Most readers were not happy when I didn’t buy into the mainstream presentation of a the widespread news reports that a letter sent on behalf of a group of investors constituting approximately $16.5 billion (per the Wall Street Journal) of $47 billion (presumably face amount) of bonds was a Really Big Deal in terms of the damage it might do to Bank of America.
Let’s start with the obvious. I’m no fan of Countrywide, in fact I was a early to criticize Bank of America’s staged purchase. I think mods are a great thing, and anything to promote mods is a plus. There is plenty of evidence that servicers behave badly, so the idea that Countrywide has behaved in ways that would make people eager to sue them is entirely credible. So you would think I’d like this case.
While the letter that the investors sent to Countrywide is laying the groundwork for litigation, any litigation is going to be more of an uphill battle and less lucrative than the breathless reports would lead you to believe. Part of this overreaction is in keeping with the excitement over similar putback litigation by the monolines, who despite their better grounds for lawsuits, is similarly overhyped. But the presence of famous names, particularly that of the New York Fed, has led this case-in-the-making to be treated as more damaging than it is likely to be.
And with misreporting to boot, no wonder Mr. Market got excited. This is from The Street.com:
The news follows reports earlier Tuesday that institutional investors are pressuring Bank of America to repurchase $47 billion worth of mortgage-backed securities.
No, sports fans, the investors group is not seeking a repurchase, but a putback of bad loans, and that is a subset of the total value of the deal. Again, recall that the investor press release reports that they hold 25% of voting interest; the Wall Street Journal puts their holdings at $16.5 billion (which we are assuming is face amount) of the total.
The underlying beef, in simple form, is that Countrywide has not done right by the investors as a servicer and and really ought to have put a lot of loans back to the originator, which in this case happens to be Countywide, but almost certainly a different legal entity (we have Countrywide as Master Servicer, identified in the documents as Countrywide Home Loans Servicing LP ;note that the Bank of New York is also a recipient of the letter but per its own assessment it not presently a target). Here is the juicy part of the letter:
1. Section 2.03(c) of the PSAs states that “Upon discovery by any of the parties hereto of a breach of a representation or warranty with respect to a Mortgage Loan made pursuant to Section 2.03(a) … that materially and adversely affects the interests of the Certificateholders in that Mortgage Loan, the party discovering such breach shall give prompt notice thereof to the other parties.” The Master Servicer has failed to give notice to the other parties in the following respects:
a. Although it regularly modifies loans, and in the process of doing so has discovered that specific loans violated the required representations and warranties at the time the Seller sold them to the Trusts, the Master Servicer has not notified the other parties of this breach;
b. Although it has been specifically notified by MBIA, Ambac, FGIC, Assured Guaranty, and other mortgage and mono-line insurers of specific loans that violated the required representations and warranties, the Master Servicer has not notified any other parties of these breaches of representations and warranties;
c. Although aware of loans that specifically violate the required Seller representations and warranties, the Master Servicer has failed to enforce the Sellers’ repurchase obligations, as is required by Section 2.03; and,
d. Although there are tens of thousands of loans in the RMBS pools that secure the Certificates, the Trustee has advised the Holders that the Master Servicer has never notified it of the discovery of even one mortgage that violated applicable representations and warranties at the time it was purchased by the Trusts.
The real significance of this move is political. It is a new front in the battle between investors and banks. However, this measure isn’t as radical as it sounds; Freddie and Fannie have been putting back certain types of bad mortgages to major banks for some time, which has led to an ongoing drain to major bank earnings. But Freddie and Fannie deals provide for relatively straightforward putback provisions. The process here is procedurally far more difficult, and establishing damages is also more cumbersome and costly. That means the odds of success and the level of any payout are likely to be lower than most assume. Note that a secondary objective is for Countrywide to accelerate its handling of delinquent loans. From the first report on this story, by Jody Shenn of Bloomberg:
If Countrywide doesn’t correct the servicing problems within a few months, her clients could have the right to pursue legal action against Bank of America, Bank of New York or both, she said. “None of the bondholders are opposed to modifications for deserving borrowers, but you’ve got to get it done” in a timely fashion, she added.
Let’s look at the major issues:
This is not going to play out quickly. The group has sent a “Notice of Non-Performance,” which is intended to start a 60 day period for Countrywide to remedy the alleged breaches. Countrywide is likely to adopt a posture of foot-dragging, for instance, by saying they need more time to conduct their review. And after that period is done, Countrywide is likely to reply that it found no (or perhaps very few) material breaches that would justify investor action. So this has to go at least a round, perhaps longer, before any litigation will be filed to declare Countrywide in default of its servicing obligations.
Any lawsuit has to pass procedural hurdles. The Bloomberg article mentioned that the investors may lack standing to sue and that is not a non-issue. The problem is that unlike Fannie and Freddie, the investors have to get to Countrywide-as-originator through Countrywide-as-servicer. Now that language from the letter seemed really strong, right? Surely there is a problem here……but wait! Consider this discussion from Subprime Shakeout:
Nevertheless, these efforts may well fail for an additional reason that was cited as a basis for Bank of New York’s refusal to comply with Patrick’s earlier request – the failure to provide evidence of a specific breach. Though Patrick’s letter is reported to identify several provisions of the relevant PSAs that it alleges were violated, it’s unclear what, if any, specific evidence Patrick has provided that would induce the trustee to act. A Bank of New York spokesman has already indicated that the trustee will not act in response to this letter, stating, “[The letter] appears to be directed to Countrywide and does not ask BNY Mellon to take any action. We will continue to perform our duties as trustee.”
Yves here. That comment was written without having had the benefit of reviewing the actual text of the letter. The problem in suing is a bit circular: you need to be able to argue specifically what sort of breaches occured, but the investors lack access to the loan information to enable them to refer to specific breaches in a lawsuit. Or to put it in a bit more legalistically, unless the pooling and servicing agreement gives the certificateholders some rights to access to the loan files, you are likely to have an impediment: you may not be able to allege specific enough breaches to get to perform discovery.
The finesse appears to be that the investors are trying to piggyback on the actions of various monolines who are also suing Countrywide and argue that because they are suing for breaches, Countrywide failed in its duties to inform them (note the monoline contracts with the originators give them much easier access to the loan level information. But the problem is these monoline claims are mere allegations; Countrywide is disputing them. Even if it had an obligation to notify the certificate holders, it seems hard to believe it would extend to having to provide the loan information, which is what the investors really need.
Rep and warranty suits tend not to produce big settlements. Even if the investor group can get access to the loan files, it has to argue its breaches on a loan by loan basis. It needs to prove that the loan defaulted not for normal credit loss reasons, such as death, disability, job loss, but due to failure to adhere to the representations and warranties made. (see ScribD for an illustrative set of reps and warranties, pages 3-6). This is brutal for both sides to pursue, so the two sides tend to settle rather than get very far with pursuing the case. And Bank of America has indicated it is well aware of this issue and will fight hard:
“It’s loan by loan, and we have the resources to deploy in that kind of review,” said Brian T. Moynihan, Bank of America’s chief executive, on a conference call to discuss the bank’s results for the third quarter.
This gives Bank of America more than a bit of a home court advantage. It can deploy comparatively cheap bank employees to do this analysis; the investors will have to use more costly experts and law firm resources.
Although the sample of my sources is limited, their experience with rep and warrranty cases is that the plaintiff only recovers 10-20% of the amount of credit losses. Now these loans were really drecky; we might assume 25%, and Bank of America has been reserving 1/3 against other rep and warranty cases (but their language says this is all over the map, so the reserves may also vary a lot by deal. It’s also possible that BofA reserves generously for this sort of litigation). But let’s do some math, and readers are invited to chip in.
Normally, you can sue only based on actual losses, not expected losses. On subprime deals, these are only running at 10% thus far (the 28%+ loss estimates for all subprime RMBS are total expected losses, only a portion of which have been realized). I’d wonder if the reason Bank of America’s reserves are so high is that there are such firm forecasts of future subprime losses that they are also reserving for a portion of expected losses. So let’s do some rough math; you can use your own assumptions for damage percentages.
The key bit is subrime losses are only 10% of original par amount. Principal paydowns are about 50% on these deals. Of the remaining 40%, the expected losses are about 40-45%.
So if you take the $16.5 billion the investors own x 10% (losses) x 33% (losses due to rep and warranty breach) = $545 million
If you assume they get a lesser percentage on expected losses, say 15% (I’ll be generous and assume 20%), then the math is:
$16.5 billion x 40% (remaining value) x (45%) expected losses x 20% (amount deemed due to rep and warranty breaches) = $594 million.
So the two together take you to just over $1 billion.
The one factor that could make these numbers much larger is if the bond amounts reported in the media are market value, not face value, then I would need to apply much higher percentages to get the right loss and expected loss amounts.
But as you can see, the multi-billion claim looks to be a stretch. Given that the attorney made a big procedural misstep on her first effort, I would not take her estimates of recoveries as seriously as I might otherwise.
Update 2:00 PM: Received this from a source who knows the bond insurer litigation (and is no fan of banks, this is not an industry friendly source):
The bond insurers currently have access to the mortgage loan data and they have not been able to prove widespread fraud (or even widespread breaches), so I am skeptical that there is really a goldmine here for investors. What he is describing is a fishing expedition. There may have been a case or two of a loan reported to be 70% LTV but it really was 100% and that would be material, but I am fairly confident that isn’t widespread.