Goldman Sachs Group, Inc. v. Arkansas Teacher Retirement System

594 U.S. 113, 210 L. Ed. 2d 347, 141 S. Ct. 1951

• Court: Supreme Court of the United States
• Decided: June 21, 2021
• Docket: 20-222
• Status: Published

Opinion

(Slip Opinion) OCTOBER TERM, 2020 1

Syllabus

NOTE: Where it is feasible, a syllabus (headnote) will be released, as is being done in connection with this case, at the time the opinion is issued. The syllabus constitutes no part of the opinion of the Court but has been prepared by the Reporter of Decisions for the convenience of the reader. See United States v. Detroit Timber & Lumber Co., 200 U. S. 321, 337.

SUPREME COURT OF THE UNITED STATES

GOLDMAN SACHS GROUP, INC., ET AL. v. ARKANSAS TEACHER RETIREMENT SYSTEM, ET AL.

CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT

No. 20–222. Argued March 29, 2021—Decided June 21, 2021 Respondent shareholders (Plaintiffs) filed this securities-fraud class ac- tion alleging that The Goldman Sachs Group, Inc., and certain of its executives (collectively, Goldman) violated securities laws and regula- tions prohibiting material misrepresentations and omissions in con- nection with the sale of securities. 15 U. S. C. §78j(b); 17 CFR §240.10b–5. Plaintiffs allege that Goldman maintained an artificially inflated stock price by repeatedly making false and misleading generic statements about its ability to manage conflicts. Under Plaintiffs’ in- flation-maintenance theory, Goldman’s alleged misrepresentations caused its stock price to remain inflated until the market reacted to the truth about Goldman’s practices—at which point Goldman’s stock price dropped and Plaintiffs suffered losses. Seeking to certify a class of Goldman shareholders harmed by reliance on Goldman’s alleged misrepresentations, Plaintiffs invoked the presumption, endorsed by the Court in Basic Inc. v. Levinson, 485 U. S. 224, that investors are presumed to rely on the market price of a company’s security, which in an efficient market will reflect all of the company’s public statements, including misrepresentations. The Basic presumption allows class-ac- tion plaintiffs to prove reliance through evidence common to the class. Goldman in turn sought to defeat class certification by rebutting the Basic presumption through evidence that its alleged misrepresenta- tions had no impact on its stock price. After an initial round of litiga- tion which resulted in a remand from the Second Circuit, the District Court certified the class based on Goldman’s failure to establish by a preponderance of the evidence that its alleged misrepresentations had no price impact. The Second Circuit authorized an appeal under Fed- eral Rule of Civil Procedure 23(f), and affirmed in a divided decision, 2 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM Syllabus

finding that the District Court’s price impact determination was not an abuse of discretion. Goldman now argues that the Second Circuit erred twice: first, by holding that the generic nature of Goldman’s al- leged misrepresentations is irrelevant to the price impact inquiry; and second, by assigning Goldman the burden of persuasion to prove a lack of price impact. Held: 1. The generic nature of a misrepresentation often is important evi- dence of price impact that courts should consider at class certification, including in inflation-maintenance cases. That is true even though the same evidence may be relevant to materiality, an inquiry reserved for the merits phase of a securities-fraud class action. See Amgen Inc. v. Connecticut Retirement Plans and Trust Funds, 568 U. S. 455. A court has an obligation before certifying a class to determine that Rule 23 is satisfied, Comcast Corp. v. Behrend, 569 U. S. 27, 35, and a court can- not make that finding in a securities-fraud class action without consid- ering all evidence relevant to price impact. See Halliburton Co. v. Er- ica P. John Fund, Inc., 573 U. S. 258, 284 (Halliburton II). The parties now accept this legal framework but dispute whether the Second Cir- cuit properly considered the generic nature of Goldman’s alleged mis- representations. Because the Court concludes that the Second Cir- cuit’s opinions leave sufficient doubt on this question, the Court remands for the Second Circuit to consider all record evidence relevant to price impact, regardless whether that evidence overlaps with mate- riality or any other merits issue. Pp. 6–9. 2. Defendants bear the burden of persuasion to prove a lack of price impact by a preponderance of the evidence at class certification. The Court has held that nothing in Federal Rule of Evidence 301 constrains the Court’s authority to change customary burdens of persuasion un- der a federal statute, see NLRB v. Transportation Management Corp., 462 U. S. 393, 404, n. 7, and the Court has exercised this authority to reassign the burden of persuasion to the defendant in other contexts. Goldman does not challenge the Court’s relevant precedents, but ques- tions whether the Court exercised this authority in establishing the Basic framework pursuant to the securities laws. The Court concludes that Basic and Halliburton II did allocate to defendants the burden of persuasion to prove a lack of price impact. As relevant here, Basic explains that defendants may rebut the presumption of reliance if they “show that the misrepresentation in fact did not lead to a distortion of price” by making “[a]ny showing that severs the link between the al- leged misrepresentation and . . . the price received (or paid) by the plaintiff.” 485 U. S., at 248 (emphasis added). Similarly, Halliburton II held that defendants may rebut the Basic presumption at class cer- Cite as: 594 U. S. ____ (2021) 3

tification “by showing . . . that the particular misrepresentation at is- sue did not affect the stock’s market price.” 573 U. S., at 279 (emphasis added). These references to a defendant’s “showing” require a defend- ant to do more than produce some evidence relevant to price impact; the defendant must “in fact” “seve[r] the link” between a misrepresen- tation and the price paid by the plaintiff. Moreover, Halliburton II’s holding that plaintiffs need not directly prove price impact to invoke the Basic presumption, 573 U. S., at 278–279, would be negated in al- most every case if a defendant could shift the burden of persuasion to the plaintiffs by mustering any competent evidence of a lack of price impact (including, for example, the generic nature of the alleged mis- representations). Thus, the best reading of the Court’s precedents as- signs defendants the burden of persuasion to prove a lack of price im- pact by a preponderance of the evidence. Even so, that allocated burden will be outcome determinative only in the rare case in which the evidence is in perfect equipoise. Pp. 9–12. 955 F. 3d 254, vacated and remanded.

BARRETT, J., delivered the opinion of the Court, in which ROBERTS, C. J., and BREYER, KAGAN, and KAVANAUGH, JJ., joined in full; in which THOMAS, ALITO, and GORSUCH, JJ., joined as to Parts I and II–A; and in which SOTOMAYOR, J., joined as to Parts I, II–A–1, and II–B. SOTOMAYOR, J., filed an opinion concurring in part and dissenting in part. GORSUCH, J., filed an opinion concurring in part and dissenting in part, in which THOMAS and ALITO, JJ., joined. Cite as: 594 U. S. ____ (2021) 1

Opinion of the Court

NOTICE: This opinion is subject to formal revision before publication in the preliminary print of the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Wash- ington, D. C. 20543, of any typographical or other formal errors, in order that corrections may be made before the preliminary print goes to press.

SUPREME COURT OF THE UNITED STATES _________________

No. 20–222 _________________

GOLDMAN SACHS GROUP, INC., ET AL., PETITIONERS v. ARKANSAS TEACHER RETIREMENT SYSTEM, ET AL. ON WRIT OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT [June 21, 2021]

JUSTICE BARRETT delivered the opinion of the Court. This case involves a securities-fraud class action filed by several pension funds against The Goldman Sachs Group, Inc., and three of its former executives (collectively, Gold- man). Plaintiffs allege that Goldman maintained an artifi- cially inflated stock price by making generic statements about its ability to manage conflicts—for example, “We have extensive procedures and controls that are designed to identify and address conflicts of interest.” Plaintiffs say that Goldman’s generic statements were false or misleading in light of several undisclosed conflicts of interest, and that once the truth about Goldman’s conflicts came out, Gold- man’s stock price dropped and shareholders suffered losses. Below, this securities-fraud class action proceeded in typ- ical fashion. Plaintiffs sought to certify a class of Goldman shareholders by invoking the presumption endorsed by this Court in Basic Inc. v. Levinson, 485 U. S. 224 (1988). The Basic presumption is premised on the theory that investors rely on the market price of a company’s security, which in an efficient market incorporates all of the company’s public 2 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM Opinion of the Court

misrepresentations. For its part, Goldman sought to defeat class certification by rebutting the Basic presumption through evidence that its alleged misrepresentations actu- ally had no impact on its stock price. After determining that Goldman had failed to carry its burden of proving a lack of price impact, the District Court certified the class, and the Second Circuit affirmed. In this Court, Goldman argues that the Second Circuit erred twice: first, by holding that the generic nature of its alleged misrepresentations is irrelevant to the price impact inquiry; and second, by assigning Goldman the burden of persuasion to prove a lack of price impact. On the first question, the parties now agree, as do we, that the generic nature of a misrepresentation often is im- portant evidence of price impact that courts should consider at class certification. Because we conclude that the Second Circuit may not have properly considered the generic na- ture of Goldman’s alleged misrepresentations, we vacate and remand for the Court of Appeals to reassess the District Court’s price impact determination. On the second ques- tion, we agree with the Second Circuit that our precedents require defendants to bear the burden of persuasion to prove a lack of price impact by a preponderance of the evi- dence. We emphasize, though, that the burden of persua- sion should rarely be outcome determinative. I A Section 10(b) of the Securities Exchange Act of 1934 and its implementing regulation, Rule 10b–5, prohibit material misrepresentations and omissions in connection with the sale of securities. 48 Stat. 881, as amended, 15 U. S. C. §78j(b); 17 CFR §240.10b–5 (2020). We have inferred from these provisions an implied private cause of action permit- ting the recovery of damages for securities fraud. Hallibur- ton Co. v. Erica P. John Fund, Inc., 573 U. S. 258, 267 Cite as: 594 U. S. ____ (2021) 3

(2014) (Halliburton II ). To recover damages, a private plaintiff must prove, among other things, a material mis- representation or omission by the defendant and the plain- tiff ’s reliance on that misrepresentation or omission. Ibid. This case concerns the element of reliance. The “tradi- tional (and most direct) way” for a plaintiff to prove reliance is to show that he was aware of a defendant’s misrepresen- tation and engaged in a transaction based on that misrep- resentation. Ibid. (internal quotation marks omitted). In Basic, however, we held that a plaintiff may also invoke a rebuttable presumption of reliance based on the fraud-on- the-market theory. 485 U. S., at 241–247. The “fundamental premise” of the fraud-on-the-market theory underlying Basic’s presumption is “that an investor presumptively relies on a misrepresentation so long as it was reflected in the market price at the time of his transac- tion.” Erica P. John Fund, Inc. v. Halliburton Co., 563 U. S. 804, 813 (2011). To invoke the Basic presumption, a plain- tiff must prove: (1) that the alleged misrepresentation was publicly known; (2) that it was material; (3) that the stock traded in an efficient market; and (4) that the plaintiff traded the stock between the time the misrepresentation was made and when the truth was revealed. Halliburton II, 573 U. S., at 268. The defendant may then rebut the presumption through “[a]ny showing that severs the link between the alleged misrepresentation and either the price received (or paid) by the plaintiff, or his decision to trade at a fair market price.” Basic, 485 U. S., at 248. Although the Basic presumption “can be invoked by any Rule 10b–5 plaintiff,” it has “particular significance in se- curities-fraud class actions.” Amgen Inc. v. Connecticut Re- tirement Plans and Trust Funds, 568 U. S. 455, 462 (2013). The presumption allows class-action plaintiffs to prove re- liance through evidence common to the class. That in turn makes it easier for plaintiffs to establish the predominance requirement of Federal Rule of Civil Procedure 23, which 4 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM Opinion of the Court

requires that “questions of law or fact common to class members predominate” over individualized issues. Fed. Rule Civ. Proc. 23(b)(3). Indeed, without the Basic pre- sumption, individualized issues of reliance ordinarily would defeat predominance and “preclude certification” of a secu- rities-fraud class action. Amgen, 568 U. S., at 462–463; see Halliburton II, 573 U. S., at 281–282. As a result, class-action plaintiffs must prove the Basic prerequisites before class certification—with one exception. In Amgen, we held that materiality should be left to the merits stage because it does not bear on Rule 23’s predomi- nance requirement. 568 U. S., at 466–468. The remaining Basic prerequisites—publicity, market efficiency, and mar- ket timing—“must be satisfied” by plaintiffs “before class certification.” Halliburton II, 573 U. S., at 276. Satisfying those prerequisites, however, does not guaran- tee class certification. We held in Halliburton II that de- fendants may rebut the Basic presumption at class certifi- cation by showing “that an alleged misrepresentation did not actually affect the market price of the stock.” Id., at 284. If a misrepresentation had no price impact, then Basic’s fundamental premise “completely collapses, render- ing class certification inappropriate.” Id., at 283. B Respondents here—whom we’ll call Plaintiffs—are Gold- man shareholders. They brought this securities-fraud class action against Goldman in the Southern District of New York, alleging violations of §10(b) and Rule 10b–5. The specific theory of securities fraud that Plaintiffs al- lege is known as inflation maintenance. Under this theory, a misrepresentation causes a stock price “to remain inflated by preventing preexisting inflation from dissipating from the stock price.” FindWhat Investor Group v. Cite as: 594 U. S. ____ (2021) 5

FindWhat.com, 658 F. 3d 1282, 1315 (CA11 2011).1 Plaintiffs allege here that between 2006 and 2010, Gold- man maintained an inflated stock price by making repeated misrepresentations about its conflict-of-interest policies and business practices. The alleged misrepresentations are generic statements from Goldman’s SEC filings and annual reports, including the following:  “We have extensive procedures and controls that are designed to identify and address conflicts of inter- est.” App. 216 (emphasis and boldface deleted).  “Our clients’ interests always come first.” Id., at 162 (same).  “Integrity and honesty are at the heart of our busi- ness.” Id., at 163 (same). According to Plaintiffs, these statements were false or mis- leading—and caused Goldman’s stock to trade at artificially inflated levels—because Goldman had in fact engaged in several allegedly conflicted transactions without disclosing the conflicts. Plaintiffs further allege that once the market learned the truth about Goldman’s conflicts from a Govern- ment enforcement action and subsequent news reports, the inflation in Goldman’s stock price dissipated, causing the price to drop and shareholders to suffer losses. After Goldman unsuccessfully moved to dismiss the case, Plaintiffs moved to certify the class, invoking the Basic pre- sumption. In response, Goldman sought to rebut the Basic presumption by proving a lack of price impact. Both parties submitted extensive expert testimony on the issue. The District Court certified the class, but the Second Cir- cuit authorized a Rule 23(f ) appeal and vacated the class-

—————— 1 Although some Courts of Appeals have approved the inflation-

maintenance theory, this Court has expressed no view on its validity or its contours. We need not and do not do so in this case. 6 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM Opinion of the Court

certification order. 879 F. 3d 474 (2018). The Second Cir- cuit held that Goldman, as the defendant, bears the burden of persuasion to prove a lack of price impact by a prepon- derance of the evidence. But it concluded that the District Court erred by holding Goldman to a higher burden of proof and by refusing to consider some of Goldman’s price impact evidence. On remand, the District Court certified the class again, finding that Goldman’s expert testimony failed to establish by a preponderance of the evidence that its alleged misrep- resentations had no price impact. The Second Circuit again authorized a Rule 23(f ) appeal and this time affirmed in a divided decision. 955 F. 3d 254 (2020). As relevant here, the Court of Appeals held that the District Court’s price im- pact determination was not an abuse of discretion. In dis- sent, Judge Sullivan concluded that “the generic quality of Goldman’s alleged misstatements, coupled with” Goldman’s expert testimony, compelled the conclusion that Goldman proved a lack of price impact. Id., at 278–279. We granted certiorari. 592 U. S. ___ (2020). II Goldman argues that the Second Circuit erred in two re- spects: first, by concluding that the generic nature of al- leged misrepresentations is irrelevant to the price impact question; and second, by placing the burden of persuasion on Goldman to prove a lack of price impact. We address these arguments in turn. A 1 On the first question—whether the generic nature of a misrepresentation is relevant to price impact—the parties’ dispute has largely evaporated. Plaintiffs now concede that the generic nature of an alleged misrepresentation often will be important evidence of price impact because, as a rule Cite as: 594 U. S. ____ (2021) 7

of thumb, “a more-general statement will affect a security’s price less than a more-specific statement on the same ques- tion.” Brief for Respondents 15; see Tr. of Oral Arg. 6–7, 59, 79. The parties further agree that courts may consider ex- pert testimony and use their common sense in assessing whether a generic misrepresentation had a price impact. See Tr. of Oral Arg. 12, 64. And they likewise agree that courts may assess the generic nature of a misrepresentation at class certification even though it also may be relevant to materiality, which Amgen reserves for the merits. See id., at 23, 65. We share the parties’ view. In assessing price impact at class certification, courts “ ‘should be open to all probative evidence on that question—qualitative as well as quantita- tive—aided by a good dose of common sense.’ ” In re Allstate Corp. Securities Litig., 966 F. 3d 595, 613, n. 6 (CA7 2020) (quoting Langevoort, Judgment Day for Fraud-on-the-Mar- ket: Reflection on Amgen and the Second Coming of Halli- burton, 57 Ariz. L. Rev. 37, 56 (2015); emphasis added). That is so regardless whether the evidence is also relevant to a merits question like materiality. As we have repeat- edly explained, a court has an obligation before certifying a class to “determin[e] that Rule 23 is satisfied, even when that requires inquiry into the merits.” Comcast Corp. v. Behrend, 569 U. S. 27, 35 (2013); see Wal-Mart Stores, Inc. v. Dukes, 564 U. S. 338, 351, and n. 6 (2011). And under Halliburton II, a court cannot conclude that Rule 23’s re- quirements are satisfied without considering all evidence relevant to price impact. See 573 U. S., at 284.2 —————— 2 We recognize that materiality and price impact are overlapping con-

cepts and that the evidence relevant to one will almost always be rele- vant to the other. But “a district court may not use the overlap to refuse to consider the evidence.” In re Allstate, 966 F. 3d, at 608. Instead, the district court must use the evidence to decide the price impact issue “while resisting the temptation to draw what may be obvious inferences for the closely related issues that must be left for the merits, including 8 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM Opinion of the Court

The generic nature of a misrepresentation often will be important evidence of a lack of price impact, particularly in cases proceeding under the inflation-maintenance theory. Under that theory, price impact is the amount of price in- flation maintained by an alleged misrepresentation—in other words, the amount that the stock’s price would have fallen “without the false statement.” Glickenhaus & Co. v. Household Int’l, Inc., 787 F. 3d 408, 415 (CA7 2020). Plain- tiffs typically try to prove the amount of inflation indirectly: They point to a negative disclosure about a company and an associated drop in its stock price; allege that the disclosure corrected an earlier misrepresentation; and then claim that the price drop is equal to the amount of inflation main- tained by the earlier misrepresentation. See, e.g., id., at 413–417; In re Vivendi, S. A. Securities Litig., 838 F. 3d 223, 233–237, 253–259 (CA2 2016). But that final inference—that the back-end price drop equals front-end inflation—starts to break down when there is a mismatch between the contents of the misrepre- sentation and the corrective disclosure. That may occur when the earlier misrepresentation is generic (e.g., “we have faith in our business model”) and the later corrective disclosure is specific (e.g., “our fourth quarter earnings did not meet expectations”). Under those circumstances, it is less likely that the specific disclosure actually corrected the generic misrepresentation, which means that there is less reason to infer front-end price inflation—that is, price im- pact—from the back-end price drop. 2 The parties do not dispute any of this. They disagree only about whether the Second Circuit properly considered the generic nature of Goldman’s alleged misrepresentations.

—————— materiality.” Id., at 609. Cite as: 594 U. S. ____ (2021) 9

Because the Second Circuit’s opinions leave us with suffi- cient doubt on this score, we remand for further considera- tion.3 On remand, the Second Circuit must take into ac- count all record evidence relevant to price impact, regardless whether that evidence overlaps with materiality or any other merits issue. B Goldman also argues that the Second Circuit erred by re- quiring Goldman, rather than Plaintiffs, to bear the burden of persuasion on price impact at class certification. Gold- man relies exclusively on Federal Rule of Evidence 301, which provides in full: “In a civil case, unless a federal statute or these rules provide otherwise, the party against whom a presump- tion is directed has the burden of producing evidence to rebut the presumption. But this rule does not shift the burden of persuasion, which remains on the party who had it originally.” According to Goldman, Rule 301 applies to the Basic pre- sumption at class certification, and, as a result, a plaintiff ’s satisfaction of the Basic prerequisites shifts only the bur- den of production to the defendant. Once a defendant dis- charges that burden by producing any competent evidence of a lack of price impact, Goldman says, the Basic presump- tion is rebutted and the plaintiff must carry the burden of persuasion to show price impact.

—————— 3 Compare 955 F. 3d 254, 268 (2020) (“Whether alleged misstatements

are too general to demonstrate price impact has nothing to do with the issue of whether common questions predominate over individual ones” (emphasis added)); id., at 270 (“The inflation-maintenance theory does not discriminate between general and specific misstatements”), with 879 F. 3d 474, 485–486 (2018) (correctly requiring the District Court to con- sider Goldman’s price impact evidence notwithstanding overlap with ma- teriality). 10 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM Opinion of the Court

We disagree. We have held that Rule 301 “in no way re- stricts the authority of a court . . . to change the customary burdens of persuasion” pursuant to a federal statute. NLRB v. Transportation Management Corp., 462 U. S. 393, 404, n. 7 (1983). And we have at times exercised that au- thority to reassign the burden of persuasion to the defend- ant upon a prima facie showing by the plaintiff. See, e.g., Teamsters v. United States, 431 U. S. 324, 359, and n. 45 (1977); Franks v. Bowman Transp. Co., 424 U. S. 747, 772– 773 (1976). Goldman does not ask us to revisit these precedents. So the threshold question here is not whether we have the au- thority to assign defendants the burden of persuasion to prove a lack of price impact, but instead whether we already exercised that authority in establishing the Basic frame- work pursuant to the securities laws. We conclude that Basic and Halliburton II did just that. Basic held that defendants may rebut the presumption of reliance if they “show that the misrepresentation in fact did not lead to a distortion of price.” 485 U. S., at 248 (empha- sis added). To do so, Basic said, defendants may make “[a]ny showing that severs the link between the alleged mis- representation and . . . the price received (or paid) by the plaintiff.” Ibid. (emphasis added). Similarly, Halliburton II held that defendants may rebut the Basic presumption at class certification “by showing . . . that the particular misrepresentation at issue did not affect the stock’s market price.” 573 U. S., at 279 (emphasis added). Goldman and JUSTICE GORSUCH argue that these refer- ences to a defendant’s “showing” refer to the defendant’s burden of production. Post, at 6–8 (dissenting opinion) (hereinafter the dissent). On this reading, Basic and Halli- burton II require a defendant merely to offer “evidence that, if believed, would support a finding” of a lack of price im- pact. Post, at 5. But Basic and Halliburton II plainly re- quire more: The defendant must “in fact” “seve[r] the link” Cite as: 594 U. S. ____ (2021) 11

between a misrepresentation and the price paid by the plaintiff—and a defendant’s mere production of some evi- dence relevant to price impact would rarely accomplish that feat.4 Accepting Goldman and the dissent’s argument would also effectively negate Halliburton II’s holding that plain- tiffs need not directly prove price impact in order to invoke the Basic presumption. 573 U. S., at 278–279. If, as they urge, the defendant could defeat Basic’s presumption by in- troducing any competent evidence of a lack of price im- pact—including, for example, the generic nature of the al- leged misrepresentations—then the plaintiff would end up with the burden of directly proving price impact in almost every case. And that would be nearly indistinguishable from the regime that Halliburton II rejected. Thus, the best reading of our precedents—as the Courts of Appeals to have considered the issue have recognized— is that the defendant bears the burden of persuasion to prove a lack of price impact. See Waggoner v. Barclays PLC, 875 F. 3d 79, 99–104 (CA2 2017) (“the phrase ‘[a]ny showing that severs the link’ aligns more logically with im- posing a burden of persuasion rather than a burden of pro- duction”); In re Allstate, 966 F. 3d, at 610–611 (“Basic said that ‘[a]ny showing that severs the link’ would be sufficient to rebut the presumption, not that mere production of evi- dence would defeat the presumption” (citation omitted)). We likewise agree with the Courts of Appeals that the de- fendant must carry that burden by a preponderance of the evidence. See Waggoner, 875 F. 3d, at 99; In re Allstate, 966 F. 3d, at 610. Although the defendant bears the burden of persuasion,

—————— 4 The dissent points out that, as a general rule, presumptions shift only

the burden of production. Post, at 2–4. We don’t disagree, but we read Basic and Halliburton II as a clear departure from that general rule. 12 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM Opinion of the Court

the allocation of the burden is unlikely to make much dif- ference on the ground. In most securities-fraud class ac- tions, as in this one, the plaintiffs and defendants submit competing expert evidence on price impact. The district court’s task is simply to assess all the evidence of price im- pact—direct and indirect—and determine whether it is more likely than not that the alleged misrepresentations had a price impact. The defendant’s burden of persuasion will have bite only when the court finds the evidence in eq- uipoise—a situation that should rarely arise. Cf. Medina v. California, 505 U. S. 437, 449 (1992) (preponderance of the evidence burden matters “only in a narrow class of cases where the evidence is in equipoise”). * * * The Second Circuit correctly placed the burden of proving a lack of price impact on Goldman. But because it is unclear whether the Second Circuit properly considered the generic nature of Goldman’s alleged misrepresentations in review- ing the District Court’s price impact determination, we va- cate the judgment of the Second Circuit and remand the case for further proceedings consistent with this opinion.

It is so ordered. Cite as: 594 U. S. ____ (2021) 1

SOTOMAYOR, J., Opinion concurring of SinOTOMAYOR part and,dissenting J. in part

GOLDMAN SACHS GROUP, INC., ET AL., PETITIONERS v. ARKANSAS TEACHER RETIREMENT SYSTEM, ET AL. ON WRIT OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT [June 21, 2021]

JUSTICE SOTOMAYOR, concurring in part and dissenting in part. I agree with the Court’s answers to the questions pre- sented, and I accordingly join Parts I, II–A–1, and II–B of the Court’s opinion. Under Basic Inc. v. Levinson, 485 U. S. 224 (1988), securities plaintiffs may demonstrate reliance by invoking the rebuttable presumption that investors rely on any misrepresentations reflected in a security’s market price. Id., at 241–247. The Basic presumption is particu- larly useful to class-action plaintiffs who, without the pre- sumption, ordinarily could not demonstrate that questions common to the class predominate over individual ones. Ante, at 3–4. Defendants, for their part, may rebut the Basic presumption by demonstrating that the alleged mis- representations did not in fact affect the security’s price. Ante, at 3. So-called “price impact” may be disproved with a variety of evidence, alone or in combination. As the Court holds today, one potentially relevant piece of evidence may be the “generic nature” of the misrepresentation. Ante, at 6–8. I do not, however, join the Court’s judgment to vacate and remand because I believe the Second Circuit “properly con- sidered the generic nature of Goldman’s alleged misrepre- 2 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM SOTOMAYOR, J., Opinion concurring of SinOTOMAYOR part and,dissenting J. in part

sentations.” Ante, at 8. On appeal, Goldman did not con- tend that the District Court improperly refused to consider the generic nature of the alleged misstatements as evidence of price impact (or lack thereof ). Instead, Goldman argued that “general statements, like those challenged here, are in- capable of impacting a company’s stock price as a matter of law” because they are “ ‘too general to cause a reasonable investor to rely upon them.’ ” Brief for Appellants in No. 18–3667 (CA2), pp. 43, 46. Goldman reasoned that “the challenged statements are incapable of maintaining infla- tion in a stock price for the same reasons that those state- ments are immaterial as a matter of law (as well as fact).” Id., at 48. The Second Circuit properly rejected Goldman’s argu- ment. The court explained that although “Goldman is not formally asking for a materiality test,” its proposed rule would “essentially requir[e] courts to ask” at the class- certification stage “whether the alleged misstatements are, in Goldman’s words, ‘immaterial as a matter of law.’ ” 955 F. 3d 254, 267 (2020). But “materiality is irrelevant at the Rule 23 stage.” Id., at 268 (citing Amgen Inc. v. Connecticut Retirement Plans and Trust Funds, 568 U. S. 455, 468 (2013)). “If general statements cannot maintain price infla- tion because no reasonable investor would have relied on them, then the question of inactionable generality is com- mon to the class.” 955 F. 3d, at 268. In declining to adopt Goldman’s proposed rule that ge- neric misstatements cannot have a price impact (as a mat- ter of law), the Second Circuit nowhere held that the ge- neric nature of an alleged misstatement could not serve as evidence of price impact (as a matter of fact). Nor did the Second Circuit refuse to consider such evidence in affirming the District Court’s finding that Goldman failed to rebut the Basic presumption. The Court nevertheless reads a hand- ful of sentences in the Second Circuit’s opinion to create Cite as: 594 U. S. ____ (2021) 3

SOTOMAYOR, J., Opinion concurring of SinOTOMAYOR part and,dissenting J. in part

“doubt” over whether the Court of Appeals refused to con- sider “all record evidence relevant to price impact.” Ante, at 9; see also ante, at 9, n. 3. But such statements must be viewed in the context of Goldman’s now-abandoned argu- ment that generic misrepresentations have no price impact as a matter of law. Take, for example, the Second Circuit’s statement that “[w]hether alleged misstatements are too general to demonstrate price impact has nothing to do with the issue of whether common questions predominate over individual ones.” 955 F. 3d, at 268. Fairly read in light of Goldman’s appellate briefing, that sentence addresses only Goldman’s argument that general statements are always per se irrelevant. That is why the Second Circuit observed several sentences later that “Goldman’s test is materiality by another name.” Ibid. At the same time, the court was careful to emphasize that defendants “may attempt to dis- prove [price impact] at class certification” even though the inquiry “resembles materiality.”* Id., at 267. In short, the Second Circuit did not address whether the generic nature of a misstatement may be used as evidence to disprove price impact for a simple reason: Goldman iden- tified no error in the District Court’s treatment of such evi- dence. Goldman did not press the argument in the Second Circuit that it now urges here, and the Second Circuit did not reject the proposition that this Court now adopts. Thus, the argument Goldman seeks to press on remand is unpre-

—————— *Indeed, in a prior appeal in this case, an earlier Second Circuit panel vacated an order of the District Court in part because it had refused to consider price-impact evidence that overlapped with materiality. See 879 F. 3d 474, 486 (2018) (holding that “[a]lthough price impact touches on materiality, which is not an appropriate consideration at the class certification stage,” courts nonetheless must consider evidence regarding “[w]hether a misrepresentation was reflected in the market price at the time of the transaction”). It is hard to imagine that the Second Circuit here was unaware of (or intended to depart from) the prior panel’s hold- ing. 4 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM SOTOMAYOR, J., Opinion concurring of SinOTOMAYOR part and,dissenting J. in part

served, and nothing in the Second Circuit’s opinion mis- states the law. Because affirmance is appropriate under these circumstances, I respectfully dissent from Part II–A– 2 of the Court’s opinion and from the judgment of the Court. Cite as: 594 U. S. ____ (2021) 1

G Opinion , J., ORSUCHof dissenting GORSUCH , J.

GOLDMAN SACHS GROUP, INC., ET AL., PETITIONERS v. ARKANSAS TEACHER RETIREMENT SYSTEM, ET AL. ON WRIT OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT [June 21, 2021]

JUSTICE GORSUCH, with whom JUSTICE THOMAS and JUSTICE ALITO join, concurring in part and dissenting in part. I join all but Part II–B of the Court’s opinion. There, the Court holds that the defendant, rather than the plaintiff, “bear[s] the burden of persuasion on price impact.” Ante, at 9. Respectfully, I disagree. We start from common ground. Basic Inc. v. Levinson, 485 U. S. 224, 245–247 (1988), sought to import fraud on the market theory from economics into securities litigation. In doing so, Basic posited two things—first, in an efficient market a company’s stock price generally reflects any pub- lic and material information about the company; second, in- vestors generally rely on a company’s stock price as an in- dicator of the firm’s true value. Ibid. Given these economic assumptions, the Court held that securities fraud plaintiffs can presumptively meet their burden of proving reliance on an alleged misrepresentation by proving four things: (1) the defendant’s alleged misrepresentation was publicly known; (2) it was material; (3) the stock traded in an effi- cient market; and (4) the plaintiff purchased the stock at the market price between the time the misrepresentation was made and the truth was revealed. See Halliburton Co. v. Erica P. John Fund, Inc., 573 U. S. 258, 277–278 (2014) 2 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM Opinion of GORSUCH, J.

(Halliburton II). The presumption of reliance not only helps a plaintiff prove one of the essential elements of a securities fraud claim. Certain class actions require that “questions of law or fact common to class members predominate over any questions affecting only individual members.” Fed. Rule Civ. Proc. 23(b)(3). So to the extent a court is able to pre- sume reliance by everyone who purchased an affected stock, Basic can help avoid individualized questions that other- wise might stand in the way of proceeding with a securities fraud action on a classwide basis. Erica P. John Fund, Inc. v. Halliburton Co., 563 U. S. 804, 809–811 (2011) (Hallibur- ton I). At the same time, Basic’s presumption of reliance has only ever been just that. Everyone accepts that, if a defend- ant undermines one of the assumptions on which it rests, the presumption dissipates. So, for example, if the defend- ant’s alleged misrepresentation did not actually affect the market price, there can be no ground for presuming anyone relied on that misrepresentation when purchasing the stock. Halliburton II, 573 U. S., at 279. Similarly, if a par- ticular plaintiff did not care about the integrity of the mar- ket price when purchasing a stock, there is no basis for pre- suming that individual’s reliance. Id., at 276. Before us, the only meaningful dispute concerns what burden a defendant bears when it comes to rebutting the Basic presumption. Does the defendant carry only a burden of production, or does the defendant sometimes carry a bur- den of persuasion? In my view, only a burden of production is involved. Start with what we have said about presumptions like Basic’s. This Court has long recognized that a “ ‘ “presump- tion” properly used refers only to a device for allocating the production burden.’ ” Texas Dept. of Community Affairs v. Burdine, 450 U. S. 248, 255, n. 8 (1981). Throughout the law, courts have sometimes created presumptions to help Cite as: 594 U. S. ____ (2021) 3

Opinion of GORSUCH, J.

plaintiffs prove their cases when direct evidence can be hard to come by. See Basic, 485 U. S., at 245. These pre- sumptions operate by allowing the plaintiff to prove only certain specified “predicate fact[s]” at the outset. St. Mary’s Honor Center v. Hicks, 509 U. S. 502, 506 (1993). If the plaintiff does so, an inference or “presumption” arises that the plaintiff has met its burden of persuasion, at least “in the absence” of some competing “explanation.” Ibid. (inter- nal quotation marks omitted). At that point, the defendant bears a burden of production to present evidence that, if “taken as true,” would “permit the conclusion” that the pre- sumption in the plaintiff ’s favor is mistaken. Id., at 509 (emphasis deleted). If the defendant produces such evi- dence, the presumption “drops from the case.” Id., at 507 (internal quotation marks omitted). “[T]he trier of fact” then “proceeds to decide the ultimate question.” Id., at 511. Throughout this whole back-and-forth process, the burden of persuasion never shifts: The “plaintiff at all times bears the ultimate burden of persuasion” to prove all aspects of its cause of action. Ibid. (internal quotation marks omit- ted). The Court has explained that nearly “all presumptions” operate in this way. Id., at 507. The Federal Rules of Evi- dence confirm the point too. Rule 301, titled “Presumptions in Civil Cases Generally,” provides that “the party against whom a presumption is directed has the burden of produc- ing evidence to rebut the presumption,” but “the burden of persuasion . . . remains on the party who had it originally.” Again, a burden of production may shift to the defendant, but never the burden of persuasion. Title VII practice offers a familiar illustration of these principles. There, the plaintiff bears the ultimate burden of proving that his employer intentionally discriminated against him because of his race or some other unlawful fac- tor. See St. Mary’s Honor Center, 509 U. S., at 511. But because direct evidence of intentional discrimination can be 4 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM Opinion of GORSUCH, J.

“elusive,” the Court has created a presumption. See id., at 506 (internal quotation marks omitted). If a plaintiff proves certain “predicate fact[s]”—for example, that he is black, that he was fired from a job for which he was qualified, and that the job remained open and was ultimately filled by a white person—an inference or presumption of intentional discrimination arises. Ibid. At that point, the defendant bears the burden of producing evidence that, if accepted as true, shows it fired the plaintiff for only legitimate business reasons. Id., at 506–507. Should that happen, the pre- sumption of intentional discrimination disappears and the trier of fact must weigh the parties’ competing proof. Id., at 510–511. None of that means the plaintiff ’s indirect evi- dence of discrimination also disappears. It simply means the trier of fact must consider any inferences arising from that indirect evidence while also considering the defend- ant’s evidence and any other proof the plaintiff submits. See id., at 511. “[A]t all times” throughout the litigation, however, the plaintiff bears the “ultimate burden of per- suading the trier of fact that he has been the victim of in- tentional discrimination.” Id., at 507–508 (brackets and in- ternal quotation marks omitted). Since the Court first started tangling with the fraud on the market theory in Basic, it has followed these traditional rules. Consistently, our decisions have “made clear” that Basic’s “presumption” of reliance is “just that.” Halliburton I, 563 U. S., at 811. Much as the Court said it created the Title VII presumption to help prove the “elusive” question of intentional discrimination, Basic said it created its pre- sumption of reliance to relieve “an unnecessarily unrealis- tic evidentiary burden” on securities fraud plaintiffs. 485 U. S., at 245. And when creating its presumption Basic ex- pressly cited Rule 301. Ibid. The process Basic outlined matches traditional under- standings too. The Court explained that a plaintiff ’s ability to prove certain “threshold facts”—about market operations Cite as: 594 U. S. ____ (2021) 5

and the publicity of the misstatement—gives rise to a “pre- sumption” of reliance. See id., at 248, and n. 27. After such a showing, the Court continued, a defendant may then pro- ceed to “rebut the presumption.” Id., at 248. Nowhere in any of this did Basic suggest the order of operations govern- ing its presumption should differ in any way from those gov- erning others commonly found in the law and subject to Rule 301. Nor is there any doubt which party has the bur- den of persuasion on the question of reliance in securities fraud cases like ours. From start to finish, the plaintiff has the burden to satisfy that essential element of its claim. Basic’s presumption of reliance thus “does not shift” any burden of persuasion—that always “remains” with the plaintiff. Fed. Rule Evid. 301; see also St. Mary’s Honor Center, 509 U. S., at 506–508. Consider how all this works in routine securities fraud cases. Once a plaintiff proves the four “predicate facts” Basic specified, see supra, at 1, a presumption of reliance attaches. At that point, the defendant bears the burden of producing evidence that, if believed, would support a find- ing that the plaintiff did not actually rely on its alleged mis- representation. As we have seen, a defendant might do so by producing evidence suggesting that its alleged misrepre- sentation did not have an impact on market price or that the plaintiff was indifferent to the alleged misrepresenta- tion. Upon such a showing, the presumption of reliance drops from the case and the trier of fact must decide the question of reliance vel non, cognizant of the fact the plain- tiff bears the burden of proving reliance like any other es- sential elements of its claim. Again, that does not mean the plaintiff ’s indirect proof disappears. A court may still infer from the Basic predicates that a particular misstatement was incorporated into the stock price and that the plaintiff relied on the integrity of that price. Both sides are free to present additional proof too. It’s simply that a court no longer must find reliance. See St. Mary’s Honor Center, 509 6 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM Opinion of GORSUCH, J.

U. S., at 511. The Court disputes none of this. It does not even try to defend on the merits its unusual suggestion that the de- fendant carries some burden of persuasion in a plaintiff ’s claim for securities fraud. Instead, the Court contends only that precedent ties our hands. Primarily, the Court points to a single clause in a single sentence in Basic observing that a defendant may rebut the presumption of reliance with “[a]ny showing that severs the link between the alleged misrepresentation” and the stock price. See ante, at 10 (quoting Basic, 485 U. S., at 248) (em- phasis deleted). The Court then splices that clause together with another clause in a preceding sentence explaining that, before Basic, lower courts had said a defendant rebuts the fraud on the market presumption by showing “that the misrepresentation in fact did not lead to a distortion of price.” Ante, at 10 (quoting Basic, 485 U. S., at 248; empha- sis deleted). But what does this prove? Surely this language confirms an important and by now familiar point: Once a defendant produces evidence that, if believed, shows that fraud on the market theory does not hold in its particular case because its alleged misrepresentation in fact failed to affect the stock price, the presumption of reliance drops away. On the Court’s reading today, however, this language doesn’t just carry that obvious meaning. We are told it also must mean that Basic intended to shift the “burden of persuasion” with respect to “price impact” to the defendant—at least “at class certification”—because the “mere production of some evi- dence relevant to price impact would rarely accomplish th[e] feat” of “in fact” “ sever[ing] the link between a misrep- resentation and the price paid” for the stock. Ante, at 10– 11 (internal quotation marks omitted; emphasis deleted). That much does not follow. Not only has this Court often said it is a mistake to parse terms in a judicial opinion with the kind of punctilious exactitude due statutory language. Cite as: 594 U. S. ____ (2021) 7

See Reiter v. Sonotone Corp., 442 U. S. 330, 341 (1979). Even read for all they are worth, the handful of words on which the Court rests its entire holding today—a “showing” that “in fact” “sever[s] the link”—cannot begin to carry the weight the Court assigns them. See ante, at 10 (emphasis deleted). These terms do not even appear together in Basic: The Court has to pluck the phrase “in fact” from one sen- tence and the phrase “[a] showing that severs the link” from another, and then combine them to create a new clause that appears nowhere in the U. S. Reports—a “showing” that “in fact sever[s] the link.” Ante, at 10 (internal quotation marks omitted). Even then, the Court’s newly handcrafted phrase does not so much as mention the terms “burden of persuasion” or “price impact.” The hard truth is that in the 30-plus years since Basic this Court has never (before) suggested that plaintiffs are relieved from carrying the burden of persuasion on any as- pect of their own causes of action. To the contrary, when discussing the presumption it created, Basic expressly ref- erenced Rule 301 and invoked its normal order of opera- tions. And this Court has long explained that presumptions “properly used” refer only to devices “for allocating the pro- duction burden,” and not the burden of persuasion. Bur- dine, 450 U. S., at 255, n. 8 (internal quotation marks omit- ted). Are we really to believe that Basic—while referencing traditional understandings embodied in Rule 301 and just seven years after Burdine—secretly meant to depart from traditional and “proper” understandings about how pre- sumptions work? Thanks to spliced clauses found in two sentences this Court has never before read that way? All while using words that carry another and much more natu- ral meaning? To state the theory is to refute it. If Basic doesn’t command today’s result, the Court offers a backup theory. Separately, it insists, Halliburton II re- quires us to shift a burden of persuasion to the defendant. Specifically, the Court points to the fact that Halliburton II 8 GOLDMAN SACHS GROUP, INC. v. ARKANSAS TEACHER RETIREMENT SYSTEM Opinion of GORSUCH, J.

reaffirmed Basic’s holding that a plaintiff need not show re- liance “directly,” but may do so “presumptively” by carrying the burden of proving the four Basic factual predicates. 573 U. S., at 278. A decision holding that the defendant merely bore the burden of producing evidence suggesting a lack of price impact at class certification, the Court now submits, “would be nearly indistinguishable from the regime that Halliburton II rejected.” Ante, at 11. That much does not follow either. Like Basic, Hallibur- ton II concerned what facts a plaintiff must produce to gen- erate a presumption of reliance. This case is about what defendants must do to rebut that presumption. Deciding one does not resolve the other. To say these issues are “in- distinguishable” is to miss the entire point of a presump- tion: It allows the plaintiff to state a prima facie case based on inference and requires the defendant to bear the burden of producing evidence in response; once the defendant does so, the presumption has served its purpose and drops from the case. At that point, the factfinder now has the benefit of evidence from both sides and must decide the case with reference to the plaintiff ’s burden of persuasion. Nothing in Halliburton II suggests a departure from these princi- ples, let alone that some burden of persuasion secretly shifts to the defendant in a plaintiff ’s claim for securities fraud. To the contrary, that decision arose in the class cer- tification context and expressly reaffirmed that “[t]he Basic presumption does not relieve plaintiffs of the burden of proving” they have satisfied “the predominance require- ment of Rule 23(b)(3).” 573 U. S., at 276. The Court has no answer to any of this. Instead, it replies only by touting the fact that two Court of Appeals decisions have read Basic and Halliburton II as it does. Ante, at 11. But this is a non sequitur. The Court does not suggest that a pair of lower court opinions represents some robust judi- cial consensus. Nor does the Court suggest those opinions free us from having to interpret the law for ourselves. After Cite as: 594 U. S. ____ (2021) 9

all, “[o]ur duty is to follow the law as we find it, not to follow rotely whatever lower courts might once have said about it.” BP p.l.c. v. Mayor and City Council of Baltimore, 593 U. S. ___, ___–___ (2021) (slip op., at 11–12). The fact remains that nothing in our prior decisions has ever placed a burden of persuasion on the defendant with respect to any aspect of the plaintiff ’s case. It is incumbent on the plaintiff to prove reliance, not the defendant to disprove it. If a major- ity of the Court today really believes some novel new bur- den of persuasion should be placed on the defendant, it ought to say so. Past decisions—by this Court or others— cannot be blamed for today’s result. Perhaps recognizing the incongruity of its conclusion, the Court goes out of its way to downplay its significance. We’re told that “on the ground” today’s holding “is unlikely to make much difference” because “[i]n most securities-fraud class actions . . . the plaintiffs and defendants submit com- peting expert evidence on price impact.” Ante, at 12. And in cases like these, “[t]he district court’s task,” according to the Court, “is simply to assess all the evidence of price im- pact” and “determine whether it is more likely than not that the alleged misrepresentations had a price impact.” Ibid. This is a curious disavowal. Obviously, the Court thinks the issue important enough to spend the time and effort to rejigger the burden of persuasion. Now, though, it says none of this matters because most cases come down to a dis- pute over evidence of price impact irrespective of the pre- sumption. The Court’s suggestion that the burden of per- suasion will “rarely” make a “difference” misses the point too. The whole reason we allocate the burden of persuasion is to resolve close cases by providing a tie breaker where the burden does make a difference. That close cases may not be common ones is no justification for indifference about how the law resolves them. Respectfully, I dissent.