Okay, fine, that’s not what the Second Circuit formally held, but to be honest, I can’t read this decision any other way.

I’ve blogged about this issue before here, here, and here.  Basically, the situation is this:  In the class action context, there is frequently an issue as to whether the named plaintiff’s own individual claims against the defendant are sufficiently similar to the claims of the rest of the class so as to allow the named plaintiff to sue in a representative capacity.  Historically, these issues have been resolved via Rule 23 of the Federal Rules of Civil Procedure, which, among other things, requires a court to decide whether there is “commonality” among the class members, whether the common issues predominate over the individual ones, whether the named plaintiffs’ claims are typical of those of absent class members, and whether the named plaintiff will serve as an adequate representative for the absent class members.  Rule 23, of course, is only invoked after there has been substantial discovery, and certification determinations under Rule 23 frequently include expert analysis.

In the wake of the mortgage crisis, more and more courts began making these determinations on the pleadings, framing the question not in terms of class certification, but in terms of whether the named plaintiff has “standing” to bring claims on behalf of absent parties, as I discussed in more detail here.  The issue has basically been that if an investment bank underwrites multiple RMBS offerings, and I buy an RMBS issued by a particular trust backed by a particular pool of mortgages, how can a court be certain that my claims are similar enough to purchasers of different RMBS issued by a different trust, backed by different mortgages, such that I should be permitted to represent those purchasers in a securities class action against the underwriter? 

Courts have been unwilling to go the traditional route and wait until a class certification hearing to make this decision; instead, they have been seeking to limit a named plaintiff’s ability to represent absent RMBS purchasers.  They have been fundamentally troubled by the idea that a purchaser of one RMBS could represent all purchasers not only of that RMBS, but of multiple other RMBS, with face values totaling in the billions of dollars.  Courts have come up with a variety of bright-line rules limiting how the class can be defined, at the pleading stage – for example, some courts have held that plaintiffs may only represent purchasers of RMBS from the same trust; others have held that plaintiffs may only represent purchasers of RMBS from the same tranche within a trust.

That orientation has spread to other kinds of claims – similar disputes have arisen in the context of false advertising, for example, where a single misrepresentation is alleged to have been plastered across multiple similar products.  (Say, a false representation that ice cream flavors are “natural,” appearing on chocolate, vanilla, and strawberry – is it necessary that the plaintiff have purchased neapolitan in order to represent absent purchasers of all three flavors?).

[More under the jump]

In my view, this dispute is why the Supreme Court ultimately dismissed the writ in the Indymac case, which was originally set to be heard this term.  In that case, the question presented was  whether the filing of a class action tolls the statute of repose with respect to absent class members.  As I discussed here, I believe the Court ultimately decided not to hear the case because it realized that it could not reach the tolling issue without delving in to the hotly contested – and unbriefed – prior issue as to whether a purchaser of one RMBS can represent absent purchasers of different RMBS in a single class action.

The latest iteration of this battle has now come in Ret. Bd. of the Policemen’s Annuity & Benefit Fund of Chi. v. Bank of N.Y. Mellon, 2014 U.S. App. LEXIS 24264 (2d Cir. N.Y. Dec. 23, 2014). 

In that case, the plaintiffs purchased RMBS issued from 26 trusts, all of which had the Bank of New York Mellon (“BNYM”) as trustee, and which contained mortgages that had been issued by Countrywide.  The plaintiffs claimed that BNYM had breached its fiduciary duties under the Trust Indenture Act, and its contractual obligations, by failing to pursue claims against Countrywide for selling nonconforming mortgages, and failing to repurchase flawed mortgages, to the 26 trusts.  The plaintiffs also sought to represent a class of purchasers who had bought RMBS from an additional 504 trusts, all of which had BNYM as trustee, and which were backed by Countrywide mortgages. 

Without allowing any discovery or expert analysis, and based solely on the pleadings, the Second Circuit held that the named plaintiffs did not have Article III standing to bring class claims on behalf of absent purchasers in the 504 additional trusts.  The Second Circuit reasoned that BNYM’s misconduct would have to be proved loan by loan, and trust by trust – such that plaintiffs who purchased from one trust would not be able to show, on a classwide basis, that BNYM had breached its duties to all of the trusts.  As the court put it, “For example, whether Countrywide breached its obligations under the governing agreements (thus triggering BNYM’s duty to act) requires examining its conduct with respect to each trust. Whether it was obligated to repurchase a given loan requires examining which loans, in which trusts, were in breach of the representations and warranties. And whether a loan’s documentation was deficient requires looking at individual loans and documents. We see no way in which answering these questions for the trusts in which Plaintiffs invested will answer the same questions for the numerous trusts in which they did not invest.”

This is, of course, nonsense.  It’s a direct contradiction of the court’s earlier decision in NECA-IBEW Health & Welfare Fund v. Goldman Sachs & Co., 693 F.3d 145 (2d Cir. 2012), which I discuss in more detail here.  In that case, the plaintiffs bought RMBS that had been issued and underwritten by Goldman Sachs, and sought to represent absent purchasers of other RMBS issued and underwritten by Goldman, alleging that all of the registration statements falsely described mortgage quality.  The Second Circuit held that so long as there was any originator overlap between the RMBS, one plaintiff (at least at the pleading stage) could represent all RMBS purchasers.  According to the Second Circuit, because the statements across each offering were so similar, if a plaintiff were able to show those statements were false for one originator whose loans were included in one trust, that would tend to prove that the statements were false for any other trust that included loans by the same originator.

In other words, in Goldman, the Second Circuit wasn’t concerned that different loans backing different RMBS might be subject to separate analysis – the only thing the court cared about in Goldman was that a plaintiff be able to show that a single originator generally conducted business badly – and that would be enough, for pleading purposes, to infer that bad loans  infected multiple RMBS trusts containing loans by the same originator.

Now, as I originally posted, that rule is both too broad and too narrow – which is precisely why the Second Circuit should not have made this a pleading issue at all, but instead should have held that this was an issue to be decided at the class certification stage.  Yet so long as that is the rule, there is no reason why that same rule should not have governed the claims in the BNMY case. 

(Imagine if one person was in charge of underwriting and repurchasing at Countrywide and that person was in a coma from 2004 to 2007 – if that’s the case, plaintiffs could prove their claims against all 530 trusts in a single deposition.  Is that what actually happened here?  I don’t know; the court didn’t allow discovery.).

The takeaway, then, is that according to the Second Circuit, class certification is to be decided on a case by case basis, on the pleadings, without discovery, based on the whim of the court, with different rules for different statutory claims.  This is utterly contrary to Rule 23, and, it seems to me, blatantly result oriented.  I think the Second Circuit was simply terrified by the numbers – 504 absent trusts, with uncountable amounts of money at stake.  But that’s simply not an appropriate consideration when making these judgments, and certainly is not a determination that should be made on  motion to dismiss.