The DEPOT, INC.; Union Club Bar, Inc.; Trail Head, Inc., Plaintiffs-Appellants, v. CARING FOR MONTANANS, INC., fka Blue Cross Blue Shield of Montana, Inc.; Health Care Services Corporation, Defendants-Appellees.
Plaintiffs are three small employers in Montana who are members of the Montana Chamber of Commerce. Defendants are health insurance companies that marketed fully insured health insurance plans to the Chamber’s members branded “Chamber Choices.” From 2006 until 2014, plaintiffs provided their employees with healthcare coverage under Chamber Choices plans, and did so based on defendants’ representations that the monthly premiums would reflect only the cost of providing benefits. But according to plaintiffs, these representations were false—defendants padded the premiums with hidden surcharges, which they used to pay kickbacks to the Chamber and to buy unauthorized insurance products.
Upon learning of these surcharges, plaintiffs filed suit against defendants, asserting two claims under the Employee Retirement Income Security Act of 1974 (“ERISA”), 29 U.S.C. § 1001 et seq., as well as several state-law claims based on defendants’ misrepresentations. The district court dismissed all of the claims, concluding that plaintiffs failed to state actionable claims under ERISA while at the same time concluding that plaintiffs’ state-law claims are preempted by ERISA. We affirm the district court’s dismissal of plaintiffs’ ERISA claims, reverse the dismissal of plaintiffs’ state-law claims, and remand.
A. Factual Background
Plaintiffs are three small businesses operating in Montana.1 The Depot, Inc. is a steakhouse; Union Club Bar, Inc. is a bar; and Trail Head, Inc. is a sporting goods retailer. During the period relevant to this lawsuit, plaintiffs were members of the Montana Chamber of Commerce. Blue Cross Blue Shield of Montana (“BCBSMT”)—an insurance company that is now known as Caring for Montanans, Inc. (“CFM”)—marketed “fully-insured” group health insurance plans to the Chamber’s employer-members known as “Chamber Choices.” Health Care Service Corp. (“HCSC”) purchased the health insurance business of BCBSMT in July 2013 and marketed the Chamber Choices plans thereafter.
From 2006 to 2014, plaintiffs enrolled in Chamber Choices plans and paid monthly premiums to defendants in exchange for health insurance coverage for their employees. Coverage for plaintiffs’ employees hinged on plaintiffs paying the required monthly premiums. According to plaintiffs, “[i]n the course of marketing Chamber Choices,” defendants represented that the premiums would be equal to the “actual medical premium”—i.e., “the cost of providing insurance benefits to covered individuals plus administrative costs” and “[not] for any purpose other than to pay for the purchased health insurance coverage.” Plaintiffs accordingly relied on that representation in choosing to participate.
All parties agree that each Chamber Choices plan constituted an “employee welfare benefit plan” subject to ERISA. 29 U.S.C. § 1002(1); see Fossen v. Blue Cross & Blue Shield of Mont., Inc., 660 F.3d 1102, 1109–10 (9th Cir. 2011). According to the Member Guide for one of the Chamber Choices plans 2 —which provides a summary of benefits available to covered employees for the relevant year—the employers (i.e., plaintiffs), not BCBSMT, were the named “plan administrator[s]” and fiduciaries under ERISA. Defendants, however, performed most of the claim management and administration duties. Plaintiffs’ role was limited to deducting monthly premiums from their employees’ wages to send to defendants for coverage and notifying defendants if an employee lost eligibility for coverage. The Member Guide also purported to allow defendants to make changes to the terms of the policy in the following modification provision:
[BCBSMT] may make administrative changes or changes in dues, terms or Benefits in the Group Plan by giving written notice to the Group and/or purchasing pool member at least 60 days in advance of the effective date of the changes. Dues may not be increased more than once during a 12-month period, except as allowed by Montana law.
The requirement that enrollees receive 60 days’ advance notice of modifications is consistent with federal and state laws governing group health plans, including plans not subject to ERISA. See 29 C.F.R. § 2590.715-2715(b) (requiring 60 days’ advance notice of “any material modification ․ in any of the terms of the plan or coverage”); Mont. Code Ann. § 33-22-107(3)(a) (requiring 60 days’ advance notice of “a change in rates or a change in terms or benefits”).
Plaintiffs allege that, while they subscribed to Chamber Choices plans, defendants unlawfully padded the premiums with two surcharges without plaintiffs’ knowledge or consent. First, from 2006 to 2014, defendants secretly embedded a surcharge into the premiums, which they used to pay kickbacks to the Chamber. These kickbacks were designed to persuade the Chamber to continue to market defendants’ plans to its members. Second, from 2008 to 2014, defendants secretly embedded an additional surcharge into the premiums that defendants used to purchase “additional insurance products that [plaintiffs] did not request or authorize.” Plaintiffs further allege that defendants took efforts to conceal these surcharges. Beginning in 2009, defendants began “channeling the kickbacks to the Chamber through an insurance agent and channeling a share of the [surcharges] into a ‘rate stabilization’ account.” And beginning in 2012, defendants began “making cryptic notations that itemized certain charges on the bills” in an effort to “reduce [their] own legal risk.”
In February 2014, the Montana Commissioner of Securities and Insurance fined BCBSMT $250,000 for illegal insurance practices under Montana law, including billing in excess of the actual medical premium and paying kickbacks to the Chamber. See Mont. Code Ann. §§ 33-18-208, 33-18-212. After the Commissioner’s findings were publicly released in March 2014, a group of Chamber Choices participants filed a class action suit against defendants in state court alleging claims of breach of fiduciary duty, breach of contract, unfair and deceptive trade practices, and unjust enrichment. Mark Ibsen, Inc. v. Caring for Montanans, Inc., 383 Mont. 346, 371 P.3d 446, 448 (2016). After defendants unsuccessfully tried to remove the case to federal court, the state trial court dismissed the lawsuit, finding that the plaintiffs’ claims were based on statutory violations and that the relevant state statute did not provide a private right of action. Id. at 448–49. The Montana Supreme Court affirmed. Id. at 455.
B. Procedural History
Plaintiffs filed this lawsuit in federal court in June 2016. In their original complaint, plaintiffs raised two ERISA claims: a breach of fiduciary duty claim under 29 U.S.C. § 1132(a)(2), and a prohibited interested-party transaction claim under 29 U.S.C. § 1132(a)(3). Plaintiffs also raised state-law claims for breach of contract, breach of fiduciary duty, breach of the implied covenant of good faith and fair dealing, negligent misrepresentation, unjust enrichment, and unfair trade practices under the Montana Unfair Trade Practices and Consumer Protection Act, Mont. Code Ann. § 30-14-101 et seq.
The district court dismissed the original complaint without prejudice. The court concluded that defendants did not satisfy ERISA’s definition of a “fiduciary” for purposes of the breach of fiduciary duty claim, and that plaintiffs were not seeking “appropriate equitable relief” as required for the prohibited transaction claim. The court also concluded that plaintiffs’ state-law claims were preempted by ERISA because they constituted “alternative enforcement mechanisms” to the prohibited transaction claim.
Plaintiffs filed their amended complaint in March 2017. As before, plaintiffs assert two claims under ERISA: a breach of fiduciary duty claim under 29 U.S.C. § 1132(a)(2), and a prohibited transaction claim under 29 U.S.C. § 1132(a)(3). Plaintiffs also assert state-law claims for fraudulent inducement, constructive fraud, negligent misrepresentation, unjust enrichment, and unfair trade practices. The district court dismissed all of the claims, this time with prejudice. The court concluded that plaintiffs failed to remedy the defects in their ERISA claims and that plaintiffs’ state-law claims (including the new fraud claims) were still preempted by ERISA. The court also concluded that plaintiffs’ allegations of fraud do not satisfy the heightened pleading requirements under Federal Rule of Civil Procedure 9(b). Plaintiffs timely appealed, and we have jurisdiction pursuant to 28 U.S.C. § 1291.
II. STANDARD OF REVIEW
On appeal, plaintiffs challenge the district court’s dismissal of their ERISA claims as well as the dismissal of their state-law claims. We review de novo a district court’s order granting a motion to dismiss for failure to state a claim under Federal Rule of Civil Procedure 12(b)(6). Santomenno v. Transamerica Life Ins. Co., 883 F.3d 833, 836 (9th Cir. 2018). To survive a motion to dismiss, the 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, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009) (quoting Bell Atl. Corp. v. Twombly, 550 U.S. 544, 570, 127 S.Ct. 1955, 167 L.Ed.2d 929 (2007) ). We will thus “affirm a dismissal for failure to state a claim where there is no cognizable legal theory or an absence of sufficient facts alleged to support a cognizable legal theory.” Interpipe Contracting, Inc. v. Becerra, 898 F.3d 879, 886 (9th Cir. 2018) (quoting L.A. Lakers, Inc. v. Fed. Ins. Co., 869 F.3d 795, 800 (9th Cir. 2017) ). And although “we accept as true all factual allegations,” we do not “accept as true allegations that are conclusory.” In re NVIDIA Corp. Sec. Litig., 768 F.3d 1046, 1051 (9th Cir. 2014). Nor do we consider factual assertions made for the first time on appeal, as “our review is limited to the contents of the complaint.” Allen v. City of Beverly Hills, 911 F.2d 367, 372 (9th Cir. 1990); see Amgen Inc. v. Harris, ––– U.S. ––––, 136 S.Ct. 758, 760, 193 L.Ed.2d 696 (2016) (per curiam).
We also review de novo a dismissal for failure to satisfy Federal Rule of Civil Procedure 9(b), which requires a party alleging fraud to “state with particularity the circumstances constituting fraud.” WPP Luxembourg Gamma Three Sarl v. Spot Runner, Inc., 655 F.3d 1039, 1047 (9th Cir. 2011) (quoting Fed. R. Civ. P. 9(b) ).
III. ERISA CLAIMS
We begin with plaintiffs’ claims under ERISA. “Congress enacted ERISA to ‘protect the interests of participants in employee benefit plans and their beneficiaries’ by setting out substantive regulatory requirements for employee benefit plans” and “an integrated system of procedures for enforcement.” Aetna Health Inc. v. Davila, 542 U.S. 200, 208, 124 S.Ct. 2488, 159 L.Ed.2d 312 (2004) (internal alterations omitted) (first quoting 29 U.S.C. § 1001(b); then quoting Mass. Mut. Life Ins. Co. v. Russell, 473 U.S. 134, 147, 105 S.Ct. 3085, 87 L.Ed.2d 96 (1985) ). Relevant here, ERISA’s enforcement scheme provides a cause of action against plan fiduciaries for breach of their fiduciary duties, 29 U.S.C. § 1132(a)(2), and a cause of action to remedy plan or ERISA violations—including prohibited interested-party transactions—with “appropriate equitable relief,” id. § 1132(a)(3). Plaintiffs bring claims against defendants under both provisions. We address each in turn.
A. Breach of Fiduciary Duty Claim
Plaintiffs claim that, by collecting and concealing the premium surcharges, defendants breached their fiduciary duties—including a duty “to act ‘solely in the interest of the participants and beneficiaries,’ ” Varity Corp. v. Howe, 516 U.S. 489, 506, 116 S.Ct. 1065, 134 L.Ed.2d 130 (1996) (quoting 29 U.S.C. § 1104(a)(1) )—and are thus liable under 29 U.S.C. § 1132(a)(2).3 But to breach a fiduciary duty, one must be a fiduciary. And here, defendants were not acting as fiduciaries when taking the action subject to plaintiffs’ complaint. We thus affirm the district court’s dismissal of plaintiffs’ breach of fiduciary duty claim.
There are two types of fiduciaries under ERISA. First, a party that is designated “in the plan instrument” as a fiduciary is a “named fiduciary.” 29 U.S.C. § 1102(a)(2). Second, ERISA provides the following definition of what is sometimes referred to as a “functional” fiduciary:
[A] person is a fiduciary with respect to a plan to the extent (i) he exercises any discretionary authority or discretionary control respecting management of such plan or exercises any authority or control respecting management or disposition of its assets, (ii) he renders investment advice for a fee or other compensation, direct or indirect, with respect to any moneys or other property of such plan, or has any authority or responsibility to do so, or (iii) he has any discretionary authority or discretionary responsibility in the administration of such plan.
Id. § 1002(21)(A); Santomenno, 883 F.3d at 837. Because a person 4 is a fiduciary under this provision only “to the extent” the person engages in the listed conduct, a person may be a fiduciary with respect to some actions but not others. Pegram v. Herdrich, 530 U.S. 211, 225–26, 120 S.Ct. 2143, 147 L.Ed.2d 164 (2000) (quoting 29 U.S.C. § 1002(21)(A) ); see Acosta v. Brain, 910 F.3d 502, 519 (9th Cir. 2018) (“[W]e must distinguish between a fiduciary ‘acting in connection with its fiduciary responsibilities’ with regard to the plan, as opposed to the same individual or entity ‘acting in its corporate capacity.’ Only the former triggers fiduciary status; the latter does not.” (internal citation omitted) ). The central question is “whether that person was acting as a fiduciary (that is, was performing a fiduciary function) when taking the action subject to complaint.” Pegram, 530 U.S. at 226, 120 S.Ct. 2143.
Plaintiffs claim that defendants were acting as fiduciaries when charging excessive premiums based on the functions described in subparagraph (i)—exercising discretion over plan management and exercising authority over plan assets.5 These provisions are distinct and therefore must be analyzed separately. See IT Corp. v. Gen. Am. Life Ins. Co., 107 F.3d 1415, 1421 (9th Cir. 1997).
1. Discretion over Plan Management
Plaintiffs first argue that, in secretly charging excessive premiums, defendants “exercise[d] ․ discretionary authority or discretionary control respecting [plan] management.” 29 U.S.C. § 1002(21)(A)(i). We disagree. Insurance companies do not normally exercise discretion over plan management when they negotiate at arm’s length to set rates or collect premiums. That is because these negotiations occur before the agreement is executed, at which point the insurer has no relationship to the plan and thus no discretion over its management.
We addressed a similar issue in Santomenno, holding that a service provider—i.e., a company that managed a self-funded ERISA plan as a third-party administrator—“is not an ERISA fiduciary when negotiating its compensation with a prospective customer.” 883 F.3d at 837. The service provider in that case performed several functions for retirement plans, including the selection of various potential investments, and its compensation was set as a fixed percentage of the assets managed. Id. at 835–36. We explained that “[a] service provider is plainly not involved in plan management when negotiating its prospective fees”; to the contrary, “at that stage ‘discretionary control over plan management lies with the trustee, who decides whether to agree to the service provider’s terms.’ ” Id. at 838 (internal alterations omitted) (quoting Santomenno ex rel. John Hancock Tr. v. John Hancock Life Ins. Co. (U.S.A.), 768 F.3d 284, 293 (3d Cir. 2014) (“John Hancock ”) ). In other words, “ ‘a service provider owes no fiduciary duty with respect to the negotiation of its fee compensation’ because ‘nothing prevent[s] the trustees from rejecting the provider’s product and selecting another service provider; the choice [i]s theirs.’ ” Id. (internal alterations omitted) (quoting John Hancock, 768 F.3d at 295). And after the contract is executed, the “service provider cannot be held liable for merely accepting previously bargained-for fixed compensation” because “the plan administrator act[s] as ‘a fiduciary only for purposes of administering the plan, not for purposes of negotiating or collecting its compensation.’ ” Id. at 840 (citations omitted).
The reasoning in Santomenno applies equally to rate-setting by insurance companies. Premium rates, like fixed compensation fees, are generally negotiated in “an arm’s length bargain presumably governed by competition in the marketplace,” Schulist v. Blue Cross of Iowa, 717 F.2d 1127, 1132 (7th Cir. 1983), which means that the insurance company is “free to negotiate its [rates] with an eye to its profits,” Srein v. Soft Drink Workers Union, Local 812, 93 F.3d 1088, 1096 (2d Cir. 1996). Because the potential purchaser of the insurance policy remains free to “reject[ ] the [insurer’s] product and select[ ] another,” the insurance company “is plainly not involved in plan management when negotiating” premium rates. Santomenno, 883 F.3d at 838 (quoting John Hancock, 768 F.3d at 295); see also Cotton v. Mass. Mut. Life Ins. Co., 402 F.3d 1267, 1278–79 (11th Cir. 2005) (“Simply urging the purchase of its products does not make an insurance company an ERISA fiduciary with respect to those products.” (citation omitted) ).
That reasoning forecloses plaintiffs’ claim here. Plaintiffs concede that the terms of the insurance agreements in this case—including the premium amounts—were negotiated at arm’s length. The agreements were thus “ ‘presumably governed by competition in the marketplace’ that specified the premium rates.” Seaway Food Town, Inc. v. Med. Mut. of Ohio, 347 F.3d 610, 618 (6th Cir. 2003) (quoting Schulist, 717 F.2d at 1132). And the alleged misconduct in this case—misrepresenting that the monthly premiums reflected only the actual medical premium—occurred when the policies were being marketed. Indeed, plaintiffs allege that defendants made these misrepresentations in an effort to “induce [p]laintiffs to buy coverage through Chamber Choices.” Although plaintiffs may not have known about the surcharges when deciding to subscribe to Chamber Choices plans, the premium amounts were fully disclosed, and plaintiffs always remained free to walk away and select another insurance company.6 Defendants exercised no discretionary control over the plan’s management at that point.
Although plaintiffs agree that insurance companies do not ordinarily act as fiduciaries when negotiating premium rates, they claim that this is not an ordinary case. Rather, according to plaintiffs, defendants possessed an ability to exercise discretion over the plan by virtue of the modification provision in the Member Guide:
[BCBSMT] may make administrative changes or changes in dues, terms or Benefits in the Group Plan by giving written notice to the Group and/or purchasing pool member at least 60 days in advance of the effective date of the changes. Dues may not be increased more than once during a 12-month period, except as allowed by Montana law.
Plaintiffs argue that this provision granted defendants an unfair “ability to unilaterally amend plan terms” upon “mere notice to the beneficiaries” and therefore “subjected [them] to ERISA fiduciary duties.”
Generally, a “company does not act in a fiduciary capacity when deciding to amend ․ a welfare benefits plan.” Curtiss-Wright Corp. v. Schoonejongen, 514 U.S. 73, 78, 115 S.Ct. 1223, 131 L.Ed.2d 94 (1995) (citation omitted). Setting that aside, the mere existence of a discretionary ability is insufficient to bestow fiduciary status if that discretion was not “exercise[d].” 29 U.S.C. § 1002(21)(A)(i). Even assuming plaintiffs’ contention that the modification provision granted defendants the ability to unilaterally amend plan terms, plaintiffs do not adequately allege that defendants ever exercised that discretion, let alone “when taking the action subject to complaint.” Pegram, 530 U.S. at 226, 120 S.Ct. 2143. Indeed, plaintiffs concede that defendants “initially imposed the surcharges in an arms-length negotiation”—i.e., before the modification provision was effective.7 Thus, even if defendants may have become fiduciaries “at some point after entering into the contracts, [they] plainly held no such status prior to the execution of the contracts” when the premium amounts were negotiated. Santomenno, 883 F.3d at 839 (internal alterations omitted) (quoting F.H. Krear & Co. v. Nineteen Named Trs., 810 F.2d 1250, 1259 (2d Cir. 1987) ).
We made this point in Santomenno. There, the plaintiffs argued that the service provider “was a fiduciary when selecting the investment options because it retaine[d] the right to delete or substitute the funds the employer ha[d] selected for the Plan.” Id. at 839 n.5. We disagreed, observing that the plaintiffs failed to “allege that [the service provider] ever exercised its discretion.” Id. We also noted that the service provider could “only alter investment options upon six months’ notice,” and that the contract “allow[ed] the employer opportunity to terminate the contract if displeased with any change.” Id. Similarly, we rejected the plaintiffs’ reliance on a contractual provision that allowed the service provider “to change [its fees] upon advance written notice,” explaining that the “plaintiffs have not alleged that [the service provider] ever changed its fees; indeed, if it did, the employer is free under the [contract] to find another provider.” Id. at 839 n.4; see id. at 841 n.8 (similar). The bottom line is that a party is a fiduciary only to the extent the party actually exercises the alleged discretionary control or authority over plan management.
Seeking to overcome this conclusion, plaintiffs argue that defendants did exercise discretion after the contracts were executed by virtue of their “decision to continue charging inflated amounts, when [they] held complete and unilateral authority to eliminate those overcharges.” This argument—which condemns defendants for failing to exercise discretion under the very provision that plaintiffs denounce as unfair—is unpersuasive. In this context, a failure to exercise discretion does not amount to an exercise of discretion within the meaning of § 1002(21)(A)(i). See Santomenno, 883 F.3d at 839 n.5. Once defendants agreed to enter into a contract with plaintiffs, defendants may have acquired fiduciary status with respect to some plan functions, see, e.g., King v. Blue Cross & Blue Shield of Ill., 871 F.3d 730, 745–46 (9th Cir. 2017), but any fiduciary status defendants may have acquired did not compel defendants to renegotiate the premium rates they had just agreed to accept. Rather, defendants were “merely accepting previously bargained-for” premiums, and the bargaining itself did not give rise to fiduciary status. Santomenno, 883 F.3d at 840 (citation omitted).8 Thus, in allegedly charging and collecting excessive premiums—which is “the action subject to complaint,” Pegram, 530 U.S. at 226, 120 S.Ct. 2143—defendants were not exercising discretionary authority over plan management.
2. Control over Plan Assets
Plaintiffs separately argue that defendants acted as fiduciaries because they “exercise[d] ․ authority or control respecting management or disposition of [plan] assets.” 29 U.S.C. § 1002(21)(A)(i). Plaintiffs contend that the premiums they paid to defendants for insurance coverage were “plan assets” and that defendants were subject to fiduciary obligations when using them. But plaintiffs’ premise is flawed. Premiums paid to an insurance company in return for coverage under a fully insured insurance policy are not “plan assets.”
ERISA defines “plan assets” to mean “plan assets as defined by such regulations as the Secretary [of Labor] may prescribe.” Id. § 1002(42). Although the Secretary has not prescribed a comprehensive regulation defining “plan assets,” 9 several circuits have followed the “consistent[ ]” position of “[t]he Department of Labor ․ that ‘the assets of a plan generally are to be identified on the basis of ordinary notions of property rights under non-ERISA law.’ ” Gordon v. CIGNA Corp., 890 F.3d 463, 472 (4th Cir. 2018) (citing U.S. Dep’t of Labor, Advisory Op. No. 93-14A, 1993 WL 188473, at *4 (May 5, 1993) ); accord Merrimon v. Unum Life Ins. Co. of Am., 758 F.3d 46, 56 (1st Cir. 2014); Hi-Lex Controls, Inc. v. Blue Cross Blue Shield of Mich., 751 F.3d 740, 745 (6th Cir. 2014); Tussey v. ABB, Inc., 746 F.3d 327, 339 (8th Cir. 2014); Edmonson v. Lincoln Nat. Life Ins. Co., 725 F.3d 406, 427 (3d Cir. 2013); Faber v. Metro. Life Ins. Co., 648 F.3d 98, 105–06 (2d Cir. 2011); In re Luna, 406 F.3d 1192, 1199 (10th Cir. 2005). We agree with this weight of authority; absent applicable regulatory guidance, plan “assets” under 29 U.S.C. § 1002(A)(21)(i) are to be identified based on ordinary notions of property rights.10
Applying that principle here, we conclude that neither plaintiffs nor their employees had a property interest in the premium payments once they were paid to defendants. Plaintiffs paid the premiums to defendants in exchange for a contractual right to receive a particular service—healthcare coverage under a Chamber Choices plan. The premiums were monthly fees that defendants collected as revenue for providing that service. Upon paying the premiums, plaintiffs had no “beneficial ownership interest” in them. Hi-Lex Controls, 751 F.3d at 745 (quoting U.S. Dep’t of Labor, Advisory Op. No. 92-24A, 1992 WL 337539, at *2 (Nov. 6, 1992) ). Instead, defendants were simply indebted to plaintiffs to provide the agreed-upon coverage. Cf. Grupo Mexicano de Desarrollo, S.A. v. All. Bond Fund, Inc., 527 U.S. 308, 319–20, 119 S.Ct. 1961, 144 L.Ed.2d 319 (1999) (“[A] general creditor ․ ha[s] no cognizable interest, either at law or in equity, in the property of his debtor, and therefore [may] not interfere with the debtor’s use of that property.”).
Plaintiffs try to equate the premiums paid to defendants with “participant contributions” made into a self-funded plan, which are generally deemed to be plan assets. Acosta v. Pac. Enters., 950 F.2d 611, 620 & n.7 (9th Cir. 1991); see 29 C.F.R. § 2510.3-102(a)(1); Advisory Op. No. 92-24A, 1992 WL 337539, at *2. But plaintiffs elide the distinction between a self-funded plan and a fully insured plan. Under a “self-funded” plan, the insurance company “acts only as a third-party administrator; the employer is responsible for paying claims [out of the employees’ contributions] and bearing the financial risk.” N. Cypress Med. Ctr. Operating Co., Ltd. v. Aetna Life Ins. Co., 898 F.3d 461, 468 (5th Cir. 2018); see Gobeille v. Liberty Mut. Ins. Co., ––– U.S. ––––, 136 S.Ct. 936, 941–42, 194 L.Ed.2d 20 (2016). Premiums paid under a self-funded plan are therefore contributions from employees earmarked and held in trust by the employer for the employees’ later benefit. See Gordon, 890 F.3d at 472; Hi-Lex Controls, 751 F.3d at 747. These employer-held contributions are therefore assets of the plan. See 29 C.F.R. § 2510.3-102(a)(1) (defining “assets of the plan” to include amounts “that a participant or beneficiary pays to an employer ․ [or] that a participant has withheld from his wages by an employer” (emphasis added) ); Advisory Op. No. 92-24A, 1992 WL 337539, at *2 (describing “participant contributions” in the context of “plans whose benefits are paid as needed solely from the general assets of the employer maintaining the plan”).
This case, however, does not involve a self-funded plan. Instead, as plaintiffs allege, all of the plans sold by defendants were “fully-insured health insurance policies.” Under a “fully insured” plan, the insurance company “acts as a direct insurer; it guarantees a fixed monthly premium for 12 months and bears the financial risk of paying claims.” N. Cypress Med. Ctr., 898 F.3d at 468. Premiums paid under a fully insured plan are not held in trust; rather, they are “fixed fee[s]” paid in exchange for the insurance company “assum[ing] the financial risk of providing the benefits promised.” Pegram, 530 U.S. at 218–19, 120 S.Ct. 2143. And as explained above, once defendants collected those fees, neither plaintiffs nor their employees maintained any sort of property interest in them. Accordingly, defendants did not exercise control over plan assets when charging or spending the allegedly excessive premiums.11
Because defendants were not exercising a fiduciary function when taking “the action subject to complaint,” Pegram, 530 U.S. at 226, 120 S.Ct. 2143, we affirm the district court’s dismissal of plaintiffs’ breach of fiduciary duty claim under 29 U.S.C. § 1132(a)(2).
B. Prohibited Transaction Claim
Next, we consider plaintiffs’ prohibited transaction claim, for which they purport to seek “appropriate equitable relief” under 29 U.S.C. § 1132(a)(3). We conclude that the relief plaintiffs seek is not equitable and accordingly affirm the district court’s dismissal of plaintiffs’ prohibited transaction claim.
Under ERISA’s prohibited transaction provisions, “[a] fiduciary with respect to a plan shall not cause the plan to engage in a transaction” with “a party in interest” for the “furnishing of ․ services” if “more than reasonable compensation is paid therefor.” 29 U.S.C. §§ 1106(a)(1)(C), 1108(b)(2).12 ERISA provides a cause of action for remedying prohibited transactions:
A civil action may be brought ․ by a participant, beneficiary, or fiduciary (A) to enjoin any act or practice which violates any provision of [ERISA Title I] or the terms of the plan, or (B) to obtain other appropriate equitable relief (i) to redress such violations or (ii) to enforce any provisions of [ERISA Title I] or the terms of the plan.
Id. § 1132(a)(3). Because § 1132(a)(3) “makes no mention at all of which parties may be proper defendants,” a party in interest—including a non-fiduciary third party—may be sued under this provision for its participation in a prohibited transaction. Harris Tr. & Sav. Bank v. Salomon Smith Barney, Inc., 530 U.S. 238, 246, 249–51, 120 S.Ct. 2180, 147 L.Ed.2d 187 (2000); see Landwehr v. DuPree, 72 F.3d 726, 734 (9th Cir. 1995). And even though the plan fiduciary is the one who “cause[d] the plan to engage in [the prohibited] transaction,” 29 U.S.C. § 1106(a)(1), the “culpable fiduciary” may still bring suit against “the arguably less culpable” party in interest because “the purpose of the action is to recover money or other property for the [plan beneficiaries],” Harris Tr., 530 U.S. at 252, 120 S.Ct. 2180 (quoting Restatement (Second) of Trusts § 294 cmt. c, at 70 (1957) ).13 Thus, a plan fiduciary may (1) seek an injunction or “other appropriate equitable relief” (2) against a “party in interest” (3) for participating in a transaction for services for which “more than reasonable compensation is paid.”
The parties and the district court all agree that the second and third components are satisfied in this case. Each defendant is a “party in interest”—which ERISA defines as “a person providing services to [a] plan,” 29 U.S.C. § 1002(14)(B)—because they provide underwriting and claim-adjudication services to the plans. And the alleged conduct in this case—imposing unreasonable charges for kickbacks and unrequested benefits—is arguably a prohibited transaction for “services” between plan fiduciaries (plaintiffs) and parties in interest (defendants) for which “more than reasonable compensation is paid.” Id. §§ 1106(a)(1)(C), 1108(b)(2). Because an injunction is not at issue here, the only dispute is whether plaintiffs are seeking “appropriate equitable relief” under § 1132(a)(3).
In a series of decisions, the Supreme Court has explained that the phrase “appropriate equitable relief” in § 1132(a)(3) “is limited to those categories of relief that were typically available in equity during the days of the divided bench (meaning, the period before 1938 when courts of law and equity were separate).” Montanile v. Bd. of Trs. of Nat’l Elevator Indus. Health Benefit Plan, ––– U.S. ––––, 136 S.Ct. 651, 657, 193 L.Ed.2d 556 (2016) (internal quotation marks omitted); see US Airways, Inc. v. McCutchen, 569 U.S. 88, 94–95, 133 S.Ct. 1537, 185 L.Ed.2d 654 (2013); Sereboff v. Mid Atl. Med. Servs., Inc., 547 U.S. 356, 361–62, 126 S.Ct. 1869, 164 L.Ed.2d 612 (2006); Great-West Life & Annuity Ins. Co. v. Knudson, 534 U.S. 204, 210, 122 S.Ct. 708, 151 L.Ed.2d 635 (2002); Mertens v. Hewitt Assocs., 508 U.S. 248, 256, 113 S.Ct. 2063, 124 L.Ed.2d 161 (1993). As a result, a plaintiff may not use § 1132(a)(3) to seek any “form of legal relief,” such as “money damages.” Great-West, 534 U.S. at 210, 122 S.Ct. 708 (internal alteration omitted) (quoting Mertens, 508 U.S. at 255, 113 S.Ct. 2063). To qualify as “equitable relief,” both “(1) the basis for the plaintiff’s claim and (2) the nature of the underlying remedies sought” must be equitable rather than legal. Montanile, 136 S.Ct. at 657 (internal alterations omitted) (quoting Sereboff, 547 U.S. at 363, 126 S.Ct. 1869).
Even if we assume that the basis for plaintiffs’ claim in this case is equitable,14 the nature of the underlying remedies sought by plaintiffs in their complaint is not equitable. Plaintiffs seek a judgment to obtain money from defendants, and “[a]lmost invariably suits seeking ․ to compel the defendant to pay a sum of money to the plaintiff are suits for ‘money damages’ ”—the “classic form of legal relief.” Great-West, 534 U.S. at 210, 122 S.Ct. 708 (citation and internal alteration omitted). Although plaintiffs attempt to characterize the relief they seek as equitable by labeling it “restitution” and “disgorgement,” we must “look to the ‘substance of the remedy sought rather than the label placed on that remedy.’ ” Mathews v. Chevron Corp., 362 F.3d 1172, 1185 (9th Cir. 2004) (citation and internal alteration omitted). We conclude that, notwithstanding the labels, plaintiffs’ requested relief is not equitable in nature.
The restitution plaintiffs seek is not equitable. The Supreme Court has drawn a “fine distinction between restitution at law and restitution in equity.” Great-West, 534 U.S. at 214, 122 S.Ct. 708. A plaintiff seeks “restitution at law” when the plaintiff cannot “assert title or right to possession of particular property” but instead seeks to “impose personal liability on the defendant” as a means of “recovering money to pay for some benefit the defendant ․ received from [the plaintiff].” Id. at 213–14, 122 S.Ct. 708 (quoting 1 Dan B. Dobbs, Law of Remedies § 4.2(1), at 571 (2d ed. 1993) (“Dobbs”) ). By contrast, a plaintiff seeks “restitution in equity, ordinarily in the form of a constructive trust or an equitable lien, where money or property identified as belonging in good conscience to the plaintiff [can] clearly be traced to particular funds or property in the defendant’s possession.” Id. at 213, 122 S.Ct. 708 (citing 1 Dobbs § 4.3(1), at 587–88); see 2 Dobbs § 6.1(3), at 12–13.
The Court has illustrated this distinction in several cases involving ERISA plans claiming entitlement to a settlement fund obtained by plan beneficiaries. In Great-West, the Court concluded that the plan was seeking legal restitution because the specifically identified fund was not in the defendants’ possession. 534 U.S. at 214, 122 S.Ct. 708. By contrast, in Sereboff, the Court concluded that the plan was seeking equitable restitution because the plan “sought ‘specifically identifiable’ funds that were ‘within the possession and control’ ” of the defendants—not recovery from the defendants’ “assets generally.” 547 U.S. at 362–63, 126 S.Ct. 1869 (citation omitted). And in Montanile, the Court faced the problem of a fund that, although specifically identified at one time, had been dissipated by the defendant on nontraceable items, leaving the plan to seek recovery “out of the defendant’s general assets.” 136 S.Ct. at 658. Observing the rule that equitable restitution must seek to recover “specifically identified funds that remain in the defendant’s possession or ․ traceable items that the defendant purchased with the funds (e.g., identifiable property like a car),” the Court held that seeking recovery out of “the defendant’s general assets” due to dissipation of the funds “on nontraceable items (like food or travel)” amounts to “a legal remedy, not an equitable one.” Id.
In this case, the district court concluded that the nature of the remedy sought by plaintiffs is not equitable, reasoning that, under Montanile, “a party cannot recover in equity unless the funds have been maintained in a segregated account.” Montanile, however, concerned dissipation and tracing, not segregation. The Court noted this distinction in dicta, observing that at least in some instances, if a defendant “commingl[es] a specifically identified fund ․ with a different fund,” the “commingling allow[s] the plaintiff to recover the amount of the lien from the entire pot of money.” Montanile, 136 S.Ct. at 661. And as plaintiffs point out, we have, in another context, permitted recovery out of commingled funds under the “lowest intermediate balance” doctrine, which “evolved from equitable principles of trusts”:
“Where a wrongdoer mingles another’s money with his own, from which commingled account withdrawals are from time to time made, there is a presumption of law that the sums first withdrawn were moneys of the tortfeasor.” ․ If the amount on deposit is depleted below the amount of the trust, however, the amount withdrawn is treated as lost, and subsequent deposits do not replenish the trust. Thus, the beneficiary is entitled to the lowest intermediate balance between the date of the commingling and the date of payment.
In re R & T Roofing Structures & Commercial Framing, Inc., 887 F.2d 981, 987 (9th Cir. 1989) (internal alterations omitted) (quoting Republic Supply Co. v. Richfield Oil Co., 79 F.2d 375, 378 (9th Cir. 1935) ); see also Schuyler v. Littlefield, 232 U.S. 707, 710, 34 S.Ct. 466, 58 L.Ed. 806 (1914) (“[W]here one has deposited trust funds in his individual bank account, and the mingled fund is at any time wholly depleted, the trust fund is thereby dissipated, and cannot be treated as reappearing in sums subsequently deposited to the credit of the same account.”). Thus, plaintiffs argue, when a specific fund is commingled with other funds in a general account, restitution is available out of the general account as long as the general account balance does not dip below the amount of the wrongfully held money.
We need not decide whether this proposition is correct because, even assuming it is, “the facts and allegations supporting that proposition [do not] appear in [plaintiffs’] complaint.” Amgen, 136 S.Ct. at 760. First, plaintiffs have not identified a “specific fund” to which they are entitled. As Montanile explains, “[e]quitable remedies ‘are, as a general rule, directed against some specific thing,’ ” not “a sum of money generally.” 136 S.Ct. at 658 (citation omitted). Plaintiffs, however, seek to recover not a specific thing but instead some unidentified portion of the many premium payments that exceeded “reasonable compensation.” The premium surcharges plaintiffs seek to recover “never existed as a distinct object or fund”; rather, they reflect “a specific amount of money encompassed within a particular fund”—the total premiums paid to defendants. Bilyeu v. Morgan Stanley Long Term Disability Plan, 683 F.3d 1083, 1093 (9th Cir. 2012). And even if the amount of the overcharges is measurable or otherwise identifiable, “[i]t is the fund, not its size, that must be identifiable.” Cent. States, Se. & Sw. Areas Health & Welfare Fund v. First Agency, Inc., 756 F.3d 954, 960 (6th Cir. 2014); see Bilyeu, 683 F.3d at 1093 (distinguishing between a specific “amount of money” and a specific “fund”). Indeed, a judgment in plaintiffs’ favor would have no connection to any particular fund whatsoever. Defendants would simply be required to pay a certain amount of money, and they could “satisfy that obligation by dipping into any pot” they like. First Agency, 756 F.3d at 960. That is restitution at law, not equity.
Second, the complaint never mentions the existence of a general account in which the ill-gotten funds (i.e., the premium surcharges) were commingled, such that the product of those funds would be traceable. Again, Montanile is clear: “[W]here a person wrongfully disposed of the property of another but the property cannot be traced into any product, the other cannot enforce a constructive trust or lien upon any part of the wrongdoer’s property.” 136 S.Ct. at 659 (emphasis and internal alterations omitted) (quoting Restatement of Restitution § 215(1), at 866 (1936) ); see also George G. Bogert et al., The Law of Trusts and Trustees § 921 (2d rev. ed. 1995) (“Bogert”) (explaining that restitution of “trust property or its product” is generally unavailable “where the proof of the beneficiary-claimant merely shows the receipt of trust property by the defendant and makes no case as to its subsequent history or its existence among the present assets of the defendant”).
Nor do plaintiffs allege that defendants’ account balance remained above the surcharge amounts for purposes of their “lowest intermediate balance” theory. Indeed, a “consequence of the lowest balance rule is that, unless there is evidence to show the amount of the low balance, the plaintiff may recover nothing at all, on the view that without such evidence, the plaintiff’s funds have not been identified in the account.” 2 Dobbs § 6.1(4), at 22. Here, the gravamen of plaintiffs’ complaint is that defendants spent the surcharges on kickbacks and unwanted insurance products. That leaves plaintiffs to simply declare in their briefs that it is “almost inconceivable” that defendants did not place the surcharges into a general account before spending them, and that the general account still exists today such that the surcharges would be traceable. But as the complaint itself explains, BCBSMT has substantially reorganized, changed its name to CFM, sold its health insurance business, and at some point has donated or will donate its assets to public charity. Thus, even if defendants placed surcharges collected between 2006 and 2014 into a general account, we certainly find it at least “conceivable” that the account no longer exists. Because “our review is limited to the contents of the complaint,” Allen, 911 F.2d at 372, we decline to entertain plaintiffs’ unpleaded theory on appeal.
Plaintiffs also purport to seek disgorgement, which they define as a money judgment equivalent in value to ill-gotten assets that were dissipated on non-traceable items. This characterization of disgorgement—which runs headlong into Montanile’s refusal to permit recovery of assets that have been dissipated “on nontraceable items,” 136 S.Ct. at 658—is unavailing.
“Disgorgement” is simply a form of “[r]estitution measured by the defendant’s wrongful gain” rather than by the plaintiff’s loss, and is often described as “an ‘accounting for profits.’ ” Restatement (Third) of Restitution and Unjust Enrichment § 51 cmt. a, at 204 (2011); see Edmonson, 725 F.3d at 419 (“[D]isgorgement and accounting for profits are essentially the same remedy.”); 1 Dobbs § 4.3(5), at 610 (“[A]ccounting for profits ․ forces the [defendant] to disgorge gains received from improper use of the plaintiff’s property or entitlements.”). And as the Supreme Court explained in Great-West, “an accounting for profits” is an additional remedy—available when the plaintiff “is entitled to a constructive trust on particular property held by the defendant”—that allows the plaintiff to “recover profits produced by the defendant’s use of that property, even if [the plaintiff] cannot identify a particular res containing the profits sought to be recovered.” 534 U.S. at 214 n.2, 122 S.Ct. 708 (citing 1 Dobbs §§ 4.3(1), 4.3(5), at 588, 608). That is also how the Court described “disgorgement” in Harris Trust—a remedy available to recover the “proceeds” from disposing of particular property as well as “profits derived” from the illicit use of that property. 530 U.S. at 250, 120 S.Ct. 2180. Given the absence of any particular property in this case, plaintiffs’ request for disgorgement is not equitable in nature.15
Plaintiffs try to erase this particularity requirement by citing several trust law treatises that explain that trust beneficiaries could sue a third-party transferee in a court of equity to obtain a “money judgment” when the ill-gotten assets cannot be traced. See, e.g., Bogert § 868; 4 Austin Wakeman Scott & William Franklin Fratcher, The Law of Trusts §§ 291.1, 291.2, at 78–79 (4th ed. 1989); Restatement (Second) of Trusts § 291 cmt. e, at 59. Indeed, plaintiffs proclaim, courts of equity had “exclusive jurisdiction” in this context, and an “exclusively equitable remedy is, by definition, a typically equitable remedy.” But the Supreme Court rejected this reasoning in Mertens, explaining that although “courts of equity had exclusive jurisdiction over virtually all actions by beneficiaries for breach of trust”—including actions for monetary relief “against third persons who knowingly participated in the trustee’s breach”—many of those actions sought what were in effect “ ‘legal remedies’ granted by an equity court.” 508 U.S. at 256, 113 S.Ct. 2063. The phrase “equitable relief” in ERISA does not mean “whatever relief a common-law court of equity could provide.” Id. at 257, 113 S.Ct. 2063. Rather, it means relief that was “typically available in equity (such as injunction, mandamus, and restitution, but not compensatory damages).” Id. at 256, 113 S.Ct. 2063. And as explained above, equitable restitution (including its disgorgement variant) generally requires specifically identifiable property or its traceable proceeds.
Plaintiffs also point to Harris Trust, but nothing in Harris Trust alters that conclusion. As the Court explained in that case, beneficiaries can recover in equity from a third-party transferee only because “equity impresse[d] a constructive trust on the property” upon its transfer. Harris Tr., 530 U.S. at 250, 120 S.Ct. 2180 (quoting Moore v. Crawford, 130 U.S. 122, 128, 9 S.Ct. 447, 32 L.Ed. 878 (1889) ). And a “constructive trust ․ may be imposed only where the plaintiff’s funds are themselves located and identified or where they are traced into other funds or property.” 2 Dobbs § 6.1(3), at 12–13 (footnotes omitted). Indeed, “the nature of the relief” that the Court “described in Harris Trust [was] a claim to specific property (or its proceeds) held by the defendant.” Great-West, 534 U.S. at 215, 122 S.Ct. 708 (emphasis added). Because plaintiffs have not identified any specific property from which proceeds or profits derived, they cannot recover the derivative remedy of disgorgement.
In sum, plaintiffs are not seeking “appropriate equitable relief” under 29 U.S.C. § 1132(a)(3). We thus affirm the district court’s dismissal of plaintiffs’ prohibited transaction claim.
IV. STATE-LAW CLAIMS
We now turn to plaintiffs’ claims under state law for fraudulent inducement, constructive fraud, negligent misrepresentation, unjust enrichment, and unfair trade practices under the Montana Consumer Protection Act. Each of these claims is based on defendants’ alleged misrepresentations to plaintiffs that the premiums charged reflected the actual medical premium amount. See generally Morrow v. Bank of Am., N.A., 375 Mont. 38, 324 P.3d 1167, 1180–85 (2014) (describing the elements of fraud, constructive fraud, negligent misrepresentation, and unfair trade practices). The district court dismissed these claims after concluding that they are preempted by ERISA and that plaintiffs’ fraud allegations are not pled with sufficient particularity for purposes of Federal Rule of Civil Procedure 9(b). Plaintiffs challenge both conclusions.
“[T]wo strands of ERISA preemption” are relevant here: (1) “express” preemption under 29 U.S.C. § 1144(a); and (2) “conflict” preemption based on 29 U.S.C. § 1132(a). Paulsen v. CNF Inc., 559 F.3d 1061, 1081 (9th Cir. 2009) (citation omitted). Addressing each strand, we conclude that ERISA does not preempt plaintiffs’ state-law claims.
1. Express Preemption
ERISA expressly preempts “any and all State laws insofar as they may now or hereafter relate to any employee benefit plan.” 29 U.S.C. § 1144(a). The text of this provision—and in particular, the phrase “relate to”—is broad. So broad, in fact, that the Supreme Court has rejected an “ ‘uncritical literalism’ in applying” it given its potentially never-ending reach. Gobeille, 136 S.Ct. at 943 (quoting N.Y. State Conference of Blue Cross & Blue Shield Plans v. Travelers Ins. Co., 514 U.S. 645, 656, 115 S.Ct. 1671, 131 L.Ed.2d 695 (1995) ). To provide some “workable standards” for determining the scope of § 1144(a), the Court has identified “two categories” of state-law claims that “relate to” an ERISA plan—claims that have a “reference to” an ERISA plan, and claims that have “an impermissible ‘connection with’ ” an ERISA plan. Id. (citations omitted); see Or. Teamster Emp’rs Tr. v. Hillsboro Garbage Disposal, Inc., 800 F.3d 1151, 1155 (9th Cir. 2015) (“A [state] law claim ‘relates to’ an ERISA plan ‘if it has a connection with or reference to such a plan.’ ” (citation omitted) ). These two categories operate separately. See Cal. Div. of Labor Standards Enf’t v. Dillingham Constr., N.A., Inc., 519 U.S. 316, 324–25, 117 S.Ct. 832, 136 L.Ed.2d 791 (1997).
We first address the “reference to” category. A state-law claim has a “ ‘reference to’ an ERISA plan” if it “is premised on the existence of an ERISA plan” or if “the existence of the plan is essential to the claim’s survival.” Hillsboro Garbage, 800 F.3d at 1155–56 (quoting Providence Health Plan v. McDowell, 385 F.3d 1168, 1172 (9th Cir. 2004) ). In this case, plaintiffs’ state-law claims are not premised or dependent on the existence of an ERISA plan. Indeed, as explained above, the alleged misrepresentations occurred prior to any plan’s existence. We thus have little difficulty concluding that plaintiffs’ state-law claims do not have an impermissible “reference to” an ERISA plan.
We reach the same conclusion with respect to the “connection with” prong. A claim has “an impermissible ‘connection with’ ” an ERISA plan if it “ ‘governs a central matter of plan administration’ or ‘interferes with nationally uniform plan administration,’ ” Gobeille, 136 S.Ct. at 943 (internal alteration omitted) (quoting Egelhoff v. Egelhoff, 532 U.S. 141, 148, 121 S.Ct. 1322, 149 L.Ed.2d 264 (2001) ), or if it “bears on an ERISA-regulated relationship,” Hillsboro Garbage, 800 F.3d at 1155 (quoting Paulsen, 559 F.3d at 1082). We look to “the objectives of the ERISA statute as a guide,” bearing in mind a “ ‘starting presumption that Congress d[id] not intend to supplant’ ․ state laws regulating a subject of traditional state power” unless that power amounts to “a direct regulation of a fundamental ERISA function.” Gobeille, 136 S.Ct. at 943, 946 (quoting Travelers, 514 U.S. at 654, 115 S.Ct. 1671).
Preventing “sellers of goods and services, including benefit plans, from misrepresenting the contents of their wares” is certainly an area of traditional state regulation that “is ‘quite remote from the areas with which ERISA is expressly concerned—reporting, disclosure, fiduciary responsibility, and the like.’ ” Wilson v. Zoellner, 114 F.3d 713, 720 (8th Cir. 1997) (quoting Dillingham, 519 U.S. at 330, 117 S.Ct. 832); see Nat’l Sec. Sys., Inc. v. Iola, 700 F.3d 65, 84–85 (3d Cir. 2012) (collecting cases holding that ERISA does not expressly preempt state-law claims against an insurer “who makes fraudulent or misleading statements to induce participation in an ERISA plan”). Moreover, plaintiffs’ state-law claims do not “bear[ ] on an ERISA-regulated relationship.” Rutledge v. Seyfarth, Shaw, Fairweather & Geraldson, 201 F.3d 1212, 1219 (9th Cir. 2000), amended, 208 F.3d 1170 (9th Cir. 2000), abrogated in part on other grounds by Davila, 542 U.S. 200, 124 S.Ct. 2488, 159 L.Ed.2d 312. Although plaintiffs’ prohibited transaction claim involves an ERISA-regulated relationship (the relationship between a fiduciary and a party in interest), that relationship is unrelated to plaintiffs’ state-law claims, which focus on the misrepresentations made by defendants while they were operating “just like any other commercial entity.” Paulsen, 559 F.3d at 1083 (citation omitted).
Defendants argue that our decision in Rutledge compels a contrary conclusion. In Rutledge, we concluded that ERISA preempted a plan participant’s state-law claims against a law firm that allegedly overcharged the plan for legal services. 201 F.3d at 1222. We explained that “a core factor leading to the conclusion that a state law claim is preempted is that the claim bears on an ERISA-regulated relationship,” id. at 1219, and one such “ERISA-governed relationship” is the relationship between plan participants and parties in interest “in the respect [t]here at issue—excessive fees,” id. at 1221–22 & n.12. We thus held that “state-law claims against a non-fiduciary for prohibited transactions ‘relate to the administration of a plan covered by ERISA,’ ” and that the allegation of excessive fees in that case was a “prohibited transaction governed by ERISA.” Id. at 1221–22 (quoting Concha v. London, 62 F.3d 1493, 1504 (9th Cir. 1995) ).
This case differs from Rutledge, however, because plaintiffs’ state-law claims are premised on defendants’ misrepresentations in negotiations, not prohibited transactions. Indeed, plaintiffs’ state-law claims could succeed even if the premiums that defendants charged constituted “reasonable compensation” under ERISA, 29 U.S.C. § 1108(b)(2), because the claims allege that defendants misrepresented the composition of the premiums in a way that induced plaintiffs to subscribe to Chamber Choices plans. The actual amount of the premiums—and whether that amount was “reasonable compensation” under ERISA—is irrelevant to plaintiffs’ state-law claims. And the misrepresentations occurred, at least initially, before plaintiffs ever agreed to subscribe to a plan. The claims thus do not “bear[ ] on an ERISA-regulated relationship,” Rutledge, 201 F.3d at 1219, because no such relationship existed when the misrepresentations were made. Plaintiffs’ state-law claims are accordingly not expressly preempted by ERISA.
2. Conflict Preemption
In addition to its express preemption provision, ERISA articulates “a comprehensive civil enforcement scheme” in 29 U.S.C. § 1132(a) that is designed “to provide a uniform regulatory regime over employee benefit plans.” Davila, 542 U.S. at 208, 124 S.Ct. 2488 (quoting Pilot Life Ins. Co. v. Dedeaux, 481 U.S. 41, 54, 107 S.Ct. 1549, 95 L.Ed.2d 39 (1987) ).16 As a result, “any state-law cause of action that duplicates, supplements, or supplants the ERISA civil enforcement remedy conflicts with the clear congressional intent to make the ERISA remedy exclusive” and is therefore barred by conflict preemption. Id. at 209, 124 S.Ct. 2488. Conflict preemption can bar a state-law claim “even if the elements of the state cause of action [do] not precisely duplicate the elements of an ERISA claim,” id. at 216, 124 S.Ct. 2488, but a state-law claim is not preempted if it reflects an “attempt to remedy [a] violation of a legal duty independent of ERISA,” id. at 214, 124 S.Ct. 2488. State-law claims “are based on ‘other independent legal duties’ ” when they “are in no way based on an obligation under an ERISA plan” and “would exist whether or not an ERISA plan existed.” Marin Gen. Hosp. v. Modesto & Empire Traction Co., 581 F.3d 941, 950 (9th Cir. 2009) (internal alteration omitted) (quoting Davila, 542 U.S. at 210, 124 S.Ct. 2488).
In this case, the duties implicated in plaintiffs’ state-law claims do not derive from ERISA; indeed, ERISA does not purport to govern negotiations between insurance companies and employers. Each of the state-law claims arises from defendants’ misrepresentations and the effect they had on plaintiffs’ decisions to subscribe to Chamber Choices plans. The legal duties at issue in these state-law claims are independent of the duties imposed by ERISA and would exist regardless of whether an ERISA plan existed. See Cotton, 402 F.3d at 1290 (finding no conflict preemption where the plaintiffs sought “damages based on fraud in the sale of insurance policies”). Put in the terms used by the district court, plaintiffs’ state-law claims are not “alternative enforcement mechanisms” to ERISA claims because ERISA does not have an enforcement mechanism that regulates misrepresentations by insurance companies. Plaintiffs’ state-law claims are thus not barred by either express or conflict preemption.
B. Rule 9(b) Particularity
Finally, we turn to the district court’s conclusion that plaintiffs “have not met the heightened pleading standard required under Federal Rule of Civil Procedure 9(b) as to their allegations of fraud.” Rule 9(b)’s particularity requirement applies to plaintiffs’ claims of fraudulent inducement and constructive fraud.17
Under Rule 9(b), a party “alleging fraud or mistake ․ must state with particularity the circumstances constituting fraud or mistake.” Fed. R. Civ. P. 9(b). To satisfy Rule 9(b)’s particularity requirement, the complaint must include “an account of the ‘time, place, and specific content of the false representations as well as the identities of the parties to the misrepresentations.’ ” Swartz v. KPMG LLP, 476 F.3d 756, 764 (9th Cir. 2007) (per curiam) (quoting Edwards v. Marin Park, Inc., 356 F.3d 1058, 1066 (9th Cir. 2004) ). In other words, the pleading “must ‘identify the who, what, when, where, and how of the misconduct charged, as well as what is false or misleading about the purportedly fraudulent statement, and why it is false.’ ” Salameh v. Tarsadia Hotel, 726 F.3d 1124, 1133 (9th Cir. 2013) (quoting United States ex rel. Cafasso v. Gen. Dynamics C4 Sys., Inc., 637 F.3d 1047, 1055 (9th Cir. 2011) ).
The complaint in this case alleges that defendants misrepresented the basis of the premiums they charged. But the complaint lacks sufficient detail with respect to the “who,” “when,” “where,” or “how.” Plaintiffs vaguely allege that defendants made these misrepresentations “[i]n the course of marketing” the plans to plaintiffs over a period of eight years—from 2006 to 2014. Plaintiffs do not allege the details of these misrepresentations, such as when defendants made them, where or how defendants made them, to whom they were made, or the specific contents of the misrepresentations.18 See Concha, 62 F.3d at 1503 (“Rule 9(b) ․ requires that plaintiffs specifically plead those facts surrounding alleged acts of fraud to which they can reasonably be expected to have access.”). We therefore agree with the district court that plaintiffs’ allegations do not state with particularity the circumstances of the alleged fraud.
Nevertheless, because we reverse the district court’s conclusion that plaintiffs’ state-law claims are preempted, we reverse the district court’s dismissal with prejudice of the state-law claims so that plaintiffs may amend their complaint to state the fraud allegations with greater particularity. See United States ex rel. Swoben v. United Healthcare Ins. Co., 848 F.3d 1161, 1167 (9th Cir. 2016). We note, however, that because we affirm the dismissal of plaintiffs’ ERISA claims, the district court is also free on remand to decline to exercise supplemental jurisdiction over the state-law claims and allow plaintiffs to bring them in state court. See Sanford v. MemberWorks, Inc., 625 F.3d 550, 561 (9th Cir. 2010) (discussing supplemental jurisdiction under 28 U.S.C. § 1367(c)(3) ).
“[R]educed to the size of a pea, this case is really about claims of fraud and misrepresentation in the sale of some [health] insurance policies.” Cotton, 402 F.3d at 1279. ERISA does not regulate such conduct, which means that plaintiffs’ ERISA claims, and defendants’ ERISA preemption defense, fail. We accordingly AFFIRM the district court’s judgment with respect to plaintiffs’ ERISA claims, REVERSE the district court’s judgment with respect to plaintiffs’ state-law claims, and REMAND for further proceedings consistent with this opinion. Each party shall bear its own costs on appeal.
AFFIRMED IN PART, REVERSED IN PART, and REMANDED.
1. Because this case comes to us on review of a motion to dismiss, we accept as true the factual allegations in the operative complaint. See Askins v. U.S. Dep’t of Homeland Sec., 899 F.3d 1035, 1038 (9th Cir. 2018).
2. Plaintiffs incorporated the Member Guide by reference in their complaint. See Santomenno v. Transamerica Life Ins. Co., 883 F.3d 833, 836 n.2 (9th Cir. 2018). According to plaintiffs, the Member Guide is representative of the Chamber Choices plans.
3. Under § 1132(a)(2), a plan participant, beneficiary, or fiduciary may bring “[a] civil action ․ for appropriate relief under [§ 1109].” Section 1109 imposes personal liability upon “[a]ny person who is a fiduciary with respect to a plan who breaches any of the responsibilities, obligations, or duties imposed upon fiduciaries by [ERISA].” 29 U.S.C. § 1109(a).
4. ERISA’s definition of “person” includes “corporation[s]” and other “association[s].” 29 U.S.C. § 1002(9).
5. Plaintiffs do not argue that defendants acted as fiduciaries under 29 U.S.C. § 1002(21)(A)(iii), which provides that a person is a fiduciary to the extent it “has any discretionary authority or discretionary responsibility in the administration of [the] plan.” Insurers generally act in a fiduciary capacity under this “plan administration” provision when making a discretionary determination about whether a claimant is entitled to benefits. See King v. Blue Cross & Blue Shield of Ill., 871 F.3d 730, 745–46 (9th Cir. 2017).
6. Plaintiffs contend that, because the surcharges were concealed, the premiums were not “definitively calculable and nondiscretionary compensation ․ clearly set forth in a contract with the fiduciary-employer.” Santomenno, 883 F.3d at 841. But the premiums were fully disclosed and negotiated; the fact that defendants did not disclose the composition of the premiums (or how they were spending them) does not mean that defendants exercised discretion in setting them.
7. Plaintiffs point to an allegation in the complaint that defendants “ ‘imposed and increased’ the non-premium surcharges ‘without providing the notice required by contract.’ ” That conclusory allegation does not indicate that defendants exercised discretion under the modification provision. Nor would such an allegation be plausible, because plaintiffs’ theory is that the modification provision granted defendants discretion to increase premiums, not “non-premium surcharges.” And if, as plaintiffs allege, they “had no way to know” that defendants were embedding surcharges into the premium amounts, any alleged increases would have had to occur before the premium amounts were agreed to. Otherwise, a mid-year increase in surcharges would have led to a mid-year increase in premiums, a fact that is not alleged in the complaint and that would have been quite obvious to plaintiffs had it occurred.
8. Defendants’ failure to exercise discretion also sets this case apart from the cases cited by plaintiffs in which an insurer actually exercised discretion that had been granted in the contract. See, e.g., Ed Miniat, Inc. v. Globe Life Ins. Grp., Inc., 805 F.2d 732, 734, 737–38 (7th Cir. 1986) (insurer “unilaterally and without justification announced ․ an abandonment of existing policy holders, by reducing the rate of return paid on account ․ and increasing premium rates to the maximum allowed by the policy”); Chi. Bd. Options Exch., Inc. v. Conn. Gen. Life Ins. Co., 713 F.2d 254, 255, 259–60 (7th Cir. 1983) (insurer “exercis[ed]” its discretion to make a “unilateral amendment [to] an annuity contract” from which the plaintiffs could not withdraw). Here, the premiums charged were based not on defendants’ discretion but instead on arm’s-length negotiations.
9. A Department of Labor regulation defines “assets of the plan” to “include amounts ․ that a participant or beneficiary pays to an employer, or amounts that a participant has withheld from his wages by an employer, for contribution ․ to the plan, as of the earliest date on which such contributions ․ can reasonably be segregated from the employer’s general assets.” 29 C.F.R. § 2510.3-102(a)(1). As explained below, this regulation is inapplicable here because the “amounts” plaintiffs seek to recover are not amounts that were “pa[id] to an employer” by plan participants but instead amounts paid by an employer to an insurance company.
10. Although we have adopted a “functional definition of what constitutes an ‘asset of the plan’ for purposes of [29 U.S.C. § 1106],” Kayes v. Pac. Lumber Co., 51 F.3d 1449, 1466–67 (9th Cir. 1995) (citing Acosta v. Pac. Enters., 950 F.2d 611, 620 (9th Cir. 1991) ), that definition assumes fiduciary status and is thus not helpful in determining whether a party is in fact a fiduciary under 29 U.S.C. § 1002(A)(21)(i).
11. In light of our holding, we express no opinion on the district court’s conclusion that the premium payments were not plan assets pursuant to ERISA’s “guaranteed benefit policy” provision, 29 U.S.C. § 1101(b)(2).
12. The text of these sections creates the prohibited transaction as follows: “Except as provided in section 1108 ․ [a] fiduciary with respect to a plan shall not cause the plan to engage in a transaction, if he knows or should know that such transaction constitutes a direct or indirect ․ furnishing of ․ services ․ between the plan and a party in interest.” 29 U.S.C. § 1106(a)(1)(C). “The prohibitions provided in section 1106 of this title shall not apply to ․ [c]ontracting or making reasonable arrangements with a party in interest for ․ services necessary for the establishment or operation of the plan, if no more than reasonable compensation is paid therefor.” Id. § 1108(b)(2).
13. In Harris Trust, the fiduciaries of a pension plan sued Salomon Smith Barney (“Salomon”), a firm that provided broker-dealer and trading services to the plan. 530 U.S. at 242, 120 S.Ct. 2180. While the plan was receiving these services, the plan also purchased worthless investments from Salomon at the direction of an investment manager that exercised “discretion over a portion of the plan’s assets” and was therefore also a plan fiduciary. Id. at 242–43, 120 S.Ct. 2180. The remaining fiduciaries sued Salomon under § 1132(a)(3) as a non-fiduciary party in interest, contending that the investment manager, acting as a fiduciary, caused the plan to engage in a prohibited transaction between the plan and a party in interest for the sale of property in exchange for plan assets. Id. at 243, 120 S.Ct. 2180 (citing 29 U.S.C. § 1106(a)(1)(A), (D) ).
14. Although we are skeptical that the basis for plaintiffs’ claim is equitable, defendants did not dispute that issue in the district court and raise it for the first time on appeal. We thus do not address that question and instead apply our “ ‘general rule’ against entertaining arguments on appeal that were not presented or developed before the district court.” Richards v. Ernst & Young, LLP, 744 F.3d 1072, 1075 (9th Cir. 2013) (citation omitted).
15. Under traditional rules of equity, an accounting for profits may be available in the absence of a constructive trust over specifically identifiable property if the defendant owed a fiduciary duty to the plaintiff and breached that duty. See Parke v. First Reliance Standard Life Ins. Co., 368 F.3d 999, 1008–09 (8th Cir. 2004); cf. CIGNA Corp. v. Amara, 563 U.S. 421, 441–42, 131 S.Ct. 1866, 179 L.Ed.2d 843 (2011) (describing a “surcharge,” which is an equitable remedy “in the form of monetary ‘compensation’ for a loss resulting from a trustee’s breach of duty, or to prevent the trustee’s unjust enrichment” (citation omitted) ). But fiduciary status is “[t]he important ingredient.” 1 Dobbs § 4.3(5), at 611 n.16; see CIGNA, 563 U.S. at 442, 131 S.Ct. 1866 (“[T]he fact that the defendant ․ is analogous to a trustee makes a critical difference.”). And for purposes of plaintiffs’ prohibited transaction claim, defendants are not fiduciaries but instead non-fiduciary third parties.
16. The “possible claims” under 29 U.S.C. § 1132(a) are “(1) an action to recover benefits due under the plan; (2) an action for breach of fiduciary duties; and (3) a suit to enjoin violations of ERISA or the [p]lan, or to obtain other equitable relief” to redress ERISA or plan violations. Bast v. Prudential Ins. Co. of Am., 150 F.3d 1003, 1008 (9th Cir. 1998) (internal citations omitted).
17. We disagree with plaintiffs’ assertion that Rule 9(b) does not apply to their constructive fraud claim because Montana’s version of the rule would not apply in state court. Rule 9(b)’s particularity requirement “is a federally imposed rule” that applies “irrespective of whether the substantive law at issue is state or federal.” Kearns v. Ford Motor Co., 567 F.3d 1120, 1125 (9th Cir. 2009) (quoting Vess v. Ciba-Geigy Corp. USA, 317 F.3d 1097, 1102–03 (9th Cir. 2003) ); see id. (rejecting the argument “that Rule 9(b) does not apply to California’s consumer protection statutes because California courts have not applied Rule 9(b) [to those statutes]”). State law is relevant only “to determine whether the elements of fraud have been pled sufficiently to state a cause of action.” Id. (quoting Vess, 317 F.3d at 1103). And because plaintiffs rely on a “unified course of fraudulent conduct” as the basis of the constructive fraud claim, the claim is at a minimum “grounded in fraud” and therefore “must satisfy the particularity requirement of Rule 9(b).” Vess, 317 F.3d at 1103–04.
18. Defendants also argue that the complaint impermissibly lumps together HCSC and CFM as “BCBSMT.” This argument lacks merit. The complaint specifically explains that “BCBSMT” refers to CFM for conduct occurring before July 2013, and to HCSC for conduct occurring after July 2013.
BYBEE, Circuit Judge: