From incentives for speed to forging signatures to falsifying documents, all that went wrong with mortgage servicing can be found in one story.
This month, The American Prospect presents a special report exploring the debate about the architecture of the U.S. housing market. The report, which takes on the fundamental question of the future of Fannie Mae and Freddie Mac, features my favorite people working in this area, among them Alyssa Katz, Marcus Stanley (from Americans for Financial Reform, who critiques HAMP), James Carr from the NCLC, and Dan Immergluck, who calls for a public option for mortgages.
My contribution to the report centers on the current foreclosure fraud crisis and what needs to be done to fix it. A key issue is the problematic nature of the dual nature of the mortgage servicing business. Currently, the business comprises both a high-volume, low-information loan processing business and also a default management business that should be low-volume and high-information, but instead relies on rapid turnover and reckless practices. I note that if the mechanisms for payment and default management in the largest lending market in the largest economy in the history of the world aren't trustworthy, there will be serious consequences. These companies should function as reliable, accountable utilities rather than businesses willing to cut corners, fake documents, and proceed with phantom referrals in order to increase margins by a tiny percentage.
My article was written before the excellent news that individual state attorneys general will be actively investigating these problems, led by New York’s Eric Schneiderman. Now word is coming out that Connecticut and Ohio are focusing on the issues, and that California and Illinois are specifically looking at Lender Processing Services (LPS).
Luckily Yves Smith of Naked Capitalism just posted a class-action lawsuit filing by shareholders against LPS, which is a fantastic read. At 200 pages the lawsuit is long, so I'll summarize the main argument as a way of reintroducing what the AGs want to find out about LPS. The lawsuit is important because many people understandably don't think that foreclosure fraud is a major, systemic issue that cuts to the core of the country's foreclosure system. They tend to see problems as exceptions in a large industry and may not believe that homeowners suffer real damages from this fraud. If there are any consequences to real people, the thinking goes, then they are likely for bond holders and servicing banks, or within different entities that created the securitization in the first place. In this view, the problem is just rich people with lawyers screwing other rich people with lawyers who have the means and incentives to respond. The lawsuit ends up making the case against these ideas.
Problems in Theory
LPS perfectly illustrates the problematic dual business model. The firm started as a technology company designed to provide software and web-based applications to automate payments. In 2008, it added a default management services wing to its business lines.
Default management is difficult to automate, and LPS executives made a series of decisions to further exacerbate this problem in a way that would increase its market domination and revenues. They gave away business free to clients and decided to generate revenue by coordinating a network of attorneys, making money through charging them fees. LPS acted as a filter between clients and attorneys handling defaults, which broke the client-attorney relationship. The firm then created a series of incentives to maximize speed over quality. As Reuters has noted:
Interviews, deposition transcripts and LPS's own records underline that the company keeps its clients happy and maximizes its own fee income by whipping law firms to gallop cases through the courts.
The law firms are on a stopwatch…the LPS Desktop system automatically times how long each firm takes to complete a task. It assigns firms that turn out work the fastest a "green" rating; slower ones "yellow" and "red" for those that take the longest.
In this reckless system, firms that churned and burned court pleadings using low-skilled clerical workers enjoyed green ratings. Those that moved more slowly got unfavorable red designations and kissed business good-bye.
LPS handled more than 50% of the industry’s residential mortgage volume. Their business model was designed to strip the legal work necessary for foreclosure to its bare minimum. LPS didn’t have to fear market pressure from consumers who don’t know if their mortgages will be securitized and certainly have no say in who will be managing payments if they are. And with the default management system working to obscure, cut corners and emphasize speed over reliability or quality, it is up to the legal system to provide a necessary check.
Problems in Practice
Obviously, that business model turned out poorly. Even worse, the attempt to maximize the rate of foreclosures does untold damages to both the system of records and to consumers. The report points out six distinct things LPS was doing wrong.
1) Documentation. In order to begin foreclosure on a home, the foreclosing entity has to show ownership, and during the boom these documents weren't correctly stored or ordered. This isn't a trivial point -- centuries of law have required strict adherence to this matter, and even more so for trust law (whose special tax provisions were necessary for the securitization structure to work). A special wing of LPS called DocX would, according to documents and testimony, recreate missing documents, missing assignments, and even an entire collateral file.
How would they do this?
CW724 explained the process by which documents such as assignments were generated at DocX. Indeed, he explained that Data Entry employees took information from scanned documents on their computer screens and entered it into LPS Desktop software to create assignments of mortgage. These employees entered data such as the loan amount, person’s name, address and a property description. Data Entry employees did not perform any analysis or verify any information; they just pulled information from one screen and entered it into another. CW7 then printed those documents through LPS Desktop and took them into the “Signing Room” at DocX, where a supervisor took the documents and handed them out to signers...
Indeed, LPS executed assignments fraught with deficiencies, including but not limited to: (1) signatures and dates after foreclosures were initiated for mortgages that should have been handed over to trusts; (2) signatures by LPS employees purporting to be officers of lenders that no longer exist; (3) incomplete or non-existent grantees or grantors such as “bogus assignee” or “bad bene”; (4) improper effective assignment dates such as “9/9/9999”; and (5) blank signature lines witnessed and notarized.
Any computer coders should note that the code prints out all 9s for dates when they are improper, yet documents went out that way anyway. This is scary, as these documents are used as proof of the amount, conditions, and terms of the loans.
2) Robosigning. This practice amounts to a document-signing sweatshop. Because LPS managed default services for such a large portion of the industry, it ended up with millions of documents to sign. From the report: "CW7 explained that each person pulled a page off the top of the stack near them, signed that page and moved it to another stack next to them. They did not appear to perform any analysis, review or verification of any details in the documents they were signing. These documents included mortgage or promissory notes, and assignments of mortgages." Given that some of these documents were recreated (i.e. faked), this is a bad sign.
3) Forgery. Employees at LPS would forge signatures. Check out the variety of signatures from Vice-President Linda Green:
4) Speed. Mirroring the incentives for lawyers, robosigners and document processors within LPS were paid for speed, often with a very high minimum number of signatures required in order to not be fired and with no penalty for errors. So when you add this all up, what kind of problem does it generate?
5) False referrals. The fifth problem, generated by the fourth one I’ve noted, concerns the abuse of important, minute details.
According to CW16, there were serious problems in the automation process that led to “phantom referrals”, when the LPS MSP software system generated “processes” or attorney referrals that did not really exist... While attorneys who were honest would review the file and realize there was not sufficient information to justify the referral, many other attorneys who were not honest or who had organizations with a lot of low-level employees handling the intake “would just file it even though created by error.” CW16 noted that the David J. Stern law firm would make fees wherever they could...
According to CW16, on top of the 20% of files with phantom referrals, approximately another 35% of files had some problems in them. Those problems varied, and included among others, an ARM that had improperly adjusted up, a failure to properly account for a borrower’s principal and interest payments, and a failure to properly attribute payments between pre-petition and post-petition that led the banks to try to collect pre-petition obligations they were not permitted to pursue.
False referrals were coupled with manipulating payments and numbers. This all undermines the sanctity of the court and the foreclosure process, harms consumers, and makes a mockery of the largest lending market in the world.
We don't know the extent of the problems outlined, and it is likely the banks themselves don't know the extent. But we know that the system is broken, and it requires a government response.
Mike Konczal is a Fellow at the Roosevelt Institute.