No “participation trophies” in class actions: Third and Fourth Circuits reject certification orders based on speculative evidence

Within a day of each other, the Third and Fourth Circuits both issued opinions overturning district court orders that certified damages classes based on flimsy proof. Though the cases differ, the common takeaway is that district courts remain obligated to conduct a rigorous analysis of class certification evidence and cannot accept generalized theories, recitations of potential valuation approaches, or bare statistical correlations.

In In re The Boeing Co. Securities Litigation, No. 25-1492 (4th Cir., July 20, 2026), the plaintiff shareholders alleged Boeing made dozens of false statements about safety and quality following airline crashes and an emergency landing incident. Plaintiffs claimed these misstatements artificially inflated or maintained stock prices, causing them to overpay and entitling them to out-of-pocket damages. But the plaintiffs never differentiated among the myriad misrepresentations alleged, nor did they commit to a price inflation vs. price maintenance theory of liability. As to damages, the plaintiffs’ expert provided a "non-exhaustive list of valuation techniques” for establishing out-of-pocket damages but failed to commit to a specific approach. Nevertheless, the district court certified a Rule 23(b)(3) damages class, ruling the plaintiffs’ out-of-pocket approach was an accepted damages methodology in securities cases and was aligned with their fraud-based liability theory.

The Fourth Circuit reversed, holding the plaintiffs and the district court failed to satisfy certification standards established by the U.S. Supreme Court in Comcast v. Behrend, 569 U.S. 27 (2013). Under Comcast, district courts must conduct a rigorous analysis to determine whether the plaintiffs’ damages methodology is non-speculative and sufficiently tied to their theory of liability. Here, rather than providing a liability theory, the plaintiffs simply described generic approaches applicable to "any securities fraud case." Compounding this problem, the plaintiffs’ expert never settled on a methodology, offering instead "a series of 'maybe[s],' 'perhaps[es]' and 'what ifs.'" Nor did an additional report from the expert—which applied a constant-percentage inflation figure based on the emergency landing incident—cure the problem, as the approach was inconsistent with the plaintiffs’ liability allegations, which involved dozens of separate misstatements covering different subjects over a multiyear period. As the Court put it, “certification orders are not like participation trophies that are handed out to everyone on the tee ball team.”

The Third Circuit's decision in In re Avandia Marketing, Sales Practices & Products Liability Litigation, No. 25-2278 (3d Cir., July 21, 2026), addressed a different but equally fundamental evidentiary gap: can statistical evidence prove class-wide causation in a pharmaceutical fraud case? The plaintiffs, third-party payors, alleged the defendant misrepresented a drug’s cardiovascular risks and benefits, causing physicians to prescribe the drug over cheaper alternatives, resulting in more reimbursements by the plaintiffs. The district court certified a Rule 23(b)(3) damages class based largely on the defendant’s marketing materials, its internal documents showing marketing campaigns increased prescriptions, and a drop in prescriptions following the release of a study that supposedly revealed the drug's true cardiac risks.

The Third Circuit vacated and remanded, finding the plaintiffs’ evidence oversimplified prescribers’ decisions and failed to establish on a class basis that prescribers relied on the allegedly fraudulent cardiovascular messaging as opposed to other factors. Simply put, plaintiffs failed to make the leap from correlation to causation. Joining the First, Second, and Ninth Circuits, the Third Circuit held that while plaintiffs in pharmaceutical fraud RICO actions may use statistical evidence to show causation, that evidence must be sufficiently robust to isolate the effects of the alleged fraud. Absent a regression analysis or similar methodology capable of distinguishing causation from mere correlation, the plaintiffs could not establish Rule 23(b)(3) predominance.

Although some district court decisions have characterized class certification disputes over the sufficiency of expert methodologies or statistical proof as “merits” issues not properly addressed at the certification phase, these Third and Fourth Circuit opinions show the Supreme Court’s “rigorous analysis” standard for class certification maintains its teeth. Accordingly, defendants facing statistical and/or expert evidence at the certification stage should aggressively evaluate and challenge such evidence through discovery, Daubert motions, competing expert opinions, certification briefing, and other available means.