulations. The private-label industry is capable of rapid change once it gets going, change that can outpace regulations. Someone has to keep an eye on it with an eye toward to the principles that should guide a fair market for consumer credit.

Ain’t Misrepresentin’

According to Wikipedia, the performers in the musical Ain’t Misbehavin’ “present an evening of rowdy, raunchy, and humorous songs that encapsulate the various moods of the era and reflect” a “view of life as a journey meant for pleasure and play.” In U.S. RMBS Roundtable: Arrangers And Investors Discuss The Role Of Representations And Warranties In U.S. RMBS Transactions, S&P does something similar with securitization. It presents the views of industry players as they try to predict and shape the future of the recently emerging private-label RMBS market, in the hopes of “achieving a healthy and sustainable RMBS market.” (2)

ACT I:  Lookin’ Good but Feelin’ Bad

The piece contains a lot of important insights, including the following point made by investors: “standardizing R&Ws would be a step towards improving the transparency and their ease of understanding. Smaller investors noted that they can be particularly limited in distinguishing R&Ws given the complexities involved.” (3)

This point encapsulates in so many words the classic market for lemons problem, famously formalized by George Akerlof.  The lemon problem leads us to ask how a buyer is to price a purchase where the buyer has less information about the product than the seller.  Because of this information assymetry, the purchaser will assume the worst about the product and offer to buy it with that in mind.

R&Ws are an attempt to overcome that problem because the RMBS arranger or the mortgage originator promises to compensate the investor for lemons that are contained with a mortgage pool securing an RMBS. Consistent with that view, investors noted that “they expected to be compensated for losses caused by origination defects, rather than legitimate life events.” (2) In other words, origination defects are the lemons that should be borne by the arranger/originator with its superior information about the mortgages. And “legitimate life events” represent the credit risk that the investors have signed up for.

ACT II:  That Ain’t Right

Arrangers and originators made the following points:

  1. [o]ne arranger indicated that the R&W process should be governed only by the contractual obligations negotiated for each deal. (2)
  2. [o]riginators have strict underwriting guidelines and said they take great care to follow those procedures before issuing a loan. Arrangers are also currently subjecting all or almost all loans to a third-party due-diligence review. (2)
  3. arrangers said that standardizing R&Ws will not be an easy task as differences between arrangers and product types will limit the degree to which R&Ws can be homogenized. (3)

These points clearly align with the interests of the seller in a market for lemons.  To restate them a bit, 1. caveat emptor; 2. arrangers and originators don’t sell lemons (!); and (3) it is too hard to come up with provisions that consistently protect investors so don’t bother trying.

ENCORE:  Find Out What They Like

S&P notes that there “was broad agreement that one of the keys to achieving a healthy and sustainable RMBS market is aligning the interests of arrangers and investors.” (2) From that broader perspective, S&P is right that the industry should work toward a state of affairs that “minimizes the cost of unknown risks and ultimately reduces losses and related litigation.” (2) Given the spate of lawsuits over reps and warranties, we had fallen shy of that mark in the past (here, for example).  It remains to be seen if the industry can get it better next time and if the incentives are aligned enough to do so.


Rakoff Rules for Monoline Insurer Against Securitizer

Judge Rakoff ruled for Assured Guaranty against Flagstar and awarded Assured more than $90 million.  The 103 page order is a pretty compelling read as it is a careful review of a battle of the experts and has lots of details about the loans in two Flagstar securitizations.  Some interesting bits:

  • the third-party due diligence providers, the Clayton Group and the Bohan Group, “were only expected to flag issues if something in the loan files appeared ‘ridiculous.’” (18)
  • the court found that many of the loans had “blatant” defects. (75)

Ultimately, the court found “that the loans underlying the Trusts here at issue pervasively breached Flagstar’s contractual representations and warranties.” (91)  The court’s findings may have unexpected relevance to other burning issues coming out of the financial crisis, such as whether these securitizations complied with the REMIC rules.