A new article by Shahien Nasiripour of Huffington Post, “Big Banks Save Billions As Homeowners Suffer, Internal Federal Report By CFPB Finds,” includes a presentation from the Consumer Financial Protection Bureau dated February 14 prepared for Tom Miller, the Iowa Attorney who is leading the 50 state attorneys general foreclosure fraud settlement negotiations.
If I were a betting person, I’d wager this document was leaked to show that the Administration and the AGs did not just make up the $20 to $30 billion settlement figure that has been bandied about as their ask, but have a sound, reasoned basis for their demand.
Unfortunately, the document simply proves that they did make up the $20 to $30 billion figure. Not only is the analysis effectively fabricated, it’s the wrong analysis. But I have to say, having been at McKinsey, it’s impressive how the use of McKinsey firm format makes a story look much more credible than it really is.
The critical part comes on the third page, “Calibrating the Size of Potential Penalties”. You’ll note it assumes that the cost of special servicing of delinquent loans would have cost 75 basis points a year more than actual costs incurred. That drives the entire analysis.
The rest is based on delinquencies at various major servicers from 2007 to the third quarter of 2010; presumably the CFPB has been able to get reasonably accurate data on that front.
Now….is this “75 basis points a year” a knowable figure, ex doing a lot of real nitty gritty work, which certainly has not taken place? We can debate whether this is the right figure, and whether the CFPB has also captured the actual costs correctly. Servicers are already losing boatloads of money; the economic model was never designed for a high level of delinquencies. Our Tom Adams has estimated that servicing now costs 125 basis points versus the banks’ typical fees of 50 basis points, plus another 30 to 50 basis points in late and junk fees.
If you take this analysis at face value, the biggest question is what standard of servicing is implied by “effective special servicing of delinquent loans”? If they mean loan modification, that’s the same as a new underwriting of a mortgage. That cannot be done through the current platform and would require new staff with different skill sets and software/systems support. So any estimates are at best finger in the air exercises. And given that some servicers are far more abusive with junk fees than others, Tom Adam’s comment above suggests that a one-size-fits-all estimate is misleading too.
But arguing over a pretty much made-up figure misses the critical point: the money the servicers saved is not even remotely the right basis for thinking about the appropriate settlement level. Settlements are based on potential liability. For instance, in 1998 the tobacco settlement, the tobacco companies agreed to pay a minimum of $206 billion over 25 years to be released from liability on Medicare lawsuits on health care costs plus private tort liability.
The saved costs bear no relationship to the banks’ legal liability for servicer-driven foreclosures, nor to the damage they have done to homeowners or broader society through their actions. It’s like basing the penalties in a robbery on the unpaid parking fees and rental costs of the car used to make the heist.
This resorting to completely irrelevant metrics results from the problem we have harped on from the onset of the settlement talks: the lack of investigations. You can’t settle what you haven’t investigated. The fact that Tom Miller has suddenly mentioned to an obscure mortgage industry rag that state banking regulators probed Ally is unpersuasive. First, Miller has a record for being less than truthful; he promised criminal investigations in no uncertain terms and has been walking that back ever since. Second, his own staff and various state attorneys general have effectively said there has been no investigation (as in they’ve at most gotten voluminous but undigested responses to subpoenas). You’d think state AGs would be aware of what their own state banking regulators were up to on a hot topic like foreclosure fraud. Third, I guarantee whatever thin “investigation” has taken place has overlooked the most important issue, and one that lay at the heart of the 2003 FTC/HUD examination of and settlement with rogue servicer Fairbanks: junk fees and misapplication of payments that push borrowers who’d otherwise be viable into foreclosure.
There’s more not to like in this document. For instance, on the second page, “Mortgage Servicing Settlement in Context,” under the column “Align Servicer Incentives,” we see the statement “Create a new trust structure outside existing RMBS which “traps cash” to align servicer and investor incentives.
Earth to base, this is a new variant on HAMP, which is pay the servicers to do mods. The only new wrinkle: taking the money from servicers and giving them the opportunity to earn it back. But as we saw with HAMP, the puny incentives provided by payments are dwarfed by the need to preserve the fictive value of over $400 billion of home equity loans and second mortgages, held by banks affiliated with the five biggest servicers. That, sports fans, means only shallow mods, when investors would prefer to take the hit of deep mods to viable borrowers, because the costs of foreclosure are even higher.
And we see the perverse program design on page 6, “Calibrating Breadth And Depth”. To be presented as some sort of success, the program needs to be able to tout large numbers of mods. Yet it is only clawing back a relatively small amount of money relative to the US negative equity hole (for starters, $480 billion on homes 50% or more underwater). So it’s going to focus on those only a little bit in negative equity land, which means it’s going to concentrate its efforts on those least in need of help. What is that going to do for the ground zeros of the housing crisis, such as Florida, Nevada, California, and Arizona? And what is it going to do to stem the losses investors are facing on foreclosures on deep negative equity homes, which from their perspective are the ones where mods make most sense? Apparently nothing.
This document looks to be rooted in Jean Baptiste Colbert’s saying: “The art of taxation consists in so plucking the goose as to get the most feathers with the least hissing”. Any number that was within hailing distance of the real damage done by foreclosure (which father of securitization Lew Ranieri was astonished to learn in 2008 was standard practice), rather than by doing mods for viable borrowers, would be a multiple of the levels under discussion here; and the servicer-driven foreclosure aspect pushes the figure higher still. This document bears the hallmarks of looking to rationalize a figure that would sound big enough to impress the public as being punitive, yet not hurt the banks at all (as page 4 demonstrates). But having a settlement designed around not damaging predators is certain to perpetuate their destructive conduct.