A rough theory you could have of the mortgage-backed securities problem goes like so:
- Banks packaged and sold a lot of mortgages to a lot of people.
- Those people were in the business of backing up the truck to buy lots of paper with a certain rating, and weren't poring through documents and doing exhaustive analytical credit work.
- So the banks cut a lot of corners.
- Then the securities lost a lot of value for reasons that were partly the banks' fault (the corner-cutting) and partly not (general economic decline), in ways that are hard to separate from each other.
- So everyone sued.
- So the banks reached settlements for large round numbers.
- With people who, remember, were not predisposed by nature to poring through documents and doing exhaustive analytical work.
So it's not a huge surprise that some people think those are lowball settlements? Here is a pretty interesting DealBook article about those people, and particularly Fir Tree Partners, who have come up with the clever idea of buying a bunch of those bonds, rejecting the settlements, and seeking larger, less-round numbers from the banks.
I always enjoy good maneuvering-by-tender-offer and I think this qualifies. Fir Tree has launched a tender offer for about $300 million or so worth1 of bonds from a handful of the 330 JPMorgan residential mortgage backed securities trusts that are part of a $4.5 billion settlement that JPMorgan reached with 21 institutional investors a while back. A fun fact about that settlement is that it's not really a settlement, and the trustees of those trusts aren't involved: It's just an agreement between JPMorgan and some investors in the bonds; once they reached the agreement they took it to the trustees. The trustees haven't yet said yes.2
Another fun fact: The big investors who reached the agreement with JPMorgan aren't required to hold on to their bonds. So they can just sell if they're offered a better deal.3 If enough people sell, then Fir Tree will have a lot of bonds and can make the trustees sue instead of settling. (Just for the handful of bonds Fir Tree is after, but I guess they might inspire copycats.) This feels like a little bit of an oversight by JPMorgan? Like, they reached a big settlement with a bunch of big bondholders, but the bondholders are free to abandon it if they get a better deal. On the other hand the investors are people like BlackRock and Pimco and Fannie Mae and Goldman Sachs Asset Management and MetLife and Prudential and TIAA-CREF so I guess you can't just tell them to stop selling bonds.
Will a lot of people sell? I mean, what do I know? The Fir Tree people think they're offering a better deal, and sure, why not. They think that JPMorgan's settlement offers a recovery of about 7.5 percent of losses (and that uncertainly and in the distant future), while their offer covers about 10 percent of losses (and in cash in February).4 Ten is more than 7.5. Sure. In any case, the bonds seem to be trading up since the offer launched, which suggests that their offer has a higher expected value than JPMorgan's.5 So, by arithmetic, lots of people should tender and then JPMorgan should be back in the soup.
But even if that happens, the soup will remain pretty soupy. DealBook points out that the New York Appellate Division seems to think that claims like Fir Tree's are time-barred. It would be awkward for them if they bought hundreds of millions worth of bonds and then found that they had no claims. (Though it would be even awkwarder for JPMorgan if it agreed to pay $4.5 billion for claims that turned out to be worth zero?6 )
Also, maybe Fir Tree is insider trading, though I don't particularly buy that.7
It's easy to feel like this is a funny system. It is 2014, you know? You could imagine approaches that would have solved this problem in 2006, or 2009, or at least 2013 when JPMorgan entered this settlement. For instance, the sophisticated investors such as Pimco and BlackRock and Goldman Sachs who bought all this RMBS in 2006 and 2007 could have done a bit more due diligence than they did. This is a tough one because, by most accounts, the offering documents were full of lies, but on the other hand they were also full of warning signs that might have popped up under more rigorous scrutiny.
Or, if BlackRock et al. really are getting lowballed now, then you'd think they would, like, try harder. They're smart guys with lots of clout; why accept less for these claims than they're worth?8
But the Fir Tree system actually seems pretty likable. Sure, there's something to be said for big investors doing some due diligence before buying massive amounts of an asset class that ends up going pear-shaped, rather than relying on banks and ratings agencies to look out for their best interests. But there's also something to be said for being able to rely on banks and ratings agencies to look out for investors' best interests. That is sort of what they're there for? There is a real value to trustworthy financial gate-keeping, and real efficiency savings if investment managers don't have to carefully scrutinize every deal for fraud. The prospect of beady-eyed hedge funds coming in years after the fact and holding banks to account for everything they get wrong in underwriting is what's supposed to keep their underwriting honest. Sadly, this effect has only limited ability to travel back in time but, you know, next time.
Similarly, sure, BlackRock and Goldman and whoever could pursue a scorched-earth fight against JPMorgan over JPMorgan's alleged misdeeds. But that has problems too. Risk, for one thing: Fir Tree is offering certainty by cashing out holders, but has no certainty that it'll make its money back. That might be a positive-expected value risk for a hedge fund, but less appealing to a big institution.
There are other reasons for the big investors to avoid a vicious fight. BlackRock is JPMorgan's biggest shareholder. (Goldman is number 22.) And BlackRock remains in the business of going around buying tons and tons of bonds every day, so it has to keep up good relationships with all the banks that sold it garbage in 2007. Fir Tree can burn some bridges if the price is right.
The other likable thing about this system is that it seems designed to get the right answer. I mean, who knows what a "right" answer would look like: There are lots of uncertainties and feedback loops and statistical problems with deciding how much of the losses on these securities are JPMorgan's "fault," and how much are losses due to economic circumstances that are the investors' problem. Most of the settlements reached so far, with governments or investors, feel a bit like picking numbers out of a hat.
Here, though, everyone's set up to fight for the right answer. And nobody's all that sympathetic. On one side you have JPMorgan, everyone's pretty mad at JPMorgan. On the other hand you have the least sympathetic possible adversary in Fir Tree. They're a hedge fund! They're not a victim: They were never deceived, and didn't actually suffer any losses. They just bought up claims from the people who did. (And, delightfully, they might have been insider trading a little when they bought those claims. From, remember, the real victims.9 ) They're cold-blooded profiteers.
So that's perfect, right? What you don't want is a system settled by a public relations balancing of the equities among parties who have repeated and interwoven dealings with each other. That's the sort of system that leads to ad hoc, round-number settlements that do things like make mortgage bond investors pay for the banks' misconduct in selling mortgage bonds to them. What you want is just a pair of unsympathetic jerks in a courtroom arguing fiercely about what the right answer is. Fir Tree has helpfully supplied the jerk.
1 That's my math and it's pretty loose. I read the tender offer document as covering bonds from twelve trusts, with an original principal amount of about $2.1 billion and a current principal amount of about $300 or so million. Fir Tree is on average offering a premium to current principal amount, though it varies from bond to bond. Anyway though that number assumes that Fir Tree owns none of the bonds (surely false, but they don't disclose how much they do own) and gets 100 percent of them in the tender. So really rough numbers here.
2 They had until January 15 to say yes, but they got an extension.
3 Section 7.01 of the settlement agreement requires them to hold enough bonds to have standing to intervene in judicial proceedings, but otherwise "this Agreement shall not restrict the right of any Institutional Investor to sell or exchange any security issued by a Settlement Trust free and clear of any encumbrance."
4 Those are their numbers, which I do not pretend that I can perfectly replicate, though I am no RMBS expert. So, sure, whatever. Importantly the JPMorgan numbers are pretty vague; 7.5 percent is Fir Tree's estimate of total settlement amount ($4.5 billion) divided by total realized and expected losses ($60 billion-ish, by their estimate) in all of the 330 trusts subject to the settlement, so maybe these trusts would have a higher recovery, I don't really know.
5 Based on casual perusal of Bloomberg HP data for a few of the securities involved. SACO 2006-3 A1, for instance, traded at 120-30 before Fir Tree launched its tender last week, and is at 132-26 now. Other SACOs seem to be up similarly. BSMF 2006-SL1 A1 is up less (just under 62 to just under 62.5). JPMAC 2006-WMC4 A2 is up about three points, from 47 and change to 50 and change. I have no idea of the depth or reliability of these numbers but the trend seems up.
There's some tolling of the statute of limitations language in the settlement agreement. Still.
7 From DealBook:
The hedge funds may also face an uncomfortable type of counterattack that could complicate their litigation. The banks’ lawyers may question whether a hedge fund trading in litigated mortgage bonds had been unduly influenced by potentially material nonpublic information that the fund gained through the discovery process in their litigation.
I guess? I mean, they're buying from people who own bonds that are being litigated, and who are sort of part of that litigation. You'd hope that someone would send those people enough information to make an informed decision about whether the settlement is a good one, and you'd think that Fir Tree wouldn't be responsible. Anyway, Fir Tree's tender docs give a high-level (two-page) overview of the thought process in concluding that JPMorgan is lowballing, while also saying, as you'd expect, "The FT Offer Prices for this Offer may or may not reflect the current market value of the Eligible RMBS Securities. Holders should not rely on the FT Offer Prices as a definitive indication of value and should conduct their own independent analysis before making any decision to participate in the Offer." But what is insider trading anyway, etc. etc. etc.
8 Of course one entirely plausible answer is that they're worth $4.5 billion and Fir Tree is just wrong and overpaying. Entirely plausible! The Times quotes a Nomura analyst saying that "The payouts will likely be higher" for Fir Tree than for the settlement, but who knows.
9 Or their successors. There's no special reason that all these bonds would be in the same hands now as they were in 2007.
To contact the writer of this article: Matt Levine at email@example.com.
To contact the editor responsible for this article: Tobin Harshaw at firstname.lastname@example.org.