This post continues our discussion of the role of “independent” foreclosure review consultant Promontory Financial Group. Here we focus on what happened, or more important, didn’t happen in Promontory’s conduct of the reviews, and how that contrasts with the staggering fees the firm is widely believed to have earned.
David Einhorn has a saying about the companies that he has shorted: “No matter how bad you think it is, it’s worse.” Promontory’s and the OCC’s performance on the foreclosure reviews epitomize this dictum.
As much as the foreclosure reviews were widely expected to be a charade, given the conflicted roles of supposedly independent review firms like Promontory, there was no reason to expect the reviews to also be an ineptly managed, costly fiasco. Promontory’s inability to man and adequately oversee the project also meant what little useful work was accomplished was done overwhelmingly by contractors and temps, who were far enough removed not to have gotten the wink and nod that no borrower harm was to be found. So individuals who have no lasting loyalty to Promontory and Bank of America, and whose confidentiality agreements are likely inoperative due to defective drafting, public policy exceptions to these agreements, and enhanced whistleblower protection under Dodd Frank, have unearthed widespread evidence at all of Promontory’s clients on the independent foreclosure reviews of borrower harm. No wonder the OCC rode to the rescue to shut the reviews down.
It is hard to overstate how badly the project at Bank of America and other sites was operated. We have spoken to multiple contractors who worked for Promontory on its OCC foreclosure reviews for PNC (Promontory was acting as “independent” consultant to Bank of America, Wells Fargo, and PNC). They also have knowledge of the process at Bank of America because some contractors moved between the Bank of America and PNC engagements and compared notes with their colleagues.
Doug Smith, former McKinsey partner and co-author of the international best selling book, The Wisdom of Teams, reviewed their accounts. His reaction:
It’s hard to imagine a more unprofessional, atrociously managed effort. In my more than three decades of working with, observing and advising teams and projects, I cannot identify a single worse example.
But the surprise was how inept Promontory proved to be in its efforts to hide how much damage was done to borrowers. The scale of the exercise, combined with the pervasive reliance on contractors and temps who were told only the official story that their mission was to find evidence of damage done to borrowers, meant they went down that path, often surprisingly far. That produced further complications as Promontory then had to stymie and redirect their efforts.
Promontory was not even remotely up to handling this foreclosure review assignment, either from a competence or an operational standpoint. And this wasn’t simply due to the scale of the project at Bank of America. The whistleblowers who worked for Promontory on the considerably smaller engagement at PNC, present a picture of complete disorder. Moreover, some contractors went from PNC to Bank of America and they indicated that some pieces of the PNC engagement that had been organized by the contractors (as opposed to Promontory) were in better shape than the work at Bank of America.
One basic problem was that Promontory had no meaningful knowledge of mortgage securitization or servicing; if you look at its areas of expertise, there’s nothing close. That put it in the dangerous position of not knowing what it did not know, and also of being dependent on its client.* While that may not seem to be much of a problem if the name of the game is to find nothing, it turns out the OCC had unwittingly required that servicers like Bank of America make a serious-looking stab at it.
As we documented in detail in our earlier posts, the result was that the work of going through six of the seven substantive tests was performed on Bank of America premises with personnel under the control of Bank of America. Promontory did provide the software with the endlessly-revised questions that the personnel at Bank of America tried to answer, along with various information guides. It visited the staff in biggest center, Tampa Bay, only four times in thirteen months, and its interaction with the people doing the review work was extremely limited. In other words, this was not a Promontory foreclosure review, it was a Promontory-decorated Bank of America foreclosure review.
By contrast, the project at PNC was modest in scale, yet it proved be well beyond the managerial capabilities of Promontory. At a bank with a comparatively small servicing portfolio, Promontory put in place a team composed almost entirely of contractors (140 to 150 when staffed up) only one Promontory employee in a managerial role, the managing director on the project, Michael Joseph.** PNC hired even more contractors to do clerical work to support this team’s efforts.
Anyone who has worked in a real organization can appreciate how insane this arrangement was. One person cannot effectively lead 150 people, particularly on a customized project operating in several locations. The only professional firm activity that routinely has such extreme ration of partners to working oars is foreclosure mills. And there it is more viable, since the work in rocket dockets is routinized.
Predictably, the contractors (who were higher level than our earlier whistleblowers) describe a project in chaos. This contractor explained how no useful work was done for the first three to four months:
Consultant D: – essentially what I witnessed in the 10 to 12 months was the fact that Promontory did not manage the project. Their effort to manage the project with any real due diligence, to me, they just, they fell short from A to Z. For example, from the time I joined the project to the time it ended, I saw the leader of the PNC part of the project two times, and the total time was less than 10 minutes….
So, in any event, to address that question, very minimal management from Promontory. Essentially what we were, we were all contract people. I had seen from the bank’s perspective and the OCC’s perspective that, you know, maybe we weren’t qualified, we didn’t have the right skills, and there was a lot of back-and-forth about that,…What I’m trying to say is that the vast majority of the people I worked with as contractors, even the reviewer level people, were competent enough to get the job done. What I saw was that Promontory – they didn’t come to the table. [Details of the types of review work done by some of the contractors] So there was borrower harm in almost every occurrence.
Yves Smith: Right. Right….
CD: As we started that review, like I said, Promontory played very little role in helping us do that, so we were essentially left to our own, our own devices, and the bank had provided us a bunch of information. PNC was very open in the beginning…But because Promontory didn’t give us any guidance, we felt we were obligated to review all of these transactions, and obviously we were, you know, given the task of finding borrower harm….we would go out and do our own research online to find, you know, the different …
YS: Applicable regulations, yeah, exactly…
CD: And our MO was essentially, “Hey, we’re going to all treat it the same way and we’re going to all include it in borrower harm when we see this and we see that, and that way if at some point in the future when Promontory catches up to us” – because, again, at this time we’re giving Promontory the benefit of the doubt. We’re just too busy. We’re ahead of them. And we said, you know, “If we find out that this shouldn’t be borrower harm, or etcetera, it should be treated this way, then we’ll know that we all treated it consistently in our conclusions.” So that was the way we proceeded. And, you know, I have to tell you we were finding significant borrower harm. So as that unfolded…..
Well, as what I just described unfolded over several weeks, and then our results were then communicated to PNC, and immediately PNC, you know, their arms went up, their eyes got big, and they started to push back. “Wait a minute!” Their first, you know, I guess, exclamation was that, “Hey, you guys aren’t supposed to be looking at all of this stuff,” because again, remind you, Promontory didn’t give us any guidelines…Because we hadn’t been given those guidelines, again, we decided as a team that we would err on the side of the borrower and then we would get explanations and let the bank, you know, have their rebuttal period, etcetera.
CD: So once they saw what we were doing, you know, they’re like, “Wait a minute. You’re supposed to only be looking at [X], not the actual integrity of [Y].” And we said, “Well, you know, Promontory said we’re here to find borrower harm. They’ve given us no other guidance.” And when that conversation took place, everything stopped.
The detailed work that was done to support the tests at Bank of America, such as matrixes with state and Fannie/Freddie/FHA/VA fee limits and HAMP mod rules, was essential for PNC to do the work properly. It clearly couldn’t afford to reinvent that wheel. So why didn’t Promontory propose paying BofA a modest license fee to use that work? Both sides would have been better off and Promontory would have cultivated a bit of good will. But aforethough was not Promontory’s long suit. This came from Consultant B:
Well, there was – one thing I can tell you, generally speaking, the planning was piss poor. Piss poor. And when you have no planning whatsoever, you have chaos (laughs) until such time as people start to figure it out. And it took them four to five, six months to really get to the point where they were starting to figure out, well how are we going to do this, and about the time we got cranking then the whole question of independence came up and then we were going to have to trash everything we’d done and start all over again and design our own process without any interference from PNC.
After concluding that there were too many individual specialized pieces of a loan review to achieve consistency across the large number of reviewers, PNC pushed an attempt to break the reviews up into individual subsets that could address particular borrower harm issues, with the intent of bringing them all together at the end. That was the plan, but then they couldn’t figure out how they were going to bring all the subsets together at the end and gave up on that approach. Then, complaints as to “lack of independence” grew louder, and the OCC and Promontory were faced with junking what limited deliverables they had after 10 months of work, and starting all over with review procedures designed and blessed by Promontory alone. While that could have been done, the design stage was going to require a considerable amount of time, energy, and beta testing to get right.
Step back and understand what that section says. After 10 months, there was virtually nothing to show for this effort. Promontory had to junk what little it had done at PNC because the work to date was insufficiently “independent”. And in fact that is what happened. The work done through October 2012 was thrown out.
Not that that mattered to Promontory:
CD: We kept saying, you know, as we approached the end of the project, we kept – our confidence that Promontory was being truthful and was really going to come through with this stuff, was diminishing, obviously, over time. To the last month, in a meeting, I actually was in a meeting where it was called out once again, “When are we going to look at fee limits?” …The last comment to come from Michael Joseph, the lead of Promontory, was, “We’re not going there.”
CD: So he finally came –
CD: He actually said in the meeting, “We’re not going there.”
CD: I – you know, so that was, that’s when it solidified it for me, that this was all by design, they never had any intentions of getting the right answers.
Another reviewer stressed that the bending-over-to-the-bank posture came not just from Promontory but also the OCC:
While the general lack of “hands on” oversight and planning by Promontory contributed mightily to the failure of the project, Promontory was compromised from the get go by the OCC’s cultural bias toward keeping their “client” happy. Review process design decisions by Promotory had to be blessed first by the OCC and then by PNC.
Obviously, any “independent” bank review that give the bank the final say is fundamentally corrupt.
So How Much did Promontory Rip Out of Bank of America?
One way to appreciate what a Guinness-record-level project management disaster the Bank of America review was is to try to understand how the fees charged related to any sort of discernible work product. The costs reported on how much the banks spent on the reviews are so stratospheric that most observers simply go into My Eyes Glaze Over mode. That serves to gloss over a very ugly fact: when you look at Bank of America, where the bank itself did the overwhelming majority of actual review work on its own nickel, with its own staff and temps, the amount that Promonotory is rumored to have charged is so excessive that it raises questions of what exactly was being done on behalf of the bank.
We’re going to identify what we call “dark matter”, the magnitude of probable billings that are beyond any generous estimate of legitimate activity. Due to the extensive whistleblower reports, we can do a pretty decent job of estimating what the work at Bank of America (the part it paid for) cost, and what hefty estimates of the Promontory costs directly related to that activity should have amounted to. We can also make a good stab at what another major undertaking that was included in the foreclosure reviews ought to have cost, based on Promontory’s own statements as to how long it was taking. We can then use those to show how much dark matter remains.
Admittedly, Promontory has never ‘fessed up to how much it billed to any of its clients on these reviews. While we could not definitively confirm the total, several informed sources indicated that the gross fees Promontory earned across its three clients, Bank of America, Wells Fargo, and PNC, were roughly $1 billion. Because we have good information on the PNC team level and what the contractor’s rates were, it’s safe to say that that Promontory’s charges to PNC were probably not over $60 million. The Wells Fargo reviews were contemplated in Promontory’s engagement letter to be somewhat smaller than the Bank of America effort. Whistleblower accounts indicate they were almost certainly smaller, since the bank was more stringent about reviewing borrower complaint letters in such a way to make it hard for any to be candidates for a real investigation (for instance, if a borrower failed to provide dates of particular incidents, the letter was rejected). So it’s reasonable to assume that Promontory earned $500 million in gross fees from Bank of America.
Recall, as we have stressed, that the six of the seven substantive tests, B through G, were performed at Bank of America sites under the bank’s control.*** Promontory performed test H, the determination of harm. Preliminary tests, which allowed it to exclude some borrowers, were also performed by Bank of America. As we’ll discuss below, Promontory helped design and provided critical components for this process, such as the frequently-revised CaseTracker program and the related information sources, such as “matrixes” of state fee limits. It is also the likely source of the “resource guides” that helped the claim reviewers understand where to look to answer questions. It also over time resolved the considerable ambiguities, gaps, and errors in these components.
Thank to extensive input from our seven Bank of America whistleblowers, we have been able to construct the staffing and pay levels at the all of the Bank of America reviews sites. We’ve set out the assumptions in Appendix III at the end of this post. Please note that our assumptions are consistently “conservative” which also means “generous”. For instance, at all the sites, temps were added over time. The later temps were brought on at lower hourly rates than the earlier temps. We’ve used the hourly rates of the early hires.
Similarly, while most of the claim reviewers and related personnel were temps, there were also Bank of America full time employees in the same roles. The full time employees all in (even when you allow for benefits) were less costly than the temps’ hourly costs plus assumed agency markups. So for simplicity’s sake, we’ve assumed temp pay levels across the board, another generous assumption.
The part that is a bit tricker to estimate is the higher level supervision costs, legitimate start-up and wind down costs, and the cost of the special information tools, which were essential but were clearly not contemplated in the engagement letter. Our assumptions include what securitization experts see as an extremely generous allowance for these components ($3 million, when they estimate a securitization law firm would have charged $500,000 plus a premium for ongoing hand-holding) as well as an allowance for off site supervision in excess of the proportion provided for in the Promontory engagement letters. If you look at Appendix III below to see the salary and billing rates assumed for these personnel, you’ll see that they are also generous.
What we get is:
$90 million of expenses at Bank of America foreclosure review sites
$5.5 million of Bank of America executive level involvement in project design, oversight, and wind-down
$22.6 million of Promontory billings for staff
$3 million additional Promontory billings for possible third party expenses for development of fee matrixes
Total: $121.1 million, of which $25.6 million was Promontory billings
Recall that even though this process was extremely inefficient (the earlier hires report periods of one to seven months, depending on what test they were assigned, or performing no useful work), by August loans were being reviewed in a fairly orderly manner, even though the system was also designed to suppress finding of harm.
For this part of the project, the test that was clearly performed by Promontory was test H, the determination of harm. We’ll return to that in a bit.
Promontory’s other major task was to review a sample of 35,000 foreclosed loans separate from the reviews requested by borrowers. When the foreclosure reviews were abruptly halted, media sources claimed the banks were only 1/3 of the way through the process. We’ll be generous and assume Promontory had gotten through half of the 35,000 loans. We’ll also assume it had a learning curve, so it wasted 50% of its time figuring out how to review the loans. So that brings us to the equivalent of 3/4 of the loans, or 26,250.
Promontory told the Tampa Bay staff it took them only 8 hours to review a loan. That is remarkable, since just doing a proper fee review took a trained Tampa Bay staffer at least 10 hours. For 10 file reviewers, there was also one Quality assurance staffer and one “Senior File Reviewer”. There was also a 7.5% senior managers for every file reviewer. We’ve omitted a few roles to keep this simple; we’ll compensate by increasing our result by 5%.
The highest figure ever reported in the press for file reviewers was $250 an hour (the contractors on PNC were paid $60 an hour and believe they were billed out at $150 an hour, so again, our assumption is generous). We’ll assume the Senior File reviewers and Quality assurance types billed at a 30% premium to the file reviewer rate or $325 an hour. We’ll assume the higher level managers were a mix of top billers ($650) and more not quite as expensive people ($400-$500 an hour) for an average rate of $500. So for every $250 an hour file reviewer, you also have .4 x $325 plus 0.75 x $500 or $417.5. You then have the 5% for ancillary people plus 15% for expenses, or $501. So call it $500. In other words, even that $250 per hour per file reviewer has all sorts of project overhead and expenses piled on top of him.
We take that 26,250 loans time the fully loaded cost of a file reviewer of $500 per hour times eight hours. That takes us to $105 million.
Now go back to that $121.1 million figure before for tests A through G. Of that, only $25.6 million was actual Promontory billings. So far, we have accounted for
$95.5 million spent by Bank of America on activities performed under its control
$130.6 million of generous but defensible Promontory billings
Compare that $130.6 million to the estimate of $500 million billed.
How can Promontory justify nearly $370 million of dark matter?
$370 million is an utterly implausible amount for the mystery test H.
There is at least one troubling way to fill the gap. Readers may recall we interviewed a firm called SolomonEdwards that advertised that it was in the “file scrubbing” business. We called one of the partners to understand what those services were about. He made clear, in his coded way, that his firm engages in document fabrication, specifically allonge fabrication (allonges are falsified to remedy the failure of the originators to transfer notes properly to the foreclosing party.****). They also do other types of document “remediation”.
At the time of our conversation, March 2012, he said he had 600 people at a bank doing “OCC consent order work”. The partner made it sound as if his firm was engaged directly by his client. But the story may be more complicated.
That bank is almost assuredly Bank of America. At PNC, some of the contractors were recruited through a company that played a “staffing” role, Hilltop, which means it acted as a body shop to Promontory. Others were hired directly by Promontory. One of the PNC consultants was also approached by SolomonEdwards to work at Bank of America under a similar “staffing” arrangement and said that 975 Solomon Edwards people were working at Bank of America. That is confirmed indirectly by a 63 member “alumni of Solomon Edwards B of A project” group at LinkedIn. So it seems clear there was a large group of contractors working through SolomonEdwards at Bank of America.
Now on the one hand, this may simply mean that SolomonEdwards was doing file review and “remediation” with a 600+ person team and on side was also acting as a body shop for Promontory. However, regardless, Promontory did bother getting Hilltop cleared for conflicts in with the OCC (Hilltop was mentioned in the PNC engagement letter) but not SolomonEdwards (in other words, neither firm was cleared for conflicts on the Bank of America conflicts by the OCC, even though Promontory did mention the names of each firm on other engagements, Hilltop for PNC and SolomonEdwards for Well Fargo. It would thus seem a matter of good order to have cleared them both for Bank of America, but that appears not to have happened).
And even if most of the 975 SolomonEdwards employees were working as contractors to Promontory, we still are in the dark as to what they could possibly have been doing, particularly since we believe their billing rates were lower than the levels we’ve assumed in our estimates above.
This is a long way of saying that even when you make allowance for high billing rates, the fact that so much of the real review “work” was actually done by Bank of America on its nickel means that there is a tremendous amount of dark matter in the Promontory billings. On the one hand, this may simply be egregious incompetence; there was certainly plenty of that on display at the far more manageable PNC project. Even so, it is extremely difficult to explain the numbers that credible sources (including people in senior oversight capacities in a position to get good intelligence) believe the Bank of America charges were with any plausible level of review related work. This strongly suggests that Promontory, either in concert with or separately, may have been aware of and was potentially involved in various file “remediation” and records-doctoring activities.
Keep in mind that we’ve provided account from the Tampa Bay contractors of document fabrication of multiple types as well as of deletions from servicer notes that appeared to be to remove incriminating evidence. So there seems to be a very strong likelihood that crimes were committed.***** The only open questions are how wide ranging this activity was, who carried it out, who else was an accessory, and how much were they paid.
*One proof: the tool used by body shops to qualify candidates for Promontory work looked for expertise in mortgage underwriting and servicing. Underwriting experience is worth very little for analyzing foreclosure issues. Similarly, having servicing people review foreclosure issues is wrongheaded. While they might have experience in the area, the whole point of the foreclosure reviews was to analyze how servicers had managed the loans. It certainly isn’t independent or a “review” to have the people check to see if their own type of work was legitimate or not. Lawyers should have been reviewing these issues.
**There were “staff attorneys” and IT people tasked to the project in Miamisburg, Ohio, the biggest site, who were also Promontory employees, but Michael Joseph was the only Promontory employee in a managerial role in the foreclosure review process. And it was light on supervision of any kind. The “underling” managers were contractors, one of which had had OCC experience, the other had worked at Countrywide.
*** In the unlikely event Promontory or the Bank of America rouses itself to protest that Promontory did perform tests B though G, we’ve provided not only detailed whistleblower reports but also documents used in the reviews that demonstrate otherwise. See posts Part IIIA and Part IIIB
**** Specifically, allonges are separate pieces of paper that are required per the Uniform Commercial Code to be “affixed” to the borrower note. They are used to create space for additional signatures when room on the note has run out. Allonges were simply unheard of prior to 2008, since it is permissible to use the sides and back of a note for signatures (notes are endorsed like a check to convey ownership to the next party [this is a simplified but adequate for this discussion]). There is no need for them in the ordinary course of business, but foreclosures have long ago ceased to resemble a legitimate business. Allonges, which remember are supposed to be affixed, have this funny way of appearing out of nowhere when a borrower about to lose his house points out that the note was never properly endorsed to the party that is trying to foreclose, and only the proper party can foreclose.
***** For starters:
18 U.S.C. § 1512(c) (as amended by Sarbanes-Oxley Act § 1102) (offense encompasses any person who corruptly “alters, destroys, mutilates, or conceals a record, document, or other object, or attempts to do so, with the intent to impair the object’s integrity or availability for use in an official proceeding”; or corruptly “otherwise obstructs, influences, or impedes any official proceeding, or attempts to do so.”).
18 U.S.C.§ 1519 (created by Sarbanes-Oxley Act § 802(a)) (offense encompasses any person who “knowingly alters, destroys, mutilates, conceals, covers up, falsifies, or makes a false entry in any record, document, or tangible object with the intent to impede, obstruct, or influence the investigation or proper administration of any matter within the jurisdiction of any department or agency of the United States or any case filed under title 11, or in relation to or contemplation of any such matter or case . . . .”).
See United States v. Trauger, No. 3 03330 (N.D. Cal. filed Sept. 25, 2003). In this complaint, the government alleges that the lead audit partner of a public company concealed the alteration of computer and hard copies of audit workpapers, as originally produced to Office of Comptroller of Currency (OCC) and then produced to the SEC, because he wanted to “beef up” the files to make it appear that he had thoroughly considered the accounting issues and available facts during the course of the audits and reviews. The partner allegedly found out how he could de-archive an audit in order to revise and then re-archive the working papers; directed a colleague to backdate the system date on his computer in order to make it appear that printed copies of the altered working papers had been created during the course of the audit; added an undated handwritten note to an original and all copies of a memorandum; and directed that the e-mail evidence of the requests to alter and delete documents itself be deleted.
Lisa Epstein worked on the Excel model.