The question of who belongs in a class action deserves a lot of think about it time. A good class definition may save class plaintiffs lots of trouble in winning certification of the class — a do-or-die event in the life of the class action.

A new ruling by the Second Circuit highlights that true fact. In re Petrobras Securities, No. 16-1914-cv (2d Cir. July 7, 2017).

Something rotten in the state of Brazil

Petrobras involved the government-owned Brazilian oil giant Petroleo Brasileiro S.A. (Petrobras). By 2015, the one-time $310 billion behemoth had shrunk into a sluggard worth just $39 billion. The cause? A “cartel” of vendors had for many years charged the entity far too much for goods and services. Cartel members kicked back excess payments to execs and government officials, inflating the carrying value of assets way above their actual worth while also pumping up the market price of Petrobras securities.

After discovery of the scheme, the prices of Petrobras securities plunged. Class action lawsuits for securities fraud ensued.

The cases landed before U.S. District Judge Jed Rakoff in Manhattan. The class-action complaints alleged fraud claims under the Securities Exchange Act of 1934 with respect to purchases of Petrobras American Depository Receipts (equity). They also asserted claims under the Exchange Act and its more dangerous sibling, the Securities Act of 1933, for purchases of Petrobras “notes” (debt) in over-the-counter transactions traceable to the notes’ initial public offerings.

Unlike the ADRs, which traded on the New York Stock Exchange, the notes changed hands not in over-the-counter transactions.

Two big issues arising from Supreme Court decisions confronted the parties and the court: an “extraterritoriality” question under Morrison v. National Australia Bank Ltd., 561 U.S. 247 (2010) and one about a “presumption of reliance” under Basic Inc. v. Levinson, 485 U.S. 224 (1988).


In Morrison (post here), the Supreme Court held that federal securities laws generally reach only “domestic” purchases and sales. Applying Morrison, the Second Circuit held that “for ‘securities that are not traded on a domestic exchange,’ a transaction is considered ‘domestic if [1] irrevocable liability is incurred or [2] title passes within the United States.'” Petrobras, slip op. at 22 (quoting Absolute Activist Value Master Fund Ltd. v. Ficeto, 677 F.3d 60, 67 (2d Cir. 2012)).


Basic addressed a key concern in securities class actions. Plaintiffs must show that members of the class relied on false information when they bought the securities. But no class action could proceed if the plaintiffs have to prove the “reliance” element on an individual basis. Individual issues would likely overwhelm ones common to the class (the making of false representations or omissions, intent to defraud, and materiality of the false information).

The Supreme Court ruled in Basic that a court may presume that all purchasers relied if they alleged that defendants withheld important facts from public disclosures and purchased at a “market” price that reflected all publicly available information.

As the Court confirmed in Halliburton Co. v. Erica P. John Fund, Inc., 134 S. Ct. 2398, 2414 (2014) (post here), the presumption applies if the security traded in a “generally efficient market”.


Judge Rakoff granted the plaintiffs’ motion to certify the case as a class action for damages. He held that common issues predominated over individual ones.

Although touching on the Morrison “domesticity” question, Judge Rakoff did not expressly tie it to the predominance issue. But he did specifically rule on the “efficient market” element of the Basic presumption, holding that on balance the plaintiffs had shown market efficiency.


Petrobras urged the Second Circuit to reverse because, it claimed, the Morrison and Basic issues made the case too messy to handle as a class action.

Speaking through U.S. District Judge Nicholas G. Giraufis, who sits in Brooklyn, the 3-0 panel made four notable rulings:

  • The court does not review grants of class certification more leniently than denials.
  • Rule 23 does not impose an “administrative feasibility” test for determining who qualifies as a member of the class (the “ascertainability” requirement the court recognized in Brecher v. Republic of Argentina, 806 F.3d 22, 24 (2d Cir. 2015)).
  • Judge Rakoff erred in not expressly addressing whether the methods for proving the domesticity of securities purchases by class members would make individual issues predominate under Rule 23(b)(3).
  • Judge Rakoff did not err when he found the market in Petrobras securities efficient in spite of shortcomings in expert evidence (event studies) aiming to show by direct means that the prices of the securities responded to disclosure of material information.

What it means

Petrobras helps class actions more than it hurts them. The panel’s rejection of a new requirement — the Third Circuit’s nightmarish “administrative feasibility” test — assures that conclusion on its own.

The court’s handing out of a passing grade on the Basic gives further cause for happiness on the plaintiff side.

Even the panel’s handling of the Morrison question went fairly well. Rather than rule that proving “domesticity” will defeat predominance of common questions, Judge Giraufis’s opinion merely vacates the certification order and remands the case for Judge Rakoff to grapple with how best to address the need to show that irrevocable liability or passage of title occurred in the United States with respect to the Petrobras notes.

Which brings us back to the importance of class definition. A class that involves only ADR buyers will have no Morrison issue; all relevant sales took place on the NYSE.

A class that also covers notes may require further evidence, possibly specific to the individual buyer. If standard forms of documentation can make the process a mechanical one, no real difficulties should arise. But the availability of documentation that clinches the domesticity question may apply to some but not all of the issuances of the notes and may apply to some types of buyers but not others.

Revising the class definition to address the practical availability-of-proof concern may therefore make a ton of sense. If redefinition makes for a smaller class that is nonetheless certifiable, the loss in size will not matter. Because the alternative could be losing all.