Reader Lisa N. pointed me to a troubling by SolomonEdwardsGroup, a company that describes itself as a “national financial services consulting and staffing firm” about its remediation services for “significant loan documentation problems.” Alert readers will recognize that this is shortly after the robosiging scandal broke.
Here are the key parts of the press release:
SEG’s teams can also be rapidly deployed across the U.S., to help banks and servicers “scrub” files and determine which foreclosures may have been tainted by incorrect loan documentation and processing issues such as robo-signing….
For instance on a recent engagement, SEG quickly deployed a 25-person team to review a single-family loan portfolio containing 5,000 loans and within six weeks brought the portfolio into compliance with investor guidelines. During another recent engagement, SEG successfully completed the same type of project involving 20,000 single-family loans tainted by fraud allegations.
Needless to say, this sounds consistent to the charges we’ve heard from borrower attorneys and have even seen at trial: that of “tah dah” documents appearing suddenly in court that solved all the problems with the evidence presented. A not that unusual case occurred last week, in Kings County, New York, where in , when the lawyers for the bank tried submitting two notes (borrower IOUs), the second attempting to remedy problems raised by the first one, each presented as the original. The judge not only ruled against the foreclosure but referred the case to the district attorney and the state attorney general.
Why the Failure to Convey Notes and Make Assignments Properly is Such a Big Deal in Mortgage Securitizations
Advanced students of “securitization fail” can skip to the next bold heading.
As we discussed in prior posts, there is considerable evidence of a widespread, perhaps pervasive, failure among the parties to mortgage securitizations to adhere to the terms of the contracts that created these deals. Specifically, they were required to transfer the notes (the borrower IOU) through multiple parties and get them to the securitization trust by a specified date. This process was laborious because each time, the note had to be signed (the term of art is “endorsed”) and the mortgage assigned (which confusingly is the lien against the home, although both professionals and laypeople often refer to the note + the mortgage, which are actually two separate instruments, as the mortgage).
While in the early days of the securitization industry, in he 1980s, it appears these procedures were adhered to, there is considerable evidence that they broke down over time.
These deviations are serious because the agreements that govern these deals, called pooling and servicing agreements, were carefully crafted to satisfy a number of legal requirements, including securities law, local real estate law, tax law (REMIC, for Real Estate Mortgage Investment Conduit, set forth as part of the 1986 Tax Reform Act), the Uniform Commercial Code, and trust law.
The PSA requires the note (the borrower IOU) to be endorsed (just like a check, signed by one party over to the next), showing the full chain of title. The minimum conveyance chain in recent vintage transactions is A (originator) => B (sponsor) => C (depositor) => D (trust).
The correct conveyance of the note is crucial, since the mortgage, which is actually the lien (this is often a cause of confusion, since in lay usage, “mortgage” refers to the the note + the lien, when they are separate instruments), is a mere accessory to the note. The lien can be enforced only by the proper note holder (the legalese is “real party of interest”). The investors in the mortgage securitization relied upon certifications by the trustee for the trust at and post closing that the trust did indeed have the assets that the investors were told it possessed.
The PSA also very clearly provided for an unbroken chain of assignments and transfers though the parties (the A-B-C-D or more cited above). The use of intermediary parties between the originator and the trust, with a “true sale” occurring at each step, was intended to create FDIC and bankruptcy remoteness. The investors (who are called the certificate holders in the PSA) did not want a creditor of a bankrupt originator to be able to seize notes back out of the trust.
Some PSAs allowed for each party to endorse in blank (as in each owner simply had to have an authorized party sign it), but the note still had to have endorsements by all the parties in the conveyance chain, while others stipulated that each endorsement had to be to the next party in the chain. However per NY trust law (and New York law was chosen in the vast majority of cases to govern the trust), the final endorsement had to be to the trust, not in blank.
The last bit, and this is the source of considerable tsuris, is that all the notes had to be conveyed to the trust by a date certain, usually 90 days after the closing of the deal or after an aggregation period. Only very limited exceptions were permitted. The reason was to conform with REMIC rules. As indicated, the overwhelming majority of trusts elected New York law to govern the trust because it is very well settled. But New York trust law is also unforgiving. Trusts can operate only as stipulated; any move that deviates from its instructions in its governing documents is deemed to be a “void act” and has no legal force.
Law professors Anna Gelpern and Adam Levitin have described PSAs not just as prototypical rigid contracts, but as Normally, if there is a significant deviation from a contract, the parties can use waivers, sometimes with penalties if one party looks to have behaved badly, to remedy the breach. But the use of New York trusts, the REMIC requirement of passivity, the way the relationship among the parties was set up (bad incentives for the servicers, demotivated trustees because they are indemnified by the servicer and have no reason to watch out of the investors, formal consent requirements among dispersed investors, when they sometimes have conflicting interests) make changing the PSA well nigh impossible.
To put it more simply: if notes were not conveyed to the trust in the stipulated time frame, it means that someone other than the trust has the right to foreclose, presumably one of the parties earlier in the securitization chain. But no one is willing to admit that since it would mean that investors had been sold what Adam Levitin has called “non-mortage backed securities.” There is no clean way for a party earlier in the securitization chain to foreclose and transfer the proceeds to the trust. So the trust HAS to be the one to make the foreclosure, at least in the minds of everyone involved with these deals, whether it actually is the right party or not.
How SolomonEdwards “Remediates” Problem Loans
I called SolomonEdwards to discuss its press release about “scrubbing” loan files and had two conversations totaling over 50 minutes with a partner in this business. What was disconcerting about this discussion what that despite his emphasis on how thorough SolomonEdwards is in inspecting loan files (its software allows it to flag hundreds of items on a file review) and how strict it is in managing conflicts (it has over 600 people working on OCC consent orders for a single bank but would never take an OCC engagement that would put it in the position of having to review its own work, which as Abigail Field, and have stressed, is taking place frequently), he seemed remarkably unaware of the differences between how you can handle a loan that a bank owns versus one that was supposed to be transferred to a trust pursuant to a PSA. When I asked specifically about whether their process was different for securitized loans versus bank owned loans, he said that there was not a great deal of differences. The partner did not refer to any of the issues discussed at length above, but instead mentioned investor guidelines and who the investors were. So it seems that they check the loan files to see whether they conformed with the representations and warranties made in the PSA (for instance, they also flag as irremediable all loans with interest rates that would have them be deemed predatory) but not the transfer and custody requirements.
SolomonEdwards apparently did and continues to do a lot of FDIC-related work, and my understanding is that a lot of procedures were developed to deal with the to transfer mortgages out of banks that had failed. Those are perfectly kosher with a non-securitized loan but you just can’t do that in a PSA context. For instance, many of these actions are tantamount to trying to transfer a defaulted loan into a REMIC trust. Not only is that a void act under New York trust law (it was never contemplated, hence the trust can’t do that) but also under REMIC, a trust cannot accept a non-performing asset (which is exactly what a dud loan is).
In fact, the SolomonEdwards conversations confirmed what we have inferred about the widespread failure of originators to convey notes to trusts properly. When I asked how they start a file review, the partner took time to stress that they often didn’t start with file reviews, that at the big volume originators, it was often a big process just to find where the loans were; “They don’t know where it is…they all have significant problems.” That should simply not be the case with a securitized loan. It is supposed to be with the trustee, either held by them or a custodian they hired (WaMu and Chase deals are an exception, they do allow for the originator to hold the notes). Trustees provided multiple certifications to the SEC that they DID have all the loans.
Similarly, while he did say that there were some loan files that could not be remediated, such as if the note was missing and the seller/servicer was bankrupt and “no longer exists.” (in fact, the example he used, Thornburg, is still around). The methods he said they used for remediation were troubling. For instance, he said if the seller/servicer were dead, they would not be able to get a replacement note (he referred to “replacement” several times, that if you went back through the assignments or the registrations in MERS to the various parties and get a “certified” copy). Notes are like checks, they are negotiable instruments. You can’t enforce a copy or use a copy to try to recreate an original. This is exactly the sort of activity that got the notorious DocX shuttered. Yet he seemed to think the use of a copy or a “replacement” adequate. But you can’t “replace” a note; it’s an original, and you need to have the borrower’s signature for it to be binding, and I can guarantee no one is getting borrowers to sign replacement notes.
Similarly, one thing that foreclosure defense attorneys have seen as a huge red flag of servicer chicanery is the use of allonges. An allonge is a separate piece of paper used for endorsements that is required by the Uniform Commercial Code to be “affixed” to the note and used for endorsements when there is no more space left on the note for signatures. Allonges were pretty much never seen until the robosigning scandal, since all the space on a note (meaning the back and the margins) can be used for endorsements.
But SolomonEdwards official said that they’ve been able to get copies of the note from the seller and have been able to “bring them forward with allonges that were re-executed.” When asked, he confirmed that they create allonges now that confirm with the transfers that they’ve found ought to have taken place, either via the PSA, MERS, or other routes. Again, in a securitized trust, that it tantamount to trying to transfer the note now and is not valid. When I pressed him on how they did that, how they got signatures from intermediary parties, he demurred and said, “I don’t want to give away too much of our secret sauce.”
He also discussed using lost note affidavits. That is permissible only on an exception basis; indeed, many deals limited how many lost note affidavits could be used. If a firm like SolomonEdwards is seeing more than a couple of missing notes on a deal, that means transfers did not happen and there is a much more fundamental problem with the securitization, potentially a contract formation failure (if no notes were transferred by the cutoff date, the trust was not formed).
The SolomonEdwards executive also made it clear that he regarded the mortgage assignments as more important than the note, which is backwards (the lien follows the note) and that they spent more time on getting them executed. He said that his firm found “missing” mortgage assignments (“they can’t be found”) to be common. Again, since the assignments had to be completed by the cutoff date, that means they are either making obviously invalid assignments, are deliberately making back-dated assignments (not kosher) or have a time machine. Yet he said there were “potentially some things that could be done with the MERS system” or “going back” and “rebuilding a title chain” to remedy these failings. We also asked if they had ever been the professional deponent on a foreclosure, since testimony is supposed to be provided by a party with personal knowledge (hearsay is not permitted) and if they were the ones who had remediated the files, they would presumably be required to testify if the borrower challenged the authenticity of documents submitted to the court. He said that had never come up.
In fact, these reviews sound like documentation theater. The partner stressed how through SolomonEdwards was and how they had software that allowed them to record up data items and capture whether a item was material or not material and then risk rate an entire loan file. They can look at up to 12000 variants (no typo) for the OCC reviews (how many they actually look at depends on the scope of the client engagement; the difference between the number of steps, as he called them, in the OCC reviews versus the typical bank engagement is because the OCC reviews include state law requirements. Needless to say, it’s a bit curious that routine forensic investigations do not include state law matters). He also stressed that they have senior teams working on these projects, 5 years average experience for the OCC work, more than that on bank work, and that on a normal engagement, they would typically spend 3 hours per file, but if a bank had serious documentation problems, it might take as long as 12 hours.
He said that a typical bank engagement would require looking at 100 to 150 items. For a 3 hour process, that’s less than two minutes an item (and remember, that includes the time to log their findings). But the reality is that there are really only 5-10 things you need to look at: Do you have an original note? Does it have all the endorsements that the PSA says it should have? Do the mortgage assignments correspond to the endorsements? Were they all completed on time?
These multi-hour investigations are fee-padding form over substance. But this sort of thing is perfect for the bank-defending OCC, since it would take someone pretty expert to penetrate the fiction that this exercise in counting trees was designed to miss the forest.
It was disconcerting to speak to someone who obviously thinks his firm is highly professional engaged in activities that include document fabrication, which is what creating allonges now amounts to. And the worst is I have no doubt SolomonEdwards is more careful than most firms in the industry. This confirms, as we have said repeatedly, that there was a massive failure in the industry to conform to the requirement of the legal agreements that it devised. And there is a very big business, now with a government seal of approval, in covering up that fact.