Securities Class Certification Requires Weighing Market-Inefficiency Evidence Beyond the Cammer/Krogman Factors

1. Introduction

In Bozorgi v. Cassava Sciences, the Fifth Circuit clarified how courts must assess market efficiency when investors seek class certification in securities-fraud litigation. The court held that the familiar Cammer/Krogman factors are analytical tools, not an exclusive checklist. A district court must also address evidence suggesting that a security’s market did not efficiently incorporate public information, including evidence of unusual price movements and constraints on short selling.

The investors alleged that Cassava Sciences, Incorporated, and executives Remi Barbier, Eric J. Schoen, and Lindsay Burns misrepresented research concerning simufilam, an experimental Alzheimer’s treatment. Their claims arose under sections 10(b) and 20(a) of the Securities Exchange Act and Rule 10b-5.

The central appellate question was procedural, not whether the alleged fraud occurred: could the district court certify an investor class after dismissing defendants’ evidence of “meme-stock dynamics” without meaningfully weighing that evidence against the plaintiffs’ showing of market efficiency?

The Fifth Circuit answered no. It vacated certification and remanded, while expressly declining to determine whether Cassava’s stock actually traded in an efficient market. The decision principally reinforces existing law rather than creating a categorical exception for meme stocks.

2. Background and Procedural History

The alleged misconduct

Cassava developed simufilam beginning in 2011. After an unsuccessful efficacy study in May 2020, its stock price declined. According to the complaint, defendants subsequently promoted supposedly promising results while concealing data manipulation, research anomalies, and conflicts of interest. Investors alleged that favorable statements in press releases, SEC filings, and scientific journals increased Cassava’s stock price, followed by declines when contrary information emerged.

These were allegations, not findings of fraud. The opinion also noted separate SEC proceedings that resulted in settlements and civil penalties; those proceedings did not resolve the certification question before the court.

Class certification and expert evidence

The proposed class covered persons and entities that purchased or otherwise acquired Cassava securities between September 14, 2020, and October 12, 2023. The investors first moved for certification in March 2024, and a magistrate judge recommended granting the motion.

Following that recommendation, defendants filed a Daubert motion seeking to exclude a plaintiffs’ expert. The district court struck the motion as untimely, denied the initial certification motion without prejudice because of changed circumstances, permitted an amended complaint, and directed renewed certification briefing without new arguments or evidence.

The existing record included competing economic reports from Dr. Steven P. Feinstein for plaintiffs and Dr. Rene M. Stulz for defendants. The district court ultimately certified the class, appointed class representatives and counsel, and accepted plaintiffs’ market-efficiency showing. Defendants obtained permission to appeal under Rule 23(f).

The intervening settlement

After the appeal began, the parties reached a binding settlement agreement. Plaintiffs argued that the settlement mooted the appeal. The Fifth Circuit rejected that argument because the proposed class settlement still required judicial approval and expressly preserved the appeal.

3. Summary of the Opinion

1. The appeal remained live. A proposed class settlement requiring Rule 23(e) approval did not eliminate the controversy before approval occurred.
2. Market efficiency required a complete evidentiary assessment. The district court could not treat the Cammer/Krogman factors as an exhaustive certification checklist.
3. The defendants’ evidence required substantive consideration. Price volatility without news and constraints on short selling could undermine informational efficiency, even without an expert expressly labeling Cassava a “meme stock” or its market “inefficient.”
4. The certification order rested on legal error. By declining to weigh contrary evidence, the district court abused its discretion.
5. The ultimate efficiency question remained open. The appellate court vacated and remanded rather than deciding that certification must be denied.

The court did not resolve defendants’ remaining certification and procedural challenges. It also granted judicial notice in part and denied it in part, accepting the relevant court filing and SEC disclosure while finding it unnecessary to determine whether a press release was judicially noticeable.

4. Legal Framework and Reasoning

A. Why reliance controlled the predominance analysis

A Rule 23(b)(3) class requires common legal or factual questions to predominate over questions affecting individual members. Securities-fraud claims present a recurring difficulty: if every investor must separately prove reliance on a particular misleading statement, individualized inquiries may overwhelm common issues.

The opinion identified six elements of a section 10(b) and Rule 10b-5 claim: a material misrepresentation or omission, scienter, a connection with a securities transaction, reliance, economic loss, and loss causation. It focused on reliance because the investors sought to establish that element through the fraud-on-the-market presumption.

Under that presumption, an investor who trades at a market price may be treated as relying indirectly on public misrepresentations incorporated into that price. The opinion described the underlying predicates as publicity, materiality, market efficiency, and trading between the misrepresentation and revelation of the truth. Its holding concerned market efficiency; it did not alter the separate rules governing materiality at certification.

For these plaintiffs’ proposed reliance theory, an efficient market supplied the link between public statements and classwide reliance. Without that link, reliance would require individualized proof, undermining predominance.

B. The Cammer/Krogman factors are tools, not substitutes for judgment

The eight familiar factors examine:

1. Average weekly trading volume relative to outstanding shares;
2. Analyst coverage;
3. Trading by market makers and arbitrageurs;
4. Eligibility to use SEC Form S-3;
5. Empirical evidence connecting unexpected corporate information with immediate price responses;
6. Market capitalization;
7. The bid-ask spread; and
8. The stock’s public float, excluding insider-held stock.

These indicators help determine whether market prices incorporate public information. They do not exhaust the relevant evidence, and they must be weighed rather than counted. Several favorable indicators cannot automatically neutralize substantial contrary evidence.

The district court therefore asked the wrong legal question when it emphasized the absence of Fifth Circuit authority requiring abandonment of these factors for a purported meme stock. Defendants did not need to establish an exception to the framework. The framework itself required consideration of evidence outside the listed factors.

C. Economic evidence mattered more than the “meme stock” label

The district court discounted Stulz’s evidence because he had not expressly concluded that Cassava was a meme stock or that its market was inefficient. The appellate court rejected that approach. His testimony that Cassava “may have been influenced by the meme phenomenon,” supported by identified economic indicators, required evaluation on its substance.

The decision thus avoids making certification depend on an imprecise market label. Calling a security a meme stock does not establish inefficiency; declining to use that label does not make evidence of inefficiency irrelevant.

D. Unexplained volatility could challenge informational efficiency

Stulz discussed Cassava’s price volatility in the absence of news and compared its behavior with recognized meme stocks, including AMC and GameStop. The relevance was functional: if prices move substantially without corresponding value-relevant information, the inference that prices efficiently reflect public information may weaken.

The court did not hold that volatility alone defeats market efficiency. Nor did it require prices always to equal a company’s objectively correct value. The relevant inquiry was whether information was being incorporated into price, not whether every price movement could be justified as fundamentally accurate.

E. Short-selling constraints could obstruct incorporation of negative information

Stulz also identified a short utilization rate exceeding 93 percent and borrowing costs reaching 67 percent during the class period. Such conditions could limit investors’ ability to trade on negative information.

Short selling can help translate unfavorable information into market prices. If borrowing stock becomes difficult or prohibitively expensive, that mechanism may function less effectively. The court treated these constraints as potentially important indirect evidence of inefficiency—not as automatic proof of it.

The percentages were factual features of the record, not numerical thresholds adopted for future cases. Their significance depended on assessment alongside the remaining evidence.

F. Legal error justified vacatur without appellate fact-finding

Class-certification decisions receive deferential abuse-of-discretion review, but deference does not protect a decision based on an incorrect legal standard. Here, the error was the district court’s failure to perform the required comprehensive assessment.

The Fifth Circuit did not choose between the experts or resolve the market-efficiency dispute itself. It returned those factual tasks to the district court. The plaintiffs retained the obligation to establish the requirements for certification; defendants did not have to supply a particular verbal formula before their contrary evidence deserved consideration.

5. Precedents Cited and Their Significance

A. Certification standards and appellate review

Kincannon v. United Airlines, Inc., 168 F.4th 713 (5th Cir. 2026)

This decision supplied the abuse-of-discretion standard and emphasized the predominantly factual character of certification inquiries. It supported leaving the ultimate weighing of market evidence to the district court, while permitting correction of the legal framework governing that inquiry.

Sampson v. United Servs. Auto. Ass'n, 83 F.4th 414 (5th Cir. 2023)

The court relied on this precedent both for Rule 23’s structure and for the principle that an erroneous understanding of governing law constitutes an abuse of discretion. It provided the immediate doctrinal basis for vacating certification despite the ordinarily deferential standard of review.

Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338 (2011)

Cited through Sampson, this case established the foundational requirement that a proposed class satisfy every Rule 23(a) condition and an applicable Rule 23(b) category. Meeting some certification requirements could not compensate for a defective predominance analysis.

B. Securities-fraud elements and presumed reliance

Erica P. John Fund, Inc. v. Halliburton Co. (Halliburton I), 563 U.S. 804 (2011)

This case directed courts to begin predominance analysis with the elements of the underlying claim. It also explained why reliance commonly determines whether securities-fraud claims can proceed collectively and described the connection between public information, market prices, and indirect investor reliance.

Matrixx Initiatives, Inc. v. Siracusano, 563 U.S. 27 (2011)

Cited through Halliburton I, this decision supplied the six-element formulation of the securities-fraud claim. Its role was foundational rather than specific to meme-stock evidence.

Basic Inc. v. Levinson, 485 U.S. 224 (1988)

This was the central substantive precedent. Its fraud-on-the-market presumption permits reliance to be established without proving that every investor personally encountered the alleged misstatements. Because that presumption rests on public information entering market prices, the district court had to examine evidence challenging that premise.

Ga. Firefighters' Pension Fund v. Anadarko Petroleum Corp., 99 F.4th 770 (5th Cir. 2024)

This Fifth Circuit decision connected the Basic presumption to Rule 23(b)(3) predominance and summarized its predicates. It located the present dispute within the circuit’s established treatment of classwide reliance.

Amgen Inc. v. Conn. Ret. Plans & Tr. Funds, 568 U.S. 455 (2013)

The opinion cited this decision for the proposition that developed markets can efficiently process public information. That proposition helped identify the relevant form of efficiency: responsiveness to information, rather than guaranteed accuracy of valuation.

Halliburton Co. v. Erica P. John Fund, Inc. (Halliburton II), 573 U.S. 258 (2014)

The court invoked this decision to distinguish inaccurate prices from the relevant relationship between statements and prices. It cautions against assuming that speculative or imperfect pricing necessarily defeats the fraud-on-the-market framework. The present holding remained focused on the adequacy of the efficiency analysis, not a separate resolution of price impact.

C. Market-efficiency methodology

Unger v. Amedisys Inc., 401 F.3d 316 (5th Cir. 2005)

This was the controlling methodological authority. It required rigorous preliminary proof, identified the familiar efficiency indicators, rejected mechanical factor-counting, and demanded consideration of evidence both for and against efficiency. The Fifth Circuit treated the district court’s approach as inconsistent with that existing command.

Cammer v. Bloom, 711 F. Supp. 1264 (D.N.J. 1989)

This case supplied the first five commonly used indicators. The present opinion preserved their usefulness while rejecting the suggestion that satisfying them necessarily completes the inquiry.

Krogman v. Sterritt, 202 F.R.D. 467 (N.D. Tex. 2001)

This decision supplied market capitalization, bid-ask spread, and float as additional indicators. Its discussion of relative price stability without news also explained why unexplained volatility could matter alongside favorable structural indicators.

In re Petrobras Sec., 862 F.3d 250 (2d Cir. 2017)

The Second Circuit’s endorsement of a holistic, totality-of-the-evidence assessment reinforced the Fifth Circuit’s rejection of an exhaustive checklist. It supplied persuasive support for evaluating the complete evidentiary picture.

Local 703, I.B. of T. Grocery & Food Emps. Welfare Fund v. Regions Fin. Corp., 762 F.3d 1248 (11th Cir. 2014)

This precedent supported flexibility in a fact-intensive inquiry, including consideration of new indicators recognized by market theory. Its significance was particularly clear where defendants alleged trading dynamics not neatly captured by traditional factors.

Bratya SPRL v. Bed Bath & Beyond Corp., 752 F. Supp. 3d 34 (D.D.C. 2024)

This persuasive district-court authority provided a close practical comparison. It explained that high volume may reflect activity unrelated to value-relevant information and that short-selling constraints can impede negative information from entering prices. The Fifth Circuit used it repeatedly to show why favorable conventional indicators must be weighed against contrary economic evidence.

Smilovits v. First Solar, Inc., 295 F.R.D. 423 (D. Ariz. 2013)

Quoted through Bratya SPRL, this case characterized short-selling constraints as indirect evidence of inefficiency. Its role was evidentiary: such constraints merit consideration without conclusively determining the outcome.

D. Derivative control-person liability

Alaska Elec. Pension Fund v. Flotek Indus., Inc., 915 F.3d 975 (5th Cir. 2019)

The opinion cited this case to explain that section 20(a) liability requires an underlying primary securities-law violation. The control-person claim therefore did not provide an independent escape from the difficulties affecting the underlying section 10(b) theory.

Southland Sec. Corp. v. INSpire Ins. Sols., Inc., 365 F.3d 353 (5th Cir. 2004)

Cited through Alaska Elec. Pension Fund, this decision supplied the rule that control-person liability is secondary and cannot exist without a primary violation. Neither precedent resolved whether any defendant was liable here.

E. Settlement, mootness, and judicial notice

Havens Realty Corp. v. Coleman, 455 U.S. 363 (1982)

This Supreme Court authority supported the conclusion that a proposed settlement does not moot an action when judicial approval is necessary before it becomes binding in the relevant sense. Anticipated approval was not equivalent to an already effective resolution.

In re Asbestos Litig., 90 F.3d 963 (5th Cir. 1996)

The opinion quoted this decision’s application of the approval-dependent settlement rule. It expressly disclosed that the decision had been vacated on other grounds, allowing readers to evaluate its procedural status rather than treating it as unqualified authority.

Ortiz v. Fibreboard Corp., 521 U.S. 1114 (1997) (mem.)

This citation identified the disposition vacating In re Asbestos Litig. on other grounds. It was not invoked as an independent substantive rule governing Cassava’s market efficiency or settlement.

In re Deepwater Horizon, 934 F.3d 434 (5th Cir. 2019)

This case supported judicial notice of the publicly filed mediation notice. Its contribution was procedural: the court could consider the relevant public record in assessing the settlement-based jurisdictional argument.

Lovelace v. Software Spectrum Inc., 78 F.3d 1015 (5th Cir. 1996)

This decision supported judicial notice of Cassava’s SEC-required Form 8-K. Because the noticed public records contained the necessary information, the court did not need to decide whether the related press release was judicially noticeable.

The court also consulted the settlement stipulation filed in In re Cassava Scis., Inc. Sec. Litig. That document was case-specific evidence, not precedent; its express preservation of this appeal reinforced the conclusion that appellate proceedings remained live.

6. Potential Impact

A. More demanding explanations in certification orders

District courts must explain how significant contrary evidence fits into their overall efficiency assessment. Reciting favorable factors and adopting one expert’s conclusion will be vulnerable if the order leaves competing evidence materially unaddressed. The obligation is meaningful evaluation, not mandatory acceptance of the defense position.

B. Greater attention to how trading mechanisms operate

The decision encourages parties to examine whether trading volume, liquidity, arbitrage, and price movements actually demonstrate information processing. An apparently active market may still face obstacles to incorporating negative information. Conversely, unusual trading does not necessarily establish an inefficient market.

C. No automatic exclusion of meme-stock investors

The opinion neither defines a legally disqualifying category of meme stocks nor holds that social-media influence defeats certification. Future disputes should turn on the evidence connecting particular trading conditions to informational efficiency, not on labels or investor stereotypes.

D. Continued certification scrutiny despite settlement

A proposed settlement does not necessarily halt a pending Rule 23(f) appeal. Because class settlement approval requires an appropriately certified class, parties cannot assume that settlement negotiations eliminate an existing certification dispute. The opinion nevertheless left the settlement’s broader effect on district-court proceedings unresolved.

E. A limited but important remand

The district court could reach the same certification result after applying the correct analysis, or it could conclude that certification is unwarranted. The Fifth Circuit resolved neither that question nor the remaining procedural objections, including challenges surrounding expert-related proceedings.

7. Complex Concepts Simplified

Predominance

Whether shared questions are sufficiently central that the claims can fairly be tried together, rather than through numerous investor-specific inquiries.

Fraud-on-the-market presumption

A means of treating investors as relying on public misinformation indirectly through the market price, without requiring each investor to have read the statement.

Informational efficiency

The market’s ability to incorporate public information into price. It does not mean that prices are always correct or speculation is absent.

Price impact versus market efficiency

Market efficiency concerns how a market generally processes information. Price impact concerns whether a particular alleged misstatement affected price. This appeal centered on the former.

Short selling and utilization

A short seller borrows shares and sells them, expecting to repurchase them at a lower price. High utilization indicates that much of the available lendable stock is already borrowed; scarce shares and high borrowing costs may impede that trading.

Scienter

The culpable state of mind required for securities fraud, distinct from merely making an inaccurate statement.

Mootness

The absence of a continuing dispute that a court can meaningfully resolve. A settlement awaiting required approval had not eliminated this dispute.

Vacatur and remand

The appellate court set aside the certification order and returned the matter for reconsideration. It did not dismiss the investors’ claims or decide the merits.

8. Conclusion

Bozorgi v. Cassava Sciences establishes a clear practical instruction: market efficiency must be assessed through the whole evidentiary record, not inferred from a favorable checklist while contrary evidence is ignored.

Its significance is methodological rather than categorical. Meme-stock dynamics and short-selling constraints may matter, but they do not automatically defeat certification. The district court must weigh their actual significance before deciding whether presumed reliance permits the case to proceed on a classwide basis.