The Servicing Fraud Settlement: the Real Game
Warning: This is a long blog post. But if you follow mortgage servicing, I think you’ll find it worth reading. Despite lots and lots of media coverage of the servicing fraud settlement, nobody seems to understand the real story that's going on. I think that this post will explain a lot.
Let's start by recapping what we know. Back in March we started hearing media reports of a proposed penalty for servicers in the $20-$30B range. Then the American Banker published a 27-page term sheet from the AGs for servicing standards. Next, Huffington Post published a 7-page CFPB powerpoint presentation. Then came the draft C&D orders and then in April, the final C&D orders (which eliminated the ridiculous "single point of contact which need not be a single person" and replaced it with "single point of contact as hereinafter defined" and then failed—quite deliberately—to define it anywhere in the document).
Now there’s another round of activity and conflicting reporting. The American Banker reported that there was a new AG term sheet proposed and that principal reductions were off the table. That turns out to be incorrect, as Shahien Nasiripour reported in the Huffington Post. The new AG term sheet that the American Banker referenced deals only with servicing standards. The American Banker assumed that this mean that principal reductions were off the table because they weren’t referenced in the term sheet. In fact they are still very much in play. They’re just in a second, separate term sheet. So now there are two separate term sheets--one covering servicing standard and another covering monetary issues/principal reductions. (Recall that the original AG term sheet did not cover the monetary issues—that was clearly for a separate document.) We are also hearing news reports that the banks are offering to settle for $5B and won’t go above $10B.
So how do we make sense out of all of this?
The short answer is that the fight is not over a piddling $5B or $10B or even $20B. The banks would buy peace in a second for $20B and servicing reform. So what does that tell us? It indicates that the negotiations are over a substantially bigger figure than $20B. And this explains everything about the banks' negotiating strategy including the recent attacks on Elizabeth Warren by the Wall Street Journal's editorial page and by Congressional Republicans on the CFPB.
Now this isn't just my theory from reading between the lines. Instead, its exactly what follows from a careful reading of the documents and the rhetoric. The key, our Rosetta Stone, as it were is the CFPB powerpoint. It hasn't gotten a lot of analysis, but careful analysis of it explains everything that's going on.
There are two important things to note in the CFPB powerpoint.
1.The CFPB powerpoint contains an analysis of how much money servicers' saved by failing to comply with the law. It concludes that they saved at least $25B, based on an assumption that it would cost 75bps more per year to service each of these loans. (One can argue about that assumption, but that’s neither here nor there.) So if servicers were simply fined $24B, it wouldn't include any actual penalty. It would only be disgorgement of wrongful profits.
I don't think anyone has really understood the significance of this number. It means that the CFPB's (frankly rather conservative) estimate is that the banks made $24B from servicing fraud. That's the largest consumer fraud in history. This isn't just some chump game with cutting corners on affidavits. It's that doing that (and lots of other bad stuff) has saved the banks $24B in costs.
2. The CFPB powerpoint contained an analysis of the cost of various levels and numbers of principal reduction modifications. This is critical--the principal reduction modifications are separate from the $24B penalty. In other words, the total cost to the banks of a settlement would not be $24B. It would be $24B in disgorgement + the principal reductions.
The grid on the CFPB powerpoint shows the costs of a range of principal reduction modifications to be done over 6 months. The axes on the grid are the number of modifications and the depth of the modifiations. There's an enormous range of costs on the grid--from $7B to $135B. In other words, the total settlement cost (putting aside the cost of implementing improved servicing standards) would be between $32B and $160B depending on the number and level of principal reduction modifications. I want to underscore, however, that the powerpoint does not indicate what the CFPB thinks is an appropriate number--and clearly that would be a negotiated issue.
Whatever that number, it's also important to recognize that not all of the cost of principal reductions would be borne by the banks. In the powerpoint, at least, the MBS investors would bear the costs of the principal reductions if NPV positive, but 2d liens (big 4 bank balance sheets) would be reduced too. But it means that the price tag for settlement being offered to the banks isn't $24B. It's substantially higher. We don't know how much higher--the powerpoint was simply showing a range of options and their costs, not recommending any particular option--but even at the low end of $7B that's a sizeable increase on top of $24B.
Recognizing the full potential cost of a settlement to the banks and how most of it could be in the form of principal reductions rather than a fine explains everything that's been happening with the negotiations and the Congressional Republicans' witchhunt against Elizabeth Warren.
If the cost of peace with the AGs and Feds was a mere $24B, the banks would settle. Remember, that’s $24B for all the banks, not $24B for any one bank. The mortgage servicing issues are an enormous drag on BoA generally, Wells knows that it is a huge litigation target in every state, Citi just wants to keep its head down, and Ally wants a clean bill of health for its IPO. The banks really want to put these issues behind them.
$5B a piece for each of the big servicers isn’t a ridiculous price for putting the issue to rest. Indeed, news reports that the banks will consider $5B-10B show that this is only a matter of haggling about price, not principle (no pun intended).
That’s why I don't think the hold up is over the $24B. Instead, the sticking point in the negotiations has got to be the additional cost of the principal reduction mods, whatever that might be. At the extreme level, if the settlement would cost a total of $160B, that would be about a third of the equity of the four biggest banks. (Compare that with the $750B in negative equity that exists.)
If I'm right about this, then everything about the banks' negotiating strategy makes sense. If the banks are willing to pay $24B, but not the additional cost of principal reduction mods, it makes sense for them to run the clock, to focus on principal reductions in their PR, and to do everything possible to minimize the role of the CFPB.
First, inflation alone will help reduce negative equity and thus the cost of principal reduction mods (unless we continue to see a double dip…which is likely, but why not take a gamble on it?).
Second, the AGs' main leverage here is the threat of litigation. But litigation would be incredibly slow, especially if the OCC has the banks' backs and raises preemption challenges at every step (as it might do in light of the consent orders). It might take 3-4 years to get to a judgment, by which time housing prices might have rebounded and lots of foreclosures would have been processed, so there wouldn't be that much negative equity outstanding to reduce through principal mods. Delay lets the banks avoid principal reduction mods.
Moreover, the banks know that the AGs can’t be too serious because they haven’t done any investigation. I’m just baffled how the AGs are conducting settlement negotiations without having done any investigation. It’s a serious problem. How can the AGs know the proper price for the settlement (high or low) without knowing what cards they hold?
Third, the one agency that could really speed along litigation is the CFPB. The part of the litigation that will take up the time would be discovery, but the CFPB could speed that up significantly through its examination power. But the CFPB can only be effective with this if it has a Director. And that explains why the banks (and the OCC) brought out the Wall Street Journal editorial page and Congressional Republicans to wage war on Elizabeth Warren, the frontrunner to be appointed CFPB Director. Make Elizabeth Warren politically toxic on whatever trumped up charges can be found (she said "advised" when she in fact recommended! She was in the room during negotiations! Gasp!). The goal is to keep the Director position vacant, so the CFPB can't move along litigation.
There was a lot of justified pushback (including from yours truly) against the attacks on Elizabeth Warren. So rather than beat up on Professor Warren, the bank strategy changed to attacking the CFPB itself under the guise of regulatory “improvement.” First, the GOP pushed several bills in the House Financial Services Committee meant to smother the CFPB in its crib. Then the GOP in the Senate said that they wouldn’t confirm any CFPB Director unless the House reforms were made, and then they started making unpleasant noises at the suggestion that there could be a recess appointment. (News flash: winning 1 house of Congress in a mid-term election ain’t an electoral mandate to do anything. You need the hat trick for that.)
Every week that is spent on negotiations that get nowhere lets the banks run the clock a little further. So the banks will try to stay at the negotiating table as long as possible without every actually conceding anything. But that's the strategy here--run the clock to avoid principal reductions. What terrible is that this strategy seems to be working--the CFPB has been shut out of negotiations because the Wall Street Journal and Congressional Republicans have made such an issue over the CFPB. (This might turn out to be short-sited, but that's another story.)
And this ties in perfectly with the PR spin: the line coming out of the banks hasn't been an objection to $25B in fines. It's been an objection to principal reductions. The hackneyed moral hazard objection has been trotted out (despite the banks' doing some principal reduction mods already) and we've had Moynihan (BoA), Stumpf (Wells), and Jaime Dimon (JPM) saying that principal reduction mods are "off the table."
Listen to the banks. Their rhetoric says it all--the game here is about the principal reduction mods, not about the servicing standards or the $24B fine. It's about the cost of the principal reduction mods. (There might be some ancillary issues like the number of mods, but it’s really gotta be about cost.)
Now lets be clear. Principal reduction mods are not about correcting robosigning. Robosigning is what's gotten the most media attention, but that's not the only issue around. There are a host of other flat out legal violations (just consider the $20M jury verdict in the Servicemembers Civil Relief Act cases to get a sense of what these violations cost-1,000 verdicts is $20B). There's another panoply of questionable, but perhaps not illegal acts (e.g., MERS issues). And if you want a doozy, how about the many loans that are endorsed like simply to Deutsche Bank as trustee, rather than Deutsche Bank as trustee for a particular trust (Deutsche is trustee for over 2,000 RMBS trusts). That didn’t fly in a North Carolina appellate court, and it wasn’t a fluke endorsement (and there are worse problems than that in terms of endorsements).
And then there's also lots of good policy reasons for pushing principal reduction modifications. Principal reduction modifications start to address the $750B in negative equity in this country and help the housing market to clear without the inefficiencies and social externalities of foreclosure. And of course principal reduction mods make the banks pay an appropriate price (and in an appropriate form) for the economic and social harms they caused with the housing bubble and foreclosure aftermath, including threatening our fundamental property title systems via corner cutting on paperwork.
Finally, consider what it means that we're even seeing an eye-popping figure like $160B. It might be out there just to push the banks toward settling. But the amount of shit that the feds and AGs must think there is to come out with a number like that down on paper (especially for an agency under as much scrutiny for an sign of going off the rails as CFPB) makes my skin crawl. I worry that we don't have any handle on just how much rot is in the system and that we've been papering over it as the stock market rebounds.
Now the banks have some legitimate concerns about principal reductions. If they are handled incompetently they will undoubtedly lead to a strategic default problem. And the banks know this—Countrywide’s settlement with the AGs’ was a paragon of foolishness. It made modification eligibility depend on delinquency status and applied prospectively. That’s virtually inviting strategic defaults.
But there are lots of ways to fix this. Here’s a simple one: principal reductions apply only retrospectively. Now there are problems with doing it just retrospectively. But prospectively, it could be applied to mortgages that fit particular criteria irrespective of default status (indeed, for mortgages that are 240 days delinquent, I’m not sure what good principal reduction is likely to do)—the biggest bang in terms of stabilizing the housing market might be to do principal reduction to homeowners who are not yet in default or to those who do strategically default because they’re the ones who are willing to gamble (and walkaway) from their homes.
Whatever the terms of a settlement, perhaps the most important question is what issues actually get settled. Obviously the AGs can’t settle for consumers or investors, and those issues are going to continue to plague the banks for some time to come. And the AGs can't bind the CFPB in terms of prospective regulation of the servicing industry. (Fat chance the banks clean up their act by July 21). But the AGs are the biggest dogs in the hunt at this point.
The banks are, of course, going to want the broadest possible settlement and if the AGs aren't careful, the banks will pull one over on them like they did on the merchants in the Wal-Mart interchange antitrust litigation settlement, where the wording was vague and then subsequently interpreted in the banks' favor.
The CFPB powerpoint also gives us a useful yardstick for measuring a final settlement. Anything less than $24B would let the banks come out ahead. Let me repeat that. Anything short of $24B means that the banks broke the law and got to keep some of the profits. If that's what the AGs settle for, it's a disgrace. $24B really has to be the baseline above which there's a settlement. I don't think the AGs are going to solve the foreclosure crisis in one fell swoop (they'd need to do some investigation to even have a shot at that), but settling for $5-$10B means that they'd let the banks keep 60%-80% of estimated illegal profits. Just keep that in mind.
Final thought: If I'm right about this, and that the number being bickered about isn’t $5 vs. $10B but something more like $40B-$60B, I worry that all hell is going to break loose. Progressives that were hating on the AG settlement for being too light on the banks, might rethink that position. And the howl we are going to hear from the right is going to be unparalleled. The idea that businesses could have done multi-billion dollars worth of harm to consumers (or the legal system) simply isn’t within the conceptual grasp of the Wall Street Journal editorial page and its ilk. The only possible explanation they have for this is a shake-down. Oh it’s going to be a fun summer.
Thanks for bearing through to the end. I hope it was worthwhile.