There has already been a lot of good commentary on the Senate hearings on the misnamed Independent Foreclosure Reviews, notably by Pam Martens. I’ve finally gotten a transcript (see at Corrente which helps in reviewing it more carefully. Since Part 2 of the hearings take place this week, I’ll focus on some key issues that haven’t gotten the attention they warrant.
First is that even though Elizabeth Warren correctly has gotten a lot of praise for her pointed and persistent questioning, Sherrod Brown and Jack Reed also did exemplary jobs.
What emerged from the hearings, particularly if you have been following the story, is the remarkably poor performance of the OCC. The only open question is to what extent that resulted from pure incompetence, and how much resulted from being so completely captured by the banks that it swallowed their PR, that hardly any homeowners had been harmed by bad servicing, and accordingly botched the design and execution of a coverup. No matter which theory you subscribe to, it paint a picture of an agency that is hopelessly bad at its job, which raises the question of why it should continue to exist.
Let’s recount some of its gaffes so far. It said initially it was shutting down the process in part because it was taking too long to get money to harmed borrowers, and it had also found few borrowers were harmed. It then provided two different figures for overall borrower harm, first 6.5%, then 4.2%, but the latter was flagged as sus by the Wall Street Journal since the reviews for JP Morgan Chase had done the easiest files first (as in no Bear or WaMu) and thus come up with an error rate of 0.6%. Warren got the OCC to admit to what critics knew, that the OCC and Fed had settled while having no idea of the actual harm. Ironically, it was the Fed that wound up doing the brunt of the defense of the non-numbers; the OCC representative said later that this wasn’t his area but reviewing 100,000 file “is not valueless”:
Sen. Warren: The problem is that the 6.5% is not accurate. Your staff admitted to us in a meeting earlier this week that the number is not based on a random sample, not on a review of these cases; it was determined based on whatever files had been reviewed by the time you shut down this process…But the question I have is, what is the right number? Is it six tenths of 1%? Is it 6.5%, 9%, 11%, 20%, 50%, 90%? If you can’t correctly tell how many people were the victims of illegal bank actions, how can you possibly decide how much money is an appropriate amount for settlement?
Mr. Ashton, Fed: Well, Senator, I think I can only reiterate that the decision that was made to accept that agreement, and we recognized that that was not a perfect agreement, was based on the delay that would have been involved in any alternative. To continue the Independent Foreclosure Review and trying to –
Sen. Warren: Mr. Ashton, I’m sorry. I understand the point about delay. But it doesn’t mean you pick a number out of the air…. I can’t believe that you are saying that the only reason the number 9 billion was settled on was so that it could be done quickly, and that you’re saying that the OCC did not have an estimate in mind of how many banks had broken the law and how many homeowners were the victims of illegal activities. Mr. Stipano, is that the case for the Federal Reserve Bank as well?
Mr. Stipano, OCC: Senator, I am not an expert on the IFR settlement. I was not directly involved in it. I can answer general questions and I’ll do my best to do that. My understanding is that when it comes to the error rate, I don’t really know how it was calculated, to be honest. There are people in the agency who do. They’re not here. I do believe that we did review a substantial number of files, admittedly –
Sen. Warren: I’m sorry. Mr. Stipano, we met with your staff –
Mr. Stipano, OCC: Yeah.
Sen. Warren: – and your staff has made clear, you did not review a random sample.
Mr. Stipano, OCC: No, it wasn’t a random sample. No.
Sen. Warren: And without a random sample, can you then generalize to the accurate number, even an estimate, of how many banks broke the law?
Mr. Stipano, OCC: Um, not – in my, my understanding is not in a statistically valid way. However –
Sen. Warren: Okay, that’s a no.
Warren also secured the bombshell admission from the consultants that none of them had been involved in preparing the schedule of how many borrowers fell in which category of harm, meaning it was cooked up by the banks and the OCC. Her summation:
Sen. Warren: All right. I just wanted to make sure, because it appears that the people who broke the law are the same people now who have determined who will be compensated from that lawbreaking. I just find this one amazing.
But the admissions that caught my eye were that the OCC had expected the reviews to take 120 days and two of the consultants had said they’d expect them to cost between $5 and $8 million. If you take 14 servicers and $2 billion previously reported, the average is $143 million. Oh, and better yet, good child Deloitte got permission before the hearing from its clients to say how much it got: $190 million from US Bank, $175 million from Citigroup, and $60 million from Suntrust. That will pressure Promontory to come clean. I’m waiting with baited breath.
Now anyone who thought a nanosecond would know there was absolutely no way these reviews could be completed in 120 days. Recall there were two efforts (more on that shortly): the reviews based on borrower letters, and the reviews based on sampling of files. And this was to meet the goal of identifying all borrowers who suffered financial harm. I am not kidding you. The purported aim was to find everyone.
Now of course, the process of developing and vetting the letters to send to borrowers and printing and sending them would easily take a month. You’d have to allow for time for them to respond. So how in a 120 day time frame was there any time to do reviews?
The sampling track was similarly obviously insane. There is no way to identify all the borrowers harmed through sampling. The two processes don’t even mesh. The rationale was that the consultants would use a multi factor, iterative sampling approach to help identify populations that needed a 100% file review (and the engagement letters were of the flavor that those populations would be the exception, not the rule). But what happens even if this process had identified, say, certain types of borrower where the harm rate was 5% or 10%? Given their goal of identifying all harmed borrowers, you’d still need to review all the files in that population for that problem (admittedly not of the entire file, just related to that issue). But why sample? You could tell mods were a disaster. You could do a PACER search and find litigation that established that there were meaningful problems with certain types of loans, like borrowers who had filed for bankruptcy.
Now there is one way to make sense of these procedures: if the purpose of the exercise was really to find very low harm numbers. Remember the genesis: that the OCC broke from the attorney general-Federal settlement negotiations to impose the consent orders in order to derail the settlement talks. So then the elaborate sampling would serve to prove that the low findings of harm on the borrower letters were warranted. This dim view is confirmed by the GAO report released last week, which showed that the banks and the consultants expected 0% errors in most categories, and at most 10% in a few.
We also have the absurdity of the OCC saying it was in charge even though the banks engaged and paid for the consultants:
Mr. Stipano, OCC: No, I don’t think that’s accurate. We did require that the contracts be submitted to us for our review and we directed the servicers to put language into the contracts that made it clear that the independent consultants were acting pursuant to our direction, not the servicers.
Sen. Reed: In reality, did the independent consultants act at your direction?
Mr. Stipano, OCC: I believe so, yes.
Sen. Reed: Can you provide us documentation to that effect?
There’s a very simple proof that the OCC is at best deluded and at worst lying. There was no consistent definition of what harm was across the servicers. So even if the consultants had miraculously completed the review, the process still would have failed in its objective of assuring that similarly situated borrowers were treated similarly.
Of course, the whistleblowers at various servicers also undermine the OCC’s claim. As Pro Publica initially found, and we documented in greater detail, Bank of America was driving its reviews of borrower letters. At PNC, the consultants got inquires from the bank, not the OCC, as to how things were going, and reacted with alarm and put the reviews on hold when it learned the consultants were actually finding harm.
The OCC also got a beating for not having written standards for selecting consultants. Sherrod Brown honed in on the fact that Promontory had conducted an incredible whitewash on behalf of Standard Chartered (curiously avoiding mentioning Promontory by name):
Sen. Brown: We’re not – you used the word pristine. We don’t expect pristine here. That sounds too difficult. But we do expect, I think, clear standards in what qualified means. And, for instance if a consulting firm, and there is one in this situation, has repeatedly been, for lack of a better term, at the scene of a crime, what would it take before they are viewed as not qualified? What if they, for instance, what if they underestimated the value of an institution’s money-laundering transactions by 250 billion dollars or presented watered-down reports to regulators? Wouldn’t – that wouldn’t be enough for disqualification?
Mr. Stipano, OCC: Well, again, I think you have to look at the total context, but I do believe this is an area where there are lessons to be learned for us and we are committed to exploring ways to do better. And, you know, maybe that results in, you know, some kind of written standards. We don’t presently have them, but I think this is an important area and we are committed to doing a better job.
Brown also pointed out that Promontory board member Alan Blinder said that the IFR was not in Promontory’s “sweet spot,” and again tried to get the OCC to explain how they had been an acceptable choice despite that. He didn’t get a very satisfactory answer.
While it is gratifying to get some important admissions from key participants in this botched cover-up, it is still disheartening to know that the problems exposed in the hearings so far only scratch the surface of both the underlying borrower harm and bad faith with which the reviews were handled. I hope the session this week manages to turn up more dirt. There is unquestionably a lot there.