Thomas Curry keeps insisting the decision to end the independent foreclosure review will deliver fatter checks to more borrowers.
"Our new approach will get more money to more people much more quickly," the Comptroller of the Currency said Feb. 13.
But where's his proof?
The government's deal with 13 servicers to end the unwieldy review of 4.2 million mortgages affected by the 2009-10 robo-signing scandal may indeed get money to borrowers faster, but it will not ensure they receive more money.
That's because the government does not know which borrowers suffered what degree of harm.
The settlement should fit the harm, but the Office of the Comptroller of the Currency is making the harm fit the settlement.
According to an amended deal revealed last week, 13 firms will pay $3.6 billion to these borrowers. The servicers themselves will slot each borrower into one of 11 categories, depending on their circumstances. The OCC has pledged to review the servicers' decisions, but borrowers will have no say and no way to appeal.
The Comptroller's Office has not yet put a dollar figure on any of the 11 categories.
Instead, the agency plans to wait to see how many borrowers the servicers slot into each category and then the agency will assign a corresponding payment. That way the money paid out will match the total committed by the servicers.
The OCC did release a matrix last June with 13 categories of harm, and the corresponding payouts ranged from $1,000 to $125,000.
The agency has condensed the number of categories to 11 and says it is now reworking the dollar payments. It won't say when the new numbers will be made public, which undercuts the agency's promise to have checks arriving in borrowers' mailboxes within weeks.
But the OCC has said that every borrower, regardless of harm, will receive some remuneration. And it has confirmed the maximum payment to any single borrower will be $125,000, but the agency has never explained how it set that ceiling.
Among the circumstances that would have earned the maximum payout under the June 2012 matrix was a borrower who was foreclosed upon even though the loan was not in default. If that happened to me, I'd want more than $125,000.
The national newspapers have dug into this story and lawmakers are demanding better, deeper explanations from federal regulators.
I think it's becoming increasingly clear that the Comptroller's Office moved to settle too quickly. The agency should have held off settling with servicers until it had more information.
The consultants who were reviewing the mortgages claim they were on the verge of giving the agency the first set of statistical data on how much harm had been inflicted when the agency pulled the plug on the review and settled with the servicers. A pilot remediation project was in the works, too.
It's natural to be skeptical of the consultants; they've all just lost lucrative work. But they did spend the past year-plus trying to figure out which foreclosures resulted from a servicer's error so, like it or not, they are in the best position to assess borrower harm.
If the Comptroller's Office had waited for the consultants' preliminary results, then at least it would have some justification for the settlement's dollar figure.
The OCC says it arrived at the cash portion of the $9.3 billion settlement through a careful calculation.
The payout was determined "by looking at the amount of likely harm under the IFR process, the remaining IC [independent consultant] costs of completing the reviews and other costs and expenses associated with continuing the process we had in place," the OCC's point man on this fiasco, Morris Morgan, told me in an interview for the first story I did on the settlement.
Unfortunately the OCC can't measure that "harm."
Under pressure from Congress to prove it has some idea how many borrowers were harmed, the OCC has subsequently said that just 4.2% of the loans had problems that required compensation be paid to the borrower.
But The Wall Street Journal, citing anonymous sources, reported last week that the figure is much higher 11% at Wells Fargo and 9% at Bank of America.
The OCC can claim the overall total is brought down by the much lower error rate at JPMorgan Chase, which the Journal reported was a meager 0.6%.
But how does that make sense? Can JPMorgan's error rate be that much lower than, say, Wells Fargo's when its portfolio included mortgages made by the failed Washington Mutual? And if the error rates among servicers were that far apart, shouldn't regulators consider whether the reviews were fundamentally flawed?
I repeat: no one knows how many mortgages had problems severe enough to require remuneration to the borrower. So the OCC cannot know that its $9.3 billion settlement will deliver more money to harmed borrowers as Curry has repeatedly claimed.
You may be wondering why I haven't mentioned the Federal Reserve Board, which signed on to this settlement, too. When Fed Chairman Ben Bernanke testified last week, he was asked about the abrupt decision to halt the reviews and settle with servicers. Like Curry, Bernanke told Congress that the settlement would deliver money to victims faster. But he stopped there and never said borrowers would be getting a bigger payout.
While the two agencies have worked together, the OCC has taken the lead since the robo-signing consent order came down in April 2011.
That order required the servicers to hire independent consultants to review foreclosures begun in the 2009-10 time frame. But the process was more expensive and less effective than anyone dreamed so late last year the Comptroller's Office began negotiating a settlement.
The first version was announced Jan. 7. It covered 10 servicers that agreed to provide $3.3 billion in cash to borrowers and another $5.2 billion in foreclosure-prevention moves such as loan modifications.
The deal was amended on Feb. 28 to include three more servicers HSBC, Goldman Sachs and Morgan Stanley. Together, they have 332,474 borrowers covered by the review and are contributing a combined $300 million in cash and $500 million in other assistance. That brought the totals to $3.6 billion in cash payments and $5.7 billion in other help.
Last week's changes revealed that servicers will get credit toward the $5.7 billion total if they provide assistance to any borrower not just the robo-signing victims of 2009-10. This assistance must be delivered by Jan. 7, 2015.
That's yet another reason why Curry is wrong when he says borrowers will receive higher payouts under this settlement.
Either the Comptroller's Office did a lousy job negotiating that part of the settlement or it just hasn't explained it very well. The New York Times and the Journal have sunk their teeth into this story and report that the credit the servicers will get for foreclosure prevention will greatly exceed their actual costs.
Unless the Comptroller's Office comes up with more answers fast, this snafu may consume the agency. And that would be bad for bankers.
What the industry and the regulators need most right now is trust. They need the public to believe that they are making sound decisions and doing the right thing.
This foreclosure settlement demonstrates a failure on both fronts.