facts alleged—comparative or not—lead to the plausible inference that the actual process used by the defendant fiduciary was flawed. With appropriate evidence, Anderson could have stated a claim by pleading a true benchmark comparison, by providing other circumstantial allegations that plausibly suggested imprudence, or by directly showing that the specific investments the Intel fiduciaries selected or the general methodologies they used were imprudent.

COUNSEL

Matthew W.H. Wessler (argued), Gupta Wessler LLP, Washington, D.C.; Neil K. Sawhney and Jessica Garland, Gupta Wessler LLP, San Francisco, California; R. Joseph Barton, The Barton Firm LLP, Washington, D.C.; Joseph Creitz, Creitz & Serebin LLP, San Francisco, California; Michael L. Murphy and Gregory Y. Porter, Bailey & Glasser LLP, Washington, D.C.; for Plaintiffs-Appellants. Juli A. Lund (argued), Daniel F. Katz, and David S. Kurtzer- Ellenbogen, Williams & Connolly LLP, Washington, D.C.; Scott P. Cooper and Jennifer L. Roche, Proskauer Rose LLP, Los Angeles, California; Myron D. Rumeld, Proskauer Rose LLP, New York, New York; for Defendants-Appellees. Jaime Santos, Goodwin Procter LLP, Washington, D.C.;

ANDERSON V. INTEL CORP. INV. POLICY COMM. 5

Jordan Bock, Goodwin Procter LLP, Boston, Massachusetts; Tara S. Morrissey and Jordan L. Von Bokern, U.S. Chamber Litigation Center, Washington, D.C.; for Amicus Curiae the Chamber of Commerce of the United States of America.

OPINION

MILLER, Circuit Judge:

Winston R. Anderson brought this putative class action under the Employee Retirement Income Security Act of 1974 (ERISA), Pub. L. No. 93-406, 88 Stat. 829 (29 U.S.C. § 1001 et seq.), against the trustees of Intel Corporation’s proprietary retirement funds. He alleged that the trustees breached their duty of prudence by investing some of the funds’ assets in hedge funds and private equity funds. He also alleged that they breached their duty of loyalty by steering retirement funds to companies in which Intel’s venture-capital arm, Intel Capital, had already invested. The district court dismissed Anderson’s claims, concluding that he had not plausibly alleged a breach of either the duty of prudence or the duty of loyalty. We affirm. I From 2000 to 2015, Anderson was an Intel employee who participated in Intel’s employee retirement plans, including the Intel 401(k) Savings Plan and the Intel Retirement Contribution Plan. Both plans are “employee pension benefit plans” subject to ERISA. 29 U.S.C. § 1002(2)(A). ERISA requires that private pension plan assets “be held in trust.” 29 U.S.C. § 1103(a). To that end, it imposes certain fiduciary duties on a plan’s trustees, two of which are relevant here. First, the trustees have a duty of prudence: They must act “with the care, skill, prudence, and diligence under the circumstances then prevailing that a prudent man acting in a like capacity and familiar with such matters would use in the conduct of an enterprise of a like character and with like aims.” Id. § 1104(a)(1)(B). Second, they have a duty of loyalty: They must “discharge [their] duties with respect to a plan solely in the interest of the participants and beneficiaries.” Id. § 1104(a)(1). Participants in Intel’s plans may choose to invest their accounts in one or more customized funds managed by the plans’ trustees. Those funds include target-date funds, which hold a mix of asset classes including stocks, bonds, and cash equivalents that are adjusted to become more conservative as the fund approaches the target retirement date, and global diversified funds, which invest in a variety of assets, including domestic and international equity funds, bonds, and short-term investments. In response to the 2008 market crash and the ensuing recession, Intel redesigned its funds so that they included not just stocks and bonds but also hedge funds and private equity funds. A hedge fund is a privately organized pooled investment vehicle that engages in active trading of various assets, often including securities and commodity futures and options contracts. A private equity fund acquires and manages companies with the goal of improving them to earn a profit when the companies are sold again. Intel told participants that its new strategy was aimed at decreasing volatility and reducing the risk of large losses during a market downturn. It also disclosed the price that participants would pay for this risk mitigation: Because of their broad

ANDERSON V. INTEL CORP. INV. POLICY COMM. 7

diversification, the funds would not compare favorably with equity-heavy funds during bull markets. In 2019, Anderson brought this action in the Northern District of California against the managers of the plans. See 29 U.S.C. § 1109(a) (making ERISA plan fiduciaries personally liable for any losses to the plan resulting from a breach of fiduciary duty); id. § 1132(a)(2) (permitting plan participants to bring a civil action for relief under section 1109). He alleged that they had breached their duty of prudence because their large allocations to hedge funds and private equity funds had “drastically departed from prevailing standards of professional asset managers.” He also alleged that they had breached their duty of loyalty by improperly favoring investments that benefited Intel Capital—Intel’s venture capital arm—at the expense of the plan participants. (He also asserted several additional claims, but because the parties agree that those claims are derivative of the claims based on breach of fiduciary duty, we do not separately discuss them.) Anderson asked the district court to certify a class consisting of all plan participants whose accounts were invested in the target-date funds or global diversified funds after October 2009. The case was subsequently consolidated with a case brought by Christopher Sulyma, another former Intel employee. The district court dismissed the complaint for failure to state a claim. The court rejected the duty-of-prudence claim because Anderson had not alleged facts sufficient to support the allegation that the funds suffered from poor performance compared to peer funds. To make such an allegation plausible, the court reasoned, Anderson would need to provide “a meaningful benchmark against which to compare the Intel Funds,” but he had “failed to allege facts that would demonstrate that [his] chosen ‘comparable funds’” were indeed meaningful benchmarks. As to the duty-of-loyalty claim, the court held that Anderson’s allegations were “devoid of plausible allegations that could show a conflict of interest or self-dealing.” The district court granted leave to amend. In the amended complaint—the operative pleading here— Anderson again asserted claims based on breach of the duty of prudence and the duty of loyalty. The amended complaint detailed how the funds underperformed allegedly comparable alternatives, including published indices like the S&P 500 and Morningstar categories of peer-group funds. It also alleged “that hedge funds and private equity pose challenges and risks beyond those posed by ‘traditional investments’ such as mutual funds” and “do not increase diversification of asset classes.” It included further detail on how the fiduciaries’ investment decisions had benefited Intel and Intel Capital. The district court again dismissed, this time with prejudice. The court concluded that Anderson still had not identified a “meaningful benchmark” against which to compare the performance of Intel’s funds. The court explained that “simply labeling funds as comparable or as in the same category as the Intel [target-date funds] and Intel [global diversified funds] is insufficient to establish that those funds are meaningful benchmarks.” The court also stated that, although Anderson had added more detail to his duty-of-loyalty allegations, the allegations were “much the same as” those of the first complaint and were insufficient to support the claim that the fiduciaries had engaged in self- dealing. Anderson appeals. We review de novo the district court’s dismissal of a complaint for failure to state a claim. Wells

ANDERSON V. INTEL CORP. INV. POLICY COMM. 9

Fargo Bank, N.A. v. Mahogany Meadows Ave. Tr., 979 F.3d 1209, 1213 (9th Cir. 2020). “To survive a motion to dismiss, a complaint must contain sufficient factual matter, accepted as true, to ‘state a claim to relief that is plausible on its face.’” Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) (quoting Bell Atl. Corp. v. Twombly, 550 U.S. 544, 570 (2007)). II We begin with Anderson’s claim that the plan trustees breached their duty of prudence. Anderson contends that the trustees acted imprudently both by initially allocating some of the plans’ assets to hedge funds and private equity funds and by failing to adjust that allocation as it became clear that hedge funds and private equity funds were producing lower returns than those available from more traditional assets like stocks and bonds. We agree with the district court that Anderson has not stated an imprudence claim under ERISA. ERISA requires plan trustees to act with the “care, skill, prudence, and diligence under the circumstances then prevailing that a prudent man acting in a like capacity and familiar with such matters would use in the conduct of an enterprise of a like character and with like aims.” 29 U.S.C. § 1104(a)(1)(B); see also 29 C.F.R. § 2550.404a-1(b)(1). “At times, the circumstances facing an ERISA fiduciary will implicate difficult tradeoffs, and courts must give due regard to the range of reasonable judgments a fiduciary may make based on her experience and expertise.” Hughes v. Northwestern Univ., 595 U.S. 170, 177 (2022). ERISA “requires prudence, not prescience.” Debruyne v. Equitable Life Assurance Soc’y of the U.S., 920 F.2d 457, 465 (7th Cir. 1990) (quoting DeBruyne v. Equitable Life Assurance Soc’y of the U.S., 720 F. Supp. 1342, 1349 (N.D. Ill. 1989)). We therefore assess “a fiduciary’s actions based upon information available to the fiduciary at the time of each investment decision and not from the vantage point of hindsight.” PBGC ex rel. St. Vincent Catholic Med. Ctrs. Ret. Plan v. Morgan Stanley Inv. Mgmt. Inc., 712 F.3d 705, 716 (2d Cir. 2013) (quoting In re Citigroup ERISA Litig., 662 F.3d 128, 140 (2d Cir. 2011)); accord In re Unisys Sav. Plan Litig., 74 F.3d 420, 434 (3d Cir. 1996) (explaining that the inquiry turns on “a fiduciary’s conduct in arriving at an investment decision, not on its results”). Specifically, we ask “whether the individual trustees, at the time they engaged in the challenged transactions, employed the appropriate methods to investigate the merits of the investment and to structure the investment.” Wright v. Oregon Metallurgical Corp., 360 F.3d 1090, 1097 (9th Cir. 2004) (quoting Donovan v. Mazzola, 716 F.2d 1226, 1232 (9th Cir. 1983)). Because we evaluate prudence prospectively, based on the methods the fiduciaries employed, rather than retrospectively, based on the results they achieved, it is not enough for a plaintiff simply to allege that the fiduciaries could have obtained better results—whether higher returns, lower risks, or reduced costs—by choosing different investments. Instead, a plaintiff must provide “some further factual enhancement” to take the claim across “the line between possibility and plausibility.” Twombly, 550 U.S. at 557. There are a “myriad of circumstances that could violate the [prudence] standard.” In re Syncor ERISA Litig., 516 F.3d 1095, 1102 (9th Cir. 2008). For example, a plaintiff can plead a breach of the duty of prudence by alleging facts that would directly show that the fiduciaries employed unsound methods in making their investment decisions. See, e.g., Appvion, Inc. Ret. Sav. & Emp. Stock Ownership Plan ex rel. Lyon v. Buth, 99 F.4th 928, 946 (7th Cir. 2024) (trustees of

ANDERSON V. INTEL CORP. INV. POLICY COMM. 11

employee stock ownership plan purchased stock in reliance on appraiser’s valuation but “were careless in failing to scrutinize [the appraiser’s] valuation methods”); Stegemann v. Gannett Co., Inc., 970 F.3d 465, 476 (4th Cir. 2020) (trustees of retirement fund allegedly did not monitor a stock fund even though “two years elapsed” during which they “received risk warnings from auditors”). Alternatively, a plaintiff can make “circumstantial factual allegations” from which the court “may reasonably ‘infer from what is alleged that the process was flawed.’” St. Vincent, 712 F.3d at 718 (quoting Braden v. Wal-Mart Stores, Inc., 588 F.3d 585, 596 (8th Cir. 2009)). When an ERISA plaintiff attempts to do so by relying on a theory that “‘a prudent fiduciary in like circumstances’ would have selected a different fund based on the cost or performance of the selected fund,” that plaintiff “must provide a sound basis for comparison.” Meiners v. Wells Fargo & Co., 898 F.3d 820, 822 (8th Cir. 2018) (quoting St. Vincent, 712 F.3d at 720); accord Matney v. Barrick Gold of N. Am., 80 F.4th 1136, 1149 (10th Cir. 2023) (“A court cannot reasonably draw an inference of imprudence simply from the allegation that a cost disparity exists; rather, the complaint must state facts to show the funds or services being compared are, indeed, comparable.”); Albert v. Oshkosh Corp., 47 F.4th 570, 581–82 (7th Cir. 2022) (“The fact that actively managed funds charge higher fees than passively managed funds is ordinarily not enough to state a claim because such funds may also provide higher returns,” so a plaintiff must offer “more detailed allegations providing a ‘sound basis for comparison.’” (quoting Meiners, 898 F.3d at 822)); Smith v. CommonSpirit Health, 37 F.4th 1160, 1166 (6th Cir. 2022) (“[P]ointing to an alternative course of action, say another fund the plan might have invested in, will often be necessary to show a fund acted imprudently . . . .”). In other words, when a plaintiff alleges imprudence based on a fiduciary’s decision to make one investment rather than an alternative, “[t]he key to nudging an inference of imprudence from possible to plausible is providing ‘a sound basis for comparison—a meaningful benchmark’—not just alleging that ‘costs are too high, or returns are too low.’” Matousek v. MidAmerican Energy Co., 51 F.4th 274, 278 (8th Cir. 2022) (quoting Davis v. Washington Univ. in St. Louis, 960 F.3d 478, 484 (8th Cir. 2020)). The need for a relevant comparator with similar objectives—not just a better-performing plan or investment—is implicit in ERISA’s text. By making the standard of care that of a hypothetical prudent person “acting in a like capacity . . . in the conduct of an enterprise of a like character and with like aims,” the statute makes clear that the goals of the plan matter. The Department of Labor regulations implementing ERISA do the same. Those regulations provide that the duty of prudence is satisfied if the fiduciary has made a determination that a chosen investment “is reasonably designed, as part of the portfolio . . . , to further the purposes of the plan, taking into consideration the risk of loss and the opportunity for gain . . . compared to the opportunity for gain . . . associated with reasonably available alternatives with similar risks.” 29 C.F.R. § 2550.404a-1(b)(2)(i) (emphasis added). Anderson has made no direct allegation about Intel’s investment-selection methods, and he attempts to show a breach of the duty of prudence only through the circumstantial route. Specifically, he argues that the decision to invest in hedge funds and private equity funds caused Intel’s funds to underperform other funds and to incur higher fees. But the district court correctly determined that

ANDERSON V. INTEL CORP. INV. POLICY COMM. 13

Anderson did not plausibly allege that Intel’s funds underperformed other funds with comparable aims. Intel clearly disclosed the aims of its funds. Disclosures for the global diversified funds explained Intel’s risk- mitigation objective, noting that assets were allocated to “provide greater downside protection in faltering markets, with the tradeoff being slight underperformance in rallying ones, as has been the case in the current bull market.” Disclosures for the target-date funds similarly made clear that the goal was to “reduce investment risk by investing in assets whose returns are less correlated to equity markets.” Notably, Intel developed its own customized benchmarks, made up of a “composite of the underlying . . . benchmarks” for each asset class included in the Intel funds, which it disclosed to plan participants and beneficiaries. Intel explained that the benchmarks had “the same asset allocation as the Fund’s target asset allocation and use[d] index returns to represent the performance of the asset classes.” But rather than presenting a comparison to Intel’s composite benchmarks or to available funds with similar risk-mitigation strategies and objectives, Anderson sought to compare Intel’s funds to equity-heavy retail funds that pursued different objectives—typically revenue generation. As the district court observed, “simply labeling funds as ‘comparable’ or ‘a peer’ is insufficient to establish that those funds are meaningful benchmarks against which to compare the performance of the Intel funds.” Anderson’s putative comparators were not truly comparable because they had “different aims, different risks, and different potential rewards.” Davis, 960 F.3d at 485. Anderson emphasizes that the duty of prudence is “derived from the common law of trusts,” Tibble v. Edison Int’l, 575 U.S. 523, 528 (2015) (quoting Central States, Se. & Sw. Areas Pension Fund v. Central Transp., Inc., 472 U.S. 559, 570 (1985)), and that “[n]o fixed formula exists for determining whether a trustee has met the standard of care,” George G. Bogert, The Law of Trusts & Trustees § 541 (3d ed. 2019). He also insists that the “appropriate inquiry will be context specific.” Hughes, 595 U.S. at 177 (quoting Fifth Third Bancorp v. Dudenhoeffer, 573 U.S. 409, 425 (2014)). That is true as far as it goes: As we have already explained, a plaintiff does not necessarily need to identify comparable funds or investments; he might, for example, make direct allegations of a breach of ERISA’s duty of prudence. We do not hold that a plaintiff must always identify a comparator when relying on circumstantial allegations of a breach of the duty of prudence. But to the extent a plaintiff asks a court to infer that a fiduciary used improper methods based on the performance of the investments, as Anderson in part does here, he must compare that performance to funds or investments that are meaningfully similar. Meiners, 898 F.3d at 822. The same reasoning holds for Anderson’s allegations that investors in the Intel-created plans incurred higher fees. As with the performance allegations, the fact that different kinds of funds with distinct objectives and approaches carried different fees does not by itself demonstrate imprudence. Anderson’s comparison to off-the-shelf funds that did not seek to mitigate risk to the same degree as Intel’s funds is not enough to show that the Intel funds’ fees were excessive to the point of imprudence. Anderson argues that there are “no meaningful comparators for the fiduciaries’ decision” because Intel’s approach “was unusual, if not unparalleled.” That argument conflates the risk-mitigation objective of the Intel funds with

ANDERSON V. INTEL CORP. INV. POLICY COMM. 15

the allocation decisions made to implement that objective. Anderson’s complaint suggests that what he is really challenging is the former: He alleges that “in pursuing a purported risk-mitigation strategy, the Intel Funds gave up the long-term benefit of investing in equity, which delivers superior returns.” But as the district court noted, “ERISA fiduciaries are not required to adopt a riskier strategy simply because that strategy may increase returns.” To the contrary, courts have routinely rejected claims that an ERISA fiduciary can violate the duty of prudence by seeking to minimize risk. See Pizarro v. Home Depot, Inc., 111 F.4th 1165, 1181 (11th Cir. 2024) (“Home Depot offered the stable value fund because it was conservative, advertised it as conservative, and benchmarked it against a conservative metric. Because the fund met the expectations set for it, the plaintiffs’ breach-of-fiduciary-duty claim relying on comparisons to other, more aggressive benchmarks fail[s].”); Ellis v. Fidelity Mgmt. Tr. Co., 883 F.3d 1, 10 (1st Cir. 2018) (rejecting an argument “that a plan fiduciary’s choice of benchmark, where such a benchmark is fully disclosed to participants, can be imprudent by virtue of being too conservative”). Anderson insists that he is challenging the implementation of the risk-minimization strategy, as opposed to the strategy itself. In that respect, his argument appears to rest on the proposition that the fiduciaries’ allocation strategy was imprudent because hedge funds and private equity funds are inherently so risky that no prudent investor with the same aims would have invested in them, or at least not in the proportions the fiduciaries selected. As Anderson puts it, Intel should have been aware of “contemporaneous reports of poor hedge-fund returns, the exorbitant expenses of hedge funds and private equity, and [their] well-recognized risks.” Anderson’s per se challenge to hedge funds and private equity investments overlooks that “the prudence of each investment is not assessed in isolation but, rather, as the investment relates to the portfolio as a whole.” St. Vincent, 712 F.3d at 717; see also California Ironworkers Field Pension Tr. v. Loomis Sayles & Co., 259 F.3d 1036, 1043 (9th Cir. 2001). ERISA requires that a fiduciary “diversify[] the investments of the plan so as to minimize the risk of large losses.” 29 U.S.C. § 1104(a)(1)(C). And the Department of Labor’s regulations contemplate that a fiduciary should act as a prudent investment manager following the principles of modern portfolio theory, which recognizes that while the individual riskiness of a particular investment cannot be eliminated, it can be managed through the diversification of investment assets. See 29 C.F.R. § 2550.404a-1(b)(1)(i)– (ii); see also DiFelice v. U.S. Airways, Inc., 497 F.3d 410, 423 (4th Cir. 2007) (“[M]odern portfolio theory has been adopted by the investment community and, for the purposes of ERISA, by the Department of Labor.” (citing 29 C.F.R. § 2550.404a-1)); Laborers Nat’l Pension Fund. v. Northern Trust Quantitative Advisors, Inc., 173 F.3d 313, 322 (5th Cir. 1999) (“Since 1979, investment managers have been held to the standard of prudence of the modern portfolio theory by the Secretary’s regulations.” (citing 29 C.F.R. § 2550.404a-1)). Indeed, in some cases, “an investment in a risky security as part of a diversified portfolio is . . . an appropriate means to increase return while minimizing risk.” DiFelice, 497 F.3d at 423. Thus, generalized attacks on hedge funds and private equity funds as a category have been rejected both by courts, see, e.g., St. Vincent, 712 F.3d at 723, and by the Department of Labor, which has opined that

ANDERSON V. INTEL CORP. INV. POLICY COMM. 17

“a fiduciary may properly select an asset allocation fund with a private equity component as a designated investment alternative for a participant directed individual account plan,” Letter to Jon W. Breyfogle from Louis Campagna, Chief, Division of Fiduciary Interpretations, Office o f R e g u l a t i o ns an d Int e rp r et at i o ns , E m pl o y ee Benefits S e c ur i t y A dm i ni s t ra t i o n , Un i t ed S t at es Department of Labor ( June 3 , 2020) , available a t https://www.dol.gov/agencies/ebsa/about-ebsa/our- activities/resource-center/information-letters/06-03-2020. It is possible that a plaintiff could make out an imprudence claim by alleging that a plan invested much more in a particularly risky class of assets than did other, comparable plans, even if investing in that asset class is not per se imprudent in smaller amounts. Cf. California Ironworkers, 259 F.3d at 1045 (holding it sufficient to allege that a fiduciary allocated nearly one third of a plan to a “highly risky investment[]” and the same fiduciary allocated only five percent and seven percent of its other plans to that same investment). But Anderson has not plausibly alleged that Intel’s specific investments were imprudent at the scale it made them. Although Intel identified the hedge funds and private equity funds in which it invested, Anderson has not alleged that those investments were particularly risky, individually or in the aggregate. Notably, the complaint suggests that the fiduciaries’ choices had their intended effects. For example, one chart in the complaint shows that hedge funds (albeit a composite index rather than the specific funds the Intel fiduciaries selected) underperformed the global stock market in “up” months, but overperformed in “down” months—precisely the tradeoff Intel had disclosed. Nor does Anderson’s “risk-adjusted” analysis suffice. He alleges that the Intel funds had a greater “risk per unit of return” than did other target-date funds. But an ERISA plaintiff cannot make incomparable funds comparable simply by using a ratio. The “risk-adjusted” analysis does not allege that any funds with comparable risk profiles and greater returns actually existed; it only speculates that if a fund with a comparable risk profile had followed the trend of other, presumably riskier, funds, it would have generated higher returns than the Intel funds did. Finally, Anderson emphasizes the liberal pleading standards of Federal Rule of Civil Procedure 8, arguing that the district court impermissibly “parsed” his chosen comparators, and improperly engaged in factfinding. But Anderson’s complaint explained that “there are considerable differences among [target-date funds] offered by different providers, even among [target-date funds] with the same target date,” so it was appropriate for the district court to consider those differences carefully. Furthermore, the district court had to assess the similarities and differences between Anderson’s chosen comparators and the Intel funds so that it could determine whether they were appropriate comparators in the first place. See Davis, 960 F.3d at 485 (“Comparing apples and oranges is not a way to show that one is better or worse than the other.”). Such analysis—even at the pleading stage—is appropriate in ERISA cases. To be sure, plaintiffs “typically lack extensive information regarding the fiduciary’s ‘methods and actual knowledge’ because those details ‘tend to be in the sole possession of [that fiduciary].’” Meiners, 898 F.3d at 822 (alteration in original) (quoting St. Vincent, 712 F.3d at 719). And a court cannot reasonably demand that plaintiffs plead “facts which tend systemically to be in the

ANDERSON V. INTEL CORP. INV. POLICY COMM. 19

sole possession of defendants.” Braden, 588 F.3d at 598. But ERISA requires plan administrators to make extensive disclosures to participants, including a summary plan description and an annual report with audited financial statements. 29 U.S.C. §§ 1022, 1023. Those disclosures give prospective plaintiffs “the opportunity to find out how the fiduciary invested the plan’s assets.” St. Vincent, 712 F.3d at 720. An ERISA plaintiff can “use the data about the selected funds and some circumstantial allegations about methods to show that ‘a prudent fiduciary in like circumstances would have acted differently.’” Meiners, 898 F.3d at 822 (quoting St. Vincent, 712 F.3d at 720); see also 29 U.S.C. § 1104(a)(1)(B). Anderson has not made such a showing. He has had access to detailed information about the Intel funds— including the identities of the hedge funds and private equity funds in which they invested—and therefore has been well positioned to find appropriate comparators or to explain why these specific investments are so inherently risky, individually or in the aggregate, that selecting them was imprudent. Nevertheless, he makes only general arguments about the riskiness and costliness of hedge funds and private equity funds without providing factual allegations sufficient to support the claim that the investments that were actually made were ill-suited to the Intel funds. The district court therefore correctly held that he failed to state a claim for breach of ERISA’s duty of prudence.

III

Anderson also claims that Intel’s fiduciaries breached their duty of loyalty. They did so, he says, by giving hedge funds and private equity funds more capital to invest in companies and startups in which Intel Capital had already invested, so as to benefit Intel Capital by reducing the risk of its investments. The district court dismissed that claim, explaining that an ERISA plaintiff asserting a breach of the duty of loyalty must “plausibly allege a real conflict of interest, rather than the mere potential for a conflict of interest.” We agree. ERISA imposes a duty of loyalty on plan fiduciaries: A fiduciary must administer plan assets “solely in the interest of the participants and beneficiaries.” 29 U.S.C. § 1104(a)(1); see also id. § 1106(b)(1) (“A fiduciary with respect to a plan shall not deal with the assets of the plan in his own interest or for his own account.”); Pegram v. Herdrich, 530 U.S. 211, 225 (2011). Like the duty of prudence, ERISA’s duty of loyalty finds its source in the common law of trusts. See Central States, 472 U.S. at 570. The Supreme Court has explained, however, that “the analogy between ERISA fiduciary and common law trustee” is imperfect because unlike a trustee at common law, “the trustee under ERISA may wear different hats,” and “a fiduciary may have financial interests adverse to beneficiaries.” Pegram, 530 U.S. at 225. For example, employers “can be ERISA fiduciaries and still take actions to the disadvantage of employee beneficiaries, when acting as employers (e.g., firing an employee for reasons unrelated to the ERISA plan).” Id. The statute requires that fiduciaries “wear the fiduciary hat when making fiduciary decisions.” Pegram, 530 U.S. at 225; see Hughes Aircraft Co. v. Jacobson, 525 U.S. 432, 443–44 (1999). But it does not prohibit “the mere act of becoming a trustee with conflicting interests.” Friend v. Sanwa Bank California, 35 F.3d 466, 469 (9th Cir. 1994); see id. (“ERISA does not expressly prohibit a trustee from having dual loyalties.”). Thus, “the potential for a conflict,

ANDERSON V. INTEL CORP. INV. POLICY COMM. 21

without more, is not synonymous with a plausible claim of fiduciary disloyalty.” Kopp v. Klein, 894 F.3d 214, 222 (5th Cir. 2018); accord Donovan v. Bierwirth, 680 F.2d 263, 271 (2d Cir. 1982) (explaining that corporate officers who also serve as trustees of the company’s retirement plans “do not violate their duties as trustees by taking action which . . . they reasonably conclude best to promote the interests of participants and beneficiaries simply because it incidentally benefits the corporation”). Anderson insists that the Intel fiduciaries’ investment in certain hedge funds and private equity funds “had the potential to benefit” Intel Capital “by allowing Intel Capital to invest in technology startups more effectively and with reduced risk.” But as the district court observed, nowhere in the complaint did Anderson allege that the Intel fiduciaries had any influence over any investment firm’s decision “to invest in one of the startups in which Intel [had already] invested.” And the mere fact that members of senior management at Intel Capital also served as members of Intel’s Investment Policy Committee does not, on its own, support an inference that such individuals acted disloyally while discharging their fiduciary duties. All Anderson presented was the potential for conflicts of interest, with nothing more. The district court was correct to hold that Anderson did not adequately plead a claim of breach of the duty of loyalty. AFFIRMED. BERZON, Circuit Judge, concurring:

I concur in full in the majority opinion. I write separately to clarify the role of comparisons and circumstantial allegations in duty-of-prudence claims. Comparison is not a pleading requirement for a breach of fiduciary claim. ERISA’s fiduciary provisions do define the legal standard of conduct using comparisons. But the statute does not require pleading an empirical comparator— in the form of a “meaningful benchmark” alternative investment or otherwise—to state a claim. There is a crucial difference between (a) comparisons that define the standard of conduct with (b) comparisons that can, but need not, be pleaded to show that the standard has been violated. I address these two different uses of comparisons in ERISA imprudence claims in turn. A. ERISA requires that a fiduciary “discharge” her duties “with the care, skill, prudence, and diligence under the circumstances then prevailing that a prudent man acting in a like capacity and familiar with such matters would use in the conduct of an enterprise of a like character and with like aims.” 29 U.S.C. § 1104(a)(1). Duty-of-prudence claims thus require, by definition, a legal comparison between the defendant fiduciary and the hypothetical “prudent man.” Crucially, this invited comparison is not a factual requirement, and so does not require pleading any facts about the “prudent man.” Instead, the comparison is a way of defining the applicable legal standard. The “prudent man” is an imaginary archetype, like the “reasonable person” in negligence law or the “bad man” imagined by Justice Holmes. See Oliver Wendell Holmes, The Path of the Law,

ANDERSON V. INTEL CORP. INV. POLICY COMM. 23

10 Harv. L. Rev. 457 (1897). 1 As the “prudent man” is not real, a plaintiff cannot plead facts that empirically demonstrate how a “prudent man” would have acted. Instead, the “prudent man” personifies the ideal of prudence and emphasizes that perfection is not required; only what is humanly attainable is expected. A plaintiff need only provide evidence from which a factfinder can determine that the investment process did not meet this standard. The upshot is that although ERISA’s standard of prudent conduct is defined by comparison, a plaintiff need not plead facts about a “prudent man”—or his investment decisions— to establish a comparator and so show that the comparative legal standard has been violated. 2 B. Even though comparative allegations are not required to state an ERISA imprudence claim, they can be useful— indeed, they are often the best way for a plaintiff to plead such a claim at the outset of a case. The reason is simple:

1 In this essay, Holmes distinguishes between morality and law by arguing that even a hypothetical “bad man” who “cares nothing for an ethical rule” will nevertheless want to “avoid being made to pay money, and will want to keep out of jail if he can.” 10 Harv. L. Rev. at 459. 2 Labor Department regulations specify that a fiduciary satisfies her duty of prudence if she has “determine[ed] . . . that [a] particular investment or investment course of action is reasonably designed . . . to further the purposes of the plan, taking into consideration the risk of loss and the opportunity for gain . . . associated with the investment or investment course of action compared to the opportunity for gain . . . associated with reasonably available alternatives with similar risks.” 29 CFR § 2550.404a-1(b) (emphasis added). This regulation requires the fiduciary to make a comparison and to evaluate the relative costs and benefits of different investments. It does not set forth a pleading requirement for plaintiffs alleging that the fiduciary breached her duty. ERISA’s duty of prudence is a standard of conduct rather than results. But plaintiffs generally will know only the outcome of a fiduciary’s decisions—which investments were selected, for example, and how those investments performed. They typically will not know details about the process by which these decisions were made—which other options were considered, or how and why certain investments were selected over alternatives. “ERISA plaintiffs generally lack the inside information necessary to make out their claims in detail unless and until discovery commences.” Braden v. Wal-Mart Stores, Inc., 588 F.3d 585, 598 (8th Cir. 2009). As a result, a plaintiff is not “required to describe directly the ways in which [defendants] breached their fiduciary duties,” Braden, 588 F.3d at 595; “a claim . . . may still survive a motion to dismiss if the court, based on circumstantial factual allegations, may reasonably ‘infer from what is alleged that the process was flawed,’” Pension Ben. Guar. Corp. ex rel. St. Vincent Catholic Med. Ctrs. Ret. Plan v. Morgan Stanley Inv. Mgmt. Inc., 712 F.3d 705, 718 (2d Cir. 2013) (alterations omitted) (quoting Braden, 588 F.3d at 596). Consequentially, plaintiffs often plead a claim using allegations that do not directly describe the fiduciary’s decision-making process but support the inference that the methods used were unwise. 3

3 The permission to state a claim with indirect allegations that support an inference of liability is not some special carveout for ERISA imprudence claims. It is an application of Federal Rule of Civil Procedure 8, under which a complaint need only include “factual content that allows the court to draw the reasonable inference that the defendant is liable for the misconduct alleged.” Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009).

ANDERSON V. INTEL CORP. INV. POLICY COMM. 25

But of course, as with any other kind of claim, indirect allegations are not the only way to state a claim. The straightforward approach is to plead facts that directly show that the fiduciary’s methods, processes, or objectives were imprudent. For example, if a plaintiff learned that a plan manager chose investments by writing the ticker symbol for each publicly traded U.S. company on a bingo ball and then drawing ten to invest in at random, the plaintiff could almost certainly plead a duty-of-prudence claim attacking this process simply by recounting these facts. A court could determine as a matter of law that no prudent investor would select investments entirely at random. Also, unlike some other rules and statutes, ERISA does not impose a heightened pleading standard for imprudence or any other claims. Cf. Fed. R. Civ. P. 9(b) (particularity standard for allegations of fraud); 15 U.S.C. § 78u-4(b)(2) (particularity standard for state-of-mind allegations supporting private class action securities claims). So, as with any other claim not required to be pleaded with special particularity, the question is simply whether the plaintiff has adequately pleaded facts, direct or circumstantial, showing that he is entitled to relief. C. What sort of indirect facts are sufficient to support an inference that a fiduciary breached the duty of prudence? There are a “myriad of circumstances that could violate the [prudent man] standard,” In re Syncor ERISA Litig., 516 F.3d 1095, 1102 (9th Cir. 2008), so there is no fixed formula for the facts from which we might infer imprudence, nor is there a specific requirement to plead a particular kind of indirect allegation to support such an inference. As the majority notes, though, bare allegations that “costs are too high, or returns are too low” are not enough to suggest that the investment process was flawed. See Maj. Op. at 12 (quoting Matousek v. MidAmerican Energy Co., 51 F.4th 274, 278 (8th Cir. 2022)). With these principles in mind, I address several kinds of indirect allegations that can support an inference of imprudence, although the list is of course not exhaustive. I first note that, although many cases in which plaintiffs have pleaded imprudence with indirect facts involve comparative allegations, comparisons are not the only form of indirect allegation that could support a claim. For example, imagine a plaintiff who had no idea how a fiduciary selected investments but knew that the fiduciary had allocated a significant portion of the plan’s assets to a new type of security backed entirely by lottery tickets. The inherent risk of that category of investment might be sufficient, even without any details about how the fiduciary selected it or any comparison to other investments or other plans, to support a claim of imprudence—and that would be so even if, against all odds, the plan purchased a winning ticket. A common way of pleading imprudence with indirect allegations is to provide comparisons that support an inference that a fiduciary’s methods were imprudent. One category of comparison is the “meaningful benchmark” comparison, championed by the district court and the majority opinion. This kind of comparison is one between individual investments that were actually available to the fiduciary. The “meaningful benchmark” language originated in Meiners v. Wells Fargo & Co., 898 F.3d 820, 822 (8th Cir. 2018). Meiners, in turn, coined the phrase in reference to the Eighth Circuit’s earlier decision in Braden v. Wal-Mart Stores. Id.

ANDERSON V. INTEL CORP. INV. POLICY COMM. 27

Braden involved an ERISA-covered employee retirement plan that allowed individual participants to direct how their assets were invested by selecting from a menu of investment options selected by the plan’s fiduciary. 588 F.3d at 589. The plaintiff alleged that the plan was large enough that it had the ability to offer as choices on this menu of investment options either retail-class or institutional shares of the same mutual funds. 588 F.3d at 590. Retail shares “charge[d] significantly higher fees than institutional shares for the same return on investment,” and the complaint included “specific allegations about the relative cost of institutional and retail shares in the funds.” Id. at 595 & n.5. Based on the allegation that the plan’s managers chose to make available the higher-cost version of an otherwise identical investment, the Eight Circuit concluded that it was reasonable to infer that the process by which the plan was managed was flawed. Id. at 596; see also Meiners, 898 F.3d at 822. Notably, in Braden, the allegation was not just that “cheaper alternative investments exist in the marketplace.” 588 F.3d at 596 n.7. Braden emphasized that such allegations would be insufficient. Id. Instead, Braden alleged that the plan managers had the option to choose between two different classes of shares in the same mutual funds, with the only difference being that one class of shares had higher fees than the other. Id. at 595–96. An investor need not peer into a crystal ball to discern, even at the outset, that selecting the more expensive of the two share classes will lead to lower returns. Because the only difference between the available investments was their varying costs, the allegations in Braden were sufficient to suggest that opting for the more expensive option was imprudent at the time the decision was made, not just in retrospect. Id. at 596. These kinds of “benchmark” comparisons to individual real-world investment options are useful in “an investment- by-investment challenge”—a theory of breach based on a fiduciary’s failure to “remove imprudent investment options” when there exist better specific alternatives. Davis v. Wash. Univ. in St. Louis, 960 F.3d 478, 484 (8th Cir. 2020). But investment-versus-investment benchmarks are just one way of providing comparative allegations that could show imprudence. A plaintiff might also support an inference of imprudence by providing a plan-level comparison rather than an individual investment-level comparison. For example, in California Ironworkers Field Pension Trust v. Loomis, Sayles & Co., plaintiffs asserted that an investment manager breached ERISA’s duty of prudence in managing an ERISA-covered employee benefit plan that had adopted “conservative investment guidelines,” seeking to “achieve decent returns with minimum market risk.” No. CV964036, 1999 WL 1457226 at *3, *6. (C.D. Cal. Mar. 26, 1999). The manager invested nearly a third of the plan’s assets in a form of mortgage-backed security called an “inverse floater.” Id. On appeal, we affirmed the district court’s conclusion that the manager breached the duty of prudence by investing so high a proportion of the plan’s assets in this single form of security, which we noted “could be highly risky.” California Ironworkers Field Pension Tr. v. Loomis Sayles & Co., 259 F.3d 1036, 1045 (9th Cir. 2001). In concluding that it was imprudent to allocate nearly a third of the plan’s assets to this one, risky kind of asset, we emphasized (as had the district court) that two other employee benefit plans managed by the same fiduciary had allocated much smaller percentages—less than 10%—to the same risky inverse floaters. 259 F.3d 1036, 1045; 1999 WL

ANDERSON V. INTEL CORP. INV. POLICY COMM. 29

1457226 at *3. In other words, we inferred imprudence based in part on a plan-versus-plan comparison rather than an investment-versus-investment comparison. In California Ironworkers, we compared plans managed by the same fiduciary and evaluated their varying allocations to risky assets. But there is no reason this logic could not extend to plans managed by different fiduciaries as well. For example, if a plaintiff showed that a fiduciary allocated a third of a plan to one kind of risky asset and asserted that several other plans with comparable aims but different managers each allocated much smaller percentages to that same asset, those allegations might similarly support an inference of imprudence. And the inference might be stronger still if the plaintiff analyzed the entire market and alleged that no comparable plan adopted a similar allocation—a plan-versus-aggregate comparison rather than a plan-versus-plan comparison. Either way, though, such a comparison operates somewhat differently than an investment benchmark comparison. A benchmark investment comparison between two otherwise-identical investments that differ on only one characteristic, like fee amount, can suggest that the fiduciary who selected the worse of the two options acted imprudently. A plan-by-plan or aggregate comparison, by contrast, can suggest imprudence by demonstrating that the fiduciary’s actions were an outlier. Deviation alone may not be enough to suggest imprudence, but coupled with some reason why the fiduciary should have known at the time the decision was made that the aberrant allocation or investment decision would be imprudent, divergence could suggest that the fiduciary’s conduct fell short of the prudent person standard. Thus, in California Ironworkers, we looked not only to the fact that the one plan’s allocation to risky inverse floaters far exceeded two similar plans’ allocations, but also to the non- comparative facts “that inverse floaters could be highly risky investments” and “that the [plan] had very conservative investment guidelines.” 259 F.3d at 1045. The foregoing discussion is not exhaustive. My point, instead, is that any set of allegations which, taken as true and viewed in the plaintiff’s favor, plausibly support an inference that a fiduciary acted imprudently is sufficient at the pleading stage. The ultimate question, absent direct allegations about the fiduciary’s investment methods, is not how other plans were managed or what other investments were available, but whether the facts alleged—comparative or not—lead to the plausible inference that the actual process used by the defendant fiduciary was flawed. * * * With appropriate evidence, Anderson could state a claim by pleading a true benchmark comparison, by providing other circumstantial allegations that plausibly suggested imprudence, or by directly showing that the specific investments the Intel fiduciaries selected or the general methodologies they used were imprudent. But I agree with the majority opinion’s conclusion that Anderson has failed to plead facts that support his claim either directly or inferentially, and so concur in full in the majority opinion.