Whistleblower: Wells Fargo Fabricated and Altered Mortgage Documents on a Mass Basis

Over the last two and a half years, Wells Fargo, like most of the major mortgage servicers, claimed that it had a “rigorous system” to insure that mortgage documents were accurate and complete. The reason this mattered was that there was significant evidence to the contrary. Foreclosure defense attorneys found repeatedly that, for securitized mortgages, the servicer or foreclosure mill attorney would present documents to the court that failed to show the borrower’s note (a promissory note) had been transferred properly to the trust. This mattered not only on a borrower level, but indicated that originators of the mortgage securitizations hadn’t bothered transferring the notes properly to the trusts that were to hold them. This raised the ugly specter of what was called “securitization fail,” that investors had been sold securities that they had been told were mortgage backed when they might in practice not be.

The robosiging scandal was merely the tip of the iceberg of mortgage and foreclosure problems that resulted from the failure to adhere to the requirements of well-settled state real estate law. The banks maintained that there was nothing wrong with mortgage ownership or with the records. All they had were occasional errors and some unfortunate corners-cutting with affidavits. If they merely re-executed all those robosigned documents, all would be well.

Wells Fargo’s own actions say the reverse. It has been doctoring documents in house for over fifteen months for borrowers who are targeted for foreclosure. It was having this sort of work done outside the bank for an unknown period of time prior to that.

A contractor who worked at a Wells Fargo facility in Minnesota reports that the bank engaged in systematic, large scale alteration of mortgage notes and fabrication of related documents in preparation for foreclosure. The procedures the bank used are questionable for a large portion of the mortgages.

A team of roughly 100 temps divided across two shifts would review borrower notes (the IOU) to see whether they met a set of requirements the bank set up. Any that did not pass (and notes in securitized trusts were almost always failed) went to another unit in the same facility. They would later come back to the review team to check if the fixes and fabrications had been done correctly.

Not only is having Wells Fargo tamper with documents in this way dubious in many cases (more detail on that shortly), but amusingly, the bank does not even appear to be terribly competent at this sort of falsification. The bank changed procedures frequently, and did not go back to redo its prior work. In addition, it regularly took loans that appear to have been endorsed properly and changed them as well. Finally, even if the procedures had been proper, the temps were required to meet such aggressive production timetables and were so laxly supervised that it seems unlikely that their work was done well.

This account confirms what foreclosure defense attorneys have reported for some time: that servicers have been engaging in document fabrication for some time. It’s not uncommon for a servicer or foreclosure mill to present “tah dah” documents that miraculously remedy the problems that homeowner attorneys have raised, sometimes resulting in clear proof of fabrication, like two different notes (borrower IOUs) having been presented to the court, each supposedly an original.

But what is striking about this practice is both the brazenness and the scale. Our source was told that Wells Fargo added a second shift to its mortgage review operation in November 2011 [update: it is likely the related doctoring activities were increased correspondingly]; he* did not know when it had been established. Bank employees claimed that some of these operations had formerly been done by outside firms and the cost of doing it in-house was much lower than the cost of doing it externally. Apparently having plausible deniability was too expensive.

We sought comment from Wells Fargo on these allegations and they declined to respond.

Description of Mortgage Doctoring Operations

The document fixing took place at 1000 Blue Gentian Road in Eagan, Minnesota, which the whistleblower described as an enormous facility, and ironically, one at which one of the 9/11 hijackers received flight training.

The whistleblower worked with a team of 50-60 temps, one of the two shifts involved in checking documents before and after the “corrections” were made. The temps came via agencies, were required to have a college degree and pass a security clearance, and were paid roughly $13.00 to $14.50 an hour for eight hours (seven hours of work + breaks). The whistleblower said very few people (under 20%) had prior experience with mortgage documentation. Since Wells has a long-standing practice of promoting temps into permanent positions, the workers had a strong incentive to perform well. Our source worked for the bank for nine months.

His unit would review mortgage documents of borrowers who were described as “in foreclosure” which he understood in practice meant they were delinquent but the foreclosure has not not been initiated. When our source arrived (spring 2012), they were in the process of doubling the work capacity of this effort. Wells Fargo beefed up in the wake of the state attorney general/Federal mortgage settlement of early 2012, evidently seeing it as a green light for more aggressive and systematic document fixing.

This team had two tasks. The first was to review documents that were delivered periodically (often daily) to make sure they were in order. The part we’ll focus on is that they would check the notes to see if the endorsements matched up against what the bank wanted them to look like. (Regular readers of this blog will recall that mortgage notes are endorsed to convey ownership, and in foreclosures, attorneys often challenge the foreclosure if the borrower note does not show a complete and unbroken chain of endorsements to the party initiating the foreclosure). The whistleblower estimated that 99.5% of the notes that he reviewed that had been securitized failed the bank’s tests, and roughly 10% to 15% of the bank owned mortgages were tagged as “fails”.

Mortgage notes that failed this review were sent to a neighboring section. Weeks later, they would come back to the same section with the corrections made, either in the form of new endorsements made to the note, or the addition of an allonge. An allonge is a separate piece of paper, attached (“affixed”) to a negotiable instrument so that more signatures can be added. They were virtually unheard of prior to the robosigining scandal, since in the normal course of business, there would be no reason to use an allonge (the margins and back of a note can be used for signatures). The people in his unit were then to check that this doctoring had been done correctly.

The work environment had a peculiar combination of regimentation and chaos. The temps were given instructions that kept changing and were inconsistent over time (and remember, this worker joined after the state/Federal mortgage settlement was final):

This was a document processing facility where we would go through the files that were already in the foreclosure pipeline, as decided by somebody else, so we would kind of source and classify each file according to, you know, various criteria. First of all, just make sure they’ve got all the parts, like the note and the mortgage and the title policy, and if they’ve got all those and they matched, then see if they’ve got the right information on them, the priority being on the, you know, the final endorsement on the note…

One of the points I was going to make was, when we originally started, the protocol was very distinct for one as opposed to the other. And then rapidly states were passing laws, is what we were told, to change it, so that the number of OD {original document] states being fewer and fewer. Then after the second and third decree there was no distinction anymore. And it seemed like we were supposed to have original documents for everything at that point. So actually a lot of my impression is that there were several things that were a little strange that changed as some of these decrees went through. So like, that’s the second one I was going to mention, is when we were first trained, the way that you treated a standard loan file and a securitized loan file were very, very different, and there was a fairly strict protocol. You had to have a continuous chain of endorsement, had to have a final endorsement to Wells Fargo or one of its affiliates, for a note to pass. But, if it was securitized, you went to this LPS database called CPI, and there would be a list of, you know, however many people had once claimed to own this file, this note. And all of a sudden the continuous chain of endorsement rule went away and you didn’t necessarily use the last one, you would just pick one out of the list that matches your last endorsement and that was good enough.

You can see how irregular this procedure was. Notice how the bank went from having the view that fewer and fewer states required a review and correction of original documents, then reversing itself and deciding all did.

Similalry, if the temps were instructed to match a note to any listed party they could find on a Lender Processing Services database (which relies on manually input data and is thus not reliable), and it was not the final party, that means they are constructing a chain of title that is at odds with the bank’s own touted system of records. If the bank were serious about even getting its fixes right, for securitized loans, it would go to the pooling & servicing agreement and see what it stipulated as the chain of title and work from there. [Update: our source clarified upon seeing the post that once the were given only actual mortgage notes to work with, they were instructed to look for a complete chain of endorsements. That’s an improvement over the previous process, but not necessarily sufficient. This is now playing on the lack of patience of judges in understanding how elaborately lawyered-up securitizations were supposed to work. A complete-looking chain might not be the proper or complete conveyance chain as set forth in the relevant PSA. This is basically looking to see if the documents look internally plausible enough to pass muster with most judges, rather than doing it correctly].

It is important to point out that it perfectly OK for the bank to transfer notes it owns (loans owned by Wells Fargo entities, including banks it acquired) any time it wants to prior to foreclosure. Where this gets dodgy is on the securitized loans. These loans were supposed to have been transferred to the securitization trust, through a series of intermediary parties, with a complete and unbroken chain of endorsements on each borrower note. These transfers were to have been completed by a specified cut-off date, with a limited period of time after that for any document clean-up. The trustees on these deals provided multiple certifications to the effect that they had the notes in good order (which would mean the trust properly owned them, that is, all the transfers had been completed as reflected, among other things, in the note being endorsed correctly).

The fact that Wells Fargo is dorking with documents on a mass basis at this late state is an indication of how little of the work that the mortgage industrial complex has kept insisting was done correctly was done at all.

And this was a high-volume operation. Back to our source:

There was a big board that would have inventory in and out for each shift on each day, but that is a little fuzzy. My recollection is that we could move anywhere between 5,000 and 11,000 files a day. A really slow day would be 3,000 for our shift and people might have to go home early. That happened a couple days a week for several weeks the last few months I was there. We generally measured the shift inventory in bins. We would have just a few bins on a slow day, but on a typical busy day there would be 25 to 35 bins full of files to go through.

I’m getting fuzzy on what our hourly targets were. For electronic files I believe we were supposed to do at least 35 or more an hour. I also remember the number 55. I can’t remember if that was a target or not. With paper files I believe we were supposed to do at least 25 an hour, although after two or three months there wasn’t so much discussion of volume and the focus was mostly on accuracy. There were many who did more than this.

These targets don’t seem to square with the daily final tallies I remember people putting in which ranged from 45-130 per person per shift. There were people who would double the target and people who were fairly below it.

Let’s take the midpoint of his 45-130 files a shift range, which is 87.5. They worked 7 hour per shift. That’s under 5 minutes a file. That is to check not only that all the basic documents were there, but also to go into the CPI database, and possibly also into a backup spread sheet if the desired information was not in CPI, and look for a match.

The objective was to have the final endorsement be “to blank” or what is more typically described as “in blank”. The whsileblower gave this example of how a note was supposed to look once it was corrected:

I was checking to see if whoever had written out the new endorsements really had copied what was in CPI word for word, letter for letter. After checking the first couple with increased scrutiny, it became clear that they had copied them absolutely verbatim, only in a new endorsement to blank.

Before it would be:

Pay to the Order of

Bear Stearns Trust, Pass through certificate holders 2003, VII.

Without Recourse
U.S. Bank

Joe Blow,
Vice President, U.S. Bank

According to our training, that would be an incomplete and therefore invalid endorsement as the
chain did not end with the final noteholder endorsing it to blank.

In order to remedy that, they would add an additional endorsement:

Pay to the Order of

Without Recourse
Bear Stearns Trust, Pass through certificate holders 2003, VII.

Billy Cobham
Vice President, Wells Fargo
By power of attorney

In this way, the note endorsed to the trust and stopping (an incomplete chain as I was taught at Wells) would be modified into a complete chain, The trust would endorse it to blank and that endorsement would be added by Wells power of attorney, I assumed, but was never directly informed, by way of its authority as servicer for the trust.

Now what is peculiar about this is that our source reports that the notes were almost always endorsed to the trust (description includes Trust Series Name, Trust Number, Year). This is not only a permissible endorsement, some legal experts think it is the only sort of final endorsement that is proper.** So Wells also appears to be expending a great deal of effort doctoring documents that may be perfectly kosher (assuming the chain of title up to the trust is unbroken, something our source was not instructed to examine).

None of the higher ups questioned the revisions to procedures:

Generally, however, the whole process was a matter of ever changing orders and flowcharts to follow. There was next to nothing in the way of explanation even if you asked. It is my impression that the work directors didn’t have the slightest idea about the bigger picture, what was going on or that there might be a problem.

And for a substantial period of time, the priority appeared to be production, not accuracy***:

They would periodically restructure the flow chart to improve productivity. There were also a group of seven or eight auditors who were hired as “team members” out of the temp pool and effectively served as managers and who even did training near the end. They were the best informed regarding the process and the most hands on. They would also be involved in fixing oversights in the process flow charts. Their primary job was auditing assigned samples of each employee’s production per week and compiling statistics on them for the managers to see. These weekly stats were released in an email every week with all employees on the shift ranked by name in terms of productivity (files per working hour), and later in terms of accuracy.

Our source stresses that the procedures became more “reasonable” over time, in terms of having more coherent internal logic and being less production-driven, but it still raises the question of the apparent failure to correct earlier documents (which were presumably used in foreclosures) and whether even the later “improved” processes were adequate or even permissible.

Troubling Legal and Practical Issues

It is not clear whether Wells Fargo could make these changes legally to private label (non-Freddie and Fannie) securitized mortgages. While our source believes that Wells may have gotten a power of attorney from the trustee to make these changes, the PSA does not appear to convey this authority to the trustee.**** And why would it? Making sure the notes were endorsed properly was something the trustee repeatedly certified it had done years ago.

A party cannot convey authority to another party that it does not possess. So these document changes may be a complete legal fail.

But even if they could be construed to be permissible, the process is clearly hugely flawed. The temps were inexperienced, and not well supervised, and under pressure to produce at unrealistic levels. They relied on a database of questionable accuracy. Procedures were changed so often and so radically that some clearly had to be wrong. And our source reports some of his colleagues waved through documents he would have failed.

So we have document doctoring on top of widespread fraud. Welcome to property rights and records in America. If you are a borrower, you have to be punctilious in living up to your contractual commitments, or you can expect to have your lender use your lapse to maximum advantage. But if you are a bank, the government and courts will cast a blind eye to virtually any error. Anyone with any sense will avoid being in debt, which will ultimately be to the detriment of commerce. But it will take the authorities a long time to recognize that their efforts to save the system rather than reform it will only weaken it further.

* We refer to all whistleblowers as male irrespective of gender.

** The overwhelming majority of mortgage securitizations elected New York for its governing law, precisely because its trust law is settled. But it is also very rigid. For a transfer to a New York trust to be valid, the assets need to be transferred to the trust, not just the trustee. However, this issue has rarely been raised in foreclosures, since it would add an large cost to hire New York trust experts to provide supporting testimony. Since pretty much all PSAs allowed for endorsement in blank, that is accepted in courts; the fight is usually over whether the chain of endorsements is complete and whether the final party is the one who is in court trying to foreclose. In fact, our source indicated: “They were almost alwaysn endorsed to a trust and then the endorsement chain would just stop there. ” So bizarrely, Wells Fargo was doctoring documents that were correct!

*** The bank apparently started emphasizing accuracy more later in 2012, but with no redo of the earlier work, this appears to (at best) be an effort to shut the gate after the horse is in the next county. And as indicated, their ideas of “accuracy” appear subject to question.

**** We contacted a securitization expert on this matter, who (not surprisingly) could not recall and did not find language in a PSA that authorized this sort of post-trust-closing endorsement. Via e-mail:

So far, this is all I could find about the trustee signing title over to the master servicer (from section 3.14 of the PSA):

Upon the occurrence of a Cash Liquidation or REO Disposition, following the deposit in the Custodial Account of all Insurance Proceeds, Liquidation Proceeds and other payments and recoveries referred to in the definition of “Cash Liquidation” or “REO Disposition,” as applicable, upon receipt by the Trustee of written notification of such deposit signed by a Servicing Officer, the Trustee or the Custodian, as the case may be, shall release to the Master Servicer the related Custodial File and the Trustee shall execute and deliver such instruments of transfer or assignment prepared by the Master Servicer, in each case without recourse, as shall be necessary to vest in the Master Servicer or its designee, as the case may be, the related Mortgage Loan, and thereafter such Mortgage Loan shall
not be part of the Trust Fund.

But notice this section relates to a “Cash Liquidation or REO Disposition” and not in preparation for commencing a foreclosure action.

One might try arguing from Section 3.01:

The Trustee shall furnish the Master Servicer with any powers of attorney and other documents necessary or appropriate to enable the Master Servicer to service and administer the Mortgage Loans. The Trustee shall not be liable for any action taken by the Master Servicer or any Subservicer pursuant to such powers of attorney or other documents.

But again, we have a problem of legitimate authority. If the note has not been conveyed to the trust properly, altering original mortgage documents arguably does not fall in the scope of servicing and administering the mortgages. Indeed, if the notes were not conveyed to the trust by the cutoff date, they are not the property of the trust and trustee lacks authority to take action. This is precisely the scenario that no one in the mortgage industry wanted examined closely, and why they’ve gone to such lengths to pretty up document trials to indicate otherwise.
Yves Smith – Naked Capitalism