After owners of a closely held corporation sold the company to its Employee Stock Ownership Plan ("ESOP"), a participant in the ESOP brought this action. The participant contended that the trustee chosen for the ESOP by the corporation breached its fiduciary duties to the ESOP and overpaid for the stock - improperly enriching the corporation's owners at the expense of its employees.

Following a multi-day bench trial, the district court issued detailed findings of fact concluding that the trustee had indeed breached its fiduciary duties, causing the ESOP to overpay for the corporation's stock by $ 29,773,250. The court entered judgment for the ESOP in that amount and awarded attorneys' fees to the participant's counsel. These appeals and cross-appeals followed. As explained within, we affirm the careful findings of the district court.

I.

To facilitate understanding of the issues here, we begin with the governing legal principles and background facts that gave rise to this suit. The parties do not challenge these principles or facts. All are more fully set forth in the comprehensive district court opinions, upon which we rely throughout. See Brundle v. Wilmington Tr. N.A. , 241 F.Supp.3d 610 (E.D. Va. 2017) (" Brundle I "); Brundle v. Wilmington Tr. N.A. , 258 F.Supp.3d 647 (E.D. Va. 2017) (" Brundle II ").

A.

The Employee Retirement Income Security Act of 1974 (ERISA) allows an employer to create an ESOP, an employee pension plan that invests primarily in the employer's stock. The employer makes contributions to the plan that are used to purchase stock in the employer's company. Because - and only because - an ESOP contribution qualifies as employee compensation, an employer can deduct the total value of its ESOP contribution from its income tax liability as an ordinary business expense. 26 U.S.C. § 404 ; 26 C.F.R. § 1.404 (a)-1(b). 1 In this way, an ESOP benefits both employees and employers by providing deferred compensation to the former and a valuable tax deduction to the latter.

ERISA imposes duties and obligations on all pension plan fiduciaries, including those of ESOPs. These duties "ensure that employees will not be left empty-handed once employers have guaranteed them certain benefits." Lockheed Corp. v. Spink , 517 U.S. 882 , 887, 116 S.Ct. 1783 , 135 L.Ed.2d 153 (1996). One such provision prohibits the fiduciary of any ERISA plan from causing a "sale or exchange ... of any property between the plan and a party in interest." 29 U.S.C. § 1106 (a)(1)(A). Absent a statutory exception, this provision would ban ESOPs because their creation necessarily requires the ESOP to purchase stock from its sponsoring employer, which is a party in interest. Congress, however, has carved out an exception to this prohibition to permit the creation of an ESOP if the stock purchase meets certain conditions. See 29 U.S.C. § 1108 (e).

To protect employees from losing the value of their earned retirement savings, the exception to the ERISA ban on party-in-interest transactions requires that an ESOP pay no more than "adequate consideration" for the employer's stock. Id. § 1108(e)(1). If an employer's "stock was not worth what the ESOP paid for it," then the ESOP paid more than adequate consideration and "the ESOP and its participants suffered a loss under ERISA." See Reich , 990 F.Supp. at 961.

ERISA does not define what constitutes "adequate consideration" under the § 1108(e) exception; the Department of Labor (DOL) has proposed, but never enacted, regulations doing so. 2 Although courts look to these regulations for guidance, the focus of the adequate-consideration inquiry rests on the conduct of a fiduciary, as judged by ERISA's "prudent man" standard of care. See Perez v. Bruister , 823 F.3d 250 , 263 (5th Cir. 2016) ; Henry III , 445 F.3d at 619 ; Chao , 285 F.3d at 437 ; Howard v. Shay , 100 F.3d 1484 , 1489 (9th Cir. 1996).

Under this standard, "ESOP fiduciaries are subject to the duty of prudence just as other ERISA fiduciaries are." Fifth Third Bancorp v. Dudenhoeffer , 573 U.S. 409 , 134 S.Ct. 2459 , 2467, 189 L.Ed.2d 457 (2014). Thus an ESOP fiduciary, like any other ERISA fiduciary, must "discharge his duties ... with the care, skill, prudence, and diligence under the circumstances then prevailing that a prudent man acting in a like capacity and familiar with such matters would use in the conduct of an enterprise of a like character and with like aims." 29 U.S.C. § 1104 (a)(1)(B).

Although these fiduciary duties "draw much of their content from the common law of trusts ... ERISA's standards and procedural protections partly reflect a congressional determination that the common law of trusts did not offer completely satisfactory protection." Tatum v. RJR Pension Inv. Comm. , 761 F.3d 346 , 357 (4th Cir. 2014) (internal quotation marks omitted). Courts apply the "prudent man rule ... bearing in mind the special nature and purpose of employee benefit plans." Id. (internal quotation marks and alterations omitted). For this reason, "[t]he fiduciary obligations of the trustees to the participants and beneficiaries [of an ERISA] plan are ... the highest known to the law." Id . at 356 (alterations in original) (quoting Donovan v. Bierwirth , 680 F.2d 263 , 272 n.8 (2d Cir. 1982) ).

Because an ESOP fiduciary that raises an affirmative defense under the § 1108(e) exception seeks to avoid ERISA liability for an otherwise prohibited transaction, the fiduciary bears the burden of proving by a preponderance of the evidence that the sale was for adequate consideration. See Elmore v. Cone Mills Corp. , 23 F.3d 855 , 864 (4th Cir. 1994) (en banc). "This burden is a heavy one." Shay , 100 F.3d at 1488 .

B.

With these principles in mind, we turn to the facts of this case.

Since its inception, Constellis Group, Inc., the closely held parent company of a group of private security subsidiaries, has offered some form of deferred compensation to its employees, who are primarily retired members of the U.S. Armed Forces. Brundle I , 241 F.Supp.3d at 614 . After initially offering a stock option program, Constellis replaced that program with a profit-sharing plan in 2010. Id. In mid-2013, looking for an "exit strategy" after having twice tried and failed to effectuate a sale of the company, the owners of Constellis (hereinafter "the Sellers") 3 began exploring the possibility of creating an ESOP to purchase Constellis. Id. at 615 .

To that end, Constellis retained CSG International, an investment banking firm. CSG reported that sale of Constellis to an ESOP would not only enable the Sellers to liquidate their shares but would save them 23.8% in federal capital gains taxes. Id. Although noting that an ESOP would also benefit employees, CSG emphasized that selling Constellis to an ESOP provided the best option to optimize the Sellers' after-tax cash return. See, e.g. , id. at 616 (noting that CSG reported to the Sellers that, having "looked at every possible scenario," it had not seen "any possible alternative that produce[d] more after tax cash" than the proposed ESOP (internal quotation marks omitted)).

CSG suggested that Constellis create what several trial witnesses characterized as a "unique" ESOP structure to effectuate the sale. This unique structure would permit the Sellers to retain de facto control of Constellis, even after the ESOP purchased all of the company stock. Rather than sell 100% of their ownership interest to the ESOP, as is the usual practice, the Sellers would sell 90% of their shares to the ESOP and then exchange the remaining 10% for "equity-like" warrants. These warrants would entitle the Sellers to buy back equity in Constellis from the ESOP at a designated price and guarantee the Sellers a majority on the board of directors. Id. at 626 .

To fund the ESOP's purchase, Constellis would contribute 24% of the purchase price and the ESOP would borrow the remainder from Constellis and the Sellers themselves. Id. at 625 . In late September 2013, Constellis decided to move forward with CSG's proposal to create this unique ESOP. Id. To maximize the Sellers' tax savings, Constellis had to complete the sale by the end of the calendar year. Id. at 622 .

On the recommendation of CSG, Constellis engaged Wilmington Trust, N.A., to serve as trustee for the ESOP. Id. at 616 . In its role as trustee, Wilmington hired Stout Risius Ross (SRR) to be the financial advisor on the ESOP's purchase of Constellis. Id. at 617 . CSG, Wilmington, and SRR maintained significant long-term business relationships, having worked together on more than twenty ESOP deals. Id. at 643 (calling "the ESOP world ... a very incestuous community") (internal quotation marks omitted); id. at 617 .

On November 12, 2013, SRR submitted a draft valuation of Constellis stock to Wilmington (the "SRR Report"). The SRR Report concluded that the fair market value for a single share of Constellis stock lay between $ 3,865 and $ 4,600 - resulting in a rounded median price per share of $ 4,235. This placed the company's worth, or its "enterprise value," within a range of $ 275 to $ 330 million. Id . at 620-21. On November 14, Wilmington met with analysts from SRR to discuss its report and valuation. At the meeting's conclusion, Wilmington set out to negotiate the ESOP's purchase of Constellis with authorization to offer a share price of $ 3,900 to $ 4,235. Id . at 622.

The next day, Wilmington submitted an opening bid on behalf of the ESOP for $ 3,900 per share. Id. at 642 . CSG, on behalf of Constellis, countered with $ 4,350 and suggested to Wilmington that the ESOP and Constellis agree on a price first and then adjust the terms of the agreement at a later date. Wilmington agreed and raised its own bid to $ 4,100. CSG then countered with $ 4,250. By the end of the day, Wilmington and CSG had settled on a share price of $ 4,235 - the very top of Wilmington's authorized negotiation range. The negotiations lasted little more than five hours.

The ESOP issued a tender offer for the shares on November 18 with a closing date of December 20. The parties finalized the term sheet on November 22. Between the conclusion of negotiations and the closing date for the transaction, SRR revised its valuation of the draft enterprise range downward from $ 275-$ 330 million to $ 275-$ 325 million, but the per share purchase price remained $ 4,235. On December 19, after meeting with SRR for half an hour to discuss SRR's valuation, Wilmington approved the purchase. Id. at 624 . On December 20, the ESOP's purchase of Constellis closed for a per share price of $ 4,235. Id. at 625 .

Subsequently, Tim P. Brundle, a former Constellis employee and ESOP participant, brought this action contending that Wilmington caused the ESOP to enter into a transaction prohibited under ERISA § 1106. 4 Brundle alleged that the $ 4,235 per share paid by the ESOP for Constellis was too high and resulted in the ESOP overpaying the Sellers for the stock. The district court initially held that the 2013 purchase of Constellis by the ESOP constituted a prohibited transaction under § 1106(a)(1). Brundle v. Wilmington Tr., N.A. , 2016 WL 6542718 , at *12 (E.D. Va. Nov. 3, 2016). Wilmington conceded as much before the district court but raised the adequate-consideration affirmative defense available under § 1108(e)(1). Id.

Following a bench trial, the district court issued meticulous factual findings and concluded on the basis of those findings that Wilmington had not established the § 1108(e) affirmative defense. Rather, the court found that Wilmington had violated its fiduciary duty to the ESOP. The district court concluded that Wilmington's fiduciary failures resulted in the ESOP overpaying for the Constellis stock by $ 29,773,250, and so awarded that amount to the ESOP as damages. Brundle I , 241 F.Supp.3d at 613 .

Brundle then moved for attorneys' fees. The court first ordered Wilmington to pay attorneys' fees pursuant to ERISA's fee-shifting provision, 29 U.S.C. § 1132 (g)(1), in the amount of $ 1,819,631.11. Brundle II , 258 F.Supp.3d at 663-69 . It subsequently awarded an additional $ 1.5 million in fees from the damages award.

On appeal, Wilmington challenges the court's liability and damages determinations and its award of $ 1.5 million in non-statutory attorneys' fees. Brundle cross-appeals, challenging as inadequate the same portion of the attorneys' fees award.

II.

Wilmington initially and principally contends that the district court erred in concluding that it violated ERISA. Following a bench trial, we review a district court's factual findings for clear error and its legal conclusions de novo. Nat'l Fed'n of the Blind v. Lamone , 813 F.3d 494 , 502 (4th Cir. 2016).

At the outset of both its principal brief and its oral argument, Wilmington emphasized that there is no claim in this case that it acted in bad faith. But of course, an ESOP participant seeking recovery from its fiduciary need not prove that the fiduciary acted in bad faith. Rather, an ESOP fiduciary is liable to the plan participants if it breached its fiduciary duties, i.e. , failed to act " solely in the interest of the participants," with the care, skill, prudence, and diligence used by a "prudent man acting in a like capacity." 29 U.S.C. § 1104 (a) (emphasis added). Thus our focus is not on Wilmington's motives (good or bad) but on whether it acted "solely in the interest" of the plan participants, id. , and "engaged in a reasoned decisionmaking process, consistent with that of a prudent man in like capacity," DiFelice v. U.S. Airways, Inc. , 497 F.3d 410 , 420 (4th Cir. 2007) (internal quotation marks omitted); accord Tatum , 761 F.3d at 356 . As the Fifth Circuit has put it, "a pure heart and an empty head are not enough." Donovan v. Cunningham , 716 F.2d 1455 , 1467 (5th Cir. 1983).

A.

The district court found that because Wilmington failed to prove by a preponderance of the evidence that the ESOP's purchase of Constellis was for adequate consideration - and no more than adequate consideration - the transaction failed to qualify for the § 1108(e) affirmative defense. 5 In challenging this holding, Wilmington heavily relies on the valuation of Constellis stock submitted by its advisor, SRR.

Expert advice, like an advisor's independent valuation, can of course serve as evidence of prudence in the discharge of an ESOP trustee's duties under § 1108(e). See Bussian v. RJR Nabisco, Inc. , 223 F.3d 286 , 300-01 (5th Cir. 2000). But such advice "is not a magic wand that fiduciaries may simply wave over a transaction to ensure that their responsibilities are fulfilled." Cunningham , 716 F.2d at 1474 . Rather, a plan trustee must at least show that it (1) "investigate[d] the expert's qualifications," (2) "provide[d] the expert with complete and accurate information," and (3) "[made] certain that reliance on the expert's advice was reasonably justified under the circumstances." Shay , 100 F.3d at 1489 . Because ERISA demands a high level of scrutiny from fiduciaries, a trustee must prove that it considered these factors "and any others relevant under the particular circumstances it faced at the time of the decision," and took care to avoid any identified conflicts of interest. See Bruister , 823 F.3d at 263 (internal quotation marks omitted).

The district court properly recognized that the above legal principles guided its inquiry. The court then issued extensive factual findings in support of its conclusion that Wilmington had "not demonstrated that its reliance on SRR's [valuation] report was 'reasonably justified' in light of all the circumstances because [Wilmington] ha[d] not shown that it thoroughly probed the gaps and internal inconsistencies in that report." Brundle I , 241 F.Supp.3d at 634 .

The court found Wilmington's decision-making process inadequate in four major respects. Id. It explained that "[a]lthough these failings independently" might not be dispositive, cumulatively they demonstrated that "Wilmington failed to meet its duty" under ERISA, id. , which, of course, includes the obligation to act "solely in the interest" of plan participants, 29 U.S.C. § 1104 (a). Wilmington challenges the district court's exhaustive findings of fact and conclusions of law as to each of the four designated failures.

1.

First, Wilmington maintains the district court erred in finding fault with Wilmington's failure to investigate SRR's omission from its report of another, much lower, valuation of Constellis stock. That valuation, termed the "McLean Report" by the parties, had been completed just months earlier by the McLean Group, an investment bank that had regularly provided annual share valuations for Constellis. Brundle I , 241 F.Supp.3d at 619, 634-35 .

Wilmington's trial witnesses contended that consideration of the McLean Report "would not have been helpful because it was prepared with a different purpose in mind - the valuation of a single share" for stock options, rather than the valuation of a controlling interest for sale. Id. at 634-35 . But the district court, relying on the uncontradicted testimony of a "highly credible" neutral expert, found that the two reports used "essentially the same methodology" despite being prepared for different purposes. Id. at 635 . Using that same methodology, the 2013 McLean Report estimated the per share value of Constellis stock as $ 1,838.11, while SRR estimated its median per share value as $ 4,235. Id. at 619 .

Yet the SRR Report, which guided Wilmington as it evaluated the ESOP transaction, did not even mention the McLean Report. And, although trial testimony established that at least one Wilmington official knew of the McLean Report, no one from Wilmington ever asked about it or questioned the vast discrepancy between two valuations of the same company performed mere months apart. Id . at 634-35. The district court concluded that a prudent fiduciary in Wilmington's position would have inquired as to SRR's omission. This was particularly so given that Constellis officers testified that the company's business prospects had not dramatically changed in the period between the two reports, and the SRR Report had referenced a different "valuation" older than McLean's. 6

Relying on trial evidence, the district court concluded that even a cursory review of the McLean Report would have raised questions as to whether the share valuation in the SRR Report was too optimistic. The court did not find that the McLean Report necessarily offered a more accurate picture of Constellis's true value, but simply that a reasonably prudent fiduciary would have asked more questions. The court found that the answers to these questions were "precisely the kinds of information that Wilmington, as trustee, should have sought and evaluated" when considering the valuation of Constellis stock. Id. at 636 . We see no reason to conclude otherwise.

2.

The second major fault identified by the district court concerned Wilmington's failure to adequately probe the reliability of financial projections prepared by Constellis management and used by SRR. Id. at 636-38 . The court found that "the record show[ed] that there were several red flags indicating that these projections were inflated, which Wilmington either ignored or downplayed." Id. at 636 . Wilmington offers both general and specific challenges to this finding.

First, Wilmington argues that Constellis management's financial projections were, in fact, reliable and so required no probing. In support of this view, Wilmington contends that because the McLean Report also utilized management's financial projections to arrive at its own valuation of Constellis - a valuation the court treated as relevant - the court should have found the projections that the SRR Report used were also reliable.

But although both McLean and SRR relied on financial projections prepared by Constellis management, they did not rely on the same financial projections. Id. at 638 . The projections that Constellis provided to SRR forecasted revenue over a much longer time period, id. at 637-38 , making them less accurate and less reliable. Further, Constellis management specifically prepared this set of projections for the purposes of SRR's valuation in connection with the ESOP sale. This created a risk that Constellis management might inflate its projections in response to incentives unique to the ESOP sale. Id. at 636 . In light of these distinctions, the district court's differing assessments of the two reports' comparative reliability were not inherently contradictory.

At a more granular level, Wilmington maintains that each of the red flags that the district court identified in Constellis's financial projections are "non-existent." Wilmington Opening Br. 33-34, 38-39. Wilmington asserts largely factual, and universally unpersuasive, arguments in support of this view.

Wilmington challenges the district court's conclusion that Wilmington should have been more critical of Constellis management's financial projections because management had a financial incentive to inflate the purchase price. The court found that this initial red flag should have alerted Wilmington to probe the SRR valuation of Constellis. The district court's finding on this issue primarily rested on the prospect of management receiving 5% of the purchase price in cash bonuses - a factual determination well supported by the record. See Brundle I , 241 F.Supp.3d at 636 .

The district court also identified as a red flag an ongoing government investigation into accounting and record-keeping errors by Constellis, which the court concluded "should have raised questions about the reliability of Constellis'[s] ... projections." Id. Again, contrary to Wilmington's suggestion, we see no basis for holding this finding clearly erroneous. A government investigation into errors of this nature surely creates a question as to the reliability of a company's financial projections, which a reasonably prudent fiduciary representing the buyers of the company should probe further. Cf. Keach v. U.S. Tr. Co. , 419 F.3d 626 , 633-37 (7th Cir. 2005) (finding ESOP trustee met fiduciary duty when it "analyzed exhaustively [the company's] financials, had conducted independent valuation analyses and had challenged the financial forecasts made by [the company's] management as too optimistic").

Still another red flag to which Wilmington, in the district court's view, did "not adequately respond was the riskiness of Constellis'[s] contract concentration." Brundle I , 241 F.Supp.3d at 637 . The district court believed that this risk called for Wilmington to apply further scrutiny to the company's financial projections. The court found, and Wilmington does not dispute, that at the time of the ESOP sale, about 70% of Constellis's revenue depended on just two government contracts. Id. at 637 . Yet Wilmington adopted a valuation that assumed Constellis stock posed significantly less risk than the stock market as a whole. Again relying on the trial testimony of a "neutral" and "highly credible" third party expert, the court explained its conclusion that Constellis's industry peers - namely, government contractors - are certainly "no less risky than the market as a whole." Brundle II , 258 F.Supp.3d at 663 . And again, the record offers ample support for this finding. See Brundle I , 241 F.Supp.3d at 637 ; Brundle II , 258 F.Supp.3d at 662-63 . 7

Wilmington also maintains that the district court unfairly characterized Constellis's "multiple sets of projections" as a red flag. Wilmington claims that the district court ignored the fact that SRR advised Wilmington that the market and Constellis had changed throughout 2013. But the court adequately addressed this argument, see Brundle I , 241 F.Supp.3d at 635-36 , and we defer to its factual finding that the changes in the market and Constellis's business prospects throughout 2013 did not justify multiple sets of projections.

Last, Wilmington contends that we should reverse the district court due to its initially erroneous understanding of the warranties Wilmington obtained to indemnify the ESOP against the ongoing government investigation. At first, the court incorrectly found that only Constellis had indemnified the ESOP, thus creating an illusory form of indemnification. But in denying Wilmington's Rule 59 motion, the court candidly addressed that error and acknowledged that, in fact, both Constellis and the Sellers had indemnified the ESOP. Brundle II , 258 F.Supp.3d at 656 . Nonetheless, the court found that its "corrected understanding [did] not alter [its] conclusions about ... Wilmington's inadequate investigation into the soundness of Constellis'[s] financial projections." Id . The court explained that, "[e]ven omitting its erroneous view of the Sellers' ... warranties," Wilmington failed to rebut the many red flags discussed above, each of which independently called for more robust due diligence by Wilmington. Id. at 656-57 . For the reasons we have explained above, we find no clear error in any of these findings and so reject this final argument from Wilmington as to Constellis management's financial projections.

3.

The district court found that Wilmington's third major failure in ensuring that its reliance on the SRR valuation was reasonably justified occurred when it "did not investigate the appropriateness of applying a 10% control premium" to Constellis stock and "fail[ed] to discount the [valuation] for lack of control." Brundle I , 241 F.Supp.3d at 638 .

Purchasers will generally pay more for "rights associated with control of the enterprise." Estate of Godley v. Comm'r , 286 F.3d 210 , 214 (4th Cir. 2002). We have defined "control" as "an interest which allows the shareholder to unilaterally direct corporate action, select management, decide the amount of distribution, rearrange the corporation's capital structure, and decide whether to liquidate, merge, or sell assets." Id. at 215 (internal quotation marks omitted).

The district court recognized that the ESOP had "powers beyond those of an ordinary shareholder." Brundle I , 241 F.Supp.3d at 627 n.23. But the court nonetheless found that the "ESOP essentially had no power to control Constellis" as a result of the transaction's unique structure. Id. at 638-39 . Accordingly, the court concluded that Wilmington erred in failing to investigate the SRR valuation, which not only failed to discount the stock for lack of control, but even added a 10% control premium to the value of the Constellis stock. The record supports both findings.

The Sellers retained the power to appoint a majority of the Constellis board, a key indicator of control. See id. at 626 . Further, the ESOP governing plan document and the ESOP trust agreement each required Wilmington to vote its shares as the Constellis board (not the ESOP) directed. See id. at 626-27 . Wilmington maintains that the ESOP could independently exercise its voting powers because the plan documents included a clause nullifying any provision that violated ERISA. But Constellis's general counsel testified at trial that the ESOP could only contravene the plan documents and exercise its independent voting authority by filing a lawsuit - the same limited relief available to a minority shareholder. The district court thus concluded that Wilmington breached its fiduciary obligation in failing to question SRR's failure to discount for lack of control. See id. at 639 . Given the record evidence that the ESOP's unique ownership structure was intentionally designed to maintain the Sellers' control over Constellis even after selling their shares, the district court's conclusion hardly constitutes clear error.

Wilmington next contends that even if the ESOP lacked control, a 10% control premium was justified by the fact that the ESOP owned 100% of Constellis shares and obtained "elements of control." Id. at 647 . Wilmington argues that SRR reasonably applied a "modest" 10% control premium given that market control premiums are often between 35-40%. But in light of the limited rights provided to the ESOP, we cannot conclude that the district court erred in finding that Wilmington should have done more to challenge the use of any control premium. The SRR Report "devoted just one sentence to articulating its decision," yet Wilmington never asked SRR about its decision to apply a control premium. Id. at 639 . For an assumption that carried such weighty consequences, we cannot say that the district court clearly erred in finding that a reasonably prudent fiduciary would have probed the issue further.

4.

The final major fault that the district court identified in Wilmington's reliance on the SRR Report was Wilmington's failure to probe why SRR consistently rounded the valuation of Constellis stock upwards. This repeatedly increased the price the ESOP had to pay for the stock. The district court concluded that because an ESOP fiduciary should be concerned with achieving the lowest possible price for the ESOP, Wilmington ought to have asked why SRR consistently engaged in the upward rounding to the ESOP's detriment. Id. at 639-40 .

Wilmington devotes only a short footnote in its opening brief to this issue and fails to address it at all in its reply brief. It thus appears to have forfeited an appellate challenge to the court's findings on this issue. See Wahi v. Charleston Area Med. Ctr.,Inc. , 562 F.3d 599 , 607 (4th Cir. 2009) (finding argument raised only in a footnote in appellant's opening brief forfeited on appeal). Perhaps this is because the district court's finding is self-evidently correct. And, as the district court noted, Brundle I , 241 F.Supp.3d at 640 , this conclusion accords with those of other courts. For example, the Sixth Circuit described a similar rounding practice as "extremely disconcerting" and noted that this conduct, although not dispositive, was illustrative of lackluster due diligence displayed throughout the transaction process. Chao , 285 F.3d at 433-34. So too here.

B.

Apart from Wilmington's four major failures to scrutinize SRR's report, the district court also found that Wilmington's conduct throughout the transaction provided further evidence that Wilmington failed to act as a prudent fiduciary solely on behalf of the ESOP participants. Although the court found that Wilmington's actions on behalf of the ESOP did not amount to bad faith, it recognized that "ERISA demand[ed] more to excuse a fiduciary from liability." Brundle I , 241 F.Supp.3d at 643 .

The district court found that Constellis's principal motivation for creating the ESOP was to further the Sellers' interests, not the interests of the ESOP participant buyers. Id. at 616 . The sale of Constellis to the ESOP enabled the Sellers to cash out their company stock, which they had tried and failed to accomplish several times, and to reap substantial tax savings. Moreover, the ESOP's unique structure enabled the Sellers to avoid the downside risk of holding equity, yet profit from any upside, while also retaining control over the company. The court found that Wilmington, in its role as the ESOP's fiduciary, simply ignored these motivations rather than investigating whether they affected the ESOP's legality under ERISA. Id. at 640-42 .

The district court additionally found that making the Sellers' interests paramount shortened the transaction timeline and rushed Wilmington's due diligence process and price and terms negotiations. To maximize the after-tax benefits to the Sellers, the ESOP sale had to conclude before the end of 2013. Accordingly, Wilmington and its agents completed their due diligence, made pricing decisions, conducted negotiations, and launched a tender offer in less than two months. Id. at 621-25 . Wilmington employees met to discuss the transaction only three times and met with Constellis management just once. Id. at 642 .

All meetings lasted less than 90 minutes and several of Wilmington's team members were absent from each. Id. And the district court found that Wilmington violated its own policy by repeatedly taking less than 48 hours between receiving a valuation report and meeting to discuss it. Id. at 621, 625 . The court concluded that this type of conduct fell well short of the diligence expected of a prudent fiduciary. Id. at 643 .

Finally, the court noted that "Wilmington's lack of engagement and willingness to negotiate so favorably with CSG may have been motivated by its significant business relationship with CSG." Id. The district court found that Wilmington, CSG, and SRR maintained extensive and lucrative business relationships and that SRR was willing to discuss its valuation strategies with CSG while negotiations were ongoing. Id . The record offers ample support for all of these findings.

C.

Wilmington also maintains that the district court erred in finding that the value of Stock Appreciation Rights (SARs) issued in connection with the ESOP's purchase of Constellis should have been deducted from Constellis's equity value for purposes of SRR's valuation. SARs are incentives that Constellis would have owed to management had the company's value reached a certain price by a particular date. Wilmington argues the value of the SARs should not have been deducted from Constellis's estimated value because they were inchoate at the time of the sale. The district court disagreed and found that, because Constellis intended for the SARs to be issued, their value should have been deducted from Constellis's equity value. Id. at 648 .

Wilmington rehashes the same argument on appeal but fails to identify any evidence that demonstrates the district court clearly erred in this factual determination. Because we discern no such evidence, we reject this claim.

D.

In support of its final attack on the district court's liability findings, Wilmington relies on the ESOP's subsequent sale of Constellis stock to ACADEMI LLC.

In early February 2014, less than two months after the ESOP transaction, ACADEMI, a major competitor in the private security market, began talks with Constellis management to acquire the company. Id. at 628 . Initially, ACADEMI offered to purchase Constellis at an enterprise value of $ 283.3 million, in a transaction that would result in $ 10 million in cash going to the ESOP. Seeking a better offer on the ESOP's behalf, Wilmington entered into negotiations with Constellis, the Sellers, and ACADEMI. Id. at 629 . Wilmington's efforts proved successful: the final transaction, which closed on July 25, 2014, effectuated a sale of all the Constellis stock to ACADEMI at an implied enterprise value of $ 288.3 million. Id. at 631 . The final transaction paid $ 20 million in cash to the ESOP - double the initial $ 10 million offer - as a result of the Sellers' agreeing to write off $ 33 million of the ESOP's debt. Id. at 629 .

Wilmington argues that the ACADEMI sale proves that the ESOP did not overpay for Constellis seven months earlier. Relying on the first prong of DOL's proposed adequate-consideration regulations, Wilmington maintains that the ESOP paid "fair market value" for the Constellis stock because ACADEMI's purchase of Constellis stock implied an enterprise value of $ 288.3 million, which was within SRR's estimated range of $ 275 to $ 330 million. Wilmington contends this purchase price is of particular significance because in the intervening time Constellis assertedly suffered several business setbacks that decreased its enterprise value. 8

This argument is unpersuasive for several reasons. The proposed DOL regulations upon which Wilmington relies provide that a trustee can prove consideration is "adequate" if it reflects the stock's "fair market value" and is "the product of a determination made by the fiduciary in good faith." Proposed Regulation Relating to the Definition of Adequate Consideration, 53 Fed. Reg. at 17,633. Of course, because the DOL never enacted the proposed regulations, they are not binding. And, as we have explained, courts look to the conduct of the trustee and whether it met its fiduciary obligations, not to whether the trustee arrived at a "fair" value. But even if we were to adopt the DOL's proposed regulations, a trustee must satisfy both parts of the test to benefit from the adequate consideration exception. See id. Assuming arguendo that the purchase was for "fair market value," we find no error in the district court's findings that Wilmington failed to prove that the share price was also "the product of a determination made by the fiduciary in good faith."

Furthermore, even if proving the ESOP paid "fair market value" sufficed to meet the requirements of § 1108(e), the facts of the ACADEMI sale do not prove that the ESOP paid only "fair market value" here. In fact, the district court found that the opposite was true. The court found that because ACADEMI, a private security company like Constellis, would benefit from synergies resulting from its purchase of Constellis, it would have been willing to pay significantly more for Constellis than the ESOP. Thus, that ACADEMI actually paid less for Constellis stock suggested a significantly lower fair market value for the stock at the time of the ESOP purchase. Of course, Wilmington's experts disagreed with this synergies analysis, but the district court found the testimony of other experts on this point more persuasive. Brundle I , 241 F.Supp.3d at 640 . The record provides no basis to reject this finding. See Hendricks v. Cent. Reserve Life Ins. Co. , 39 F.3d 507 , 513 (4th Cir. 1994) ("An appellate court should be especially reluctant to set aside a finding based on the trial court's evaluation of conflicting expert testimony.").

In sum, the district court found that the ACADEMI sale did not constitute a meaningful comparator. Discounting the additional synergy value, the $ 33 million debt write-off by the Sellers, and an expected tax refund from the sale, the court found the ACADEMI purchase suggested that Constellis's fair market value was closer to $ 200 million - almost $ 100 million less than the enterprise value implied by the ESOP transaction. Thus, the court found, the ACADEMI sale could not support Wilmington's argument that the ESOP paid no more than adequate consideration for Constellis. Brundle I , 241 F.Supp.3d at 640 . We see no clear error in this conclusion.

Additionally, we note that, despite paying more for Constellis stock, the ESOP did not obtain nearly the level of control that ACADEMI acquired. While the ESOP obtained only some elements of control, ACADEMI acquired absolute and uncontested control. Had the district court applied the "standard" 35-40% control premium Wilmington's experts suggested was reasonable, Constellis's fair market value would fall somewhere between $ 145 million and $ 150 million, almost half the value of the stock purchased by the ESOP only months earlier.

III.

Having rejected all of Wilmington's challenges to the district court's findings as to its liability, we turn to the court's damages award. The district court concluded that Wilmington's fiduciary failures inflated the price the ESOP paid for Constellis stock by $ 29,773,250. "We review factual findings relating to the calculation of damages for clear error." Simms v. United States , 839 F.3d 364 , 368 (4th Cir. 2016).

A.

To calculate the loss to an ESOP and compensate it for a fiduciary's ERISA violation, a court typically subtracts the stock's fair market value, as determined by the court, from the inflated price paid by the ESOP. See Bruister , 823 F.3d at 265 (collecting cases) . The district court precisely followed this approach. The court used as a baseline the comprehensive damages estimate provided by Brundle's expert - the only expert to offer one. 9 The court then considered the flaws in this valuation raised by Wilmington and made "appropriate adjustments based on the evidence adduced at trial and the relative credibility of the experts." Brundle I , 241 F.Supp.3d at 645 .

Brundle's expert estimated Wilmington's lack of due diligence had caused the ESOP to overpay for the Constellis stock by almost $ 103,862,000. But the district court did not accept this figure. Instead, it independently analyzed each of Wilmington's purported failings, considered the relevant trial testimony, and then awarded $ 29,773,250 - or less than a third of the expert's estimation - as damages. Id. at 645-49 . The following chart summarizes the district court's damages findings:

Error Difference between fair market value and inflated price paid by ESOP (Damages) Reliance on Management's $4,325,000 Growth Projections Use of .07 Beta $2,936,000 Inclusion of Control Premium $8,186,000 Failure to Include Lack of $9,715,250 Control Discount Stock Appreciation Rights $1,611,000 Rounding $3,000,000 Total Damages $29,773,250

Wilmington did not provide the district court with an alternative damages calculation or offer any evidence that the court could use to devise its own. At oral argument before us, Wilmington conceded as much. Accordingly, it has forfeited its challenge to the methodology used by the district court when calculating the damages award. See Exxon Shipping Co. v. Baker , 554 U.S. 471 , 487 n.6, 128 S.Ct. 2605 , 171 L.Ed.2d 570 (2008) ("[W]aiver and forfeiture rules ... ensure that parties can determine when an issue is out of the case, and that litigation remains, to the extent possible, an orderly progression.").

Moreover, Wilmington's principal basis for urging reversal of the damages award consists of a repetition of its challenge to the district court's liability findings. Given our affirmance of all the district court's liability findings - that Wilmington failed to adequately investigate and probe SRR as to the omission of the McLean Report, the use of Constellis management's growth projections, the 0.7 beta, the 10% control premium, improper rounding up, and SARs - that challenge fails.

B.

Wilmington additionally contends that the district court erred in declining to offset its damages calculation by the $ 20 million in cash that the ESOP received as a result of the ACADEMI sale. According to Wilmington, "the district court's failure to" do so "resulted in a clear windfall for the ESOP." Wilmington Opening Br. at 53. As the district court ably explained, see Brundle II , 258 F.Supp.3d at 660-62 , this claim is meritless.

ERISA expressly requires a fiduciary to "make good to [an ESOP] any losses" resulting from a given breach of duty. 29 U.S.C. § 1109 (a) ; cf. Restatement (Second) of Trusts § 205 cmt. e ("If the trustee is authorized to purchase property for the trust, but in breach of trust he pays more than he should pay, he is chargeable with the amount he paid in excess of its value."). Thus, in cases involving overpayment for ESOP assets, courts generally compute restitution damages exactly as the district court did here: by deducting the fair market value of the stock from the amount the ESOP actually paid.

Wilmington insists that the ESOP's subsequent and unrelated sale of the stock to ACADEMI requires a different result. This misapprehends the nature of the breach. If Wilmington had breached its duty by purchasing Constellis stock at all, then a damages award consisting of the purchase payments offset by any subsequent profits would indeed return the participants to the pre-purchase status quo.

But Wilmington breached here by overpaying for the Constellis stock. Principles of restitution therefore entitle the ESOP and its participants to compensation for the loss from the overpayment . Any subsequent gains involving the stock, which the ESOP would have obtained regardless of the overpayment, have no bearing on that loss. For this very reason, courts have uniformly rejected analogous offsets to damage awards in overpayment cases. See, e.g. , Bruister , 823 F.3d at 270-71 (rejecting offset to account for unpaid or forgiven ESOP debt and noting that "[e]very court to consider this question" agreed); Henry v. U.S. Trust Co. of Cal., N.A. , 569 F.3d 96 , 98-100 (2d Cir. 2009) (" Henry V ") ("If an investor pays $ 100 for 20 shares of stock and later sells those shares back to the original seller for $ 25, the result is not that the investor paid only $ 75 for the shares.").

Even so, Wilmington asserts that the events in this case resulted in "pure gain" to the ESOP, justifying a deviation from this rule. It claims the ESOP "made $ 20 million for the nominal cost of carrying $ 200 million in non-recourse debt for a few months," because "[j]ust seven months" after the prohibited transaction, "the ESOP exchanged its debt-burdened shares to ACADEMI for $ 20 million in cash." Wilmington Opening Br. at 55-56. We disagree. As a preliminary matter, the burden imposed in additional debt from a $ 30 million overpayment is far from "nominal," even for "just" seven months. See Henry V , 569 F.3d at 99 n.4 (explaining "the obvious fact that the assumption of indebtedness has immediate legal and economic consequences even before the borrower begins to repay the debt").

Moreover, if Wilmington seeks to portray the ACADEMI sale as an "exchange" in which the ESOP's pre-sale debt had no effect on its post-sale assets, this is both illogical and untrue. Had the ESOP not initially overpaid for the stock by $ 30 million, it would have shouldered tens of millions of dollars less in debt; in turn, it would presumably have been tens of millions of dollars better off after being compensated for the stock in the ACADEMI sale. Put differently, although the ESOP benefited despite the breach, it would have benefited at least as much - but ended in a far better position - without the breach. In fact, had the ESOP initially paid a fair (lower) value for the Constellis stock and therefore carried less debt, it would have entered the ACADEMI negotiations in a stronger bargaining position and may have profited further still.

Wilmington rejects this conclusion, stressing that the Sellers also agreed to write off more than $ 30 million of the ESOP's debt, which it describes as an "effective[ ] reduc[tion]" of the 2013 purchase price. Wilmington Opening Br. at 56. But once again, this independent act has no bearing on the ESOP's loss.

Wilmington unpersuasively contends that Henry III supports its argument. That case addressed very different facts. There, the Second Circuit held that - where sellers fully and voluntarily repaid an ESOP after the IRS found overpayment, but before a court entered a damages judgment - a duplicative recovery for the same overpayment might "result in a windfall." Henry III , 445 F.3d at 624. Wilmington does not even argue that the Sellers' write-off here was meant to remedy the ESOP's overpayment. Nor could it: unlike the midstream repayment in Henry , the write-off here occurred more than a year prior to the initiation of any legal proceedings alleging Wilmington breached its fiduciary duties. As Wilmington concedes, the write-off was part of a negotiated agreement to close the ACADEMI sale, a deal wholly unrelated to the issues raised here. Again, it is reasonable to assume that the ESOP would have negotiated equally favorable (if not more favorable) terms had it paid a fair value in the first instance and thus entered the ACADEMI discussions with substantially less debt.

In sum, we find no error in the district court's damages award.

IV.

Finally, we examine the district court's award of attorneys' fees. The court awarded Brundle's counsel $ 1,819,631.11 in statutory fees and an additional $ 1.5 million in fees from the ESOP's damages judgment. No party has presented a substantive challenge regarding the award of attorneys' fees under the fee shifting statute. 10 Brundle challenges the $ 1.5 million in additional fees (payable from the damages awarded to the ESOP) as too small; Wilmington and Constellis, as fiduciaries for the ESOP, challenge the award as legally improper. We review that award for abuse of discretion, mindful that our review "is sharply circumscribed, and a fee award must not be overturned unless it is clearly wrong."

Berry v. Schulman , 807 F.3d 600 , 617 (4th Cir. 2015) (internal quotation marks omitted).

A.

We first consider Brundle's claim that the district court abused its discretion by not awarding his counsel, Bailey & Glasser, LLP (B&G), a total of $ 9.9 million in fees under the firm's one-third contingent fee agreement with Brundle himself. The district court initially suggested that this arrangement was indeed controlling. See Brundle II , 258 F.Supp.3d at 671-72 . But the court held its fee decision in abeyance to consider the views of plan participants and ultimately based its award on equity rather than contract. We can hardly find an abuse of discretion given that Brundle has no right to have the ESOP fund his own fee agreement with B&G.

Brundle contends that 29 U.S.C. § 1132 (a)(2) - which empowers ESOP participants and beneficiaries to bring suit on behalf of an ESOP - additionally authorizes him to enter into a legal services contract that binds the ESOP. Brundle musters no caselaw to support his interpretation of the statute, and for good reason. While § 1132(a)(2) provides Brundle the right to sue on behalf of the ESOP, neither it nor any other statute authorizes him to unilaterally bind the ESOP to a contract or fee agreement. For this very reason, the vast majority of suits on behalf of ERISA plans take the form of class actions, in which counsel can seek fees only after a class has been certified and class members have been given the opportunity to object or opt out. Though B&G has a valid contingent fee agreement with Brundle, see Venegas v. Mitchell , 495 U.S. 82 , 110 S.Ct. 1679 , 109 L.Ed.2d 74 (1990), it does not have a fee agreement with the ESOP.

As a result, only equity provides any possible basis for a fee award in addition to the statutory fees awarded.

B.

After the district court received the views of some plan participants, it issued a final judgment awarding B&G $ 1.5 million in "addition[ ]" to the $ 1.8 million in statutory fees - not as a contractual right, but as "a partial common fund award." 11 Wilmington and Constellis assert that a trio of ERISA provisions precluded the court from awarding any common fund recovery. Unlike most challenges to fee awards, we review de novo questions regarding a court's legal authority. See, e.g. , Country Vintner of N.C., LLC v. E & J. Gallo Winery, Inc. , 718 F.3d 249 , 254 (4th Cir. 2013).

First, Wilmington and Constellis contend that any common fund award here would violate ERISA's "anti-alienation" and "exclusive benefit" provisions. See 29 U.S.C. § 1056 (d)(1) (providing that benefits under retirement plans cannot be "assigned or alienated"); id. § 1103(c)(1) (directing all "assets of a plan" to be "held for the exclusive purposes of providing benefits to participants in the plan and their beneficiaries and defraying reasonable expenses of administering the plan"). But the district court properly rejected this argument, concluding that because the plan administrator did not yet possess the damages award - in other words, the court still had jurisdiction over the funds and had not distributed them to the ESOP - it could not be considered a plan asset. Brundle II , 258 F.Supp.3d at 670 . Of course, if the ESOP had received payment of the award, those funds would have been subject to ERISA's duties and protections. See WilmingtonShipping Co. v. New England Life Ins. Co. , 496 F.3d 326 , 333 (4th Cir. 2007). But where a fund is still under the jurisdiction of the district court, the court may award fees from that fund without running afoul of ERISA's anti-alienation or exclusive benefit requirements. See Kickham Hanley P.C. v. Kodak Ret. Income Plan , 558 F.3d 204 , 213 (2d Cir. 2009) ; Lynn v. CSX Transp., Inc. , 84 F.3d 970 , 975 (7th Cir. 1996).

Next, Wilmington and Constellis argue that ERISA's statutory fee-shifting provision displaces the common fund doctrine. See 29 U.S.C. § 1132 (g)(1) (authorizing prevailing plaintiffs to recover reasonable attorneys' fees from defendants). They urge us to follow the Seventh Circuit in "limiting the common-fund approach to cases outside the scope of a fee-shifting statute." Pierce v. Visteon Corp. , 791 F.3d 782 , 787 (7th Cir. 2015) ; accord Haggart v. Woodley , 809 F.3d 1336 , 1358-59 (Fed. Cir. 2016) (largely adopting Pierce rule). 12

Several other circuits, however, have disagreed with this approach. See McKeage v. TMBC, LLC , 847 F.3d 992 , 1003 (8th Cir. 2017) ("[T]he presence of a fee-shifting statute precludes a common fund award only when that result is required by the statutory scheme involved."); In re Bluetooth Headset Prods. Liab. Litig. , 654 F.3d 935 , 941 (9th Cir. 2011) ("The award of attorneys' fees in a class action settlement is often justified by [a] common fund or statutory fee-shifting[,] ... and sometimes by both."); Staton v.Boeing Co. , 327 F.3d 938 , 967-69 (9th Cir. 2003) (permitting common fund recovery even where statutory fees may be available); Cty. of Suffolk v. Long Island Lighting Co. , 907 F.2d 1295 , 1327 (2d Cir. 1990) ("In our view, fee-shifting statutes are generally not intended to circumscribe the operation of the equitable fund doctrine.").

To determine the proper course here, we consider the roots of the common fund doctrine. The common fund principle derives from the equitable doctrines of quantum meruit, Central Railroad & Banking Co. v. Pettus , 113 U.S. 116 , 5 S.Ct. 387 , 28 L.Ed. 915 (1885), and unjust enrichment, Trustees v. Greenough , 105 U.S. 527 , 532, 26 L.Ed. 1157 (1881). It allows a court to award "a reasonable attorney's fee" to "a litigant or a lawyer who recovers a common fund for the benefit of persons other than himself or his client ... from the fund as a whole." Boeing Co. v. Van Gemert , 444 U.S. 472 , 478, 100 S.Ct. 745 , 62 L.Ed.2d 676 (1980). The rationale for the doctrine is that when one benefits from a lawsuit without contributing to its costs, he has been unjustly enriched at the successful litigant's or lawyer's expense. Mills v. Elec. Auto-Lite Co. , 396 U.S. 375 , 392, 90 S.Ct. 616 , 24 L.Ed.2d 593 (1970). Courts remedy this inequity by shifting a proportional share of reasonable attorneys' fees onto these unjustly enriched beneficiaries.

The Supreme Court has applied the "strong and uniform ... background rule" of the common fund doctrine "in a wide range of circumstances as part of [its] inherent authority." US Airways, Inc. v. McCutchen , 569 U.S. 88 , 104, 133 S.Ct. 1537 , 185 L.Ed.2d 654 (2013) (collecting cases). We agree with Brundle that it applies here. Scores of unnamed ESOP participants benefited substantially from this lawsuit, while B&G bore the entirety of the costs and risks. Equity thus demands that the enriched participants pay a proportional share of reasonable attorneys' fees.

To be sure, courts often, but not invariably, award common fund fees when a procedural obstacle, such as the terms of a settlement, bars statutory fees that are otherwise available. 13 But it does not follow that all statutory fee provisions necessarily displace the doctrine or render it superfluous. To the contrary: although both the common fund doctrine and fee-shifting statutes facilitate suit by plaintiffs who may otherwise lack the means to hire counsel, they operate in distinct ways and serve distinct purposes.

A common fund recovery places the plaintiffs' cost of litigation on the recovering beneficiaries of a lawsuit , whereas a fee-shifting statute places this burden on the losing party . Cf. Venegas , 495 U.S. at 90 , 110 S.Ct. 1679 (differentiating between "what the losing defendant must pay" to the plaintiff as attorneys' fees under a fee-shifting statute and "what the prevailing plaintiff must pay his lawyer"). A fee-shifting statute thus creates an exception to the "American rule" that "the prevailing litigant is ordinarily not entitled to collect a reasonable attorneys' fee from the loser." Alyeska Pipeline Serv. Co. v. Wilderness Soc. , 421 U.S. 240 , 247, 95 S.Ct. 1612 , 44 L.Ed.2d 141 (1975). In contrast, the common fund doctrine - though often similarly described as an exception - better accords with the American rule by holding the beneficiaries of a judgment or settlement responsible for compensating the counsel who obtained the judgment or settlement for them. See Staton , 327 F.3d at 967 .

Furthermore, the purpose of the common fund doctrine, unlike a fee-shifting statute, is to avoid unjust enrichment. Thus, although a statutory fee award against a defendant generally may not include a contingency enhancement, see City of Burlington v. Dague , 505 U.S. 557 , 565-67, 112 S.Ct. 2638 , 120 L.Ed.2d 449 (1992), courts routinely impose enhanced common fund awards to compensate counsel for litigation risk at the expense of beneficiaries who did not shoulder this risk, see, e.g. , In re Wash. Pub. Power Supply Sys. Secs. Litig. , 19 F.3d 1291 , 1300 (9th Cir. 1994) ("There is nothing unfair about contingency enhancements in common fund cases ...."); Florin , 34 F.3d at 564-65 (similarly holding that "the policy considerations ... that militate for forbidding risk multiples in statutory fee-shifting cases have little force in common fund cases").

From these distinctions, it follows that a "reasonable" fee payable by a defendant to compensate the prevailing plaintiff's counsel is not necessarily identical to a "reasonable" fee owed by a recovering beneficiary to plaintiff's counsel, particularly where the contingency risk to plaintiff's counsel is substantial. Cf. Venegas , 495 U.S. at 85, 90 , 110 S.Ct. 1679 (affirming $ 406,000 contractual fee, paid by prevailing plaintiff to his counsel, as not "unreasonable" even though "reasonable" statutory fee, paid by losing defendants, was only $ 117,000); McKeage , 847 F.3d at 1004 (emphasizing "a distinction between the amount a losing party may be required to pay" under fee-shifting provision in settlement "and the amount a class member should equitably be required to pay" under common fund doctrine).

Consequently, we find unpersuasive the Seventh Circuit's reasoning in Pierce . In limiting the common fund doctrine to cases "outside the scope of a fee-shifting statute," the Pierce court reasoned first and foremost that "[c]ommon-law doctrines yield to statutes." 791 F.3d at 787 . But the Supreme Court has instructed that courts must "presum[e] that Congress intended to retain the substance of the common law." Samantar v. Yousuf , 560 U.S. 305 , 320 n.13, 130 S.Ct. 2278 , 176 L.Ed.2d 1047 (2010). We recognize, of course, that legislative displacement of the common law does not require "the same sort of evidence of a clear and manifest congressional purpose demanded for preemption of state law." Am. Elec. Power Co., Inc. v. Connecticut , 564 U.S. 410 , 423, 131 S.Ct. 2527 , 180 L.Ed.2d 435 (2011) (alterations and internal quotation marks omitted). Even so, a statute that shifts a reasonable fee onto a losing defendant does not, without more, "speak[ ] directly" to congressional intent to displace recovery from a common fund. Id. at 424, 131 S.Ct. 2527 (internal quotation marks omitted).

The Pierce court additionally emphasized that "fee-shifting statutes are designed to ensure that the victims retain full compensation," 791 F.3d at 787 , a concern of indisputable importance. The common fund doctrine, however, balances this critical interest against the concern of unjust enrichment, which ERISA's fee-shifting provision does not purport to address.

Finally, echoing the Pierce court, Wilmington and Constellis stress that B&G was entitled to, and received, a "reasonable" attorneys' fee under statute. However, as we have explained, what is "reasonable" as a fee from a losing defendant may well differ from what is "reasonable" as a fee from an individual directly benefited by counsel's efforts. We particularly note that a "reasonable" fee under ERISA's attorneys' fee provision (unlike a common fund award) precludes any compensation for contingency risk. See Dague , 505 U.S. at 565-67 , 112 S.Ct. 2638 . But, we have previously underscored, contingent fees serve an important function in providing under-resourced litigants access to counsel and the courts. In re Abrams & Abrams, P.A. , 605 F.3d 238 , 245-47 (4th Cir. 2010). Conflating statutory attorneys' fees with common fund recoveries would squarely contravene this policy.

We therefore agree with the Second, Eighth, and Ninth Circuits that a statutory fee-shifting provision (or an award of fees under such a provision) does not, as a matter of law, automatically preclude an award under the common fund doctrine. Accordingly, the district court retained discretion to award supplemental attorneys' fees from the common fund.

C.

Having established the district court's authority to award supplemental fees, we turn to Brundle's contention that the court abused its discretion in awarding its counsel too little from the common fund.

Before setting the amount of its common fund award, the district court fashioned an ad hoc process analogous to class settlement procedures under Rule 23. Brundle II , 258 F.Supp.3d at 671-72 . The court ordered Brundle's counsel, B&G, to engage in a good faith effort to notify the approximately 1,800 former ESOP participants of its proposed contingency fee award. After B&G did so, the court received 59 letters objecting to a one-third contingency award. As a result of the objections, the court decided to award B&G only $ 1.5 million as a supplemental fee payable from the damages award, rather than granting counsel a full one-third ($ 9.9 million) of the ESOP's total damages recovery.

In reviewing the district court's award, we note that Brundle had available a procedural mechanism designed to carefully balance the interests of counsel, plaintiffs, and other similarly situated beneficiaries from the start - namely, class certification pursuant to Rule 23. To be fair, Brundle and another plaintiff, Halldorson, did initially seek class certification, which the district court denied. But Brundle opted not to timely appeal this denial. And although Halldorson initially sought our review, he later withdrew his appeal. See supra n.4.

Had Brundle pursued class certification and prevailed, B&G could have secured attorneys' fees from the class as a whole. Although a properly certified class might not have agreed to the same one-third contingency fee that Brundle did, the district court might well have awarded a fee significantly higher than the firm's ultimate recovery here. 14

Brundle, represented by B&G, instead opted to forego Rule 23's procedural strictures - thus avoiding "the standards set for the protection of absent class members." Amchem Prods., Inc. v. Windsor , 521 U.S. 591 , 621, 117 S.Ct. 2231 , 138 L.Ed.2d 689 (1997). In light of this choice, the district court's decision to limit the common fund award was not inequitable, as illustrated by the objections from some of the ESOP's participants. Nearly 60 of the 1,800 participants objected to the proposed common fund award, explaining that they had never been informed of the case status, never agreed to B&G's representation, and never authorized the fund to pay a one-third fee. These process concerns would have been avoided upfront had Brundle persisted in his pursuit of class certification. But because he and his counsel abandoned Rule 23's class-protective procedures, equity cuts against awarding B&G the full one-third contingency fee.

The district court properly weighed B&G's efforts on behalf of the ESOP - including contingency risk - against the objections from ESOP participants to determine a proper common fund award. Because we cannot say this balancing was clearly wrong, see Berry , 807 F.3d at 617 , we affirm the court's fee award in full.

V.

For the foregoing reasons, the judgment of the district court is in all respects

AFFIRMED .