Amicus Curiae Brief — NVIDIA Corporation, et al., Petitioners v. E. Ohman J:or Fonder AB, et al.

Supreme Court briefApr 4, 2024

Ask Donna

What actually matters in this document.

Text

No. 23-970

In the Supreme Court of the United States

——————

NVIDIA CORP. and JENSEN H UANG,

Petitioners,

E. OHMAN J:OR FONDER AB and S TICHTING

PENSIOENFONDS PGB,

Respondents.

——————

ON PETITION FOR WRIT OF CERTIORARI

TO THE UNITED STATES COURT OF APPEALS

FOR THE NINTH CIRCUIT

——————

BRIEF OF PROFESSOR JOSEPH A.

GRUNDFEST AS AMICUS CURIAE

IN SUPPORT OF PETITIONER

——————

CHRISTOPHER G. MICHEL

Counsel of Record

CASEY J. ADAMS

BRENNA LEDVORA

ABRAHAM MOUSSAKO

QUINN EMANUEL URQUHART

& SULLIVAN, LLP

1300 I Street, N.W.

Suite 900

Washington, D.C. 20005

(202) 538-8308

christophermichel@

quinnemanuel.com

Counsel for Amicus Curiae

QUESTIONS PRESENTED

Whether plaintiffs seeking to allege scienter

under the Private Securities Litigation Reform Act of

1995 (PSLRA), 15 U.S.C. § 78u-4(b), based on allegations about internal company documents must plead

with particularity the contents of those documents.

Whether plaintiffs can satisfy the PSLRA’s

falsity requirement by relying on an expert opinion to

substitute for particularized allegations of fact.

TABLE OF CONTENTS

Page

QUESTIONS PRESENTED ......................................... i

TABLE OF AUTHORITIES ....................................... iii

INTEREST OF AMICUS CURIAE ............................ 1

INTRODUCTION AND SUMMARY OF

ARGUMENT ...................................................... 2

ARGUMENT ................................................................. 4

II.

Quantitative Metrics Demonstrate That

the Circuit Splits Identified by the

Petition Are Highly Consequential ................. 4

Methods for Quantifying the

Significance of Circuit Splits................. 5

Quantifying This Petition’s Splits ........ 8

The Ninth Circuit Erred ................................. 10

The Expert Report is Not a

“Fact” ..................................................... 10

The Expert Report Does Not

Support a “Strong Inference of

Scienter” ................................................ 14

The Expert Report is Not Pled

“With Particularity” ............................. 17

These Errors Are Significant .............. 19

CONCLUSION ........................................................... 21

iii

TABLE OF AUTHORITIES

Page

Cases

Alexander v. Sandoval,

532 U.S. 275 (2001)............................................... 19

Ark. Pub. Emps. Ret. Sys. v. BristolMyers Squibb Co.,

28 F.4th 343 (2d Cir. 2022) .................................. 12

In re Ashworth, Inc. Sec. Litig.,

No. 99CV0121-L(JAH), 2001 WL

37119391 (S.D. Cal. Dec. 3, 2001) ....................... 12

Blue Chip Stamps v. Manor Drug

Stores,

421 U.S. 723 (1975)............................................... 19

Boukadoum v. Acelyrin, Inc.,

No. 2:23-cv-09672 (C.D. Cal. Mar.

26, 2024) ............................................................ 4, 13

Cent. Bank of Denver v. First Interstate

Bank of Denver,

511 U.S. 164 (1994)............................................... 20

Daubert v. Merrell Dow Pharms., Inc.,

509 U.S. 579 (1993)............................................... 13

DeMarco v. DepoTech Corp.,

149 F. Supp. 2d 1212 (S.D. Cal.

2001) ................................................................. 12, 13

Egbert v. Boule,

596 U.S. 482 (2022)............................................... 19

Fin. Acquisition Partners LP v.

Blackwell,

440 F.3d 278 (5th Cir. 2006) .......................... 12, 13

Halliburton Co. v. Erica P. John Fund,

Inc., 573 U.S. 258 (2014) ........................................ 1

Janus Cap. Grp., Inc. v. First Derivative

Traders, 564 U.S. 135 (2011) ............................... 19

Lerner v. Northwest Biotherapeutics,

273 F. Supp. 3d 573 (D. Md. 2017) ...................... 12

Merrill Lynch, Pierce, Fenner & Smith

Inc. v. Dabit, 547 U.S. 71 (2006).............. 13, 19–20

Novak v. Kasaks,

216 F.3d 300 (2d Cir. 2000).................................. 20

Ong v. Chipotle Mex. Grill, Inc.,

294 F. Supp. 3d 199 (S.D.N.Y. 2018) .................. 12

In re Silicon Graphics Inc. Sec. Litig.,

183 F.3d 970 (9th Cir. 1999) ................................ 20

Slack Technologies, LLC v. Pirani,

598 U.S. 759 (2023)................................................. 1

Stoneridge Inv. Partners, LLC v. Sci.Atlanta, 552 U.S. 148 (2008)................................ 19

Tellabs, Inc. v. Makor Issues & Rts.,

Ltd., 551 U.S. 308 (2007) ..................... 4, 10, 14–20

In re Under Armour Sec. Litig.,

409 F. Supp. 3d 446 (D. Md. 2019) ...................... 12

Statutes and Rules

15 U.S.C. § 78u-4(b) ..................................................... 2

15 U.S.C. § 78u-4(b)(2) ............................................... 10

15 U.S.C. § 78u-4(b)(1), (2)(A) ............................. 11, 17

Rule 10b-5 ............................................................... 6, 19

Sup. Ct. R. 10(a) ........................................................... 5

Other Authorities

Andrew Gelman & Eric Loken, The

Statistical Crisis in Science, Science,

102 AMERICAN SCIENTIST 460 (2014)................... 15

Joseph A. Grundfest, Quantifying the

Significance of Circuit Splits in Petitions for Certiorari: The Case of Securities Fraud Litigation, (Rock Center for Corporate Governance at

Stanford University Working Paper,

No. 254), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4768231 ...................................... 1–2, 5–7

Joseph A. Grundfest, Disimplying Private Rights of Action Under the Federal Securities Laws: Commission’s

Authority, 107 HARV. L. REV. 961

(1994) ..................................................................... 19

Joseph P. Simmons, Leif D. Nelson &

Uri Simonsoh, False-Positive

Psychology: Undisclosed Flexibility

in Data Collection and Analysis

Allows Presenting Anything as

Significant, 22(11) PSYCHOLOGICAL

SCIENCE 1359 (2011) ............................................. 16

Raphael Silberzahn et al.,

Corrigendum: Many Analysts, One

Data Set: Making Transparent How

Variations in Analytic Choices Affect

Results, 1(4) A DVANCES IN METHODS

AND PRACTICES IN P SYCHOLOGICAL

SCIENCE 337 (2018) ............................................... 15

INTEREST OF AMICUS CURIAE

Joseph A. Grundfest is the William A. Franke Professor of Law and Business (Emeritus) at Stanford

Law School and a senior faculty member at the Rock

Center on Corporate Governance. He was a Commissioner of the Securities and Exchange Commission

from 1985 to 1990. Professor Grundfest has taught

securities law for decades, published extensively on

the subject in leading law reviews (including the Harvard, Yale, and Stanford Law Reviews), and submitted amicus briefs to this Court in significant securities cases, such as Slack Technologies, LLC v. Pirani,

598 U.S. 759 (2023), and Halliburton Co. v. Erica P.

John Fund, Inc., 573 U.S. 258 (2014).

Professor Grundfest also authored a recent working paper, Quantifying the Significance of Circuit

Splits in Petitions for Certiorari: The Case of Securities Fraud Litigation (Rock Center for Corporate Governance at Stanford University Working Paper, No.

254),

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4768231 (hereinafter “Grundfest”), proposing new metrics for assessing the significance of circuit splits in securities fraud litigation. These metrics

are directly applicable to this case. Professor Grundfest accordingly has a strong interest in the Court’s



Amicus affirms that no counsel for a party authored this brief

in whole or in part, and no one other than amicus or his counsel

made a monetary contribution intended to fund the preparation

or submission of the brief. Counsel of record for all parties received notice at least 10 days prior to the due date of the intention of amicus to file this brief.

consideration of the petition, which he submits should

be granted.

INTRODUCTION AND

SUMMARY OF ARGUMENT

The petition for certiorari in this case presents

important questions of both methodology and substance. The methodology addresses the commonly occurring question of how the Court should analyze the

assertion of a conflict among federal courts of appeals

(i.e., a “circuit split”). Here, petitioners identify two

related circuit splits regarding the proper interpretation of the pleading standards in the Private Securities Litigation Reform Act of 1995 (PSLRA), 15 U.S.C.

§ 78u-4(b).

Litigants typically approach the analysis of a circuit split as a matter of simple “nose counting” (e.g.,

asking how many circuits are on each side of the asserted split). However, as detailed by amicus in a recently released paper, see Grundfest, supra, reliable

databases describing securities fraud litigation enable a calculation of the relative significance of each

circuit in the overall market for class action securities

fraud litigation. That calculation in turn enables a

more refined approach to determining whether an asserted circuit split warrants this Court’s review.

Specifically, amicus proposes two metrics that

help inform a petitioner’s claim to space on this

Court’s docket. First, the Court can examine the aggregate share of the overall market implicated by the

conflict; a split between circuits accounting for only

5% of the total market is self-evidently less important

than a split between circuits accounting for 50% of the

market. Second, the Court can examine the relative

share of the market on each side of the conflict. A

split is worthier of review if the circuits on each side

account for roughly equal shares, as opposed to lopsided shares in which one side might be dismissed as

a relatively insignificant outlier.

Applying these metrics to securities fraud litigation yields a powerful insight, likely consistent with

the Court’s intuition and experience: a circuit split

that includes the Second Circuit (accounting for 37%

of the market) on one side, and the Ninth Circuit (accounting for 23% of the market) on the other side, is

far more likely to be worthy of this Court’s review

than a split involving any other configuration of circuits.

The two questions presented implicate circuit

splits fitting that description. The first question involves a total market share of 86% (with the First and

Ninth Circuits accounting for 25% of the total market

on one side and the Second, Third, Fifth, Seventh, and

Tenth Circuits accounting for 61% on the other side).

The second question involves a total market share

of 64% (with the Ninth Circuit accounting for 23% of

the total market share on one side and the Second and

Fifth Circuits accounting for 41% of the total market

share on the other side).

The questions are additionally “certworthy” because the Ninth Circuit’s decision is legally wrong,

and badly so. As this Court has long recognized, implied private rights under federal securities laws expose defendants to particularly vexatious forms of litigation. Accordingly, Congress enacted the PSLRA to

impose stringent pleading requirements on private

securities plaintiffs and to reduce the probability that

meritless claims proceed past the pleading stage. But

the Ninth Circuit ignored both the statutory text and

purpose of the PSLRA by allowing paid expert opinion

testimony to support allegations of securities fraud.

As elaborated further by the petition and the dissent below, the Ninth Circuit’s holding defies the

PSLRA in at least three ways: it allows plaintiffs to

plead opinions masquerading as facts; it circumvents

the requirement to negate other equally compelling

inferences of scienter; and it undermines the requirement to plead claims with particularity. See Tellabs,

Inc. v. Makor Issues & Rts., Ltd., 551 U.S. 308, 313–

14 (2007); id. at 334 (Alito, J., concurring).

If this decision stands, it will amplify the kind of

vexatious litigation that Congress intended the

PSLRA to constrain. Indeed, a new securities fraud

complaint relying on an expert opinion was filed just

last week in the Ninth Circuit, in the wake of the misguided decision below. Am. Compl. at ¶¶ 66–67, 261–

262, Boukadoum v. Acelyrin, Inc., No. 2:23-cv-09672

(C.D. Cal. March 26, 2024). This is a paradigmatic

case for this Court’s review, and the petition should

be granted.

ARGUMENT

Quantitative Metrics Demonstrate That the

Circuit Splits Identified by the Petition Are

Highly Consequential

The petition presents two questions for consideration. The metrics identified in amicus’s recent paper

demonstrate that the circuit splits are highly consequential in the area of private securities litigation and

are worthy of this Court’s review.

Methods for Quantifying the Significance of Circuit Splits

The existence of a circuit split is one of the most

important factors influencing the grant of certiorari.

Sup. Ct. R. 10(a); see Grundfest, supra, at 12–13. But

not all circuit splits are created equal. Some involve

circuits that resolve a very large percentage of litigation that raises the question presented for this Court’s

review. Other splits engage circuits that only rarely

address the question presented. The most common

method of describing the significance of circuit splits

before this Court is to simply count the number of circuits that split, regardless of how often (or seldom)

those circuits confront the underlying question.

Consider a hypothetical split among four circuits.

If each circuit’s geography generates only 1% of the

litigation implicated by the question presented, then

resolving that question for these four circuits affects

only 4% of the market. But if each circuit generates

20% of the litigation implicated by the question presented, then resolving the identical question for these

circuits affects 80% of the market. All else being

equal, a split implicating 80% of the market is more

worthy of this Court’s attention than an identical split

implicating just 4% of the market.

Amicus has recently developed novel techniques

for quantifying the economic significance of circuit

splits with greater precision than simple “nose counting.” See Grundfest, supra, at 4–5. As described below and elaborated further in amicus’s paper, the

quantitative significance of a circuit split can be expressed through two metrics.

The first metric is the “aggregate circuit split

share,” which measures the total “market share” of all

circuits that have split on either side of the question.

It is here calculated by averaging several statistics

quantifying the total number and value of federal

class action securities fraud cases in the various circuits. These measures include the number of cases

filed, the number of settlements reached, the total

value of the settlements, more aggressive and conservative estimates of the plaintiffs’ alleged damages,

and mentions of key securities fraud legal terms (Rule

10b-5 and Section 11). The measures are then used

to generate an equal-weighted average for each circuit.1

Significantly, the aggregate circuit split share

metric confirms that the Second and Ninth Circuits

dominate the market for class action securities fraud

litigation. 2 Based on an equal-weighted average of all

seven metrics, these two circuits alone represent approximately 60% of federal class action securities

fraud litigation activity. The aggregate circuit split

share of the remaining circuits is as follows: First Circuit (2%); Third Circuit (12%); Fourth Circuit (3%);

Fifth Circuit (4%); Sixth Circuit (4%); Seventh Circuit

For more details on the calculation of these metrics, see

Grundfest, supra, at 16–17.

See Grundfest, supra, at 23 for the full table of metrics.

Because the metrics are very highly correlated, the use of an

equal-weighted average does not bias the result. See Grundfest,

supra, at 17.

(5%); Eighth Circuit (2%); Tenth Circuit (3%); Eleventh Circuit (4%); D.C. Circuit (1%). 3

The second metric is the “split ratio,” which describes whether a split is caused by outlier circuits

with relatively low aggregate circuit split shares, or

whether the split instead reflects a disagreement

among circuits with comparable circuit shares.

Consider a hypothetical split between two circuits

in which the decision giving rise to the petition is from

a circuit with 10% of the relevant market, while the

circuit with an opposing view has a 40% share. The

corresponding split ratio can be expressed as 10%–

40%.4 A split with that ratio is relatively lopsided,

with the decision that gives rise to the petition representing a comparative outlier. By contrast, if the two

circuits in that hypothetical split instead each had

shares of 25%, the split ratio would be 25%–25%, illustrating an even division.

The takeaway from these two metrics is straightforward. All other factors equal, the Court should

generally be more inclined to grant petitions that present splits with higher aggregate circuit split shares

and relatively even split ratios. Those cases present

See Grundfest, supra, at 23.

This statistic can also be expressed as 20%–80%, where 20 is

the percentage of the total market of cases represented by the

opinion giving rise to the petition and circuits that agree with

that opinion (10 / (10+40) = 0.20) and 80 is the percentage of the

total market of cases represented by circuits with the opposing

view (40 / (10+40) = 0.80). In the second expression, the values

always sum to 100%. See Grundfest, supra, at 17, for a detailed

discussion of this alternative calculation method.

questions of broader national significance as to which

the circuits are more evenly divided.5

Quantifying This Petition’s Splits

The two questions presented here implicate circuit

splits whose quantitative metrics suggest they are important and should be resolved by the Court.

The first question presented is “[w]hether plaintiffs seeking to allege scienter under the PSLRA based

on allegations about internal company documents

must plead with particularity the contents of those

documents.” Pet. at (i). This implicates a circuit split

involving most federal class action securities fraud litigation. Five circuits hold that litigants seeking to

plead scienter based on internal company documents

must plead their contents with particularity. Those

circuits, with their respective average shares, are the

Second (37%), Third (12%), Fifth (4%), Seventh (5%),

and Tenth (3%). In contrast, the First (2%) and Ninth

(23%) Circuits hold that plaintiffs do not need particularized allegations about the documents’ specific

contents. Id. at 15–20.

The Court might be somewhat more likely to grant cert to

resolve lopsided splits if the circuit on the “short” side of such a

split is the one issuing the decision that gives rise to the petition.

That is, the Court may prefer to review the decisions of outlier

circuits when presented with such a decision, but may have less

inclination to review the decisions of circuits on the “long” side

of a split. While recognizing that this preference (and no doubt,

many other nuances) may exist, the methodology described here

does not attempt to expressly account for it.

The aggregate circuit split share for the first question is thus 86%, the sum of the circuit shares on either side of the split. The split ratio is expressed either as 25%–61% or 29%–71%. The split is thus not

driven by outlier circuits with de minimis shares of

litigation on securities fraud. This split ratio and the

high aggregate circuit split share present a powerful

pragmatic argument for granting the petition as to

the first question.

The second question presented is “[w]hether plaintiffs can satisfy the PSLRA’s falsity requirement by

relying on an expert opinion to substitute for particularized allegations of fact.” Id. at (i). The decision below created this circuit split, which divides the Ninth

Circuit (23%) on one side of the question, and the Second (37%) and Fifth (4%) Circuits on the other side.

Id. at 27–28. The aggregate circuit split share here is

64%, and the split ratio is 23%–41% or 36%–64%.

Simply “nose counting” the number of split circuits

understates the pragmatic significance of this split.

While only three circuits are involved, compared to

the seven affected by the first split, almost two-thirds

of the market (64%) is engaged in this second split.

As with the first question presented, the split is not

caused by a rogue outlier with low market share, but

rather is between the Second and Ninth Circuits, the

circuits most important to the resolution of federal

class action securities fraud claims.

In sum, both questions implicate circuit divisions

with high aggregate circuit split shares: 86% for the

first question and 64% for the second. The split ratios

in both instances confirm that the split is not caused

by rogue circuits with small shares. Both splits also

divide the Second and Ninth Circuits, which are the

most important circuits when litigating class action

securities fraud claims. These quantitative metrics

strongly support granting the petition as to both questions.6

II. The Ninth Circuit Erred

In addition to implicating two consequential circuit splits, the decision below also commits (at least)

three significant errors of law. Under the PSLRA, a

securities fraud complaint must “state with particularity [] the facts constituting the alleged violation”

and the “‘facts giving rise to a strong inference that

the defendant acted with the required state of mind.’”

Tellabs, 551 U.S. at 313–14 (quoting 15 U.S.C. § 78u4(b)(2)). The PSLRA thus requires that allegations

(1) are of facts, (2) support a strong inference of scienter, and (3) are pled with particularity. The Ninth

Circuit’s decision ignores each of these requirements

and, if allowed to stand, will eviscerate protections essential to the PSLRA’s operation.

The Expert Report is Not a “Fact”

The Ninth Circuit errs in treating the plaintiffs’

expert report—the Prysm Report—as a fact, when it

is actually an opinion. This report was prepared by a

“consulting firm” retained by plaintiffs to offer specific

numerical estimates of NVIDIA’s crypto-related revenues, based on inferences and exogenous analysis.

Compl. ¶¶ 143, 147, 153, No. 18-cv-07669 (N.D. Cal.

Moreover, the circuits giving rise to the conflict are on the

short end of the split, which may make review even more

attractive to the Court.

May 13, 2020), ECF No. 149. This consulting firm

never claims to have set foot within NVIDIA, to have

had access to any non-public NVIDIA information, or

to have first-hand knowledge of any facts related to

NVIDIA at all. The report is a “post hoc analysis by

… an outside expert that relied on generic market research and unreliable or undisclosed assumptions to

reach its revenue estimates.” Ohman v. NVIDIA

Corp., 81 F.4th 918, 947 (9th Cir. 2023) (Sanchez, J.,

dissenting). On its face, it is not a “fact” of the sort

that the PSLRA requires to support of securities

fraud. See 15 U.S.C. § 78u-4(b)(1), (2)(A) (the complaint must “state with particularity all facts” supporting the belief “why the statement is misleading”

and “facts giving rise to a strong inference” of scienter) (emphasis added).

The expert report was nonetheless essential to the

Ninth Circuit’s conclusion that plaintiffs adequately

alleged securities fraud. To support allegations of

misrepresentation, the panel emphasized that the

crypto-related revenues publicly announced by

NVIDIA’s CEO differed from the expert’s opinion. See

Ohman, 81 F.4th at 933–34. To support scienter allegations, the panel concluded that the CEO “would

have known” about the revenue estimates later

reached by the expert’s post-hoc analysis. See id. at

940. While the panel purported to rely on other allegations in reaching its conclusion—another market

analyst whose “assumptions” and “sources of information” were not described; generalized statements

from employees; and broad market “events”—none of

these sources provided the critical revenue estimates.

Id. at 932, 954.

No other circuit would have allowed these post-hoc

revenue estimates by outsiders to plead securities

fraud.7 See Ark. Pub. Emps. Ret. Sys. v. Bristol-Myers

Squibb Co., 28 F.4th 343, 354 (2d Cir. 2022); Fin. Acquisition Partners LP v. Blackwell, 440 F.3d 278, 286

(5th Cir. 2006)); see also Pet. at 27–28. Many district

courts, too, would have rejected plaintiffs’ claims. See,

e.g., In re Under Armour Sec. Litig., 409 F. Supp. 3d

446, 454–55 (D. Md. 2019) (“Expert opinions generated for purposes of supporting Plaintiffs’ theories in

a [securities fraud complaint] do not warrant the assumption of truth.”); Ong v. Chipotle Mex. Grill, Inc.,

294 F. Supp. 3d 199, 222 (S.D.N.Y. 2018) (refusing to

consider “any conclusory allegations in the [securities

fraud complaint] that are based on the [expert opinion]”).8

It is not even clear that any court would admit the

plaintiffs’ expert report at a later stage of litigation.

The complaint nowhere alleges that the report’s

methodology has any analogue in the peer reviewed

literature, never describes the authors’ qualifications

beyond holding PhDs, and does not suggest that the

Even the First Circuit, which aligns with the Ninth Circuit in

allowing nonparticularized descriptions of internal corporate

documents to support scienter, did not rely on this type of outside

expert opinion to establish what those documents purportedly

contained. See In re Stone & Webster, Inc., Sec. Litig., 414 F.3d

187, 206–11 (1st Cir. 2005); see also Pet. at 22–23.

See also Lerner v. Nw. Biotherapeutics, 273 F. Supp. 3d 573,

590 (D. Md. 2017); In re Ashworth, Inc. Sec. Litig., No.

99CV0121-L(JAH), 2001 WL 37119391, at *3 (S.D. Cal. Dec. 3,

2001); DeMarco v. DepoTech Corp., 149 F. Supp. 2d 1212, 1222

(S.D. Cal. 2001).

authors have published any research in peer-reviewed journals that is rationally related to the expert

report’s analysis.

See Daubert v. Merrell Dow

Pharms., Inc., 509 U.S. 579, 592–93 (1993).

Thus, the panel not only accepted opinion evidence

as fact, but it accepted potentially inadmissible opinion evidence as fact. These evidentiary complications

are exactly why courts have refused to allow such

opinions to masquerade as facts in order to satisfy the

PSLRA’s pleading standards. See, e.g., Blackwell, 440

F.3d at 285–86 (“[A]llowing plaintiffs to rely on an expert’s opinion in order to state securities claims requires a court to ‘confront a myriad of complex evidentiary issues not generally capable of resolution at the

pleading stage’…. [and] might require ruling on the

expert’s qualifications.” (quoting DeMarco, 149 F.

Supp. 2d at 1221)).

The Ninth Circuit’s decision to accept paid opinion

evidence as fact conflicts with the PSLRA’s express

goal of combating abusive securities lawsuits. See

Merrill Lynch, Pierce, Fenner & Smith Inc. v. Dabit,

547 U.S. 71, 81 (2006). The decision below will invite

a flood of complaints that rely on paid outside experts

to craft post-hoc analyses, with no grounding in firsthand knowledge, alleging corporate insiders must

have known the facts that the experts think they

should have known. In a recent example, plaintiffs

cited an expert report to support allegations of fraud

and scienter. See supra Am. Compl. at ¶¶ 66–67, 261–

262, Boukadoum v. Acelyrin, Inc. (No. 2:23-cv-09672).

As we are already seeing, well-resourced plaintiffs

will be able to buy their way out of the PLSRA’s pleading requirements. This is assuredly not what Congress intended when drafting the PLSRA.

The Expert Report Does Not Support a

“Strong Inference of Scienter”

The Ninth Circuit errs again in finding that this

expert opinion supports a “strong inference” of scienter. Tellabs holds that, when assessing whether a

complaint adequately alleges a “strong inference,”

courts “must engage in a comparative evaluation,”

and consider “competing inferences rationally drawn

from the facts alleged.” 551 U.S. at 314. A plaintiff

satisfies the PSLRA’s “[e]xacting pleading requirements” only if the inference of scienter is “cogent and

at least as compelling as any opposing inference one

could draw from the facts alleged.” Id. at 313, 324.

Despite relying on an expert opinion to find scienter,

see Ohman, 81 F.4th at 940, the decision below never

engages in Tellabs’ comparative evaluation by asking

whether other revenue estimates would be “as compelling” as those offered by that expert. 551 U.S. at

314.

Nor was it even possible for the Ninth Circuit to

engage in this comparative analysis. The complaint

below asserts that the expert report’s approach was

“conservative,” that its “third party data sources”

were “credible,” and that its analysis was “rigorous.” 9

Compl. ¶¶ 147–152c. But Tellabs never concludes it

is sufficient that a complaint is conservative, credible,

rigorous, or any other synonym for “plausible.” The

As discussed below, these allegations do not describe the

expert’s analysis with sufficient particularity to conclude that it

was, in fact, conservative, credible, or rigorous. See infra at 17–

18.

requirement is, instead, far more exacting. The plaintiff’s inference of fraud must be “at least as cogent” as

any other possible inference. Tellabs, 551 U.S. at 314.

But the complaint below never supports any inference

that the expert’s sources are more credible than other

sources, or that its methodologies are superior to

other methodologies that also generate revenue estimates. The complaint therefore cannot allege that

the expert’s conclusion is “at least as compelling as

any opposing inference” because it never addresses

any opposing inferences.

The complaint’s silence on this point implicates a

larger analytic issue now attracting attention among

statisticians: the problem raised by the “garden of

forking paths.” See, e.g., Andrew Gelman & Eric

Loken, The Statistical Crisis in Science, 102 AMERICAN SCIENTIST 460 (2014). When testing a hypothesis—such as whether NVIDIA’s crypto-related revenues were higher than represented—researchers

make many decisions, including which datasets to use

and which statistical techniques to apply. All of these

choices can affect the ultimate conclusion. To illustrate, a recent study compared the conclusions of numerous research teams, who had “analyzed the same

data set to answer the same research question,” and

found that there was no consensus in either the selection of statistical techniques or outcome. See Raphael

Silberzahn et al., Corrigendum: Many Analysts, One

Data Set: Making Transparent How Variations in Analytic Choices Affect Results, 1(4) A DVANCES IN METHODS AND PRACTICES IN P SYCHOLOGICAL SCIENCE 337,

338, 343–47 (2018); see also Gelman & Loken, supra,

at 464 (suggesting that “choices in analysis and interpretation are data dependent and would have been

different given other possible data”).

Moreover, researchers (consciously or unconsciously) tend to make analytic decisions that will

support their desired result. See Joseph P. Simmons,

Leif D. Nelson & Uri Simonsoh, False-Positive Psychology: Undisclosed Flexibility in Data Collection

and Analysis Allows Presenting Anything as Significant, 22(11) PSYCHOLOGICAL SCIENCE 1359, 1360

(2011). This problem is amplified if researchers are

compensated and know that their work has value to a

paying client only if it supports a client’s desired conclusion.

To make the problem concrete in the context of this

case, assume that there are ten different credible databases and that there are ten different methodologies that can be used to estimate the effects of cryptomining demand. For simplicity, assume that the combination of these ten databases and ten methodologies

yields the possibility of one hundred different forms of

analysis—ten different analyses of ten different databases. Assume further that the plaintiffs’ expert report is a legitimate analysis of one of these one hundred possibilities. It is, as statisticians would say, one

path in a garden of forking paths. But what of the

other 99 paths that other equally competent analysts

might have followed? Tellabs commands analysis of

this broader question, but plaintiffs fail to even recognize the challenge.

Thus, even if the expert report is credible, that is

not enough. The complaint pleads no facts suggesting

that the report’s analysis is at least as credible as opposing conclusions that can be reached by other experts addressing the same question.

The Expert Report is Not Pled “With

Particularity”

The Ninth Circuit’s third error is that it ignores

the PSLRA’s “particularity” requirement, 15 U.S.C. §

78u-4(b)(1), (2)(A). This error demonstrates the prescience of Justice Alito’s concurrence in Tellabs. Justice Alito explains that the plain language of the

PSLRA requires that “a strong inference” must arise

only from facts stated “with particularity.” Tellabs,

551 U.S. at 334 (Alito, J., concurring). “It follows that

facts not stated with the requisite particularity cannot be considered in determining whether the stronginference test is met.” Id.

Justice Alito’s concern is motivated by dicta in the

majority opinion stating that “‘omissions and ambiguities’ merely ‘count against’ inferring scienter, and

that a court should consider all allegations of scienter,

even nonparticularized ones, when considering

whether a complaint meets the ‘strong inference’ requirement.” Id. This interpretation, as Justice Alito

emphasizes, would “undermine[] the particularity requirement’s purpose of preventing a plaintiff from using vague or general allegations in order to get by a

motion to dismiss.” Id. Justice Alito thus cautions

against holistic forms of analysis that consider a combination of factors, some or all of which are not pled

with sufficient particularity, as evidence supporting a

strong inference of scienter.

But that is precisely the form of analysis that dominates the opinion below. Even though the expert

opinion was the only source of the critical revenue estimates relied upon by the Ninth Circuit, the complaint “fail[ed] to describe [the expert]’s assumptions

and analysis with sufficient particularity to establish

a probability that its [revenue] conclusions are reliable.” Ohman, 81 F.4th at 953 (Sanchez, J., dissenting)

(citation omitted); see also Pet. at 29–30 (describing

the expert’s questionable assumptions). Among other

flaws, the complaint failed to describe the “proprietary analytic models” used to estimate NVIDIA’s market share, which was one step in the expert’s ultimate

revenue estimate. Id. at 953–54. “[W]ithout knowing

the basis for this input, one cannot ascertain the reliability of the output.” Id. at 954.

But instead of excluding the expert opinion because of a lack of particularity, the panel leaned heavily on it. And, in an attempt to bolster the defectively

pled expert report, the Ninth Circuit pointed to other

generalized allegations—none of which confirmed the

expert’s specific revenue estimates. See Ohman, 81

F.4th at 932. Thus, rather than relying on particularized allegations, the Ninth Circuit relied on a combination of “vague or general allegations.” Tellabs, 551

U.S. at 334 (Alito, J., concurring), and thereby

“stripped [the particularity requirement] of all meaning.” Id. This decision will permit plaintiffs to “circumvent” the PSLRA’s “important” protections

against abusive litigation—just as Justice Alito predicted seventeen years ago. Id. This Court’s review

is warranted to clarify this critically important point

in private securities litigation.

D. These Errors Are Significant

The Ninth Circuit’s errors, if uncorrected, have the

potential to dramatically expand private securities

litigation. The private right of action for securities

fraud, arising under Section 10(b) of the Securities

Exchange Act of 1934 and SEC Rule 10b-5, is implied,

not express. Janus Cap. Grp., Inc. v. First Derivative

Traders, 564 U.S. 135, 142 (2011); see also Joseph A.

Grundfest, Disimplying Private Rights of Action Under the Federal Securities Laws: The Commission’s

Authority, 107 HARV. L. REV. 961, 985–94 (1994). As

this Court has frequently held, such an implied right

must be given “narrow scope.” Janus, 564 U.S. at 145;

see, e.g., Stoneridge Inv. Partners, LLC v. Sci.-Atlanta,

552 U.S. 148, 165 (2008) (“Concerns with the judicial

creation of a private cause of action caution against

its expansion.”); see also Egbert v. Boule, 596 U.S. 482,

503 (2022) (Gorsuch, J., concurring) (“To create a new

cause of action is … a power that is in every meaningful sense an act of legislation…. It has no place in

federal courts charged with deciding cases and controversies under existing law.”); Alexander v. Sandoval,

532 U.S. 275, 286 (2001) (“private rights of action to

enforce federal law must be created by Congress”).

A narrow construction is particularly warranted

here because private securities litigation “presents a

danger of vexatiousness different in degree and in

kind from that which accompanies litigation in general.” Dabit, 547 U.S. at 81 (quoting Blue Chip

Stamps v. Manor Drug Stores, 421 U.S. 723, 739

(1975)); see, e.g., Tellabs, 551 U.S. at 313 (“Private securities fraud actions, however, if not adequately contained, can be employed abusively to impose substan-

tial costs on companies and individuals whose conduct conforms to the law.”); Cent. Bank of Denver v.

First Interstate Bank of Denver, 511 U.S. 164, 189

(1994) (“Litigation under 10b–5 thus requires secondary actors to expend large sums even for pretrial defense and the negotiation of settlements.”).

The PSLRA was enacted expressly to respond to

this problem. The PSLRA’s novel and stringent

pleading requirements were specifically designed to

reduce the probability that low-quality allegations

would support meritless federal securities law claims.

See Dabit, 547 U.S. at 81 (noting the PSLRA was “targeted at perceived abuses of the class-action vehicle”

in securities litigation, including “nuisance filings,

targeting of deep-pocket defendants, vexatious discovery requests,” and “extortionate settlements”); see

also Novak v. Kasaks, 216 F.3d 300, 306 (2d Cir. 2000)

(the PSLRA was “motivated in large part by a perceived need to deter strike suits wherein opportunistic private plaintiffs file securities fraud claims of dubious merit in order to exact large settlement recoveries”); In re Silicon Graphics Inc. Sec. Litig., 183 F.3d

970, 988 (9th Cir. 1999) (“Congress enacted the

PSLRA to put an end to the practice of pleading fraud

by hindsight.”) (quotation marks and citation omitted).

But here, the Ninth Circuit ignores these pleading

requirements, contravening the PSLRA’s text and

purpose. Allowing expert opinion testimony to masquerade as fact, while avoiding the comparative analysis commanded by Tellabs and evading the statute’s

particularity requirement, will expand, not narrow,

the scope of an implied private right of action, while

encouraging abusive litigation of the kind that the

PSLRA was enacted to limit.

CONCLUSION

The Court should grant the petition for certiorari.

Respectfully submitted.

CHRISTOPHER G. MICHEL

Counsel of Record

CASEY J. ADAMS

BRENNA LEDVORA

ABRAHAM MOUSSAKO

QUINN EMANUEL URQUHART

& SULLIVAN, LLP

1300 I Street, N.W.

Suite 900

Washington, D.C. 20005

(202) 538-8308

christophermichel@

quinnemanuel.com

Counsel for Amicus Curiae

April 4, 2024

This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.