By David Dayen, a lapsed blogger, now a freelance writer based in Los Angeles, CA. Follow him on Twitter @ddayen
Anyone paying a smattering of attention justifiably raised a skeptical eyebrow at the Office of the Comptroller of the Currency’s assurances to Congress that the Independent Foreclosure Reviews revealed hardly any borrower harm from servicer malfeasance. First of all, from the reporting we know, OCC just dropped this 4.2% error rate number without supporting information from the loan files. Second, the error rate contrasted wildly with what Yves uncovered in her superlative series on Bank of America reviews. Most critically, if the reviews were finding no borrower harm, there would have been no real reason to ditch them. They would have reinforced the bank-supported view that foreclosure fraud was simply overblown, would have silenced critics, would have reduced bank exposure to payouts from the settlement and probably in future litigation. This would have been well worth the expense of paying out another couple billion to the consultants, for the banks to get an on-the-record fact pattern establishing their relative innocence.
That’s what makes the bombshell story from the Wall Street Journal of all places, so damaging to the entire cover story, particularly for the OCC. If true, this actually establishes that the agency at least massaged the truth to Congress, if they didn’t outright lie:
Some 6.5% of files reviewed unveiled errors requiring compensation, officials at the Office of the Comptroller of the Currency said in January. They later revised the error rate to 4.2% after requesting new data, raising the total number reviewed to roughly 100,000 files.
But a breakdown of the information provided to the regulator shows that more than 11% of files examined for Wells Fargo & Co. and 9% of those for Bank of America Corp. had errors that would have required compensation for homeowners, said people who have reviewed the figures. A narrower sample of files—representing cases selected by outside consultants—showed error ratios of 21% for Wells Fargo and 16% for Bank of America, the people said.
The OCC findings appear skewed by the outsize contribution of one bank, J.P. Morgan Chase & Co., which reported an error rate far below rivals that oversaw a much larger universe of loans.
(Incidentally, you have to love the Wells Fargo flak’s claim that the error rate “does not provide conclusive information about actual financial harm.”)
The reason OCC could publish their 4.2% number is that JPMorgan reported just a 0.6% error rate, according to these anonymous sources. They also completed far more reviews than their confreres. PNC Bank reported rates over 20%, but they only delivered a small share of the reviews (Yves may have more on that bit, from her reporting she’s basically gobsmacked that PNC managed to deliver any reviews at all),. How lucky for OCC that the one bank which presented a near-perfect record submitted the most reviews!
It doesn’t pass the smell test at all that you would get such a wide discrepancy, and in particular that JPMorgan Chase would show itself to be such a precision servicer. I remember attending a NACA event where they gave out Jamie Dimon’s cell phone number to homeowners, urging them to call and harass him over the abuse his bank had exhibited on borrowers. They were seen as uniformly the worst by a wide margin.
But of course, as ably documented in this space, the entire concept of real numbers from the reviews is fanciful. When you talk to the actual reviewers, as WSJ did for their story, you get error rate numbers as high as 80%. And the files themselves were kept in such disarray that it borders on impossible to grade them at all. The numbers reported by the banks to OCC merely reflect how successful they were at controlling the review process and limiting the finding of borrower harm. This window into how Deloitte handled the JPMorgan reviews tells you everything you need to know about their numbers:
Two Deloitte employees who performed the review for J.P. Morgan in a Brooklyn office building said workers were encouraged by supervisors to examine pools of loans they knew would be less time-consuming or error-prone as they tried to hit loan quotas.
One of these employees said that at an event last year known in the Brooklyn office as “March Madness,” Deloitte officials encouraged reviewers to avoid problematic loans originated by EMC Mortgage, a troubled mortgage lender J.P. Morgan acquired in 2008.
A Deloitte spokesman said that “because of confidentiality constraints, we are not at liberty to discuss details of the engagement. We fully stand behind the quality of our work.” Deloitte had a team solely devoted to reviewing EMC loans and the consultant did not impose any quotas, said another person familiar with the review.
I guess if you pick and choose what loans to review, I suppose you can get yourself to 0.6%.
If you want to actually gauge the level of servicer abuse, you might go back to one of the only legitimate examinations from a reviewer not paid by banks: the one released as part of the AG settlement by the HUD Inspector General. That review included this infamous statistic, from the aforementioned “super-servicer” JPMorgan Chase:
For Chase, we also reviewed 36 affidavits for foreclosures in judicial States to determine whether the amounts of borrowers’ indebtedness were supported. Chase was unable to provide documentation to support the amounts of borrowers’ indebtedness listed on the affidavits for all except four. When we reviewed the four affidavits, three were inaccurate. Specifically, the amounts of the borrowers’ late charges and accumulated interest did not reconcile with the information in Chase’s mortgage servicing system.
Emphasis mine. So depending on who you believe, either OCC or the HUD IG, servicers either generated error rates of 4.2% or 97.2%. You know, somewhere in there.
This serves as yet another black eye for OCC, which clearly tried to throw a curve by Congress by highlighting the absurdly low error rate figure. Thomas Curry may have inherited this mess, but he’s starting to look as bad as his predecessors. Hopefully the nascent Congressional investigation from Elizabeth Warren and Elijah Cummings will look unkindly on the fertilizer OCC is trying to spread.