DIANA GRIBBON MOTZ, Circuit Judge:

This Employee Retirement Income Security Act ("ERISA") case returns to us for a third time. The beneficiaries of an ERISA retirement plan appeal the judgment, issued after a full bench trial, that the fiduciary's breach of its duty of procedural prudence did not cause the substantial losses in the retirement plan resulting from the sale of non-employer stock funds. We had previously remanded the case to the district court so that it could apply the correct legal standard for determining loss causation, but we expressed no opinion as to the proper outcome of the case. On remand, applying the correct standard, the court found that the fiduciary's breach did not cause the losses because a prudent fiduciary would have made the same divestment decision at the same time and in the same manner. For the reasons that follow, we affirm.

I.

A.

We begin with a brief review of the events leading up to this lawsuit. Our previous opinions and those of the district court set forth the facts of this case in far greater detail. See generally Tatum v. RJR Pension Inv. Comm. (Tatum IV), 761 F.3d 346, 351-55 (4th Cir. 2014); Tatum v. R.J. Reynolds Tobacco Co. (Tatum II), 392 F.3d 636, 637-39 (4th Cir. 2004); Tatum v. R.J. Reynolds Tobacco Co. (Tatum V), No. 1:02-cv-00373, 2016 WL 660902, at *2-*12 (M.D.N.C. Feb. 18, 2016); Tatum v. R.J. Reynolds Tobacco Co. (Tatum III), 926 F.Supp.2d 648, 651-69 (M.D.N.C. 2013).

In 1999, RJR Nabisco, Inc. decided to spin off the company's food business, Nabisco, from its tobacco business. Tatum V, 2016 WL 660902, at *2. The company provided several rationales for this decision. But the critical motivating force seems to have been that tobacco litigation against RJR adversely affected the value of Nabisco stock. Id. at *3.

Prior to the spin-off, participants in RJR Nabisco's ERISA-covered retirement plan ("the Plan") could hold stock in their retirement accounts in two funds related to the company: the RJR Nabisco Common Stock Fund and the Nabisco Common Stock Fund. Id. at *1. "After the spin-off, the RJR Nabisco Common Stock Fund was divided into two separate funds: the Nabisco Group Holdings Common Stock Fund ..., which held the stock from the food business, and the RJR Common Stock Fund, which held the stock from the tobacco business." Tatum IV, 761 F.3d at 379. "Thus, as a result of the spin-off, there were two funds holding exclusively Nabisco stock: the Nabisco Common Stock Fund, which existed prior to the spin-off, and the Nabisco Group Holdings Common Stock Fund, which was created as a result of the spin-off" (collectively the "Nabisco Funds"). Id. at 379 n.1. RJR, as the Plan administrator, concluded that, given the spin-off, "it would be inappropriate to hold stock in what was to become a non-related company." Tatum V, 2016 WL 660902, at *4. Accordingly, RJR informed Plan participants that the Nabisco Funds would be frozen on the date of the spin-off and divested within six months. Id. at *5.

The spin-off occurred on June 14, 1999. Id. At that time, RJR amended the Plan documents to freeze the Nabisco Funds. This meant that Plan participants could make no additional investments in the Funds, but participants could sell their shares in the Funds. Id. at *1, *5. In the months immediately following the spin-off, Nabisco stock sharply declined in value. Tatum IV, 761 F.3d at 353. In October 1999, RJR held a series of meetings to discuss the possibility of reversing its decision to divest the Funds, but ultimately decided to continue as planned. Id. Also in October, RJR experienced a significant adverse verdict in the tobacco litigation, which caused both RJR and Nabisco stock to fall. Tatum V, 2016 WL 660902, at *7. In November, RJR attempted to amend the Plan documents to indicate that the Plan would no longer offer the Nabisco Funds.1 Id. at *9-*10. During this time, Nabisco stock continued to decrease in value. Id. From the date of the spin-off until January 31, 2000, when the divestment occurred, the price of the Nabisco Group Holdings common stock fell 60%, and the price of the Nabisco common stock fell by almost 30%. Id. at *11.

Two months after the divestment, on March 30, 2000, investor Carl Icahn made an unsolicited bid to take over Nabisco, which allowed Nabisco to pursue corporate restructuring without tax consequences. Id. Nabisco Group Holdings rejected the Icahn offer in April, "but announced that it would explore all alternatives to maximize shareholder value, effectively putting [Nabisco Group Holdings] on the auction block." Id. By December 2000, Philip Morris had acquired Nabisco for $55 per share and RJR had bought Nabisco Group Holdings for $30 a share. Id. By the time the sales were completed, the price of both companies' stock had increased dramatically; Nabisco Group Holdings common stock increased 247% and Nabisco common stock increased 82%. Id.

B.

In May 2002, Richard Tatum, an RJR employee, brought this action against RJR and various RJR committees on behalf of himself and others similarly situated. He alleged that the defendants, as fiduciaries of the Plan, "breached their ... duties under ERISA by eliminating Nabisco stock from the Plan on an arbitrary timeline without conducting a thorough investigation." Tatum IV, 761 F.3d at 355. Tatum further charged "that their fiduciary breach caused substantial loss to the Plan because it forced the sale of the Plan's Nabisco Funds at their all-time low, despite the strong likelihood that Nabisco's stock prices would rebound." Id.

The district court initially concluded that the Plan documents required divestment of the Nabisco Funds and granted RJR's motion to dismiss. On appeal, we held "that the Plan documents did not mandate divestment of the Nabisco Funds, and thus did not preclude Tatum from stating a claim against the defendants for breach of fiduciary duty." Id. (citing Tatum II, 392 F.3d at 637). Accordingly, we reversed the judgment of the district court and remanded the case to it.

The district court then conducted a four-week bench trial, which involved the testimony of numerous fact and expert witnesses and a mountain of documents. Upon consideration of this evidence, the court issued a detailed opinion containing extensive factual findings. The court held that RJR (1) breached its fiduciary duty of procedural prudence when it decided to remove the Nabisco Funds from the Plan "without undertaking a proper investigation into the prudence of doing so," (2) "bore the burden of proving that [this] breach did not cause the alleged losses to the Plan," and (3) had, despite its breach, met its burden by showing that a hypothetical prudent fiduciary could have made the same decision. Id. (quoting Tatum III, 926 F.Supp.2d at 651).

On appeal, we agreed with the district court that RJR had breached its duty of prudence as an ERISA fiduciary, and, as such, bore the burden of proving the losses were not attributable to the breach. Id. at 351. However, we vacated the district court's judgment because we concluded that our precedent requires a breaching fiduciary to prove by a preponderance of the evidence that a hypothetical prudent fiduciary would have made the same decision. Id. at 351, 365, 368 (citing Plasterers' Local Union No. 96 Pension Plan v. Pepper, 663 F.3d 210, 218 (4th Cir. 2011)). We instructed the court "to review the evidence to determine whether RJR has met its burden of proving by a preponderance of the evidence that a prudent fiduciary would have made the same decision." Id. at 368. That review, we noted, might or might not lead the court to the same ultimate conclusion. This determination was for the district court on remand; we expressed no view as to the proper resolution of the case.

On remand, the court issued another thorough opinion. In it, the court explained why in its view RJR had met its burden of showing by a preponderance of the evidence that a hypothetical prudent fiduciary would have divested the Nabisco Funds in the manner and at the time RJR did. Tatum V, 2016 WL 660902, at *26. Tatum filed the instant appeal.

II.

Tatum initially maintains that the district court disregarded our mandate, a question we consider de novo. See S. Atl. Ltd. P'ship of Tenn., LP v. Riese, 356 F.3d 576, 583 (4th Cir. 2004). Tatum argues that, in holding that a prudent fiduciary would have made the same decision as RJR made here, the court focused too much on risk, ignored the plain language of the Plan documents, failed to consider the testimony of Professor Thomas Lys, and did not explain why a prudent fiduciary would have divested the Nabisco Funds at the time that RJR did. We disagree.

As to the district court's asserted singular focus on risk, we explained in Tatum IV that "while the presence of risk is a relevant consideration in determining whether to divest a fund held by an ERISA plan, it is not controlling." 761 F.3d at 367. We see no indication that the district court impermissibly rested its analysis solely on risk. Even a cursory review of its opinion reveals that the court also considered value and expected returns, the diversity of the Plan's investments, the requirements of the Plan documents, and the timing of the divestment. The court additionally took account of the nature of the Plan as a long-term savings vehicle and the role of a fiduciary in managing beneficiaries' assets. Tatum V, 2016 WL 660902, at *13, *25. Had the court considered only risk, it would have erred. But it did not. Rather, the court acted entirely within our mandate in addressing risk as a "relevant consideration" along with other factors.

Next, Tatum takes issue with how the district court treated the Plan documents. Tatum contends that the court failed to give adequate consideration to the fact that these documents, absent the invalid November amendment, required that the Nabisco Funds remain frozen. In Tatum IV, we explained that, while the terms of the Plan documents cannot "trump the duty of prudence," the "plan terms, and the fiduciary's lack of compliance with those terms, inform a court's inquiry as to how a prudent fiduciary would act under the circumstances." 761 F.3d at 367 n.16. We directed the district court on remand "to factor into its causation analysis RJR's lack of compliance with the governing Plan document." Id. at 368.

The district court did precisely that. The court found that "it is more likely true than not that had a prudent fiduciary reviewed the information available to it at the time, including Plan documents, public disclosures, analysts' reports and associated research as to their significance, and newspaper articles, it would have decided to divest the Nabisco Funds at the time and in the manner as did RJR." Tatum V, 2016 WL 660902, at *23 (emphasis added). Moreover, the court recognized that RJR had not amended the Plan documents but noted that no evidence suggested that RJR intentionally disregarded Plan language or that Plan participants were unaware of the planned divestment. See id. at *24. The court then concluded that, given these facts, failure to effectuate the Plan amendment did not "affect determination of the substantive issues addressed by a prudent fiduciary." Id. at *25.2

Tatum's assertion that the court refused to consider the testimony of Professor Lys similarly lacks merit. We instructed the court to consider Professor Lys's testimony "as to what constituted a prudent investment decision." Tatum IV, 761 F.3d at 368 n.17. The court did so — dedicating an entire paragraph to Professor Lys's testimony as to how a prudent fiduciary would research an investment decision and citing Professor Lys's testimony on additional issues over a dozen times. See, e.g., Tatum V, 2016 WL 660902, at *12-*13, *18, *20-*21, *23. The district court ultimately found Professor Lys's testimony less persuasive than that of RJR's experts, id. at *13, but the court certainly considered Professor Lys's testimony.

Finally, Tatum claims that the district court violated our mandate by failing to weigh "the timing of the divestment as part of the totality-of-the-circumstances inquiry."3 But the district court's straightforward opinion belies any argument that it refused to consider the timing of the divestment. Relying on expert testimony, the court determined that a prudent fiduciary would have made the decision to divest the non-employer funds (the Nabisco Funds) and would have completed the divestment when RJR did. Id. at *23.4

In sum, the district court did not disregard our mandate. That conclusion, of course, does not resolve this appeal, or even whether the court correctly evaluated risk, Plan language, Professor Lys's testimony, and the timing of the divestment. Tatum also argues that the court erred in its evaluation of these factors, and we now turn to those arguments.

III.

Tatum principally contends that in evaluating the evidence the district court applied two incorrect legal standards. The bulk of this argument rests on the assertedly critical distinction between investment decisions and divestment decisions. Tatum maintains that because the court failed to recognize this distinction, it ignored "factors relevant to a divestment decision." Tatum also claims that the district court, despite its explicit statements to the contrary, applied the improper "could have" standard that we rejected in Tatum IV. We review a district court's factual findings for clear error and its legal conclusions and application of the law to the facts de novo. Tatum IV, 761 F.3d at 357.

A.

We can dispose of the "could have versus would have" challenge quickly. Tatum maintains that the district court must have improperly used the "could have" standard because that is the only standard that permitted the court to hold for RJR.5

That argument is wishful thinking. In remanding the case, we explicitly recognized the possibility that, using the correct "would have" standard, the district court might find that RJR had met its burden. Tatum IV, 761 F.3d at 368. We explained that "[p]erhaps, after weighing all of the evidence, the district court will conclude that a prudent fiduciary would have sold employees' existing investments at the time and in the manner RJR did because of the Funds' high-risk nature, recent decline in value, and RJR's interest in diversification." Id. If the trial evidence had permitted the court to find only that a prudent fiduciary would not have divested the Nabisco Funds, we would not have remanded the case at all. See, e.g., Breeden v. Weinberger, 493 F.2d 1002, 1012 (4th Cir. 1974) (explaining that remand is unnecessary "where the record does not contain substantial evidence to support [the] decision ... under the correct legal standard").

B.

Tatum's remaining contention as to the district court's asserted adoption of an incorrect legal standard proceeds in two parts. First, Tatum contends that the district court improperly failed to require a more demanding justification for a fiduciary's divestment decisions than for its investment decisions. Second, he argues that the court's failure to recognize this allegedly critical distinction led the court to assess the trial evidence incorrectly.

Tatum maintains that a fiduciary needs a compelling reason to divest, while the decision to invest requires less critical motivation. In support of this argument, Tatum offers only ERISA's directive that a court examine a fiduciary's decision in light of "an enterprise of a like character and with like aims," 29 U.S.C. § 1104(a)(1)(B) (2012).

The statute itself does not define this phrase. The Supreme Court, however, has explained that an ERISA "enterprise of a like character and with like aims" is broadly defined as "an enterprise with ... the `exclusive purpose' ... [of] `providing benefits to participants and their beneficiaries' while `defraying reasonable expenses of administering the plan.'" Fifth Third Bancorp v. Dudenhoeffer, ___ U.S. ___, 134 S.Ct. 2459, 2468, 189 L.Ed.2d 457 (2014) (quoting 29 U.S.C. § 1104(a)(1)(A)(i), (ii)). Given the Supreme Court's expansive definition, we can hardly hold that a trial court commits legal error when it considers evidence related to prudent investment decisions to determine the propriety of a hypothetical prudent fiduciary's divestment decisions.6 Tatum does not cite a single court that has so held, and we refuse to do so in this case.

Because the district court did not err in refusing to require a more compelling reason for divestment decisions than for investment decisions, what remains of Tatum's argument on this issue is a factual dispute over whether a prudent fiduciary would have refrained from divesting when RJR did and instead waited to see if the Nabisco Funds rebounded. The evidence presented at trial simply does not compel the conclusion that a prudent fiduciary would have done so.

At base, the district court had to determine whether RJR had shown by a preponderance of the evidence that a prudent fiduciary would have acted as RJR did. Tatum IV, 761 F.3d at 368. Contrary to what Tatum seems to imply, this does not mean that all the evidence must support RJR's decision for RJR to prevail. See Preponderance of the Evidence, Black's Law Dictionary (10th ed. 2014) (defining "preponderance of the evidence" as "evidence that has the most convincing force; superior evidentiary weight that, though not sufficient to free the mind wholly from all reasonable doubt, is still sufficient to incline a fair and impartial mind to one side of the issue rather than the other ... however slight the edge may be"). Nor does it mean that the district court must recite all the evidence it considered in reaching its holding. In a case like this one, in which the trial transcript exceeds 3,000 pages and the trial exhibits fill multiple binders, such a requirement would render the district court's already heavy burden almost impossible.

Relying heavily on his experts' testimony, Tatum particularly focuses on the timing of the divestment, arguing that a prudent fiduciary would not have divested the Nabisco Funds at the time that RJR did. Review of the record persuades us that the district court carefully considered the experts' conflicting views as to the timing of the divestment and then explained why it found RJR's experts more persuasive than Tatum's.

As the court observed, RJR's experts testified that a prudent fiduciary would have concluded that retaining the Nabisco Funds exposed beneficiaries to a high risk of future loss and reduced the diversity of the Plan. Tatum V, 2016 WL 660902, at *13-*14. This is so because of the general risk inherent in single-stock funds, especially those of a non-employer, and the idiosyncratic risk of the Nabisco Funds owing to the fact that the Nabisco stock's price fluctuations were highly correlated with the tobacco litigation. Id. Indeed, in the months after the spin-off, the Nabisco stock precipitously declined in price after a court awarded tobacco litigation plaintiffs a substantial verdict against RJR. Id. at *7, *13. This decline occurred in spite of the indemnification agreements between RJR and the Nabisco companies, and along with this continued litigation risk came "a non-negligible risk of bankruptcy" for Nabisco. Id. at *6-*7, *14. Given these facts, RJR's experts did not believe a prudent fiduciary would have concluded that the Nabisco stock presented the opportunity to achieve the extraordinary returns that might offset the stock's high risk of loss. Id. at *21-*22.

As the district court recognized, Tatum's experts reached a very different conclusion. In particular, Tatum relied on the testimony of Dr. Alan Biller, who opined that a prudent fiduciary would not have divested the Nabisco Funds. In Dr. Biller's view, divestment of a single, non-employer stock fund was justified only by "massive fraud in the company" or "[e]vidence or reason to think that the company was likely to go bankrupt," neither of which he believed existed in this case.

The court explained that it found Dr. Biller's testimony unpersuasive because Dr. Biller "did not follow his own guidance when he reached his opinions." Id. at *15. Dr. Biller had initially testified that a prudent fiduciary would consider not only analyst reports, but also "company filings with the SEC, news reports, and general commentary on the state of the company and the industry when determining whether or not to eliminate an investment option from a plan." Id. But, as the district court noted, on cross-examination Dr. Biller conceded that before forming his opinion he did not read Nabisco's financial statements or SEC filings detailing litigation liabilities, and he did not independently research news reports on the company or on the largest of the tobacco litigation cases. Id. Instead, Dr. Biller based his opinion solely on the analysts' reports and limited news stories that Tatum provided to him. Id. at *15-*16.

Aside from contradicting his own advice, Dr. Biller's heavy reliance on analysts' reports was, in the district court's view, especially problematic because the record evidence did "not suggest analyst recommendations provided meaningful investment direction." Id. at *20. Rather, the trend among analysts at the time was to recommend buying and "the number of buy recommendations [for Nabisco stock] was not statistically different than the proportion of buy recommendations for the typical stock." Id. at *21. Our review of the record substantiates the district court's assessment of the analysts' reports. For example, Professor Richard McEnally, an expert for RJR, testified that "analyst ratings and change in analyst ratings for [Nabisco stock] did not provide any meaningful basis for investment decisions from June 1999 to January 31, 2000." Id. at *20. Similarly, Dr. John Montgomery, another expert for RJR, testified that favorable analysts' reports were not associated with more favorable returns than less favorable reports, especially for larger companies like Nabisco. Id. Given this evidence and the fact that Dr. Biller's testimony was based on a narrow universe of questionable information, the court committed no clear error in rejecting Dr. Biller's conclusions.

In sum, the district court weighed all the evidence presented and came to a reasonable conclusion. The court had discussed some of this evidence in its previous opinion, but at that time it had analyzed the evidence against the less stringent "could have" standard, which made it impossible to discern whether RJR had met its burden under the "would have" standard. See Tatum IV, 761 F.3d at 368 ("[W]e must conclude that application of the incorrect legal standard may have influenced the court's decision."). Applying the more rigorous "would have" standard, the district court necessarily assessed all the evidence more critically.7 It did not err in doing so.

IV.

As his final attempt to persuade us to the contrary, Tatum contends that proper application of the efficient market hypothesis and the Supreme Court's holding in Fifth Third required the court to conclude that a prudent fiduciary would not have divested at the time and in the manner RJR did. But Tatum misunderstands the efficient market hypothesis and misreads Fifth Third. Neither compel a finding against RJR.

A.

An efficient market processes all public information about a stock and adjusts the stock's price accordingly. See Tatum V, 2016 WL 660902, at *16. Because this process occurs quickly, the efficient market hypothesis predicts that typical investors cannot exploit public information prior to the market adjusting the price. Id. (citing both parties' experts' explanation of the efficient market hypothesis). For this reason, investors in stocks trading in an efficient market generally cannot expect above-market, let alone extraordinary, returns. Id. As the Supreme Court explained in Amgen Inc. v. Conn. Ret. Plans & Tr. Funds, ___ U.S. ___, 133 S.Ct. 1184, 1192, 185 L.Ed.2d 308 (2013), "Few investors in [efficient] markets, if any, can consistently achieve above-market returns by trading based on publicly available information alone, for if such above-market returns were readily attainable, it would mean that market prices were not efficiently incorporating the full supply of public information."

Based on expert testimony, the district court found that Nabisco stock was publicly traded in an efficient market. The district court confirmed, consistent with its finding that the market was efficient, that extraordinary returns from the stock were highly unlikely. Although Tatum contends that in doing so the court faltered, he offers no argument on appeal that the market for Nabisco stock was inefficient. Moreover, he concedes that extraordinary returns are not possible in an efficient market.

B.

Nonetheless, Tatum claims that the district court's efficient market analysis is at odds the Supreme Court's reasoning in Fifth Third. There, the Court applied the efficient market hypothesis to a claim alleging that a fiduciary should have known the market was overvaluing a stock. In deeming such claims "implausible," the Supreme Court explained that, absent "special circumstances,"8 when evaluating publicly traded stocks, a fiduciary may prudently rely on the market price as "the best estimate of the value of the stocks... that is available to him." See 134 S.Ct. at 2471 (quotations marks and citation omitted). This is so because an efficient market adjusts the price of a stock in response to all public information about the stock.

Applying this principle, the district court found that a hypothetical prudent fiduciary would have properly relied on the market price of the Nabisco stock as a correct estimate of the stock's present value. See Tatum V, 2016 WL 660902, at *17-*18, *25-*26. The district court's conclusion is entirely consistent with Fifth Third and its explication of the efficient market theory hypothesis.

Tatum misreads Fifth Third to hold that a hypothetical prudent fiduciary is not justified in divesting a stock based on public information about risk. But the Court in Fifth Third held no such thing. It merely held that a fiduciary is not required to divest a high-priced stock based on public information that shows a risk of price decrease. Fifth Third, 134 S.Ct. at 2471-72. This is because, in an efficient market, a fiduciary can rely on the market price to reflect the public information about risk of loss, even if, in the beneficiaries view, the market valuation is not properly accounting for the true risk of loss.9 Id. Applied to the facts of this case, Fifth Third teaches that a prudent fiduciary would have relied on the low market price of the Nabisco stock as the current value of the stock. This conclusion provides no help to Tatum, and in fact offers support for RJR's contention that a prudent fiduciary would have chosen to divest the Nabisco Funds.

C.

Tatum misreads Fifth Third in yet another way. He claims it supports his assertion that because the low market price of the Nabisco stock reflected the risk of loss posed by the tobacco litigation, the district court erred by giving risk too much consideration as compared to Nabisco's positive attributes as a company. Tatum argues that the court's analysis was "internally inconsistent" with its application of the efficient market hypothesis. He asserts that "[b]ecause all of this information [about the tobacco litigation] was publicly disclosed before the forced divestment ..., the efficient market hypothesis requires the conclusion that these publicly disclosed future risks were already factored into the price of the Nabisco Funds." According to Tatum, "By concluding, as it did here, that positive public information did not justify expectations of future returns while negative public information justified concerns about future loss, the court committed clear error." This argument rests on flawed economic theory and a refusal to appreciate a fiduciary's role.

Fifth Third instructs that, in an efficient market, all public information, negative and positive, is reflected in a stock's market price. See 134 S.Ct. at 2472. Tatum concedes as much, but he inexplicably argues that, in considering what a prudent fiduciary would have done, the district court should have considered the positive attributes of Nabisco but not the litigation risk. This is assertedly so because the litigation risk had already been reflected in the market price of the Nabisco stock. But of course, Nabisco's positive attributes had also already been reflected in the price of the Nabisco stock. At trial, the evidence confirmed this very point and that the low price of the stock indicated that the negative information dominated the market's valuation in spite of the positive information. See Tatum V, 2016 WL 660902, at *21-*22. No case suggests that a fiduciary or a court assessing fiduciary duties should give the positive information extra consideration when the market did not do so. Such an artificial advantage for some public information at the expense of other public information flies in the face of the efficient market hypothesis.10

Although Tatum cites to Fifth Third as support for this argument, that case concerned allegations that the market overvalued a stock and that a fiduciary should have known the stock was overvalued because of public information. 134 S. Ct. at 2472. The Supreme Court did not address the loss causation analysis, which requires consideration of more than the value of a stock in determining what a prudent fiduciary would have done, and the Court certainly did not hold that a prudent fiduciary would not consider risk.

In fact, ERISA directs that prudent fiduciaries act "with the care, skill, prudence, and diligence under the circumstances then prevailing that a prudent man acting in a like capacity and familiar with such matters would use in the conduct of an enterprise of a like character and with like aims." 29 U.S.C. § 1104(a)(1) (B). And there was general agreement among experts at trial that a prudent fiduciary has a duty to consider both risk and expected returns as they affect the overall Plan and its goals. Tatum V, 2016 WL 660902, at *13.

Neither the evidence at trial nor any case supports the view that the way an efficient market processes information in setting the stock's price captures fully and accurately the goals of an ERISA plan or a fiduciary's duty. See id. § 1104(a)(1) (C); Restatement (Third) of Trusts § 90 cmt. f(1) (Am. Law Inst. 2007) ("[F]iduciaries... ordinarily have a duty to seek ... the lowest level of risk and cost for a particular level of expected return — or, inversely, the highest return for a given level of risk and cost."). Simply following the market's pricing of a stock at a given time cannot satisfy a fiduciary's duty "to conduct a regular review of its investment," Tibble v. Edison Int'l, ___ U.S. ___, 135 S.Ct. 1823, 1827-28, 191 L.Ed.2d 795 (2015), which necessarily involves judgment beyond the price. See id. at 1828 (explaining that a fiduciary must monitor investments "to ensure that they are appropriate"). The market, in incorporating the tobacco litigation risk into the Nabisco stock's price, could not consider how that risk affects the "character" or composition of the Plan, the Plan's "aims," or the fiduciary's duty to administer the Plan in the interest of the beneficiaries and comply with the Plan Documents. See 29 U.S.C. § 1104(a)(1) (B). A prudent fiduciary would have considered all of these factors.

Tatum insists that the court erred in balancing the stock's risk of loss against the possibility of extraordinary returns because such returns are not possible in an efficient market. This is precisely the point. The district court explained that a prudent fiduciary would have balanced the increased risk of loss that the Nabisco Funds brought to the Plan — risk reflected in the low stock price, but also the risk inherent in their lack of diversity within the Plan and the Nabisco stock's high correlation with RJR's battered stock — against the Funds' likely average returns. See Tatum V, 2016 WL 660902, at *16. The court found that a prudent fiduciary would have concluded the Nabisco Funds' expected returns did not justify the increased risk of loss to the Plan,11 especially because ERISA requires that a fiduciary diversify plan assets to minimize risk of loss. Id. at *25 (citing 29 U.S.C. § 1104(a)(1)(C)). This conclusion was well-supported by the record and accords with the efficient market hypothesis.

In short, the district court's findings and analysis entirely accord with the efficient market hypothesis and Fifth Third.

V.

As appealing as they may be to the specific and unusual facts of this case, Tatum's arguments create an impossible standard for fiduciaries. In Tatum's view, a hypothetical prudent fiduciary would have concluded that the Nabisco stock was undervalued and that some unknown event would occur to increase its value unexpectedly or to cease its precipitous decline so that it would increase in value.

But Tatum offered no evidence that anyone anticipated the post-divestment Icahn offer or that Nabisco would have then pursued the corporate restructuring that ultimately increased the price of the Nabisco stock. Indeed, as the district court noted, only three days before the eventual post-divestment Icahn offer, Salomon Smith Barney reported, with no mention of a potential Icahn offer, that Nabisco stock might be strong long-term but was currently depressed, volatile, and "impacted by investor concerns" about litigation. See id. at *12. The record evidence provides sufficient support for the conclusion that a prudent fiduciary would have determined that the Nabisco Funds increased the Plan's overall risk of loss and that this risk was unlikely to be justified by high returns.

This is not to say that in retrospect divesting the Nabisco Funds when the Nabisco stock was at a record low was desirable. But the fact that the price of the Nabisco stock was at a record low is apparent only in hindsight, and the decreased price of the Nabisco stock is but one factor to be considered among the totality of the circumstances. Had there been evidence that the Nabisco stock showed some resistance to RJR's tobacco litigation losses, or that Icahn's offer was all but assured, or that a prudent fiduciary would not have amended the Plan Documents, the district court, analyzing these very different facts under the more demanding "would have" standard, might well have concluded that a prudent fiduciary would have waited to divest the Funds. But Tatum offered no such evidence.

One can easily hypothesize a different lawsuit challenging divestment timing involving the same spin-off and sale of the Nabisco stock. In that case, as in this one, the price of the Nabisco stock would steadily and precipitously decline in the months after the spin-off. In that case, as in this one, Nabisco would maintain its strong fundamentals and excellent long-term prospects. And in that case, as in this one, the majority of analysts would advise investors to hold or even buy the Nabisco stock. In view of these facts, in this hypothetical case, rather than divesting in January, RJR would choose to defer divestment of the Nabisco Funds. But in the hypothetical case, instead of an unexpected offer from Carl Icahn, in late March RJR would receive a string of massive verdicts in the numerous tobacco litigation cases, causing the price of the Nabisco stock to plummet. By December, instead of trading at $30 to $50 a share, the Nabisco stock in our hypothetical case would be trading at $0.30 to $0.50 a share. It is hard to argue that the fiduciaries in our hypothetical case made the better choice. Rather, as the hypothetical illustrates, much of Tatum's timing evidence depends on facts not known (and not foreseeable) when the divestment occurred. Yet he argues that a prudent fiduciary would have anticipated these facts.

The "would have" standard does not demand such an impossible feat. Having a standard in which the fiduciary is held liable regardless of whether an outcome is foreseeable is akin to having no standard at all, eliminating the purpose of the loss causation analysis. Neither ERISA nor the efficient market theory requires that fiduciaries shoulder such burden or absorb such risk. As the Supreme Court explained in Fifth Third, a fiduciary cannot be required to predict the future. See 134 S.Ct. at 2472. For this very reason, our previous opinion left open the possibility that, when the district court applied the correct, more rigorous "would have" standard, the court would find that RJR had met its burden. Given the substantial record evidence supporting the district court's holding, we cannot conclude that the district court erred.

VI.

Accordingly, the judgment of the district court is

AFFIRMED.

DIAZ, Circuit Judge, dissenting:

I.

The last time this case was before us we said that "[t]he district court's task on remand will be to review the evidence to determine whether RJR has met its burden of proving by a preponderance of the evidence that a prudent fiduciary would have made the same decision." Tatum v. RJR Pension Inv. Comm. (Tatum IV), 761 F.3d 346, 368 (4th Cir. 2014). Though "we expressed no opinion as to the proper outcome of the case," Majority Op. at 556, we did give specific instructions for adhering to our mandate. Among other things, we directed the district court to "consider all relevant evidence, including the timing of the divestment, as part of a totality-of-the-circumstances inquiry." Tatum IV, 761 F.3d at 368. We also said that "the district court erred by failing to factor into its causation analysis RJR's lack of compliance with the governing Plan document," which "required the Nabisco Funds to remain as frozen funds in the Plan." Id. at 367-68.

Simply put, the district court did not adhere to our mandate. In particular, the court failed to explain whether a hypothetical prudent fiduciary would have made the same decisions that RJR did with respect to the timing of the divestment and the fiduciary's disregard for the governing Plan document, both of which we described in our previous opinion as "extraordinary circumstances." Id. at 368.

I respectfully dissent.

II.

A.

1.

I begin with the timing of RJR's divestment of the Nabisco Funds. As my colleagues note, "RJR informed Plan participants that the Nabisco Funds would be frozen on the date of the spin-off and divested within six months." Majority Op. at 556 (citing Tatum v. R.J. Reynolds Tobacco Co. (Tatum V), No. 1:02-cv-00373, 2016 WL 660902, at *5 (M.D.N.C. Feb. 18, 2016)). Because the timing issue is so important to this case, it's worth recounting the surrounding facts in more detail.

A "working group" of RJR employees made the initial decision to divest the Nabisco Funds at a March 1999 meeting. Tatum IV, 761 F.3d at 352. But it had no authority to do so. Id. "According to testimony from members of the working group, the group spent only thirty to sixty minutes considering what to do with the Nabisco Funds in RJR's 401(k) plan." Id. With regard to the timeline for divesting, a member of the working group testified that "[t]here was a general discussion, and different ideas were thrown out, would three months be appropriate, would a year be appropriate, and everybody got very comfortable with six months," though the group did not choose a particular date. Tatum V, 2016 WL 660902, at *4. The "rationale for" the divestment timeline was "to give employees notice and allow them to reallocate their funds." Id. at *26.

The working group's recommendation to divest in approximately six months was reported back to different RJR executives, each of whom agreed with the group's proposed timeline. Months passed, the spin-off occurred on June 14, 1999, and Nabisco's stock "declined precipitously in value." Tatum IV, 761 F.3d at 353. Then, in October 1999, RJR executives "met to discuss possible reconsideration of the decision made by the working group in March to sell the Nabisco Funds." Id. "They decided against changing course, however, largely because they feared doing so would expose RJR to liability from employees who had already sold their shares of the Nabisco Funds in reliance on RJR's prior communications." Id. Ultimately they chose January 31, 2000, as the day on which divestment would occur because "[t]he closest month-end after the six-month freeze period announced on June 14, 1999, would have been December 31, 1999," and they were concerned that "potentially widespread computer problems ("Y2K") ... could interfere with the liquidation of the" Nabisco Funds. Tatum V, 2016 WL 660902, at *8.

2.

To adhere to the mandate, the district court needed to determine whether a hypothetical prudent fiduciary would have divested on the same timeline as the one described above. The court correctly recognized that its inquiry was twofold:

(1) whether a fiduciary who properly investigated the prudence of divesting the Nabisco Funds would have made the decision that the Nabisco Funds be removed from the Plan approximately six months after the spin-off and (2) whether a fiduciary, having chosen to divest the Nabisco Funds in six months, under the duty to continually monitor the prudence of the divestment decision, would have decided to go forward with the divestment as planned in light of any changes to be considered.

Id. at *26.

Yet, in addressing the first stage of the inquiry, the district court failed to apply these standards. After reiterating its prior finding that "the timeline for divestment was ... chosen arbitrarily and with no research other than asking those at the meeting their own experience with single stock funds," the district court nonetheless determined that "once the decision was made to divest the stock, six months was a period long enough to give notice to all plan participants and coordinate the process of divestment." Id. (internal quotation marks omitted). "Thus," the court concluded, "RJR's six month time frame ... was indeed within a reasonable time frame." Id. In support of this conclusion, the court referenced its findings of fact that "[e]xperts for both sides ... agreed that once the decision to divest was made, a period long enough to give notice to all plan participants" and otherwise effectuate the divestment's procedures "would be less than a year," and that "once a decision is made to eliminate an investment option in the future, a minimum of six months is needed to allow participants the opportunity to reallocate their investments." Id. at *23 (internal quotation marks omitted).

That analysis doesn't adhere to the mandate for two reasons. First, the district court's analysis of the first prong of its inquiry relies entirely on the unsupported assumption that the time required to complete a divestment is the only factor relevant to choosing a divestment's timeline — the court discusses no other factors. See id. at *26. Similarly, the only rationale that the court found for the March 1999 working group's decision was to give employees notice and an opportunity to exit the fund at any time. Id. at *4, *26.

The court's approach also implies that other factors, including the urgency with which divestment needs to occur and the underlying reasons for divesting, are not relevant to a prudent fiduciary's calculus. That doesn't square with our reasoning in Tatum IV. There, we affirmed the district court's previous holding that RJR breached its duty of procedural prudence based in part upon the court's finding that the March 1999 working group did not consider "maintaining the stock in a frozen fund indefinitely, making the timeline for divestment longer, ... any other strategy to minimize a potential immediate loss to participants or any potential opportunity for gain," or "the purpose of the spin-off." Tatum IV, 761 F.3d at 359-61 (quoting Tatum v. R.J. Reynolds Tobacco Co. (Tatum III), 926 F.Supp.2d 648, 678, 680 (M.D.N.C. 2013)).

On remand, however, the district court effectively started from the premise that a hypothetical prudent fiduciary would have used the same analytical process as these imprudent fiduciaries, and then worked backwards from there to justify the imprudent fiduciaries' choice as a reasonable one. That's different from showing why a hypothetical prudent fiduciary conducting an independent analysis would have arrived at the same decision.

Second, the district court failed again to answer the question (which it restated verbatim) whether a hypothetical prudent fiduciary would have decided to divest approximately six months after the spin-off. And, of course, the divestment timeline is one of the "extraordinary circumstances" of this case. Id. at 368.1 Instead, the district court answered the question we rejected in Tatum IV — whether a hypothetical fiduciary could have decided to divest approximately six months after the spin-off. Cf. id. at 365 ("We would diminish ERISA's enforcement provision to an empty shell if we permitted a breaching fiduciary to escape liability by showing nothing more than the mere possibility that a prudent fiduciary `could have' made the same decision."). The district court said that "six months was a period long enough to give notice to all plan participants and coordinate the process of divestment," and that the "six month time frame ... was indeed within a reasonable time frame." Tatum V, 2016 WL 660902, at *26 (emphasis added). This language — "within a reasonable time frame" and "long enough" — smacks of "could have" rather than "would have."

And finally, if there were any doubt, one need only look to the evidence that the district court pointed to in support of its "would have" conclusion. It identified a lower bound for how long a plan participant needed to reallocate funds (six months) and a length of time universally agreed upon as sufficient to effectuate a divestment (one year). Id. at *23. Then, relying only on the former, the court concluded that six months was "long enough." Id. at *26. But absent a finding that the divestment timeline needed to be as short as possible, it makes no sense to come to a conclusion about what someone would have done based upon how quickly someone could have done it.

In short, we directed the district court on remand to answer the question whether a hypothetical prudent fiduciary would have decided to divest approximately six months after the spin-off. Regrettably, the court again answered the question we rejected — whether a hypothetical fiduciary could have done as RJR did.

B.

With respect to RJR's lack of compliance with the governing Plan document, to adhere to our mandate the district court needed to answer the question whether a hypothetical prudent fiduciary faced with the same governing Plan document would have ignored its language. See Tatum IV, 761 F.3d at 367. As my colleagues note, "[i]n November [1999], RJR attempted to amend the Plan documents to indicate that the Plan would no longer offer the Nabisco Funds." Majority Op. at 557. RJR did not properly ratify the amendment (the "November Amendment"), however, "rendering it void." Id. at 557 n.1. "Thus, at the time of the divestment, the Plan documents required that the Nabisco Funds remain frozen." Id.; see also Tatum IV, 761 F.3d at 353 n.2 ("No party challenges th[e] ruling [that the November Amendment was invalid] on appeal.").

In its opinion, the district court began the section titled "Requirements of Plan Documents" by acknowledging that the Plan documents required that the Nabisco Funds remain frozen and that we "found that it was error for [the district court] not to `factor into its causation analysis RJR's lack of compliance with the governing Plan document.'" Tatum V, 2016 WL 660902, at *24 (quoting Tatum IV, 761 F.3d at 368). By "causation analysis," we referred to "objective prudence," which exists "if a hypothetical prudent fiduciary would have made the same decision." Tatum IV, 761 F.3d at 364 (internal quotation marks omitted).

Notwithstanding our mandate that the district court focus on the "objective" in its causation analysis, the court proceeded to explain why "[t]here is no issue of Plan participants being unaware of the action anticipated nor of committee members intentionally trying to act in disregard of the Plan documents," and why there is no issue "whether any of the committee members would have hesitated to sign the consent had it been properly circulated." Tatum V, 2016 WL 660902, at *24. Indeed, the court devoted two full paragraphs to the intentions and thought processes of RJR employees Robert Gordon, Ann Johnston, McDara Folan, and Kenneth Lapiejko, none of whom is a hypothetical prudent fiduciary. See id.

The court then made its only conclusion about what a hypothetical fiduciary would have done: "Once the November Amendment document had been placed in the record, it is unlikely that a fiduciary would later find the need to look for minutes of a meeting or a consent in lieu of meeting formally adopting it." Id. The court finished by stating that "[f]ailure to obtain consent in lieu of meeting does not affect determination of the substantive issues addressed by a prudent fiduciary." Id. at *25.

That analysis doesn't adhere to the mandate because it fails to answer the mandate's question whether a hypothetical prudent fiduciary would have ignored the Plan's language.2 Rather, the district court answered the inapposite question whether, given that the actual imprudent fiduciaries took all of the actions that they did, a hypothetical prudent fiduciary would have looked for evidence of formalities adopting the November Amendment.

Instead of factoring into the causation analysis the Plan's requirement that the Nabisco Funds remain as frozen funds in the Plan, the district court found a reason to not factor that requirement into the causation analysis. Worse yet, the court's reason for doing so was that the actual imprudent fiduciaries' actions and intentions rendered the Plan's language irrelevant. Tatum IV made clear that the November Amendment's validity was not up for debate on remand, but the district court did everything short of re-litigating the issue.

Moreover, the district court's characterization of the issue as failure to obtain consent in lieu of meeting rather than failure to follow the Plan's requirement that the Nabisco Funds remain as frozen funds in the Plan is not faithful to our mandate. I understood our characterization in Tatum IV of the "requirements of the governing Plan document" as an "extraordinary circumstance[]," 761 F.3d at 368, and the mandate as a whole to mean that the Plan's requirement was an issue of substance to be tackled by the district court in the causation analysis. The court instead transformed the extraordinary into the ordinary by downplaying the flouting of the Plan's language as an issue of whether RJR's imprudent fiduciaries executed more-or-less sufficient procedures to remove that language from the Plan.

* * *

I cannot agree that the district court adhered to our mandate. Accordingly, I would vacate the judgment and remand for further proceedings.

I respectfully dissent.3