By Michael Olenick, creator of FindtheFraud, a crowd sourced foreclosure document review system (still in alpha). You can follow him on Twitter at @michael_olenick or read his blog, Seeing Through Data
A post last week reviewed a botched foreclosure for a mortgage loan in Ace Securities Home Equity Loan Trust 2007-HE4 dismissed with prejudice, meaning that the foreclosure cannot be refilled; a total loss for investors. Next, we reviewed why the trust has not yet recorded the loss despite the six month old verdict.
As an experiment, I gave my six year-old daughter four quarters. She just learned how to add coins so this pleased her. Then I told her I would take some number of quarters back, and asked her how many I should take. Her first response was one – smart kid – then she changed her mind to two, because we’d each have two and that’s the most “fair.” Having mastered the notion of loss mitigation and fairness, and because it’s not nice to torture six year-old children with experiments in economics, I allowed her to keep all four.
When presented with a similar question – whether to take a partial loss via a short-sale or principal reduction, or whether to take a larger loss through foreclosure – the servicers of ACE2007-HE4 repeatedly opt for the larger losses. While the dismissal with prejudice for the Guerrero house is an unusual, the enormous write-off it comes with through failure to mitigate a breach – to keep overall damages as low as possible – is common. When we look more closely at the trust, we see the servicer again and again, either through self-dealing or laziness, taking actions that increase losses to investors. And this occurs even though the contract that created the securitization, a pooling and servicing agreement, requires the servicer to service the loans in the best interest of the investors.
Let’s examine some recent loss statistics from ACE2007-HE4. In May, 2012 there were 15 houses written-off, with an average loss severity of 77%. Exactly one was below 50% and one, in Gary, IN, was 145%; the ACE investors lent $65,100 to a borrower with a FICO score of 568 then predictably managed to lose $94,096. In April, there were 23 homes lost, with an average loss severity of 82%, three below 50%, though one at 132%, money lent to a borrower with an original FICO score of 588.
Of course, those are the loans with finished foreclosures. There are 65 loans where borrowers missed at least four consecutive payments in the last year with yet there is no active foreclosure. Among those are a loan for $593,600 in Allendale, NJ, where the borrower has not made a payment in about four years, though they have been in and out of foreclosure a few times during that period. It’s not just the judicial foreclosure states; a $350,001 loan in Compton, CA also hasn’t made a payment in over a year and there is no pending foreclosure.
There is every reason to think the losses will be higher for these zombie borrowers than on the recent foreclosures. First, every month a borrower does not pay the servicer pays the trust anyway, though the servicer is then reimbursed the next month, mainly from payments of other borrowers still paying. This depletes the good loans in the trust, so that the trust will eventually run out of money leaving investors holding an empty bag. And on top of that, when the foreclosure eventually occurs, the servicer also reimburses himself for all sorts of fees, late fees, the regular servicing fee, broker price opinions, etc. Longer times in foreclosure mean more fees to servicers. Second, the odds are decent that the servicers are holding off on foreclosing on these homes because the losses are expected to be particularly high. Why would servicers delay in these cases? Perhaps because they own a portfolio of second mortgages. More sales of real estate that wipe out second liens would make it harder for them to justify the marks on those loans that they are reporting to investors and regulators. Revealing how depressed certain real estate markets were if shadow inventory were released would have the same effect.
These loans will eventually end up either modified or foreclosed upon, but either way there will be substantial losses to the trust that have not been accounted for. Of course, this assumes that the codes and status fields are accurate; in the case of the Guerreros’ loan the write-off – with legal fees for the fancy lawyers who can’t figure out why assignments are needed to the trust – is likely to be enormous. How much? Nobody except Ocwen knows, and they’re not saying.
Knowing that an estimated loss of 77%, is if anything an optimistic figure, even before we get to the unreported losses on the Guerrero loan, it seems difficult to understand why Ocwen wouldn’t first try loss mitigation that results in a lower loss severity. If they wrote-off half the principal of the loan, and decreased interest payments to nothing, they’d come out ahead.
Servicers give lip service to the notion that foreclosure is an option of last resort but, only when recognizing losses, do their words seem to sync with their behavior. But it’s all about the incentives: servicers get paid to foreclose and they heap fees on zombie borrowers, but even with all sorts of HAMP incentives, they don’t feel they get paid enough to do the work to do modifications. Servicers are reimbursed for the principal and interest they advance, the over-priced “forced placed insurance” that costs much more and pays out much less than regular insurance, “inspections” that sometimes involve goons kicking in doors before a person can answer, high-priced lawyers who can’t figure out why an assignment is needed to bind a property to a trust, and a plethora of other garbage fees. They’re like a frat-boy with dad’s credit-card, and a determination to make the best of it while dad is still solvent.
Despite the Obama campaign promise to bring transparency to government and financial markets, the investors in trusts remain largely unknown, so we’re not sure who bears the brunt of the cost of Ocwen’s incompetence in loss mitigation (to be fair Ocwen is not atypical; most servicers are atrocious). But, ACE2007-HE4 has a few unique attributes allowing us to guess who is affected.
ACE2007-HE4 is named in a lawsuit filed by the Federal Housing Finance Agency (FHFA), which has sued ACE, trustee Deutsche Bank, and a few others citing material misrepresentations in the prospectus of this trust. As pointed out in the prior article, both the Guerreros’ first and second loans were bundled into the same trust – so there were definitely problems – though the FHFA does not seem to address that in their lawsuit.
With respect to ACE2007-HE4, the FHFA highlights an investigation by the Financial Industry Regulatory Authority (FINRA), which found that Deutsche Bank “‘continued to refer customers to its prospectus materials to the erroneous [delinquency] data’”even after it ‘became aware that the static pool information underreported historical delinquency rates.”
The verbiage within the July 16, 2010 FINRA action is more succinct: “… investors in these 16 subsequent RMBS securitizations were, and continue to be, unaware that some of the static pool information .. contains inaccurate historical data which underreported delinquencies.” FINRA allowed Deutsche Bank to pay a $7.5 million fine without either admitting or denying the findings, and agreed never to bring another action “based on the same factual findings described herein.”
Despite the finding and the fine, FINRA apparently forgot to order Deutsche Bank to knock off the conduct, and since FINRA did not reserve the right to circle back for a compliance check maybe Deutsche Bank has the right to produce loss reports showing whatever they wish to.
It is unlikely that Deutsche Bank had trouble paying their $7.5 million fine since the trust included an interest swap agreement that worked out pretty well for them. Note that these swap agreements were a common feature of post 2004 RMBS. Originators used to retain the equity tranche, which was unrated. When a deal worked out, that was nicely profitable because the equity tranche would get the benefit of loss cushions (overcollateralization and excess spread). Deal packagers got clever and devised so-called “net interest margin” bonds which allowed investors to get the benefit of the entire excess spread for a loan pool. The swaps were structured to provide a minimum amount of excess spread under the most likely scenarios. But no one anticipated 0% interest rates.
From May, 2007, when the trust was issued, to Oct., 2007, neither party paid one another. In Nov., 2007, Deutsche Bank paid the trust $175,759.04. Over the next 53 months that the swap agreement remained in effect the trust paid Deutsche Bank $65,122,194.92, a net profit of $64,946,435.88. Given that Deutsche traders were handing out t-shirts reading “I’m Short Your House” when this trust was created, I can see why they’d bet against steep interest rates over the next five years, as the Federal Reserve moved to mitigate the economic fallout of their mischievousness with low interest rates.
In any event, getting back to Fannie Mae and Freddie Mac (the FHFA does not disclose which), one of the GSEs purchased $224,129,000 of tranche A1 at par; they paid full freight for this fiasco. Since this trust is structured so that losses are born equally by all A-level tranches once the mezzanine level tranches are destroyed by losses, which they have been, to find the party taking the inflated losses you just need to look in the nearest mirror. Fannie and Freddie are, of course, wards of the state so it is the American taxpayer that gets to pay out the windfall to the Germans. In this we’re like Greece, albeit with lousier beaches and the ability to print more money.
If the mess with the FHFA and FINRA were not enough, ACE2007-HE4 is also an element in the second 2007 Markit index, ABX.HE.AAA.07-2, a basket of tranches of subprime trusts that – taken as a whole – show the overall health of all similar securities. This is akin to being one of the Dow-Jones companies, where a company has its own stock price but that price also affects an overall index that people place bets on. Tranche A-2D, the lowest A-tranche, is one of the twenty trusts in the index. Since ACE2007-HE4 is structured so that all A-tranches wither and die together once the mezzanine level tranches are destroyed it has the potential to weigh in on the rest of the index. Therefore, the reporting mess – already known to both the FHFA and FINRA – stands to be greatly magnified.
The problems with this trust are numerous, and at every turn, the parties that could have intervened to ameliorate the situation failed to take adequate measures.
First there is the botched securitization, where a first and second lien ended up in the same trust. Then, there is failure to engage in loss mitigation, with the result that refusing to accept the Guerrero’s short-sale offers or pleas for a modification, resulting in a more than 100% loss. Next, there is defective record-keeping related to that deficiency and others like it. And the bad practices ensnarled Fannie /Freddie when they purchased almost a quarter billion dollars of exposure to these loans. Then there’s the mismanaged prosecution by FINRA, where they did not require ongoing compliance, monitoring, or increasing fines for non-compliance. There’s the muffed FHFA lawsuit, where the FHFA did not notice either the depth of the fraud, namely two loans for the same property in the same trust, and that the reporting fraud they cited continues. I’m not sure if the swap agreement was botched, but you’d think FINRA and the FHFA would and should do almost anything to dissolve it while it was paying out massive checks every month. Finally, returning full circle, there’s the fouled up foreclosure that the borrowers fought only because negotiations failed that resulted in a the trust taking a total loss on the mortgage plus paying serious legal fees.
It is an understatement to say this does not inspire confidence in any public official, except Judge Williams, the only government official with the common sense to lose patience with scoundrels. We’d almost be better off without regulators than with the batch we’ve seen at work.
US taxpayers would have received more benefit by burning dollar bills in the Capitol’s furnace to heat the building than we received from bailing out Fannie, Freddie, Deutsche Bank, Ocwen, and the various other smaller leaches attached to the udder of public funds. We could and should have allowed the “free market” they worship to work its magic, sending them to their doom years ago. That would have left investors in a world-o-hurt but, in hindsight, that’s where they’re ending up anyway with no money left to fix the fallout. It is long past time public policy makers did something substantive to rein in these charlatans.
My six year-old daughter understands the concept of limiting losses to the minimum, and apportionment of those losses in the name of fairness. Maybe Tim Geithner should take a lesson from her about this “unfortunate” series of events, quoting Judge Williams, before wasting any more money that my daughter will eventually have to repay.