UNPUBLISHED

                          UNITED STATES COURT OF APPEALS
                              FOR THE FOURTH CIRCUIT


                                        No. 17-1485


WILLIAM L. PENDER; DAVID L. MCCORKLE,

            Plaintiffs - Appellants,

and

ANITA POTHIER; KATHY L. JIMENEZ; MARIELA ARIAS; RONALD R.
WRIGHT; JAMES C. FABER, JR., On behalf of themselves and on behalf of all
others similarly situated,

            Plaintiffs,

v.

BANK OF AMERICA CORPORATION; BANK OF AMERICA, NA; BANK OF
AMERICAN PENSION PLAN; BANK OF AMERICA 401(K) PLAN; BANK OF
AMERICA CORPORATION CORPORATE BENEFITS COMMITTEE; BANK
OF AMERICA TRANSFERRED SAVINGS ACCOUNT PLAN,

            Defendants - Appellees,

and

UNKNOWN PARTY, John and Jane Does #1-50, Former Directors of NationsBank
Corporation and Current and Former Directors of Bank of America Corporation &
John & Jane Does #51-100, Current/Former Members of the Bank of America
Corporation Corporate Benefit; PRICEWATERHOUSE COOPERS, LLP;
CHARLES K. GIFFORD; JAMES H. HANCE, JR.; KENNETH D. LEWIS;
CHARLES W. COKER; PAUL FULTON; DONALD E. GUINN; WILLIAM
BARNETT, III; JOHN T. COLLINS; GARY L. COUNTRYMAN; WALTER E.
MASSEY; THOMAS J. MAY; C. STEVEN MCMILLAN; EUGENE M.
MCQUADE; PATRICIA E. MITCHELL; EDWARD L. ROMERO; THOMAS M.
RYAN; O. TEMPLE SLOAN, JR.; MEREDITH R. SPANGLER; HUGH L.
MCCOLL; ALAN T. DICKSON; FRANK DOWD, IV; KATHLEEN F.
FELDSTEIN; C. RAY HOLMAN; W. W. JOHNSON; RONALD TOWNSEND;
SOLOMON D. TRUJILLO; VIRGIL R. WILLIAMS; CHARLES E. RICE; RAY
C. ANDERSON; RITA BORNSTEIN; B. A. BRIDGEWATER, JR.; THOMAS E.
CAPPS; ALVIN R. CARPENTER; DAVID COULTER; THOMAS G. COUSINS;
ANDREW G. CRAIG; RUSSELL W. MEYER-, JR.; RICHARD B. PRIORY;
JOHN C. SLANE; ALBERT E. SUTER; JOHN A. WILLIAMS; JOHN R. BELK;
TIM F. CRULL; RICHARD M. ROSENBERG; PETER V. UEBERROTH;
SHIRLEY YOUNG; J. STEELE ALPHIN; AMY WOODS BRINKLEY;
EDWARD J. BROWN, III; CHARLES J. COOLEY; ALVARO G. DE MOLINA;
RICHARD M. DEMARTINI; BARBARA J. DESOER; LIAM E. MCGEE;
MICHAEL E. O'NEILL; OWEN G. SHELL, JR.; A. MICHAEL SPENCE; R.
EUGENE TAYLOR; F. WILLIAM VANDIVER, JR.; JACKIE M. WARD;
BRADFORD H. WARNER,

             Defendants.


Appeal from the United States District Court for the Western District of North Carolina, at
Charlotte. Graham C. Mullen, Senior District Judge. (3:05-cv-00238-GCM)


Argued: March 21, 2018                                             Decided: June 5, 2018


Before KEENAN, WYNN, and FLOYD, Circuit Judges.


Affirmed by unpublished opinion. Judge Wynn wrote the majority opinion, in which Judge
Floyd joined. Judge Keenan wrote a dissenting opinion.


ARGUED: Julia Penny Clark, BREDHOFF & KAISER, PLLC, Washington, D.C., for
Appellants. Carter Glasgow Phillips, SIDLEY AUSTIN LLP, Washington, D.C., for
Appellees. ON BRIEF: Eli Gottesdiener, GOTTESDIENER LAW FIRM, PLLC,
Brooklyn, New York; Thomas D. Garlitz, THOMAS D. GARLITZ, PLLC, Charlotte,
North Carolina, for Appellants. Irving M. Brenner, MCGUIREWOODS LLP, Charlotte,
North Carolina; Anne E. Rea, David F. Graham, Tacy F. Flint, Steven J. Horowitz,
SIDLEY AUSTIN LLP, Chicago, Illinois, for Appellees.


Unpublished opinions are not binding precedent in this circuit.



                                            2
WYNN, Circuit Judge:

       This Employment Retirement Income Security Act of 1974 (“ERISA”) case returns

to the Court for a third time. See Pender v. Bank of Am. Corp., 

(4th Cir.

2015); McCorkle v. Bank of Am. Corp., 

(4th Cir. 2012). Plaintiffs, a class of

current and former employees of Bank of America and certain of its predecessors

(collectively, with the Bank’s Pension Plan, the “Bank”), seek an equitable accounting for

any profits accruing to the Bank resulting from its unlawful transfer of the balances of

Plaintiffs’ 401(k) Plan accounts into the general account of the Bank’s Pension Plan.

. In 2015, this Court ruled that the district court erred in dismissing

Plaintiffs’ accounting action, and remanded the case to the district court to determine

whether the Bank retained any profit as a result of the unlawful transfers and its use of the

transferred funds. 

370.

       On appeal, as it did before the district court, the Bank advances a simple, if

somewhat surprising, argument—that the Pension Plan’s investment strategy for the

unlawfully transferred funds, which was developed and implemented by the Bank’s trained

asset managers, performed far worse than Plaintiffs’ investment strategies, as reflected in

their 401(k) account investment allocations. Because Plaintiffs’ investment allocations

outperformed the Bank’s investment strategy—and the Pension Plan was responsible for

making up any shortfall between the performances of the Bank’s investment strategy and

Plaintiffs’ allocations—the Bank maintains that it did not profit from the transfers. After

conducting a four-day bench trial, during which the parties presented fact and expert

testimony and evidence, the district court agreed with the Bank and, therefore, dismissed

                                              3
Plaintiffs’ action as moot. Pender v. Bank of Am. Corp., No. 3:05-CV-00238, 

, at *23 (W.D.N.C. Apr. 27, 2017). For the reasons that follow, we affirm.



                                             I.

       In 1998, the Bank amended its 401(k) Plan to provide eligible participants with the

opportunity to transfer their account balances to the Bank’s defined-benefit Pension Plan.

. Once transferred to the Pension Plan, beneficiaries could continue

to allocate their account balances among various investment options. 

However,

unlike with balances held in the 401(k) Plan, which were actually invested in the selected

investment options, beneficiaries who elected to transfer their accounts to the Pension Plan

would have only notional (or hypothetical) accounts—the Bank could invest the

beneficiaries’ account balances however it saw fit. 

for beneficiaries’

agreement to transfer their balances to the Pension Plan and the use of the beneficiaries’

funds, the Bank guaranteed that such beneficiaries’ account balances would not fall below

the amount in their account at the time of the transfer. 

offered the transfer

option because it believed it could obtain a higher return with the beneficiaries’ money

than the beneficiaries were obtaining. Many beneficiaries elected to transfer their 401(k)

Plan account balances to the Pension Plan, with beneficiaries in aggregate transferring

nearly $2 billion in their 401(k) Plan balances to the Pension Plan for the Bank’s use.

       In 2005, the Internal Revenue Service (the “IRS”) concluded that the transfers

violated ERISA’s “anti-cutback provision,” 29 U.S.C. § 1054(g)(1), which bars plan

amendments from decreasing a participant’s “accrued benefit.” 

The IRS found,

                                             4
and this Court later agreed, that stripping beneficiaries of the 401(k) Plan’s “separate

account feature” deprived beneficiaries of a meaningful benefit because it subjected plan

participants to the risk that the Bank would invest the transferred assets poorly, and

therefore lack sufficient funds to satisfy all of the returns a beneficiary obtained in his

notional investment account. 

(“[T]he Bank’s promise that the value of the

transferred funds will not decrease below a certain threshold—even if, for example, it

invests Pension Plan assets poorly and loses money—is not the same as actually not

decreasing the account balance.”).

       In 2008, the IRS reached a closing agreement 1 with the Bank, pursuant to which the

Bank “(1) paid a $10 million fine to the U.S. Treasury, (2) set up a special purpose 401(k)

plan, (3) . . . transferred Pension Plan assets that were initially transferred from the 401(k)

Plan to the special-purpose 401(k) plan,” and (4) made additional payments to certain plan

participants whose hypothetical return in their notional account was less than a defined

amount. 

The Bank completed those transfers in 2009. 

a

result of the transfers to the special-purpose 401(k) plan and the additional payments to

certain plan participants, all Plaintiffs’ current account balances are at least as large as they

would have been had the funds in Plaintiffs’ accounts actually been invested in accordance

with their notional allocations. Pender v. Bank of Am. Corp., No. 3:05-CV-238, 

, at *10 (W.D.N.C. Aug. 19, 2013), rev’d on other grounds, 

.



       1
         A closing agreement is a binding agreement finally and conclusively settling—i.e.,
closing the file as to—a tax issue between the IRS and a taxpayer.

                                               5
       Around the same time that the IRS began to take action, Plaintiffs filed a variety of

equitable and statutory claims related to the transfers. 

. All but

one of those claims were dismissed and are not at issue in this appeal. 

n.4, 177. Plaintiffs’ lone remaining claim is premised on the Bank’s violation of the

anti-cutback provision. In 2013, the district court dismissed that claim on grounds that the

remedial provisions in the IRS closing agreement rendered such claims moot because it

restored the 401(k) Plan’s separate account feature. Pender, 

, at *5–9.

       This Court reversed, explaining that Plaintiffs suffered a legally cognizable ongoing

injury if the Bank retained a profit as a result of its unlawful transfer of the 401(k) Plan

balances to the Pension Plan, and its investment of those balances. 

–65. In reaching that conclusion, this Court held that Plaintiffs could pursue relief

under ERISA Section 502(a)(3), which authorizes a plan beneficiary to obtain any

“appropriate equitable relief” to redress “any act or practice which violates” certain ERISA

provisions, including the anti-cutback provision. 

This Court further concluded

that the “accounting for profits” sought by Plaintiffs is one form of “equitable relief”

available under Section 502(a)(3). 

       On remand, the district court conducted a four-day bench trial to determine

“whether, after it restored the separate account feature and paid a $10 million fine to the

IRS, the Bank nevertheless profited from its transfer strategy.” Pender, 

,

at *4. At trial, as they do on appeal, Plaintiffs and the Bank offered distinct approaches to

determining whether the Bank retained a profit as a result of the transfer of the

beneficiaries’ 401(k) Plan account balances to the Pension Plan.

                                             6
       On the one hand, Plaintiffs focused on the undisputed fact that the transferred 401(k)

Plan balances were “pooled” or “commingled” with—rather than segregated from—the

funds of the Pension Plan. 

that, as a matter of black-letter equity law,

when improperly obtained funds are commingled with other funds, a plaintiff is entitled to

a share of the returns on all of the commingled funds in proportion to the unlawfully

obtained assets’ share of the commingled fund as a whole. 

end, one of

Plaintiffs’ experts, Lawrence Deutsch, calculated—and the Bank does not dispute—that

the Pension Plan as a whole had a cumulative rate of return of 28.6% over the 1998 to 2009

period when the 401(k) Plan assets were pooled with the Pension Plan assets. See 

proportionate share of the Pension Plan’s retained profit based on that return

rate to the 401(k) Plan participants who transferred their funds to the Pension Plan, Deutsch

calculated that the Bank retained $379 million in profit, including accrued interest, from

the unlawful transaction. 

Bank did not introduce its own analysis using

Plaintiffs’ proportionate-share-of-the-whole methodology, a Bank executive testified—in

accordance with Deutsch’s analysis—that the Pension Plan as a whole outperformed the

beneficiaries’ notional allocations during the relevant time period.

       By contrast, the Bank asserted—and the district court found—that the Bank relied

on a different, contemporaneously documented, “Investment Strategy” for the 401(k)

balances transferred to the Pension Plan than for the remaining funds in the Pension Plan.

*8. In particular, “[t]he core of the [Pension] Plan’s Investment Strategy was to

invest the assets used to fund the [transferred 401(k) accounts] more heavily in equities

than participants invested their hypothetical accounts, on the theory that equities would be

                                              7
expected to outperform fixed income options over the long term. The [Pension] Plan did

this by matching or ‘hedging’ participant equity investments with [Pension] Plan equity

investments and investing approximately 60% of participant fixed income investments in

equities.” 

       The Bank’s expert, Dr. Russell Wermers, sought to determine whether the

transferred balances returned a profit to the Bank under the Investment Strategy. To do so,

Wermers analyzed the performance of three asset classes included in the Pension Plan’s

general investment portfolio: domestic equities, international equities, and fixed income

assets. Pender, 

, at *7. Wermers used three benchmark indices to

estimate the returns of those asset classes over the relevant time period—the Russell 3000

for domestic equities, the MSCI EAFE Index for non-U.S. equities, and the Lehman

Brothers Aggregate Bond Index for fixed income assets. 

Pension Plan

did not necessarily invest in these indices, the Investment Policy identified these indices as

benchmarks for their respective asset classes, and directed the investment managers to

“[e]qual or exceed the return of the benchmark, net of fees, at a comparable level of risk.”

J.A. 841.

       Wermers found that the two equity indices (Russell 3000 and MSCI EAFE) declined

during the relevant period, whereas the bond index increased by 81.1%. Pender, 

, at *7. Because (1) the Investment Strategy had the effect of overweighting equity

investments by matching 401(k) Plan participants’ equity allocations and treating their

fixed income investments as part of the Pension Plan’s general investment account and (2)

equities performed worse than fixed income assets during the relevant time, Wermers

                                              8
opined that the transfer strategy must have resulted in a loss to the Pension Plan—in the

amount of the difference in performance between the Bank’s (underperforming) allocation

and the notional allocations by 401(k) Plan participants, which were less heavily weighted

toward equities. 

to Wermers’ opinion, the Bank also introduced contemporaneously

recorded and maintained spreadsheets, which, on an aggregate basis, tracked “participants’

hypothetical equity and fixed income investments” and compared those returns to the

returns realized by Pension Plan’s Investment Strategy. 

Based on these

spreadsheets, a Bank executive responsible for monitoring the Pension Plan’s investments,

David Andreasen, calculated that the Investment Strategy of overweighting equity

investments resulted in a net investment loss to the Bank. 

Andreasen

testified that the Investment Strategy for the transferred balances yielded a return of 3.5%

over the relevant period, compared to a 16.5% aggregate return on Plaintiffs’ notional

investments, leading to a loss of $149 million. Put differently, Andreasen concluded that

the Bank’s investment allocation performed far worse than the beneficiaries’ notional

allocations. Adding the costs associated with the IRS closing (such as the IRS penalty and

the costs of creating and transferring funds into the special purpose 401(k) plan),

Andreasen testified that the transfer resulted in a cumulative loss of $272 million. 

“[b]oth the Plaintiffs[’] and the Defendants[’] experts . . . presented a

coherent and facially plausible story for their parties” and none of their testimony was

“contradicted by objective evidence,” the district court nonetheless found that

“Defendants’ experts provided evidence at trial that is more credible than the testimony

                                             9
provided by the Plaintiffs’ experts.” 

In rendering this finding, the district court

determined that Deutsch’s approach was “not the appropriate measure of profits due to the

transfer,” because it focused on the performance of the Pension Plan as a whole, not profits

on the transferred 401(k) balances attributable to the Investment Strategy. 

(“The Bank’s profit from the transfer (if any) is best measured using the returns from

the Investment Strategy that the [Pension] Plan actually used to fund the [transferred 401(k)

Plan balances].”). Put differently, the district court concluded that use of Plaintiffs’ pro

rata or “proportionate-share-of-the whole” approach “would be inappropriate because it

would produce ‘profits’ having nothing to do with the transfers and therefore is contrary to

the purpose of this inquiry.” 

       The district court further determined that “[t]he appropriate way to determine

whether there was a profit retained as a result of its investment strategy applied to the

transferred assets is to look at the returns attributable to that ‘investment strategy.’” 

(internal quotation marks omitted); see also 

(stating that in

assessing a claim for an accounting for profits a court must “‘reach the best approximation

it can under the circumstances’ of the profit attributable to the conduct at issue” (quoting

Restatement (Third) of Restitution & Unjust Enrichment § 51 cmt. g (2011))). The court

said it embraced the Bank’s “attribution” approach because “Plaintiffs’ use of total

[Pension] Plan returns would confer an inappropriate windfall on participants, act as a

penalty and otherwise be inequitable.” 

Applying the Bank’s attribution

approach and crediting Wermers’ and Andreasen’s analyses, the district court held that the



                                             10
Bank did not retain any profit as a result of the transfer and, therefore, dismissed Plaintiffs’

claim for an accounting for profits as moot. 

Plaintiffs timely appealed.



                                              II.

       On appeal, Plaintiffs argue that the district court reversibly erred in relying on the

Bank’s Investment Strategy to determine whether the Bank profited from the unlawfully

transferred funds—and therefore denying Plaintiffs equitable relief—rather than

calculating all profits accruing to the Pension Plan during the course of the commingling

of the funds and awarding Plaintiffs a proportionate share of those profits. 2 When, as here,

a district court renders a decision after a bench trial, “we review the district court’s factual

findings for clear error and its legal conclusion de novo.” F.T.C. v. Ross, 

,

894 (4th Cir. 2014); see also Fed. R. Civ. P. 52. Likewise, we review a district court’s

decision to award or deny “equitable relief for abuse of discretion, accepting the court’s

factual findings absent clear error, while examining issues of law de novo.” Dixon v.

Edwards, 

, 710 (4th Cir. 2002); see also Griggs v. E.I. Du Pont de Nemours

& Co., 

, 385–86 (4th Cir. 2001) (“leav[ing] it to the sound discretion of the




       2
        Plaintiffs further argue that this Court should direct the district court to conclude
that Deutsch correctly determined the Bank’s profit from the transaction and that the Bank
is not entitled to seek certain set-offs from or assert equitable defenses to that
determination. As detailed below, we conclude that the district court did not reversibly err
in denying Plaintiffs equitable relief. Accordingly, we need not—and thus do not—address
these additional arguments.

                                              11
district court” to determine whether, and in what form, an award of equitable relief under

Section 502(a)(3) was “appropriate”).

       To resolve Plaintiffs’ appeal, we must address two questions: (A) whether the

district court was required to follow Plaintiffs’ proposed “proportionate-share-of-the-

whole” approach and, if not, (B) whether the district court reversibly erred in relying on

the Bank’s attribution approach in determining that the Bank did not profit from the

transfer.

                                              A.

       As to the first question, the proportionate-share-of-the-whole approach advanced by

Plaintiffs finds substantial support in Restatements, treatises, and case law. For example,

the Restatement (First) of Trusts provides that “[w]here the trustee wrongfully mingles

trust property with his individual property in one indistinguishable mass” and “exchanges

the mingled mass for other property” that “becomes more valuable than the mingled mass

with which it is acquired, the beneficiary is entitled to a proportionate share of the property,

and thus to secure the profit which arises from the transaction.” Restatement (First) of

Trusts § 202 cmt. h (1935); see also Restatement (Second) of Trusts § 202 (same);

Restatement (First) of Restitution § 209 cmt. a (1937) (“The person whose money is

wrongfully mingled with money of the wrongdoer does not thereby lose his interest in the

money, although the identity of his money can no longer be shown, but he acquires an

interest in the mingled fund. His interest is such that he is entitled in equity to claim a

proportionate share of the mingled fund or a lien upon it.” (emphasis added)); 

210.



                                              12
       Likewise, authoritative legal commentators support the proportionate-share-of-the-

whole approach. See, e.g., 2 Dan B. Dobbs, Law of Remedies § 6.1(4), at 16–17 (2d ed.

1993) (“[W]hen the defendant uses the entire commingled fund to purchase property . . .

the plaintiff is not entitled to a constructive trust on the entire property purchased, but he

is entitled to a trust for a share in the property proportionate to his share in the fund.”);

Austin W. Scott, The Right to Follow Money Wrongfully Mingled with Other Money, 27

Harv. L. Rev. 125, 127 (1913) (“[W]here the claimant’s money is mingled with that of the

wrongdoer, and is therefore only partly instrumental in earning the profit[,] [t]he claimant

should be entitled to a share of the profit, in so far as his property contributed to earning

the profit.”).

       Both this Court and the Supreme Court also have endorsed use of the proportionate-

share-of-the-whole approach to determine the profit obtained by a defendant as a result of

its use of unlawfully commingled funds. See, e.g., Henkels v. Sutherland, 

,

302 (1926) (“Since the proceeds resulting from the sale of Henkels’ property have been

commingled with the proceeds of other sales and thus invested, an account must be taken

to ascertain the average rate of interest received by the Treasury upon all the proceeds

invested and thereupon . . . a proportionate allocation made in respect of the proceeds

belong to Henkels for the period of their investment.”); MacBryde v. Burnett, 

, 900 (4th Cir. 1942) (“[I]f trust funds are mingled with personal funds of a trustee, the

whole is impressed with a trust until separation of the trust property can be made, and that

the trust [beneficiaries are] entitled to a proportionate part of the profits realized by the

trustee in dealings with the fund in which the trust funds are mingled.”).

                                             13
       And other circuits also have applied the proportionate-share-of-the-whole approach

in such circumstances. See, e.g., Provencher v. Berman, 

, 570 (1st Cir. 1983)

(Breyer, J.) (explaining that when “a ‘conscious wrongdoer’ . . . uses commingled funds to

buy property, . . . the innocent party can choose either to enforce a lien on the property for

the value of the estate’s funds or” claim “the proportionate share of the real estate”);

Bartlett & Co., Grain v. Commodity Credit Corp., 

, 409 (8th Cir. 1962) (“The

amount of the actual yield of the bills is known, and the claim of Commodity for the period

now in question should be limited to its pro rata share of the yield.” (emphasis added));

Bird v. Stein, 

, 177 (5th Cir. 1958) (Wisdom, J.); Marcus v. Otis, 

, 150 (2d Cir. 1948) (A. Hand, J.) (“[W]here a wrongdoer mingles his own funds with

other funds which he has misappropriated . . . he is liable only for a proportionate part of

the profits realized based upon the ratio of the amount of money he misappropriated to the

original commingled mass.”).          Additionally, at least one court has applied the

proportionate-share-of-the-whole methodology in an analogous ERISA case. Rochow v.

Life Ins. Co. of N.A., 

, 429–30 (6th Cir. 2013) (concluding that a district court

did not abuse its discretion in ascertaining profits to be disgorged when unlawfully obtained

funds were commingled with defendant’s general assets by applying “the principle that

where funds are not traceable, an appropriate remedy is to order disgorgement of a

proportionate share of the wrongdoer’s profits”), rev’d on rehearing on other grounds 

(6th Cir. 2015) (en banc).

       Notwithstanding the proportionate-share-of-the-whole approach’s widespread

application, see 

(describing the proportionate-share-of-the-

                                             14
whole approach as “virtually universal”), a few courts took—and continue to take—other

approaches in determining whether, and to what extent, a defendant profited from the use

of unlawfully obtained, and mingled, money, see, e.g., Parke v. First Reliance Standard

Life Ins. Co., 

, 1009 (8th Cir. 2004) (holding, in ERISA breach of fiduciary

duty case, that the plaintiff was entitled to disgorgement of profits in the form of interest

because the fiduciary “‘gains’ from the wrongful withholding of the plaintiff’s benefits

even if the plaintiff does not prove specific financial profit. In particular, the defendant

receives a benefit from having control over the money”); In re Mowrey’s Estate, 232

N.W.82, 86 (Iowa 1930) (requiring executor, who commingled estate funds with personal

funds to obtain mortgages, to pay interest on commingled funds at the statutory rate).

       That the proportionate-share-of-the-whole approach appears to have been widely, if

not universally, embraced by courts and commentators does not necessarily mean,

however, that the district court was required to follow that approach in this case—the

question this Court must resolve. As to that question, Plaintiffs contend that the district

court was required to apply the proportionate-share-of-the-whole approach in this case for

two reasons.

       First, Plaintiffs argue that “[b]ecause ‘courts of equity must be governed by rules

and precedents no less than courts of law,’ . . . the Supreme Court has been insistent that

any time equity has already developed a specific rule for dealing with a recurring fact

pattern, equity courts are forbidden from ‘exercis[ing] [their] background equitable

powers’ . . . to engage in ‘ad hoc equitable departures.’” Appellants’ Br. 28 (second and



                                             15
third alterations in original) (quoting Lonchar v. Thomas, 

, 323–24, 327

(1996)). Lonchar, however, does not bear the weight Plaintiffs claim.

       Without question, Lonchar supports the proposition that, for a variety of compelling

reasons—predictability, uniformity, and fairness, to name a few—courts generally should

follow equitable rules, like the proportionate-share-of-the-whole methodology. 

(“[E]quitable rules that guide lower courts reduce uncertainty, avoid unfair

surprise, minimize disparate treatment of similar cases, and thereby help all litigants.”).

But Lonchar dealt with a meaningfully distinct question—“whether a federal court may

dismiss a first federal habeas petition for general ‘equitable’ reasons beyond those

embodied in the relevant statutes, Federal Habeas Corpus Rules, and prior precedents.” 

That question turned on a complex regulatory and statutory scheme that specifically

addressed the relevant issue and did not expressly confer equitable authority to resolve that

question. 

       By contrast, ERISA Section 502(a)(3), under which Plaintiffs seek relief, expressly

empowers courts to invoke their equitable authority and determine whether equitable relief

is “appropriate.”   29 U.S.C. § 1132(a)(3).       More significantly, the Supreme Court

subsequently recognized that, notwithstanding Lonchar’s statement that courts of equity

“must be governed by rules and precedents no less than the courts of law,” the “exercise of

a court’s equity powers . . . must be made on a case-by-case basis.” Holland v. Florida,

, 649–50 (2010) (emphasis added) (internal quotation marks omitted). “In

emphasizing the need for flexibility and avoiding mechanical rules . . . we have followed

a tradition in which courts of equity have sought to relieve hardships which, from time to

                                             16
time, arise from a hard and fast adherence to more absolute legal rules, which, if strictly

applied, threaten the evils of archaic rigidity.” 

(internal quotation marks

omitted). Accordingly, neither Lonchar nor Supreme Court precedent requires courts to

invariably follow equitable rules.

       Second, Plaintiffs argue that even if district courts generally retain discretion as to

the application of equitable rules, ERISA Section 502(a)(3) does not afford district courts

any such discretion. Appellants’ Br. 30–31. In support of their position, Plaintiffs

emphasize this Court’s holding in Pender that “[t]he Supreme Court has interpreted the

term ‘appropriate equitable relief,’ as used in Section 502(a)(3), to refer to ‘those categories

of relief that, traditionally speaking (i.e., prior to the merger of law and equity [in 1938])

were typically available in equity.’” 

(quoting CIGNA Corp. v.

Amara, 

, 439 (2011)). As detailed above, Plaintiffs are correct that their

proportionate-share-of-the-whole approach appears to have been the predominant way of

conducting an accounting of profits when unlawfully obtained funds were commingled

with other funds. But the language from Amara upon which Pender relied dealt with what

forms of “equitable relief” were available—e.g., restitution, disgorgement, accounting for

profits, etc.—not with whether a court was required to award a particular form of relief—

or calculate such relief in a particular way—the relevant question here.

       Both the language of the statute and case law contradict Plaintiffs’ claim that “courts

in ERISA cases cannot rely on their judgment to devise relief that is fair, reasonable, and

‘equitable’ in the particular circumstances of the case.” Appellants’ Br. 30–31 (emphasis

in original). To begin, whereas Pender, and the Supreme Court cases upon which it relied,

                                              17
focused on the meaning of “equitable relief” in Section 

, that

provision also requires that the award of such relief be “appropriate,” indicating that a court

has the power to deny such relief (even if it is a form of equitable relief available under

Section 502(a)(3)), if it deems such relief not “appropriate” under the particular facts of the

case. To that end, after concluding that the remedies sought in Amara were “equitable

relief,” the Supreme Court remanded the case because it was unclear “whether the District

Court will find it appropriate to exercise its discretion under § 502(a)(3) to impose that

remedy on remand.” 

(emphasis added).

       Likewise, in Griggs v. E.I. DuPont de Nemours & Co., this Court held that “even if

the redress sought by a beneficiary under ERISA Section 502(a)(3) is a classic form of

equitable relief, it must be appropriate under the 

(emphasis retained). And in McCravy v. Metropolitan Life Insurance Co., this Court

considered whether a breach of fiduciary action seeking equitable relief in the form of

estoppel and surcharge was available under ERISA Section 502(a)(3). 

, 182

(4th Cir. 2012). After concluding that estoppel and surcharge actions were “traditionally

available in [pre-merger] courts of equity”—and therefore available under Section

502(a)(3)—we remanded the case to the district court, stating: “Whether [the plaintiff’s]

breach of fiduciary duty claim will ultimately succeed and whether surcharge is an

appropriate remedy under Section [502(a)(3)] in the circumstances of this case are

questions appropriately resolved in the first instance before the district court.” 

82 (emphasis added). Accordingly, this Court has held that even if a form of equitable



                                              18
relief is available under Section 502(a)(3), a district court has discretion to deny such relief

if the court deems such relief inappropriate under the particular facts of the case.

       Other circuit courts have reached the same conclusion. For example, the Third

Circuit held that “the term ‘appropriate’ . . . confer[s] discretion on district courts, sitting

as courts of equity, to limit equitable relief by doctrines and defenses traditionally available

at equity.” Nat’l Sec. Sys., Inc. v. Iola, 

, 101–02 (3d Cir. 2012). Applying that

rule, the Third Circuit upheld a district court’s award of only partial disgorgement,

notwithstanding that the traditional equitable rule afforded full disgorgement, because

some of the unlawful payments at issue had been passed on to innocent third-parties. 

Security Systems court ruled that because courts sitting in equity are entitled

“to fashion relief tailored to the unique circumstances of a case,” the district court did not

abuse its discretion in limiting the disgorgement when it found that the facts of the case so

warranted. 

       Likewise, the Seventh Circuit has explained that “the fact that a transaction is

prohibited under ERISA does not necessarily mandate a remedy, although it is a very

dangerous area for trustees to explore, let alone attempt to exploit.” Etter v. J. Pease

Constr. Co., Inc., 

, 1009 (7th Cir. 1992) (emphasis added). “Rather, the

decision to impose a remedy lies within the court’s discretion and should be in tune with

the case’s realities.” 

marks omitted); see also Gearlds v. Entergy

Servs., Inc., 

, 452 (5th Cir. 2013) (“We leave to the district court the

determination whether Gearlds’s breach of fiduciary duty claim may prevail on the merits

and whether the circumstances of the case warrant the relief of surcharge.” (emphasis

                                              19
added)); McDonald v. Pension Plan of NYSA-ILA Pension Tr. Fund, 

, 161 (2d

Cir. 2003) (“[Section 502(a)(3)] does not require district courts to grant particular relief;

rather, it affords district courts the discretion to fashion appropriate equitable relief.”). By

contrast, Plaintiffs cite no authority, nor have we found any, holding that a district court is

barred from declining to award equitable relief under Section 502(a)(3)—even if the

requested form of equitable relief is available under that statute—if it determines the award

of such relief would be inappropriate under the facts of the case.

       Accordingly, ERISA Section 502(a)(3) did not require the district court to award

Plaintiffs relief based on the proportionate-share-of-the-whole methodology. Rather, the

district court retained discretion to consider other approaches in determining whether

equitable relief was “appropriate” under the particular facts of the case.

                                              B.

       Having concluded that the district court was not required to follow the

proportionate-share-of-the-whole approach in determining whether, and to what extent, the

Bank profited from the unlawful transfers, we next must decide whether the district court

permissibly exercised its discretion in determining equitable relief was not appropriate in

this case. Under the governing deferential standard of review, we uphold the district

court’s decision denying Plaintiffs equitable relief.

       The district court’s decision rested on extensive factual findings, none of which

Plaintiffs challenge on appeal as clearly erroneous.          Those factual findings reflect

contemporaneous Bank records maintained in the ordinary course of business outlining the

Bank’s Investment Strategy, which the district court found treated the transferred 401(k)

                                              20
balances differently than the other funds in the Pension Plan’s general account. Pender,

, at *8. The district court further found that the contemporaneously

documented Investment Strategy, to which the Bank adhered, required the Pension Plan

“to invest assets to fund [the transferred 401(k) account balances] in a higher concentration

of equities than participants invested their hypothetical accounts.” 

The district

court also found that the Bank contemporaneously tracked, on a monthly basis, the

performance of the transferred 401(k) balances, separate and apart from the performance

of the remaining funds in the Pension Plan’s general account. 

*13. And the

district court found—and Plaintiffs do not dispute—that the returns for the transferred

401(k) balances realized under “the Investment Strategy were . . . less than participants’

hypothetical returns.” 

       Based on these undisputed factual findings, the district court repeatedly asserted it

would not be “appropriate” to award Plaintiffs equitable relief under the proportionate-

share-of-the-whole approach because that approach would not measure whether any profits

accrued to the Bank “due to the transfer.” 

see also, e.g., 

(“[A]s a matter

of equity, the Court finds that [the proportionate-share-of-the-whole] methodology is

inappropriate and inferior to calculating profit based on the actual Investment Strategy

utilized with respect to the [transferred funds].”); 

Again emphasizing the

distinct, contemporaneously documented Investment Strategy for the transferred funds, the

district court further determined it would be inappropriate to apply the proportionate-share-

of-the-whole approach because “doing so would have the effect of being a penalty, and,

conversely, would create a windfall for Plaintiffs, because much of what would be captured

                                              21
as ‘profits’ under such a methodology would be investment returns the Plan would have

realized in any event regardless of the transfer.” 

       It was within the district court’s discretion to determine that awarding Plaintiffs

equitable relief using the proportionate-share-of-the-whole methodology would be

inappropriate in this case. The proportionate-share-of-the-whole approach was designed

to address situations in which a defendant mingles unlawfully obtained funds with money

of his own so that the “whole forms one indistinguishable mass[;] . . . it can no longer be

identified.” Scott, supra at 125.

       Here, the district court did not clearly err in determining that the extensive

contemporaneous evidence outlining the Investment Strategy for the unlawfully transferred

funds and separately tracking the performance of the funds invested under that strategy

made it possible to “identif[y]” the performance of the unlawfully mingled funds, 

application of the proportionate-share-of-the-whole methodology

inappropriate in this particular case, cf. Sheldon v. Metro-Goldwyn Pictures Corp., 

, 406 (1940) (“Where there is a commingling of gains, [the wrongdoer] must abide

the consequences [and disgorge all commingled gains], unless he can make a separation

of the profits so as to assure the injured party all that justly belongs to him.” (emphasis

added)).   Put differently, the extensive contemporaneous evidence identifying the

performance of the unlawfully commingled funds provided the district court with an

adequate factual basis to deviate from the proportionate-share-of-the-whole methodology,

which courts widely apply to assess whether, and to what extent, a wrongdoer profits from

unlawfully commingled funds.

                                             22
       Likewise, courts and commentators have cautioned against awarding a plaintiff

equitable relief, and disgorged profits in particular, to the extent doing so would amount to

a windfall or penalize a defendant. 

408 (explaining that “[e]quity is concerned

with making a fair apportionment so that neither party will have what justly belongs to the

other” and therefore does not permit “inflict[ing] an unauthorized penalty”); Griggs, 

, 385 (4th Cir. 2001) (stating that an ERISA plaintiff seeking equitable relief was

“not entitled to a windfall”); Restatement (Third) of Restitution and Unjust Enrichment §

51 (2011) (“[T]he unjust enrichment of a conscious wrongdoer . . . is the net profit

attributable to the underlying wrong. The object of restitution in such cases is to eliminate

profit from wrongdoing while avoiding, so far as possible, the imposition of a penalty.”);

1 Dobbs, supra § 4.5(3), at 642 (“Even the willful wrongdoer should not be made to give

up that which is his own; the principle is disgorgement, not plunder.”). In light of these

authorities, which the district court explicitly referenced, the district court did not abuse its

discretion in determining that use of the proportionate-share-of-the-whole methodology to

award Plaintiffs equitable relief would amount, in this specific case, to a windfall to

Plaintiffs, and inappropriately penalize the Bank. Pender, 

, at *14, *16,

*19. In particular, there is no dispute that as a result of the transfers and payments required

by the IRS closing agreement, Plaintiffs’ current 401(k) account balances are at least as

large as they would have been had the funds in their accounts actually been invested in

accordance with their notional allocations. Pender, 

, at *10. And the

contemporaneous records introduced by the Bank, which separately tracked the

performance of the transferred funds, provided the district court with a factual basis to

                                               23
determine that the Bank did not profit from the transaction and that any further payment to

Plaintiffs would serve only to penalize the Bank.

       The extensive evidence of the distinct, contemporaneously documented Investment

Strategy credited by the district court sets this case apart from the “hypothetical example”

set forth by our colleague in dissent. See post at 27–28. Under that example, “the investor

never explains how he could maintain a separate investment strategy that benefits only his

share of the commingled funds.” 

district court credited the Bank’s evidence

documenting the separate investment strategy and establishing that the returns accruing the

unlawfully commingled funds invested pursuant to that strategy could be separately

tracked. Plaintiffs have not challenged those factual findings as clearly erroneous, and

therefore we have no basis to conclude that the district court abused its discretion in relying

on those findings to deny Plaintiffs equitable relief.

       That we conclude that the district court did not abuse its discretion in determining

that equitable relief was not appropriate in this case does not mean that we necessarily

would have rendered the same judgment were we addressing the question in the first

instance. Having access to the additional funds obtained through the unlawful transfers

may have impacted the Bank’s investment strategy for the Pension Plan as whole, even if

it sought to hedge its risk by mirroring the 401(k) beneficiaries’ equity investments,

meaning that the benefits accruing to the Pension Plan may not have been entirely

independent of the losses related to the 401(k) balances, as the attribution approach

assumed. Additionally, as the Eighth Circuit recognized, a “defendant receives a benefit

from having control over the money” even if a profit cannot be demonstrated. Parke, 

. And the attribution approach advanced by the Bank, and relied on by the

district court, is generally applied in situations when a defendant’s skill or ingenuity

independently contributed to profits obtained, in part, by a defendant’s unlawful use of

another’s property. See, e.g., Sheldon v. Metro-Goldwyn Pictures 

–

08. By contrast, here the Bank maintains that its lack of skill in investing the transferred

funds—as evidenced by the Bank’s Investment Strategy yielding significantly lower

returns than Plaintiffs’ notional allocations—warranted application of the attribution

approach.

       Nonetheless, in light of the extensive contemporaneous records documenting the

Investment Strategy for the transferred funds and tracking the performance of those

funds—which provided the district court with an adequate factual basis to find that the

performance of the transferred funds could be separately identified—it was within the

district court’s discretion to decline to award equitable relief based on the proportionate-

share-of-the-whole approach.

       In affirming the district court’s exercise of its discretion, we do not—as our

dissenting colleague suggests—“depart” from our prior holding that if “‘Section

204(g)(1)[] . . . is to have any teeth, the available remedies must be able to reach situations

like the one this case presents, i.e., where a plan sponsor benefits from an ERISA violation,

but plan participants—perhaps through luck or agency intervention—suffer no monetary

loss.’” Post at 29 (quoting 

, 364-65). On the contrary, we simply

determine that the district court did not clearly err in finding that—based on the

contemporaneous records documenting the Investment Strategy for the transferred funds

                                              25
and tracking the performance of those funds—the Bank did not benefit from the violation.

In determining that the district court did not clearly err, we in no way retreat from this

Court’s previous holding that Section 502(a)(3) entitles plan participants to an accounting

for profits attributable to an ERISA violation, even if the participants suffered no monetary

harm from the violation. ERISA does not allow a plan sponsor to wrongfully use plan

participant funds for the sponsor’s benefit. In such circumstances, the plan sponsor must

disgorge its ill-gotten benefit to plan participants.



                                              III.

       For the foregoing reasons, we affirm the judgment of the district court.

                                                                                AFFIRMED




                                              26
BARBARA MILANO KEENAN, Circuit Judge, dissenting:

       I agree with the majority that a court examining an ERISA violation is not required

to apply a proportionate-share-of-the-whole approach when a wrongdoer has profited from

the use of commingled funds. A court’s exercise of its equitable powers must be made on

a fact-specific basis, rather than by imposing absolute outcomes without regard to factual

context. See Holland v. Florida, 

, 649-50 (2010); see also 29 U.S.C.

§ 1132(a)(3) (allowing participant to bring claim for “appropriate equitable relief”).

However, based on the Bank’s particular conduct here, I depart from the majority’s holding

allowing the Bank to profit by using the plaintiffs’ money for the Bank’s own purposes. In

my view, the district court plainly abused its discretion in permitting the Bank to profit

from its own misconduct, which even the Bank concedes was illegal.

       A simple example easily illustrates the Bank’s self-dealing logic, which the majority

accepts today. Imagine that you agree to give $100 to an investor who holds $900 of other

funds, so that he can invest the commingled funds of $1,000 and guarantee you a minimum

return on your investment of 5%. When the $1,000 in commingled funds yields an overall

25% gain, for a total of $1,250, you naturally expect to earn $25 in profit on your

investment of $100, yielding a total payment to you of $125.

       To your dismay, however, the investor pays you only the guaranteed $105. He tells

you that, unfortunately, his investment strategy with respect to your $100 share of the

single pot of commingled funds failed miserably in the market, causing him to suffer a loss

of 10% on your $100 investment. And, astoundingly, he says that he is entitled to all the

remaining profit from using your commingled funds because his separate strategy for the

                                            27
other commingled funds in the single pot was far more successful. Remarkably as well,

the investor never explains how he could maintain a separate investment strategy that

benefits only his share of the commingled funds, but not yours. Instead, the investor, who

is far more sophisticated and powerful in the marketplace than you are, simply tells you

how fortunate you are to have received the guaranteed 5% gain, declares victory, and

pockets the entire remaining profit from investing the commingled funds.

       This hypothetical example captures the essence of the Bank’s wrongdoing and

windfall in this case. 3 Thus, the example illustrates the fallacy of the district court’s

reasoning when it rejected the proportionate-share-of-the-whole approach, a longstanding

equitable principle. As we explained in the context of a trust in MacBryde v. Burnett, 

(4th Cir. 1942), when “funds are mingled with personal funds of a trustee, the

whole is impressed with a trust until separation of the trust property can be made,” and the

beneficiaries ordinarily should receive “a proportionate part of the profits realized.” 

       When there has been commingling of funds by a wrongdoer, most courts have

applied the proportionate-share-of-the-whole approach when formulating appropriate

equitable relief. See Henkels v. Sutherland, 

, 302 (1926); Provencher v.


       3
          I recognize that the plaintiffs here were informed prior to the transfer that they
would earn either the greater of their hypothetical investments or the original balance of
their transferred assets. But this initial expectation does not bear on fashioning a remedy
for the Bank’s wrongdoing. The issue before us is whether the district court abused its
discretion in rejecting the plaintiffs’ request for disgorgement of the Bank’s profits
obtained from using the plaintiffs’ assets.


                                             28
Berman, 

, 570 (1st Cir. 1983); Bartlett & Co., Grain v. Commodity Credit

Corp., 

, 409 (8th Cir. 1962); Marcus v. Otis, 

, 150 (2d Cir. 1948).

And, as the majority observes, the Sixth Circuit has applied this approach in the context of

an ERISA violation involving commingled funds. Rochow v. Life Ins. Co. of N. Am., 

, 429-30 (6th Cir. 2013), rev’d on rehearing on other grounds, 

(6th

Cir. 2015) (en banc). Nevertheless, in the face of this compelling case law and our own

decision in MacBryde, the majority swims against the strong tide of equity.

       In our previous decision in this case, we remanded the matter for the district court

to consider the plaintiffs’ claim under ERISA Section 502(a)(3) for an accounting of profits

after the Bank unlawfully transferred the plaintiffs’ 401(k) Plan investments (the

transferred assets) into the Bank’s Pension Plan. Pender v. Bank of Am. Corp., 

, 358, 364-65 (4th Cir. 2015). We explained that in filing suit against the Bank, the

plaintiffs sought “profits generated using assets that belonged to them,” reduced by the

amount the Bank already had paid the plaintiffs pursuant to the IRS settlement. 

And, importantly, we emphasized that if “Section 204(g)(1)[ ] . . . is to have any teeth, the

available remedies must be able to reach situations like the one this case presents, i.e.,

where a plan sponsor benefits from an ERISA violation, but plan participants—perhaps

through luck or agency intervention—suffer no monetary loss.” 

It seems that

the majority now has departed from our prior admonition.

       Somehow, the Bank convinced the district court on remand that the plaintiffs were

lucky to have been paid anything, because the Bank’s “investment strategy” with respect

to the plaintiffs’ portion of the commingled funds had failed. However, the strategy only

                                             29
failed with respect to amounts equal to the transferred assets. See Nickel v. Bank of Am.

Nat’l Trust & Sav. Ass’n, 

, 1138 (9th Cir. 2002) (discussing disgorgement

and stating that “[m]oney is fungible. Once in the bank’s accounts . . . the specific sums

taken from the trusts could never be identified again.”). The undisputed evidence showed

that nearly $3 billion of transferred assets were pooled, indistinguishably, with the Pension

Plan assets into one “pot” worth about $9 billion. And this “pot” profited by a margin of

28.6% during the relevant period, despite one failed aspect of the overall investment

strategy.

       Under its strategy, the Bank had intended to enhance the Plan’s “pot” by investing

amounts equal to the transferred assets more heavily in equities than the plaintiffs

themselves would have invested. In that case, the Bank would pay the plaintiffs guaranteed

minimum investment earnings, and pocket the additional earnings. See 

-59. Because the equities failed to perform as expected, the Bank claimed that the

plaintiffs’ portion of the commingled funds had shrunk, even though the “pot” actually

profited by about $379 million, including accrued interest.

       In concluding that these profits from the commingled funds were not attributable to

the Bank’s “transfer strategy,” the district court answered the wrong question. See Pender

v. Bank of Am. Corp., No. 3:05-CV-00238, 

, at *4 (W.D.N.C. Apr. 27,

2017). The court should have focused instead on the question articulated in MacBryde,

namely, what was the proportionate share of the profits made by investing the plaintiffs’

portion of the 

.



                                             30
       Although the district court appeared to couch its decision as a credibility

determination, Pender, 

, at *4, in reality, the decision merely reflected

the court’s rejection of an established equitable remedy in favor of preserving the Bank’s

profit margin. Accordingly, we are not presented with an issue of competing facts that we

review for clear error. See Dixon v. Edwards, 

, 710 (4th Cir. 2002) (reviewing

a district court’s award of equitable relief for abuse of discretion and findings of fact for

clear error). Moreover, “even if a district court applies the correct legal principles to

adequately supported facts, the discretion of the trial court is not boundless and subject to

automatic affirmance,” when we have “a definite and firm conviction that the court below

committed a clear error of judgment” in its conclusion. Westberry v. Gislaved Gummi AB,

, 261 (4th Cir. 1999); see Huskey v. Ethicon, Inc., 

, 158 (4th Cir.

2017). In my view, the district court committed a clear error of judgment and abused its

discretion in refusing to order the Bank to disgorge the wrongful gains the Bank reaped.

       As we explained in our prior decision, “ERISA borrows heavily from the language

and the law of trusts.” 

. And, under those principles, the Bank

should be required to pay the plaintiffs the profits “which in equity and good conscience

belonged” to the plaintiffs, rather than to use those profits to enhance the Bank’s bottom

line. See 

(quoting 1 D. Dobbs, Law of Remedies § 4.3(5), p. 608 (2d ed. 1993)).

Therefore, I cannot abide the decision by the district court and the majority to allow the

Bank to profit lavishly from its wrongful use of the plaintiffs’ money.




                                             31