Steves and Sons, Inc. v. Jeld-Wen, Inc.

988 F.3d 690

Opinion

PUBLISHED

UNITED STATES COURT OF APPEALS FOR THE FOURTH CIRCUIT

No. 19-1397

STEVES AND SONS, INC., Plaintiff − Appellee,

and

SAMUEL STEVES; EDWARD STEVES; JOHN G. PIERCE, Counter Defendants – Appellees, v.

JELD-WEN, INC., Defendant – Appellant,

and

UNITED STATES OF AMERICA, Amicus Supporting Appellee.

Appeal from the United States District Court for the Eastern District of Virginia, at Richmond. Robert E. Payne, Senior District Judge. (3:16-cv-00545-REP)

Argued: May 29, 2020 Decided: February 18, 2021

Before DIAZ, FLOYD, and RUSHING, Circuit Judges.

Affirmed in part, vacated in part, and remanded by published opinion. Judge Diaz wrote the opinion, in which Judge Floyd and Judge Rushing joined. Judge Rushing wrote a concurring opinion.

ARGUED: Paul D. Clement, KIRKLAND & ELLIS, Washington, D.C., for Appellant. Taylor Mayly Owings, UNITED STATES DEPARTMENT OF JUSTICE, Washington, D.C., for Amicus United States of America. Benjamin Joseph Horwich, MUNGER, TOLLES & OLSON, LLP, San Francisco, California, for Appellees. ON BRIEF: Erin E. Murphy, C. Harker Rhodes IV, Erin E. Cady, KIRKLAND & ELLIS LLP, Washington, D.C., for Appellant. Lewis F. Powell III, John S. Martin, Maya M. Eckstein, R. Dennis Fairbanks, HUNTON ANDREWS KURTH LLP, Richmond, Virginia; Marvin G. Pipkin, PIPKIN LAW LLP, San Antonio, Texas; Kyle W. Mach, Emily C. Curran-Huberty, San Francisco, California, Glenn D. Pomerantz, Ted Dane, Kuruvilla J. Olasa, MUNGER, TOLLES & OLSON LLP, Los Angeles, California, for Appellees. Makan Delrahim, Assistant Attorney General, Andrew C. Finch, Principal Deputy Assistant Attorney General, Michael F. Murray, Deputy Assistant Attorney General, Kristen C. Limarzi, Kathleen Simpson Kiernan, Antitrust Department, UNITED STATES DEPARTMENT OF JUSTICE, Washington, D.C., for Amicus United States of America.

3

DIAZ, Circuit Judge:

This case arises from JELD-WEN, Inc.’s acquisition of a competitor, CMI, in 2012.

Four years later, one of JELD-WEN’s customers, Steves and Sons, Inc., filed this suit challenging the merger. After a trial, a jury found that the merger violated the Clayton Antitrust Act and that Steves was entitled to treble damages. The district court then granted Steves’s request to unwind the merger, and plans to hold an auction for the merged assets after this appeal.

The district court held another trial before a different jury on JELD-WEN’s countersuit against Steves for trade secret misappropriation. The court allowed the three individuals at the center of JELD-WEN’s allegations—Steves’s two owners and one of its employees—to intervene in the case. After the jury ruled for Steves on most of JELD- WEN’s claims, the court entered judgment for the intervenors, even though JELD-WEN had brought no claims against them.

JELD-WEN now appeals various aspects of the district court’s rulings, the bulk of which we affirm. On the one hand, the court properly declined to grant JELD-WEN judgment as a matter of law on whether Steves demonstrated antitrust injury. The court also acted within its discretion by excluding certain evidence from the antitrust trial and by ordering JELD-WEN to unwind the merger, rejecting JELD-WEN’s laches defense in the process. The court properly found that equitable relief under the Clayton Act, 15 U.S.C. § 26, was appropriate because the merger created a significant threat that Steves will go out of business in 2021. And JELD-WEN hasn’t shown that the court’s jury instructions in the trade-secrets trial were improper.

4

On the other hand, we vacate the jury’s award of future lost profits to Steves in the antitrust trial (which was meant to be a backup remedy in case divestiture doesn’t pan out) because that issue isn’t ripe. The injury on which the future lost profits award was premised can’t occur until September 2021, and the Clayton Act requires a plaintiff seeking damages—as opposed to equitable relief—to “show actual injury,” Cargill, Inc. v. Monfort of Colo., Inc., 479 U.S. 104, 111 (1986) (contrasting 15 U.S.C. § 15 with 15 U.S.C. § 26). We also vacate the district court’s entry of judgment for the intervenors in the trade-secrets case because JELD-WEN brought no claims against them.

I.

This case concerns the American “doorskin” market and events that took place between 2012 and 2016. But we must first explore the pre-2012 doorskin market.

A.

Most doors used in homes in the United States are “molded doors,” which are made by placing a wood frame and a solid or hollow core between two “doorskins” that make up the front and back of the finished door. Doorskins are made from fibrous materials (such as wood chips or sawdust) that are combined with wax or resin and then molded by a metal die into paneled designs and textures.

Steves and JELD-WEN sell molded doors. JELD-WEN also makes doorskins, some of which it uses in its own doors, and some of which it sells to other door manufacturers (the “Independents”), including Steves. Since the Independents don’t make

5

their own doorskins, they must buy them from doorskin suppliers. As of 2015, there were six Independents in this country, of which Steves was the largest.

From 2001 through 2012, there were three doorskin manufacturers in the United States: JELD-WEN, Masonite, and CMI. In 2012, Masonite had 46% market share, JELD- WEN had 38%, and CMI had 16%. All three were vertically integrated, meaning that each made their own molded doors and also sold doorskins to the Independents.

To make doors, Steves must either buy its doorskins on the “spot market” (i.e., the market for one-off purchases) or enter a long-term supply contract with a manufacturer. From 2003–2010, Steves and JELD-WEN had a long-term agreement that covered 90% of Steves’s doorskin purchases. This deal fell apart in 2010, after JELD-WEN raised its prices in response to regulatory changes. For the next two years or so, Steves bought doorskins on the spot market from Masonite and CMI, who offered Steves low prices to try to woo it into entering a long-term supply agreement.

Instead, Steves signed another long-term agreement with JELD-WEN in May 2012 (the “Supply Agreement”). In this deal, Steves committed to purchasing at least 80% of its doorskins from JELD-WEN, with one exception: if another supplier offered a price at least 3% below JELD-WEN’s, and JELD-WEN refused to match, Steves could purchase any quantity of doorskins from that supplier. Important here, prices under the Supply Agreement varied annually based on JELD-WEN’s costs. The agreement also contained quality assurances, which required JELD-WEN to reimburse Steves for damages resulting from defective doorskins, and provided for alternative dispute resolution procedures before either party could sue the other.

6

The Supply Agreement expired in December 2019, but would automatically renew for successive seven-year terms unless either party terminated it. Steves could terminate it for any reason upon two years’ notice, and JELD-WEN could do so upon seven years’ notice. Steves could also end it immediately if JELD-WEN gave notice of termination. JELD-WEN’s then-CEO told Steves that the company viewed this as a “life time [sic] deal.” J.A. 1594.

Immediately after Steves entered into the Supply Agreement, Masonite stopped selling Steves any doorskins and cancelled existing orders.

B.

Meanwhile, JELD-WEN set its sights on acquiring CMI. CMI produced doorskins and “trim board” products at a plant in Towanda, Pennsylvania. CMI’s business was profitable until 2007, and trial evidence showed that the Independents benefited from the competition between the three American doorskin manufacturers. But CMI began struggling once the housing bubble burst. After pouring their own money into the company in 2011 to keep it afloat, CMI’s owners decided to sell the business.

Many bidders showed interest, including JELD-WEN, Masonite, and Steves.

Among these, CMI identified four or five “serious prospective buyers” (which didn’t include Steves), J.A. 3447; narrowed its consideration to JELD-WEN and Masonite; and chose JELD-WEN in early 2012. Steves learned of the merger in April 2012, just before it signed the Supply Agreement.

The CMI merger reduced the number of American doorskin manufacturers from three to two. Recognizing that this could present antitrust issues, JELD-WEN didn’t notify

7

the Department of Justice’s Antitrust Division of the merger until July 2012, after it entered into long-term supply contracts with Steves and two other large Independents. JELD-WEN hoped that these contracts, by appearing to protect the Independents from price increases or refusals to sell, would allay concerns about the merger’s potential anticompetitive effects.

When it learned of the merger, the Justice Department quickly opened an investigation and reached out to Steves, who responded that it didn’t oppose the merger. The Department closed its investigation in September 2012, and JELD-WEN and CMI consummated the merger in October 2012. JELD-WEN spent significant time and resources integrating the Towanda plant’s doorskin and trim-board manufacturing processes over the next few years.

C.

Steves had issues with JELD-WEN almost immediately after signing the Supply Agreement. Starting in mid-2012, Steves noticed quality issues with JELD-WEN’s doorskins. Internal JELD-WEN documents from this time suggest that other Independents were complaining about its doorskins, too.

JELD-WEN nonetheless increased the prices that it charged Steves in 2013, 2014, and 2015. Trial evidence showed that JELD-WEN’s costs declined in each of those years, which meant that it should have reduced Steves’s prices (per the Supply Agreement), not raised them. According to Steves’s expert, JELD-WEN charged Steves almost eight percent more than what the Supply Agreement allowed in these years. Other JELD-WEN customers without a supply agreement endured even greater price increases. Additionally,

8

JELD-WEN released two new doorskin styles (the “Madison” and “Monroe” styles) in late 2012 and charged Steves more for them than the contract allowed, asserting that they were outside the scope of the agreement. Steves protested these maneuvers, but continued buying doorskins from JELD-WEN and ultimately agreed to pay more for the new styles.

Internal documents from 2013 suggest that JELD-WEN understood that the merger gave it added leverage in contract negotiations with the Independents. Specifically, a December 2013 email reflects JELD-WEN’s belief that the CMI merger left the Independents with “few options” for doorskin suppliers, J.A. 1647, and a memorandum drafted by one of JELD-WEN’s investors states that the merger “made us and Masonite the only two manufacturers of facings [i.e., doorskins] in North America, which over time will improve our pricing power,” J.A. 1644.

The parties’ relationship deteriorated further after JELD-WEN hired a new CEO, Kirk Hachigian, in early 2014. That April, Hachigian told Steves that, in his view, the Supply Agreement didn’t adequately compensate JELD-WEN for its capital improvements to its doorskin facilities. He suggested that JELD-WEN might exercise its seven-year termination option. Meanwhile, JELD-WEN’s quality issues persisted, and it also tightened its policy for reimbursements for defective doorskins. 1 Hachigian insisted that

1 JELD-WEN made two specific changes to its reimbursement policy. First, it stopped reflexively satisfying Steves’s reimbursement requests, instead scrutinizing them closely to ensure that a refund was warranted. And second, it stopped reimbursing Steves for the cost of doors that couldn’t be sold because of defective doorskins, instead refunding only the cost of the doorskin.

9

he wanted to reach a new agreement with Steves, though, so the two sides continued to negotiate through 2014 and early 2015.

In May 2014, Masonite announced that it would stop selling doorskins to the Independents. At a public presentation, a Masonite executive explained that this was “the right strategic call” and would “make sure that there are some effective barriers to entry within the [molded-door] space.” J.A. 1657. He also opined that Masonite and JELD- WEN would maintain their doorskin duopoly because of such barriers to entry, and that the continued survival of the Independents was “less likely going forward.” Id.

That August, the Steves co-owners began exchanging emails about how they should explore an antitrust claim if JELD-WEN terminated the Supply Agreement. The month after that, Hachigian gave notice of termination, effective September 2021. JELD-WEN continued to supply Steves with doorskins under the agreement in the meantime, and still does so today.

Steves then reached out to Masonite, who refused to negotiate a long-term contract, instead offering to sell on the spot market at much higher prices than it had before 2012. Steves rejected that offer. Steves also reached out to foreign suppliers, but they offered only a fraction of the designs and sizes that Steves needed and had quality and reliability issues. According to Steves, it had no choice but to keep buying doorskins from JELD- WEN.

In January 2015, JELD-WEN again increased prices for the upcoming year. JELD-

WEN and Steves then had a contentious meeting, which ended with each of them threatening to make the other’s lives miserable until their contract expired.

10

JELD-WEN internal documents from 2015 and 2016 suggest that it planned to increase its doorskin prices, end contracts with customers, and ultimately stop selling to the Independents altogether over the next few years. JELD-WEN hoped that this would “kill off a few” of the Independents, J.A. 1803, and allow it to “[i]ncrease [its molded] door market share,” J.A. 1834. Another email from Hachigian indicated that JELD-WEN would “exit all the Steves business” in 2021, when the Supply Agreement would be terminated. J.A. 1857.

D.

Meanwhile, Steves considered whether it could build its own doorskin-

manufacturing plant. In March 2015, it hired a former JELD-WEN executive, John Pierce, as a consultant. Pierce gave Steves information about JELD-WEN’s finances and manufacturing processes, which JELD-WEN alleges were trade secrets.

Steves ultimately concluded that building a plant likely wasn’t feasible. But Steves didn’t abandon its interest in building one, and it was continuing to explore the possibility when the trial in this case began.

E.

In March 2015, Steves triggered the dispute resolution procedures built into the Supply Agreement. But face-to-face conferences and mediation with JELD-WEN failed to produce a settlement. At the second conference, Steves raised antitrust concerns to JELD-WEN for the first time. That September, Steves presented JELD-WEN with a draft complaint raising contract and antitrust claims. Over the next nine months, the parties

11

signed a series of agreements that recited their desire to resolve their dispute without litigating.

In December 2015, Steves asked the Justice Department to reexamine whether the CMI merger had been anticompetitive. The Department did so, but closed its investigation in April 2016 without acting. After JELD-WEN refused to sign another agreement not to litigate, Steves filed this action in June 2016.

II.

This litigation has unfolded in many stages, which we summarize below.

A.

Steves brought breach-of-contract claims and an antitrust claim under § 7 of the Clayton Act. The Act forbids mergers whose effect “may be substantially to lessen competition,” 15 U.S.C. § 18, and authorizes treble damages for private plaintiffs who are injured by such mergers, id. § 15(a).

To support its contract claims, Steves alleged that JELD-WEN breached the Supply Agreement by overcharging Steves for doorskins of poor quality. And the basis of Steves’s antitrust claim was that the merger gave JELD-WEN too much power in the doorskin market, which emboldened it to charge higher prices, offer inferior products and customer service, and eventually try to “kill off” Steves by refusing to sell it doorskins. Steves sought (1) past damages for JELD-WEN’s alleged breaches and (2) future damages on the theory that Steves would lose access to doorskins and go out of business after September 2021, when the Supply Agreement ended.

12

In addition to damages, Steves sought to force JELD-WEN to unwind the merger and divest the Towanda plant. Divestiture is the customary form of relief in Clayton Act § 7 cases because (among other reasons) it’s “simple, relatively easy to administer, and sure.” California v. Am. Stores Co., 495 U.S. 271, 281 (1990) (cleaned up).

Courts have often ordered divestiture in suits brought by the government. See, e.g., Saint Alphonsus Med. Ctr.-Nampa Inc. v. St. Luke’s Health Sys., Ltd., 778 F.3d 775, 792– 93 (9th Cir. 2015). The Clayton Act also authorizes divestiture in private suits when it’s “appropriate in light of equitable principles,” Am. Stores, 495 U.S. at 285, to protect plaintiffs from “threatened loss or damage by a violation of the antitrust laws,” id. at 280 (quoting 15 U.S.C. § 26). But private suits seeking divestiture are rare and, to our knowledge, no court had ever ordered divestiture in a private suit before this case.

JELD-WEN moved to dismiss Steves’s antitrust claim, arguing that Steves had not plausibly alleged any antitrust injury and that its claim for injunctive relief was barred by laches. The district court denied the motion and JELD-WEN’s subsequent summary judgment motion repeating those arguments.

After learning in discovery that Steves had hired Pierce as a consultant, JELD-WEN filed trade-secrets misappropriation counterclaims against Steves. The district court severed these counterclaims for a separate trial before another jury, over JELD-WEN’s objection.

13

B.

The district court held a jury trial as to Steves’s damages claims first, with the understanding that, if the jury found that the merger was anticompetitive, the court would then hold separate proceedings on the equitable claims. 2 To prevail on its antitrust claim, Steves had to prove that the merger’s effect “may be substantially to lessen competition,” 15 U.S.C. § 18, and that the merger inflicted upon Steves a loss that “reflect[s] the anticompetitive effect” of the merger, Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 489 (1977).

“A burden-shifting analysis applies to consider the merger’s effect on competition.”

United States v. Anthem, Inc., 855 F.3d 345, 349 (D.C. Cir. 2017). First, Steves had to “establish a presumption of anticompetitive effect by showing that the transaction [led] to undue concentration” in the American doorskin market, which is typically done by comparing the market’s concentration before and after the merger. See id. (cleaned up). JELD-WEN could then rebut that presumption either by (1) discrediting Steves’s market- concentration evidence (i.e., showing that it doesn’t reflect the true level of concentration) or (2) showing that, even if Steves’s market-concentration evidence is credible, it inaccurately predicts the merger’s probable effect on competition. See id. One way to do the latter is via the “weakened-competitor” defense: that is, by showing that the acquired

2 Where a case involves legal and equitable claims that share common factual issues, the Seventh Amendment requires a jury trial on the legal claims first, with the jury’s factual findings binding the court as to the equitable claims. See Dairy Queen, Inc. v. Wood, 369 U.S. 469, 479 (1962).

14

company (here, CMI) would have been too weak to affect competition and that there were no competitively preferable alternatives to a merger with the acquiring company (here, JELD-WEN). See FTC v. Univ. Health, Inc., 938 F.2d 1206, 1221 (11th Cir. 1991).

The jury heard testimony from several lay witnesses, including executives for both parties, CMI’s former CEO, Masonite’s CEO, and executives from a private equity firm invested in JELD-WEN. The jury also heard from four expert witnesses, including Steves’s expert economist, who opined that the high concentration in the doorskin market made the merger presumptively anticompetitive.

The expert based this opinion not only on his belief that that JELD-WEN’s enhanced market power enabled it to raise prices after 2012, but also on the merger’s placement on the Herfindahl-Hirschman Index, a common measure of market concentration that is calculated by summing the squares of each firm’s share of the relevant market, Anthem, 855 F.3d at 349. According to guidelines issued by the Justice Department and the Federal Trade Commission, a market with an Index above 2,500 is highly concentrated, and a merger that increases such a market’s Index by more than 200 points is presumptively anticompetitive. Id.; see Dept. of Justice & Fed. Trade Comm’n, Horizontal Merger Guidelines § 5.3 (Aug. 19, 2010) (the “Guidelines”). While courts aren’t bound by the Guidelines, they’re “a helpful tool, in view of the many years of thoughtful analysis they represent, for analyzing . . . mergers.” Anthem, 855 F.3d at 349.

Here, the doorskin market’s Index was 3,820 before the merger and about 5,000 after the merger, a roughly 1,200-point increase—six times the threshold for presumed illegality. Cf. FTC v. H.J. Heinz Co., 246 F.3d 708, 716 (D.C. Cir. 2001) (noting that a

15

510-point increase “create[d], by a wide margin, a presumption that the merger w[ould] lessen competition”).

Over the course of the trial, the district court made several evidentiary rulings against JELD-WEN under Federal Rule of Evidence 403. First, it excluded evidence that the Justice Department had investigated the CMI merger in 2012 and late 2015 without acting, reasoning that such evidence might mislead the jury. Relatedly, while the court admitted Steves’s 2012 statement that it didn’t object to the merger and 2015 statement that the prices Steves was paying JELD-WEN had been flat, it excluded evidence that Steves made the statements to the Justice Department. Second, the court excluded evidence of CMI’s pre-merger financial distress. In the court’s view, this evidence couldn’t support a weakened-competitor defense because CMI’s doorskin business had done well before the merger and CMI had competitively preferable alternatives to merging with JELD-WEN (e.g., a sale to another bidder or a financial restructuring). Letting such weak evidence in would distract the jury, the court reasoned. And third, the court admitted evidence that Steves possessed information about how to build a doorskin plant, but forbade evidence that Steves may have acquired it by allegedly misappropriating JELD-WEN’s trade secrets.

The jury ruled for Steves on all claims. On the contract claims, the jury found that JELD-WEN had breached the Supply Agreement by overcharging Steves for doorskins. The jury awarded Steves $9.9 million in damages, in line with Steves’s expert’s computations. Additionally, the jury awarded Steves $2.2 million on its claim that JELD- WEN had breached the Supply Agreement by changing its reimbursement policies, but the

16

court later vacated that award and granted JELD-WEN judgment as a matter of law on that claim.

On the antitrust claim, the jury found that the merger violated § 7 of the Clayton Act; that this violation caused Steves to suffer an antitrust injury; and that Steves proved both past damages and future lost profits. With respect to past damages, the jury awarded Steves about $12.1 million, the same amount that it had awarded for the contract claims. 3 The jury also awarded Steves $46.4 million in future lost profits on the theory that Steves would lose access to doorskins and thereby collapse after the Supply Agreement terminates in September 2021. The jury appeared to arrive at the amount of $46.4 million by relying on one of Steves’s expert witnesses, who calculated the lost profits by extrapolating Steves’s recent annual profits over the eight-year period from September 2021 to September 2029. After trebling these awards, the court awarded Steves $36.4 million in past damages, plus $139.4 million for future lost profits, on its antitrust claim.

The court thereafter denied JELD-WEN’s motion for judgment as a matter of law on the antitrust claim and rejected its argument that Steves’s future damages were too speculative to award.

3 The court didn’t reduce the antitrust award by $2.2 million, as it had done with the contractual-damages award, because sufficient evidence supported the jury’s finding that JELD-WEN wouldn’t have tightened its reimbursement policies absent the merger.

17

C.

Next, the district court considered Steves’s claims for equitable relief. Steves’s primary request was that the court compel JELD-WEN to divest the Towanda plant. Steves also asked the court to impose several “conduct remedies” to complement divestiture. 4 To obtain equitable relief, Steves had to demonstrate:

(1) that it had suffered an irreparable injury; (2) that remedies available at law, such as monetary damages, were inadequate to compensate for that injury; (3) that, considering the balance of hardships between the plaintiff and defendant, a remedy in equity was warranted; and (4) that the public interest would not be disserved by a permanent injunction.

See eBay Inc. v. MercExchange, LLC, 547 U.S. 388, 391 (2006) (applying these “well- established principles” of equitable relief to the Patent Act). The Clayton Act tweaks the first factor by authorizing equitable relief where a plaintiff shows “a significant threat” of irreparable antitrust injury, even if the injury hasn’t happened yet. Zenith Radio Corp. v. Hazeltine Rsch., Inc., 395 U.S. 100, 130 (1969); cf. Cargill, 479 U.S. at 122 (clarifying that the threat must be one of antitrust injury).

1.

The court heard three days of testimony and two days of argument on the equitable claims. On the issue of irreparable injury, Steves presented evidence of the company’s

4 These conduct remedies included: requiring JELD-WEN to transfer the equipment and intellectual property used at Towanda to its buyer; enabling the buyer to retain Towanda’s current employees; ordering the buyer to offer Steves an eight-year supply agreement at prices based on the current Supply Agreement; allowing other Independents to terminate their supply agreements with JELD-WEN without penalty; and permitting JELD-WEN to purchase as many doorskins as it needed from the divested entity for two years before capping the number that JELD-WEN could buy annually thereafter.

18

importance to the Steves family, who had owned it for 150 years, and to its over 1,000 employees. As to the balance of hardships, the parties disputed the extent to which divestiture would harm JELD-WEN. And as to the public-interest factor, the parties disputed whether Towanda would be viable after divestiture. JELD-WEN also argued that Steves’s four-year delay in challenging the merger precluded it from relief under the laches defense.

The Justice Department also filed a statement of interest in the equitable proceedings. It asserted that divestiture is generally, but not always, the best remedy for an anticompetitive merger. The Department said that it considers five factors when evaluating whether a divestiture would be proper:

(1) whether the divestiture assets are sufficient to create a business that will replace lost competition; (2) whether the divestiture buyer has the incentive to compete in the relevant market; (3) whether the divestiture buyer has the business acumen, experience, and financial ability to compete in the relevant market in the future; (4) whether the divestiture itself is likely to cause competitive harm; and (5) whether the asset sale is structured to enable the buyer to emerge as a viable competitor.

J.A. 970.

In the Justice Department’s view, it was difficult to assess the final four factors without knowing who would buy Towanda. Further, letting Steves (who had expressed its intent to bid for Towanda) purchase the plant might not maximize competition. Doing so, the Department maintained, would create another vertically integrated doorskin supplier, rather than an independent one, and might disadvantage Steves’s competitors in the molded-door market. Relying on the Department’s statement, JELD-WEN contended that it would be improper for the court to order divestiture without first identifying a buyer.

19

2.

The district court granted the request for divestiture. It analyzed Steves’s request as follows, relying heavily on precedent from our sister circuits because we haven’t had occasion to speak on many of these issues.

First, the court determined that “divestiture should be ordered when it is the most effective way of restoring the substantially lessened competition brought about by the merger at issue and where its collateral consequences can be mitigated.” Steves & Sons, Inc. v. JELD-WEN, Inc., 345 F. Supp. 3d 614, 650 (E.D. Va. 2018). With those principles in mind, the court proceeded to apply the four-factor test set forth in eBay.

In the district court’s view, Steves satisfied the first two factors because its 150-

year-old family-owned business would likely collapse after September 2021 without equitable relief, and such a loss can’t be measured purely in monetary terms. At factor three, the court found that the threat to Steves’s survival outweighed JELD-WEN’s hardships, which—while significant—could be mitigated by ordering the divested entity to sell JELD-WEN as many doorskins as it needed for two years. And at factor four, the court found divestiture to be in the public interest because it would restore competition to the doorskin market by creating a third supplier. Alternative equitable remedies wouldn’t do, the court reasoned, because:

Divestiture would once again place three domestic doorskin suppliers in the doorskin market. Nothing in the record points to how that could be accomplished short of divestiture. Neither party has posited an alternative.

Although the Court could solve Steves’ supply problem by ordering JELD-

WEN to supply Steves’ requirements for a long term, that alternate remedy would not restore competition in the industry as a whole. And, the record

20

proves that the lessened competition has adversely affected the Independents other than Steves. So simply securing a long-term supply for Steves would not aid those manufacturers.

Even if the Court could order JELD-WEN to sell, for a period of time, to Steves and the other Independents at the prices that prevailed before JELD-

WEN secured new prices in 2014 and 2015, there still would be only one domestic supplier willing to sell to the Independents other than on a spot basis. And, there would be no structure in place to foster competition after the Court-ordered prices expired.

Steves & Sons, 345 F. Supp. 3d at 668.

The district court further rejected JELD-WEN’s argument that it should identify and vet a proposed buyer before ordering divestiture. Instead, the court fixed on a two-step process: order divestiture first, and then—if affirmed on appeal—arrange an auction with the help of a special master. The court reasoned that the Supreme Court had approved of such a process in Brown Shoe Co. v. United States, 370 U.S. 294, 309–10 (1962); that the uncertainty of a pending appeal might chill prospective buyers; that Towanda was likely to attract many bidders because it had done so in 2012 and had more value now; and that JELD-WEN may challenge the details of the divestiture process in a later appeal. 5 Finally, the district court rejected JELD-WEN’s laches defense. Laches required JELD-WEN to show both (1) that Steves unreasonably delayed in bringing suit and (2) that its delay prejudiced JELD-WEN. See PBM Prods., LLC v. Mead Johnson & Co., 639 F.3d 111, 121 (4th Cir. 2011). JELD-WEN showed neither, in the court’s view.

5 The court also granted most of the conduct remedies that Steves requested. JELD-

WEN’s briefs don’t address the propriety of these remedies.

21

As to delay, the district court found that Steves’s nearly four-year lag in bringing suit after the merger was reasonable because: (1) Steves couldn’t reasonably have been aware that it had suffered antitrust injury until August or September 2014, once Hachigian demanded an increase in prices and terminated the Supply Agreement; and (2) from then until 2016, Steves diligently exhausted alternative remedies—i.e., other supply sources and alternative dispute resolution—before suing JELD-WEN.

The court noted that JELD-WEN’s prior CEO had told Steves in 2012 that they had a “life time [sic]” deal, J.A. 1594, and that JELD-WEN didn’t back off from that sentiment until 2014. And, said the court, Steves’s pursuit of alternative solutions before suing shouldn’t be held against it because public policy supports such efforts. Further, it noted that courts often use the Clayton Act’s four-year statute of limitations for damages claims as a guideline for analyzing laches defenses to injunctive claims. See, e.g., Oliver v. SD- 3C LLC, 751 F.3d 1081, 1085–86 (9th Cir. 2014). That cut against a finding of unreasonable delay, the court explained, as Steves sued less than four years after the merger and less than two years after it became aware of its antitrust injury.

As to prejudice, the court found that JELD-WEN kept investing heavily in Towanda even after Steves presented its draft complaint (which included an antitrust claim) in September 2015. In the court’s view, this undercut JELD-WEN’s position that Steves’s delay had affected JELD-WEN’s behavior. Further, the court observed that JELD-WEN had already recovered its investment in Towanda and made a considerable profit, so it hadn’t suffered any economic harm from Steves’s (reasonable) delay.

22

D.

There was no rest for these weary litigants, as the district court also held parallel proceedings on JELD-WEN’s counterclaims, brought under the federal Defend Trade Secrets Act, 18 U.S.C. § 1836(b), and the Texas Uniform Trade Secrets Act, Tex. Civ. Prac. & Rem. Code Ch. 134A.

While the trade-secrets trial was pending, JELD-WEN sued Pierce and the two Steves co-owners (collectively, the “Intervenors”) in Texas state court on essentially the same allegations that supported its counterclaims. In response, the Intervenors moved to intervene as counter-defendants in this case under Fed. R. Civ. P. 24(b)(1)(B), which permits intervention by anyone with “a claim or defense that shares with the main action a common question of law or fact.”

The district court granted their motion in order to “bring all implicated parties together in one forum and help avoid inconsistent rulings between [the district court] and the court presiding over the Texas case.” Steves & Sons, Inc. v. JELD-WEN, Inc., 323 F.R.D. 553, 562 (E.D. Va. 2018). The issues in this case “might well have preclusive effect in the Texas case,” the court reasoned. J.A. 439. Despite the intervention, JELD-WEN didn’t amend its counterclaims to seek relief against the Intervenors, and it agreed to a Final Pretrial Order, jury instructions, and a verdict form that omitted any mention of triable issues involving the Intervenors.

The trial followed. JELD-WEN argued that Steves willfully and maliciously misappropriated sixty-seven of its trade secrets. Two of the court’s jury instructions are at issue in this appeal. The first related to trade secret 23, which “sets out a variety of factors

23

that must be monitored together to mitigate a condition [in doorskins] known as pre-cure,” 6 J.A. 1934, and is the only “combination” trade secret that JELD-WEN asserted. 7 The court instructed the jury that

as to trade secret 23, and only trade secret 23, I would like to point out that a trade secret . . . may be comprised of several elements. And a trade secret may exist if some, or even all, of its individual elements are public, provided that the trade secret as a whole remains confidential . . . .

But that applies only to trade secret 23. That does not apply to any of these other trade secrets.

J.A. 2991 (emphasis added). In the second instruction at issue, the court advised the jury that in order for a misappropriation to be “willful and malicious” under both federal and Texas law, the defendant must have an “intent . . . to cause injury or harm” to the plaintiff. J.A. 2992–93.

The jury determined that only eight of JELD-WEN’s alleged trade secrets (including trade secret 23) were protectable and misappropriated, and that Steves’s appropriation wasn’t willful and malicious (which would have yielded exemplary damages and attorney’s fees). Consequently, the jury awarded JELD-WEN $1.2 million in compensatory damages.

6 Pre-cure occurs when the surface of a doorskin isn’t consolidated with the rest of the skin before the doorskin is cured (i.e., air-dried). This may lead the surface to separate from the rest of the doorskin, causing problems in the finished door.

7 As its name implies, a “combination” trade secret consists of “a combination of characteristics and components, each of which, by itself, is in the public domain, but the unified process, design, and operation of which, in unique combination, affords a competitive advantage and is a protectable secret.” AirFacts, Inc. v. de Amezaga, 909 F.3d 84, 96 (4th Cir. 2018) (cleaned up).

24

E.

In March 2019, after all proceedings were over, the district court entered a final judgment meant to give Steves its preferred remedies while avoiding inconsistent or double recovery. As relevant here, the court:

1. Awarded Steves $36.4 million in trebled past damages on its antitrust claim, in lieu of contract damages for the same injury;

2. Ordered divestiture of Towanda, but provided that if the divestiture didn’t go through for any reason, JELD-WEN should pay Steves $139.4 million in treble antitrust damages for future lost profits; and that if those damages were set aside on appeal, JELD-WEN should pay Steves $9.9 million on its contractual claims;

3. Ordered Steves to pay JELD-WEN $1.2 million on the trade-secrets counterclaims, and also entered judgment as a matter of law for the Intervenors on those claims.

4. Ordered JELD-WEN to continue operating Towanda under a special master’s supervision pending appeal of the divestiture order.

This appeal followed. The Justice Department filed an amicus brief, arguing that (1) laches doesn’t categorically bar divestiture in a private suit filed after a merger is consummated, particularly where (as here) the plaintiff cooperated with the Department’s review before suing; and (2) there’s no evidentiary significance to the Department’s choice not to challenge the underlying merger, as there are many reasons why the Department might make that choice.

III.

JELD-WEN raises eight issues on appeal: (1) whether Steves suffered antitrust injury; (2) whether it was required to show “antitrust impact”; (3) whether the district court

25

erred in excluding evidence under Rule 403; (4) whether divestiture was the proper remedy; (5) whether Steves’s claim for future lost profits was ripe; (6) whether certain jury instructions in the trade-secrets trial were proper; (7) whether the court’s entry of judgment for the Intervenors was appropriate; and (8) whether we should reassign this case on remand. We address each of these in turn.

A.

First, JELD-WEN argues that the district court erred in denying it judgment as a matter of law on Steves’s antitrust claim for past damages because Steves didn’t prove an antitrust injury, a necessary element of standing in the antitrust context. According to JELD-WEN, Steves’s injury was a purely contractual one: It paid more and got less than the Supply Agreement called for.

Whether antitrust injury occurred is a question for the jury to decide, Int’l Wood Processors v. Power Dry, Inc., 792 F.2d 416, 431 (4th Cir. 1986), and we must uphold the jury’s finding unless no reasonable jury could have reached that conclusion, see Int’l Ground Transp. v. Mayor of Ocean City, 475 F.3d 214, 218 (4th Cir. 2007). 8

8 If we were to find that Steves has yet to suffer an antitrust injury, we would vacate the jury’s past-damages award, but not the district court’s divestiture order. This is because the divestiture order is premised on a threatened injury: Steves’s potential loss of access to doorskins in September 2021, which would drive it out of business. Such a loss would undoubtedly be an antitrust injury, as “competitors suffer antitrust injury when they are forced from the market by exclusionary conduct” like a refusal to sell. See Viamedia, Inc. v. Comcast Co., 951 F.3d 429, 482 (7th Cir. 2020).

26

1.

To bring a private antitrust suit, a plaintiff must have antitrust standing. Novell, Inc.

v. Microsoft Corp., 505 F.3d 302, 310 (4th Cir. 2007). One component of antitrust standing is “antitrust injury,” id. at 311, which requires that the plaintiff’s loss “reflect the anticompetitive effect” of the defendant’s conduct, Brunswick, 429 U.S. at 489. Antitrust injury encompasses two concepts: (1) the “causal connection” between the plaintiff’s injury and an antitrust violation, and (2) whether the plaintiff’s injury “was of a type that Congress sought to redress in providing a private remedy for violations of the antitrust laws.” Novell, 505 F.3d at 311.

The Supreme Court’s decision in Brunswick illustrates the concept of antitrust injury. There, three bowling centers claimed that a bowling-equipment manufacturer had violated § 7 of the Clayton Act by buying and operating over 200 bowling centers around the country that otherwise would have defaulted, including ones in the same markets as the plaintiffs, making the defendant the nation’s largest operator of bowling centers by far. Id. at 480–81. The plaintiffs sought damages, arguing that they would have realized more profit had the manufacturer not acquired the defaulting centers. Id. at 481.

The Court held that the plaintiffs hadn’t proved an antitrust injury for two reasons.

See id. at 488–89. First and foremost, they were complaining that the acquisitions “preserved competition, thereby depriving [the plaintiffs] of the benefits of increased [market] concentration.” Id. at 488. And second, the plaintiffs would’ve suffered “the identical loss[,] but no compensable injury[,]” if the acquired centers had stayed in business

27

without being bought by the defendant. Id. at 487. Thus, the plaintiffs’ injury didn’t “reflect” any “anticompetitive effect” of the challenged acquisitions. Id. at 489.

Cases that involve both contractual and antitrust claims present a unique challenge.

While a single act can both breach a contract and cause antitrust injury, we must ensure that Steves’s antitrust claim isn’t “simply a contract claim masquerading as a candidate for treble damages.” SAS of P.R., Inc. v. P.R. Tel. Co., 48 F.3d 39, 44 (1st Cir. 1995). We do this primarily by considering whether Steves would have suffered “an identical loss” if JELD-WEN had breached the Supply Agreement absent the merger. See Brunswick, 429 U.S. at 487; see also Valley Prods. Co. v. Landmark, 128 F.3d 398, 404 (6th Cir. 1997) (finding no antitrust injury because “the alleged antitrust violation was simply not a necessary predicate to the plaintiff’s injury”); SAS, 48 F.3d at 45 (finding no antitrust injury because the plaintiff “would have been no less damaged” if the defendant had breached the contract without committing an antitrust violation). We also consider whether competition remained “a factor in determining” how JELD-WEN treated Steves, even after they entered into their contract. See Orion Pictures Distrib. Corp. v. Syufy Enters., 829 F.2d 946, 949 (9th Cir. 1987).

2.

We agree with the district court that a reasonable jury could have found that Steves suffered antitrust injury.

28

For starters, the CMI merger hindered Steves’s access to other doorskin suppliers, which the Supply Agreement would have otherwise protected. 9 Had JELD-WEN breached the contract without merging with CMI, Steves could obviously have bought doorskins from CMI. And, viewing the evidence in Steves’s favor (as we must), it could’ve also bought from Masonite.

Specifically, a reasonable jury could have found that, if not for the merger, Masonite would’ve taken up Steves’s invitation to negotiate a contract in 2014, when Steves wanted to end its agreement with JELD-WEN. Indeed, Masonite had actively sought such a contract before the merger and stopped selling to Steves in 2012 only because of Steves’s long-term agreement with JELD-WEN. It was only after the merger, once the two doorskin suppliers shared an incentive to kill off the Independents, that Masonite stopped selling to the Independents altogether for the express purpose of killing them off and forming a duopoly with JELD-WEN.

The ability to buy from either CMI or Masonite would have mitigated Steves’s damages from JELD-WEN’s breaches. But the merger foreclosed that option. So, the loss that Steves suffered is greater than—not “identical” to, see Brunswick, 429 U.S. at 487— what it would have suffered from breaches of the Supply Agreement absent the merger.

9 The Supply Agreement was designed to protect that ability by allowing Steves to do three things: purchase up to 20% of its doorskins from other suppliers for any reason; buy any amount from other suppliers if they beat JELD-WEN’s pricing by at least 3%; and terminate the contract immediately once JELD-WEN gave notice of termination.

29

The merger also weakened the competitive pressures on JELD-WEN to provide good customer service beyond its contractual duties. For example, the contract didn’t require JELD-WEN to supply high-quality products, maintain a liberal reimbursement policy, or come within 3% of other suppliers’ prices. Competition incentivized JELD- WEN to do those things, and the merger reduced that incentive. Indeed, after the merger, JELD-WEN’s product quality declined, it became stingier with reimbursements, and it raised its prices without fear of being undercut by another supplier. Thus, competition remained “a factor in determining” how JELD-WEN treated Steves, such that JELD- WEN’s worsened products and service do “reflect the anticompetitive effect” of the merger. See Orion, 829 F.2d at 949 (cleaned up).

We also note that JELD-WEN intended to harm Steves in a way that “the antitrust laws were intended to prevent.” Brunswick, 429 U.S. at 489. Specifically, JELD-WEN sought to leverage its enhanced market power to hurt its customers, including Steves. And that intent is relevant to our antitrust-injury analysis. Novell, 505 F.3d at 316. JELD- WEN’s contractual breach should not shield it from liability for the very injury that it meant to inflict.

JELD-WEN protests that the district court didn’t rely on these rationales in affirming the jury’s finding of antitrust injury. But that doesn’t matter. “[W]e may affirm on any ground revealed in the record,” Strawser v. Atkins, 290 F.3d 720, 728 n.4 (4th Cir. 2002), and Steves presented its evidence for these theories to the jury, showing that the merger blocked its access to other suppliers and caused a decline in JELD-WEN’s product quality and customer service.

30

Further, Steves directed us to this basis for affirming in its brief, allowing JELD-

WEN to respond in its reply brief and at oral argument. In doing so, JELD-WEN also complains that the theories of antitrust injury that we now rely on to affirm the jury’s verdict don’t align with how the antitrust damages award was calculated. Specifically, the antitrust damages are calculated in the same way as the contract damages, but under Steves’s theories of antitrust injury, the two should have been calculated differently. JELD-WEN believes that this mismatch indicates that Steves didn’t suffer an antitrust injury.

Not so. At most, the alleged mismatch means that Steves’s damages award was too large. But JELD-WEN doesn’t appeal the jury’s calculation of past damages. Instead, it argues that Steves didn’t demonstrate antitrust injury, an issue that “must be distinguished from calculation of damages,” Newton v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 259 F.3d 154, 188 (3d Cir. 2001), as amended (Oct. 16, 2001). We decline to consider sua sponte whether Steves’s damages award should be amended to correlate with its antitrust injuries. See United States v. Sineneng-Smith, 140 S. Ct. 1575, 1579 (2020) (explaining that courts should “decide only questions presented by the parties”).

B.

JELD-WEN also contends that, in addition to antitrust injury, Steves had to prove “antitrust impact.” Appellant’s Br. at 28. To do so, according to JELD-WEN, Steves had to construct a hypothetical market in which the merger never happened and show how it would have been better off therein. And Steves failed to do that, JELD-WEN insists, because it didn’t try to quantify the price of doorskins in this hypothetical market. Rather,

31

Steves showed only that JELD-WEN’s profit margins would be smaller, i.e., that its prices would be closer to its costs (whatever those costs might be).

This argument fails on several levels. First and foremost, “antitrust impact” is merely a synonym for “injury” that courts use in the class-action context. Comcast Corp. v. Behrend, 569 U.S. 27, 30 (2013) (stating that “the existence of individual injury resulting from the alleged antitrust violation” is “referred to as ‘antitrust impact’”); accord In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305, 311 (3d Cir. 2008), as amended (Jan. 16, 2009). It’s not a requirement distinct from antitrust injury, as JELD-WEN suggests. See Novell, 505 F.3d at 310 (listing the factors that determine whether a plaintiff has antitrust standing, which include antitrust injury but not “antitrust impact”).

Further, JELD-WEN doesn’t appear to have asked for a jury instruction or a question on the verdict form regarding antitrust impact. And it doesn’t challenge any of the jury instructions in the antitrust trial on appeal. So, it waived this issue.

The real thrust of JELD-WEN’s argument is that, without the merger, the market price for doorskins would have been higher than what Steves actually paid—leaving Steves worse off—because the merger cut JELD-WEN’s costs. Read liberally, this argument speaks to whether the merger caused Steves’s injuries, which is a component of antitrust injury. But Steves’s theories of antitrust injury aren’t based on the market price of doorskins, but rather on its contractual damages being exacerbated and on JELD-WEN’s inferior customer service. So, it wasn’t required to prove what that price would be had the merger never happened. That distinguishes this case from Concord Boat Corp. v. Brunswick Corp., the primary case cited by JELD-WEN, where the plaintiffs’ expert

32

sought to quantify their damages via a pricing comparison to a hypothetical market. See 21 F. Supp. 2d 923, 926 (E.D. Ark. 1998) (“Dr. Hall testified that in order to quantify damages, he had to consider what the market would have looked like without the conduct.” (cleaned up)), rev’d, 207 F.3d 1039 (8th Cir. 2000). 10 Instead, Steves could prove causation by demonstrating that the merger (1) kept it from buying from other suppliers, thereby exacerbating its contract damages, and (2) disincentivized JELD-WEN from offering quality products and customer service. A reasonable jury could find that Steves succeeded in its proof.

Of course, the hypothetical-market inquiry is relevant to whether the merger was anticompetitive. JELD-WEN could have rebutted Steves’s prima facie case by convincing the jury that the merger didn’t harm pricing competition. But the jury found that the merger was anticompetitive, and JELD-WEN doesn’t dispute the reasonableness of that finding.

C.

JELD-WEN also challenges three of the district court’s decisions to exclude evidence from the antitrust trial under Rule 403. JELD-WEN insists that these errors merit vacatur of the jury’s verdicts on the antitrust claims.

Exclusion under Rule 403 is only proper when the probative value of evidence “is substantially outweighed by a danger of . . . unfair prejudice, confusing the issues,

10 Indeed, the opinions in Concord Boat Corp. don’t use the term “antitrust impact”

or suggest that it’s a standing requirement, as JELD-WEN asserts. Rather, they discuss whether the defendant’s conduct was anticompetitive and whether the plaintiffs’ expert’s testimony about his pricing comparison was properly admitted at trial. See Concord Boat Corp., 207 F.3d at 1055–57; Concord Boat Corp., 21 F. Supp. 2d at 925–28, 934.

33

misleading the jury, undue delay, wasting time, or needlessly presenting cumulative evidence.” Fed. R. Evid. 403. Nonetheless, district courts have “broad discretion” in making such determinations, which we may overturn only “under the most extraordinary circumstances, where that discretion has been plainly abused.” United States v. Udeozor, 515 F.3d 260, 265 (4th Cir. 2008) (cleaned up). We find that no such abuse occurred here.

1.

First, JELD-WEN criticizes the exclusion of evidence related to the Justice Department’s investigations of the merger. Specifically, the district court forbade evidence that the Department had twice investigated the merger without challenging it. The court also permitted evidence that Steves had stated (in 2012) that it didn’t object to the merger and (in 2015) that the prices that it was paying JELD-WEN had been flat, while barring evidence that these statements were made to the Justice Department. The court limited JELD-WEN to asking Steves’s witnesses whether Steves had made “official statement[s]” to that effect. J.A. 1881.

We conclude that the district court acted within its discretion. The Department’s decision not to pursue the matter isn’t probative as to the merger’s legality because many factors may motivate such a decision, including the Department’s limited resources. See In re High Fructose Corn Syrup Antitrust Litig., 295 F.3d 651, 664 (7th Cir. 2002) (Posner, J.) (explaining that, for this reason, defendants couldn’t use the Department’s decision not to bring an enforcement action “to show that, because the Justice Department has not moved against the alleged . . . price-fixing conspiracy, there must not have been one”); In re Carbon Black Antitrust Litig., No. 03-CV-10191-D, 2005 WL 2323184, at *1 (D. Mass.

34

Sept. 8, 2005) (“Needless to say, the defendants will not be permitted to introduce evidence on the merits that the closing of the [Department’s] investigations is somehow evidence that no [price-fixing] conspiracy exists.”).

And in general, a defendant may not use an enforcement authority’s “decision not to take action as a sword because inaction on the part of the government cannot be used to prove innocence.” Static Control Components, Inc. v. Lexmark Int’l, Inc., 749 F. Supp. 2d. 542, 556 (E.D. Ky. 2010). In short, evidence of the Department’s decision could have misled the jury into thinking that the Department deemed the merger to be legal “when no such determination ha[d] been made.” See United States v. Candelaria-Silva, 166 F.3d 19, 35 (1st Cir. 1999) (affirming the exclusion of evidence of a prosecutor’s dismissal of a prior criminal case, which the defendants wanted to introduce to show that their conduct was legal).

Similarly, the jury didn’t need to know to whom Steves’s made its statements.

Indeed, admitting that evidence might have misled the jury by calling attention to the Department’s decision not to challenge the merger.

2.

JELD-WEN also protests the exclusion of evidence regarding CMI’s pre-merger financial distress, which JELD-WEN claimed was central to its weakened-competitor defense. Specifically, JELD-WEN wanted to show that CMI lost money each year from 2008 to 2011; that its owners loaned it $36 million to keep it afloat; that it owed an additional $16.7 million to third-party lenders; and that it lost two of its largest customers in 2010, when Masonite bought them. This evidence was meant to demonstrate that, if

35

CMI hadn’t merged with JELD-WEN, CMI would’ve been a “price taker” unable to charge less than the other two suppliers, meaning that the merger didn’t impede competition. Appellant’s Br. at 67.

The district court excluded the evidence because, in the court’s estimation, it couldn’t support the weakened-competitor defense as a matter of law and thus was irrelevant. That defense requires a showing that “the acquired firm’s weakness[] . . . cannot be resolved by any competitive means” and “would cause that firm’s market share to reduce to a level that would undermine the [plaintiff’s] prima facie case.” Univ. Health, 938 F.2d at 1221. A defendant can show this “only in rare cases,” id., which is why the weakened- competitor defense has been described as “the Hail-Mary pass of presumptively doomed mergers,” ProMedica Health Sys., Inc. v. FTC, 749 F.3d 559, 572 (6th Cir. 2014); cf. FTC v. Arch Coal, Inc., 329 F. Supp. 2d 109, 153–57 (D.D.C. 2004) (one of the rare cases in which the weakened-competitor defense succeeded).

The defense has been attempted many times (usually without success) in the context of motions for injunctions or appeals of Federal Trade Commission orders, where Rule 403 doesn’t apply. See, e.g., ProMedica, 749 F.3d at 571–72. Before this case, however, the defense had never been asserted in a jury trial, so the district court made its evidentiary ruling in uncharted waters.

We conclude that the district court’s ruling was, at worst, harmless error. JELD-

WEN’s weakened-competitor defense would have failed, for two reasons. First, CMI had options besides merging with JELD-WEN that would have preserved competition in the doorskin market. In particular, it could have sold itself to one of the two or three other

36

bidders who made serious offers (not including JELD-WEN and Masonite). This would have kept a third competitor in the doorskin market. So, CMI’s issues could have been “resolved by” means that would’ve preserved competition, Univ. Health, 938 F.2d at 1221, precluding the defense.

Second, CMI’s market share wouldn’t have fallen to a level that could have undermined Steves’s prima facie case. As explained above, the merger caused a 1,200- point increase in the doorskin market’s Herfindahl-Hirschman Index, where an increase of over 200 points (in an already highly concentrated market) triggers a presumption of illegality. To rebut that overwhelming presumption via the weakened-competitor defense, JELD-WEN had to show that CMI’s market share would have dropped from 16% down to about 3% absent the merger, such that the Index would have remained within 200 points of its current score. 11 See FTC v. ProMedica Health Sys., Inc., No. 3:11-CV-47, 2011 WL 1219281, at *58 (N.D. Ohio Mar. 29, 2011) (explaining that the defendant had to show “an imminent, steep plummet in [the acquired company’s] market share . . . such that market concentration falls below levels that trigger the presumption of anticompetitive harm”).

11 To arrive at the 3% figure, we assume that Masonite (who had 46% market share in 2012) and JELD-WEN (who had 38%) would capture equal amounts of the market share that CMI lost. Based on that assumption, if CMI’s share had dropped from 16% to 3%, then Masonite and JELD-WEN would have each captured an additional 6.5%, giving them 52.5% and 44.5%, respectively. 52.52 + 44.52 + 32 = 4,745.50, which is about 250 points less than the doorskin market’s current Index. Under that same assumption, if CMI’s share had dropped to 2%, the market’s Index would’ve been 4,838 (within about 160 points of its current score). So, to prove the weakened-competitor defense, JELD-WEN had to show that CMI’s market share would have fallen to between 2% and 3%.

37

JELD-WEN’s evidence falls far short of proving that. CMI’s doorskin business remained profitable until the merger, and the housing market vastly improved between the merger and the time of trial. 12 So, unlike in cases where the weakened-competitor defense has succeeded, CMI had “convincing prospects for improvement.” Arch Coal, 329 F. Supp. 2d at 157. And, while CMI had lost major customers before the merger, there’s no indication that its market share was dropping precipitously, let alone that it could have fallen to 2%.

Of course, in assessing the merger’s effect on competition, the jury wasn’t obliged to credit the Index. See Anthem, 855 F.3d at 349. There are many ways by which JELD- WEN could have tried to rebut the Index’s presumption of illegality—for example, by arguing that the merger’s real-world effects were benign. But “[t]he weakness of the acquired firm is only relevant if the defendant demonstrates that this weakness undermines the predictive value of the [plaintiff’s] market share statistics.” Univ. Health, 938 F.2d at 1221. Because no reasonable jury could find that the proof of CMI’s weakness did that, the evidence was irrelevant.

JELD-WEN complains that the jury should have been allowed to weigh this evidence, and that Steves could have pointed out its flaws to the jury. But the district court had a duty—which was especially salient in a long and complex trial like this one—to keep

12 The district court, of course, was considering a different economic climate than the one we face now in light of COVID-19. Throughout this opinion, we review the district court’s analysis based on the facts that were before it, without considering how the housing market has changed in the meantime.

38

out weak evidence that carried a substantial risk of undue delay, confusing the issues, and misleading the jury. By determining that CMI’s financial distress fit that description, the court acted within its discretion, as admitting the evidence could have prompted a “trial within a trial” about CMI’s prospects in 2012. See Fry v. Rand Constr. Corp., 964 F.3d 239, 249 (4th Cir. 2020) (finding that a trial court properly excluded evidence under Rule 403). And, in any event, the court’s ruling was harmless because the evidence was legally insufficient to support the weakened-competitor defense. See Fed. R. Civ. P. 61.

3.

Finally, JELD-WEN asserts that it should have been permitted to tell the jury (at the antitrust trial) about Steves’s alleged trade-secrets misappropriation. According to JELD- WEN, the steps Steves took to get these trade secrets showed how serious it was about building a doorskin plant, which was allegedly relevant to whether Steves could solve its supply issues by 2021 and thus avoid the injury on which its claim for future lost profits is based.

We think that the district court acted properly in letting the jury hear about the information Steves obtained, but excluding evidence of how Steves obtained it. This allowed the jury to make an informed judgment as to Steves’s ability (or lack thereof) to construct its own doorskin plant. Alerting the jury to the trade-secret allegations would have risked confusion and unfair prejudice to Steves.

***

In short, none of the Rule 403 rulings that JELD-WEN challenges merit vacatur of the antitrust verdict.

39

D.

Next, JELD-WEN attacks the divestiture order on the grounds that the district court improperly denied its laches defense and misapplied the factors governing equitable relief. We address laches first.

1.

Laches is a defense to a divestiture request. See Am. Stores, 495 U.S. at 296. For the defense to succeed, JELD-WEN must prove both (1) that Steves unreasonably delayed in bringing suit and (2) that Steves’s unreasonable delay prejudiced JELD-WEN. See PBM Prods., 639 F.3d at 121. The district court found that JELD-WEN satisfied neither element, and we review that finding for abuse of discretion. Id. at 120.

As to unreasonable delay, JELD-WEN makes three points. None is persuasive.

a.

First, JELD-WEN contends that a nearly four-year delay after a merger’s consummation is presumptively unreasonable. We disagree. Laches turns on “the particular circumstances of the case,” White v. Daniel, 909 F.2d 99, 102 (4th Cir. 1990), militating against a singular focus on a merger’s closing date. And we measure delay not from the date of the challenged action, but from when the plaintiff “discovers or with reasonable diligence could have discovered the facts giving rise to his cause of action,” id.,

40

and was “able to pursue a claim,” Ray Commc’ns, Inc. v. Clear Channel Commc’ns, Inc., 673 F.3d 294, 301 (4th Cir. 2012). 13 Some courts have relied on laches to dismiss post-consummation challenges to mergers. See Midwestern Mach. Co. v. Northwest Airlines, Inc., 392 F.3d 265, 277 (8th Cir. 2004); Fed. Home Loan Bank Bd. v. Elliott, 386 F.2d 42, 54 (9th Cir. 1967); Antoine L. Garabet, M.D., Inc. v. Antomonous Techs. Corp., 116 F. Supp. 2d 1159, 1171–73 (C.D. Cal. 2000). None of the plaintiffs in those cases, however, offered a good excuse for their delay. See Midwestern Mach. Co., 392 F.3d at 277 (“Midwestern produced no reasonable justification for the eleven-year delay in filing suit.”); Elliott, 386 F.2d at 54 (stating that the plaintiffs “were on notice of all material facts” three months before they filed suit, yet “they deliberately chose to wait until the merger had been effectuated”); Garabet, 116 F. Supp. 2d at 1172–73 (explaining that the plaintiffs “failed to exercise proper diligence in the pursuit of their claim(s)”). So, those cases don’t support a singular focus on the date that a merger is consummated.

Nor is such a focus warranted by the hardships of unwinding a completed merger.

While those hardships factor into the prejudice stage of the laches analysis, they don’t obviate our need to consider whether the plaintiff’s delay was unreasonable. And even if a defendant’s laches defense fails, it can still prevent divestiture by showing that the

13 If there is a presumptive deadline for Clayton Act injunctive claims, it’s four years after the challenged action, tracking that statute’s limitations period for damages claims. See, e.g., Oliver, 751 F.3d at 1085–86. This is because, in applying laches, courts of equity “usually act or refuse to act in analogy to[] the statute of limitations relating to actions at law of like character.” See King v. Richardson, 136 F.2d 849, 862 (4th Cir. 1943) (cleaned up). And here, Steves filed suit less than four years after the merger closed.

41

balance of hardships (one of the four equitable factors) tips in its favor. See Ginsburg v. InBev NV/SA, 623 F.3d 1229, 1235–36 (8th Cir. 2010) (dismissing a divestiture claim because of the effect of the plaintiff’s delay on the balance of hardships); Taleff v. Sw. Airlines Co., 828 F. Supp. 2d 1118, 1123–24 (N.D. Cal. 2011) (same). Thus, there’s no need for the hardships of unwinding a merger to bleed into our review of whether Steves’s delay was reasonable.

b.

JELD-WEN’s second argument is that Steves had notice of its injury right after the merger was announced and thus shouldn’t have waited until fall 2014 to pursue relief. Specifically, Steves knew that the merger, by removing a competitor from the market, would hinder it from buying doorskins from other suppliers and weaken JELD-WEN’s incentive to provide good service. Further, Masonite stopped selling Steves any doorskins in 2012, so for those next two years, Steves knew that its only option was to buy from JELD-WEN—yet Steves didn’t seek relief.

Again, we disagree with JELD-WEN. It’s true that Steves knew about the two injuries that support its past-damages claim in 2012. But Steves lacked notice of the threatened injury on which its divestiture claim is based—its potential loss of access to doorskins in 2021—until 2014, when JELD-WEN indicated that it was terminating the Supply Agreement and Masonite announced that it would stop selling to the Independents entirely. Before then, Steves’s access to doorskins was contractually protected for the foreseeable future. The Supply Agreement was set to renew perpetually, and JELD- WEN’s CEO had referred to it as a “life time [sic]” deal, J.A. 1594.

42

Moreover, Masonite had previously sought a long-term agreement with Steves, so Steves had reason to believe that it had a fallback if its relationship with JELD-WEN soured. That fallback vanished in 2014, when Masonite announced its strategy to kill off the Independents. JELD-WEN’s notice of termination and Masonite’s announcement are key “facts giving rise to [Steves’s] cause of action,” which Steves couldn’t have discovered before 2014. See White, 909 F.2d at 102.

The injuries that Steves suffered prior to 2014 wouldn’t have supported a divestiture claim. Absent the threat to its survival that emerged only then, Steves couldn’t have shown any of the first three eBay factors—a threatened irreparable injury, inadequacy of legal remedies, and that the balance of hardships tipped in its favor—because its earlier injuries were compensable by money damages (as evidenced by the award that Steves received in this case). Cf. Nat’l Viatical, Inc. v. Universal Settlements Int’l, Inc., 716 F.3d 952, 957 (6th Cir. 2013) (“[T]he general rule is that a plaintiff’s harm is not irreparable if it is fully compensable by money damages.” (cleaned up)).

“Logic dictates that ‘unreasonable delay’ does not include any period of time before” Steves was “able to pursue a claim.” Ray Commc’ns, 673 F.3d at 301. And, as the Supreme Court has explained, laches doesn’t require a plaintiff to “sue soon, or forever hold [their] peace.” Petrella v. Metro-Goldwyn-Mayer, Inc., 572 U.S. 663, 682 (2014) (cleaned up). In other words, a plaintiff need not challenge an illegal act immediately after it happens; it may wait until it “can estimate whether” the act threatens it with irreparable harm. See id. at 682–83. Thus, it was reasonable for Steves to wait to pursue relief until

43

2014, when it learned that the merger threatened its access to doorskins (and thus its survival) after September 2021.

c.

JELD-WEN’s last argument about delay is that Steves lacks a good excuse for not seeking divestiture between 2014 and 2016. But evidence supports the district court’s finding that Steves spent that time diligently exhausting its alternative remedies. Specifically, Steves reached out to Masonite and foreign suppliers, explored building its own doorskin plant, engaged in settlement talks and mediation with JELD-WEN, and asked for (and cooperated with) a Justice Department investigation. Moreover, between September 2015 and June 2016, JELD-WEN signed a series of agreements with Steves reciting their mutual desire to settle their dispute. “It would disserve the strong policy in favor of nonjudicial dispute resolution if [a] defendant successfully could assert that [a] . . . period of settlement attempts”—i.e., efforts to find nonjudicial remedies—“contributes to the establishment of laches,” particularly when the defendant has expressed a desire to settle. Piper Aircraft Corp. v. Wag-Aero, Inc., 741 F.2d 925, 932 (7th Cir. 1984).

In short, the district court didn’t abuse its discretion in finding that Steves’s delay was reasonable, and thus properly denied JELD-WEN’s laches defense. As JELD-WEN didn’t prove unreasonable delay, we need not address whether the delay prejudiced JELD- WEN.

2.

JELD-WEN also contends that the district court misapplied each of the four equitable factors, which are as follows (as applied to a Clayton Act suit):

44

A plaintiff must demonstrate: (1) that it [faces a significant threat of]

irreparable [antitrust] injury; (2) that remedies available at law, such as monetary damages, are inadequate to compensate for that injury; (3) that, considering the balance of hardships between the plaintiff and defendant, a remedy in equity is warranted; and (4) that the public interest would not be disserved by a permanent injunction.

See eBay, 547 U.S. at 391; see also Cargill, 479 U.S. at 122 (holding that a Clayton Act plaintiff “must show a threat of antitrust injury” to warrant injunctive relief); Zenith, 395 U.S. at 130 (indicating that the threat of antitrust injury must be “significant”). Steves had the burden of proving all four factors, and JELD-WEN insists that it proved none.

We review an “award of equitable relief for abuse of discretion,” Solis v. Malkani, 638 F.3d 269, 274 (4th Cir. 2011) (cleaned up), including a decision to enjoin a merger under the Clayton Act, Anthem, 855 F.3d at 352–53. “A district court abuses its discretion when it relies on incorrect legal conclusions or clearly erroneous findings of fact, or otherwise acts arbitrarily or irrationally in its ruling.” SAS Inst., Inc. v. World Programming Ltd., 874 F.3d 370, 385 (4th Cir. 2017) (cleaned up).

a.

We address the first two eBay factors together, as the district court did. They required Steves to “demonstrate a significant threat,” Zenith, 395 U.S. at 130, of an irreparable antitrust injury that damages couldn’t cure, see eBay, 547 U.S. at 391; Cargill, 479 U.S. at 122. According to the district court, Steves’s potential collapse after September 2021 was such an injury. JELD-WEN now asserts that this injury could be remedied by either monetary damages or equitable relief that is less drastic than divestiture.

45

We affirm the district court’s finding that Steves’s threatened collapse couldn’t be repaired by money damages. The permanent loss of a business, with its corresponding goodwill, is a well-recognized form of irreparable injury. See, e.g., Warren v. City of Athens, 411 F.3d 697, 711–12 (6th Cir. 2005). Of course, not every company’s failure will warrant equitable relief. For example, the dissolution of a new enterprise, or one that’s not very important to its owner (who may own many companies), may be reparable by damages. See, e.g., DFW Metro Line Servs. v. Sw. Bell Tel. Co., 901 F.2d 1267, 1269 (5th Cir. 1990) (“The lost goodwill of a business operated over a short period of time is usually compensable in money damages.”).

But here, Steves has been family-owned for 150 years. The right to continue a multi-generational family business “is not measurable entirely in monetary terms; the [Steveses] want to sell [doors], not to live on the income from a damages award.” See Semmes Motors, Inc. v. Ford Motor Co., 429 F.2d 1197, 1205 (2d Cir. 1970) (Friendly, J.). Thus, the district court didn’t abuse its discretion in finding that damages couldn’t repair Steves’s threatened injury. 14 A tougher question is whether the court should have chosen a different equitable remedy. As a “general rule,” “injunctive relief should be no more burdensome to the defendant than necessary to provide complete relief to the plaintiffs.” Madsen v. Women’s

14 The district court’s award of future damages as an alternative—which we vacate for unrelated reasons, as explained below—doesn’t undercut its irreparable-harm finding. As the district court recognized, while a damages award would be better than no relief at all, it wouldn’t fully repair Steves’s injury because it wouldn’t keep Steves in business.

46

Health Ctr., Inc., 512 U.S. 753, 765 (1994) (cleaned up). Here, the court acknowledged that it could protect Steves by ordering JELD-WEN to supply Steves’s requirements at fair prices going forward. But the court chose divestiture instead, reasoning that it was needed to restore competition in the doorskin market and that a conduct remedy would provide only temporary relief before eventually expiring. In JELD-WEN’s view, the availability of a less drastic remedy forecloses divestiture, and the court erred by considering what would best promote competition.

We disagree with JELD-WEN for two independent reasons. First, the district court reasonably found that a conduct remedy would only protect Steves temporarily. After it expired, “there would be no structure in place to foster competition.” Steves and Sons, 345 F. Supp. 3d at 668. The threat to Steves’s survival would persist, as there would be only two American doorskin manufacturers, each of whom would be vertically integrated. Without divestiture, “the threat to [Steves] inherent in the [merger] would [not] cease in the foreseeable future.” See Zenith, 395 U.S. at 131. Thus, a conduct remedy wouldn’t give Steves “complete relief,” Madsen, 512 U.S. at 765 (cleaned up), against the “threatened loss or damage” for which it seeks divestiture, 15 U.S.C. § 26.

Second, JELD-WEN’s argument conflicts with Clayton Act principles. The Act authorizes injunctive relief in private suits “not merely to provide private relief, but to serve as well the high purpose of enforcing the antitrust laws”—i.e., protecting competition. Am. Stores, 495 U.S. at 284 (cleaned up). Indeed, private enforcement of antitrust laws is an “integral part of the congressional plan for protecting competition,” id., so courts may fashion equitable remedies with that broader purpose in mind. A remedy that helped only

47

Steves wouldn’t promote competition in the doorskin market, conflicting with the principle that antitrust law protects competition, not competitors, Brunswick, 429 U.S. at 488.

Further, if courts were required to choose the remedy least burdensome to the defendant—rather than the one that best promotes competition—conduct remedies would be the norm because they generally burden defendants less. But that would go against Congress’s policy judgment that divestiture is “the remedy best suited to redress the ills of an anticompetitive merger,” Am. Stores, 495 U.S. at 285, as well as the principle that conduct remedies are disfavored because they “risk excessive government entanglement in the market,” Saint Alphonsus, 778 F.3d at 793; accord ProMedica, 749 F.3d at 573.

It’s telling that in American Stores—the seminal case finding that divestiture is authorized in private suits—the Supreme Court listed antitrust standing, laches, and unclean hands as the unique obstacles facing private plaintiffs. See 495 U.S. at 296. Conspicuous in its absence from that list is the equitable principle that JELD-WEN clings to now, which would be the greatest obstacle of all.

In sum, the district court didn’t abuse its discretion by finding that Steves’s threatened injury was irreparable by money damages or conduct remedies, or by considering which equitable remedy would best promote competition.

b.

The third eBay factor required the district court to “balance the hardships” that divestiture would cause JELD-WEN against those which Steves would suffer in its absence. See 547 U.S. at 391. JELD-WEN maintains that the district court gave short shrift to its hardships, improperly faulted JELD-WEN for not quantifying them, and

48

assigned too much weight to Steves’s “speculative” claim that it won’t survive absent divestiture, Appellant’s Reply Br. at 15.

Again, we conclude that the district court acted within its discretion. Record evidence supports the court’s finding that Steves faces collapse without injunctive relief. If neither JELD-WEN nor Masonite sell Steves doorskins, Steves will have no recourse. Such a “significant possibility” that a party “would be driven out of business” carries great weight when balancing hardships. See Buffalo Courier-Express, Inc. v. Buffalo Evening News, Inc., 601 F.2d 48, 58 (2d Cir. 1979) (Friendly, J.).

The court also analyzed the evidence of JELD-WEN’s hardships thoroughly. It recognized that divestiture would cost JELD-WEN a great deal financially and would reduce its doorskin output. But JELD-WEN could weather these hardships, the court found, because it was much larger and more diversified than Steves. For instance, JELD- WEN had several other doorskin plants and had just opened a new one. None of them were running at full capacity, and one wasn’t being operated at all and could be reactivated within two years. And the court could ease JELD-WEN’s transition by ordering Towanda’s new owner to supply JELD-WEN’s doorskin requirements for two years. JELD-WEN’s hardships, while significant, were outweighed by the “far more serious harm” facing Steves. Steves and Sons, 345 F. Supp. 3d at 662. We see no reversible error in this analysis.

Additionally, the district court discounted some of JELD-WEN’s claimed hardships as too speculative. For example, JELD-WEN posited that divestiture could cause layoffs at Towanda and at its other plants. In the court’s view, this was both unsupported by the

49

record and illogical because Towanda’s new owner would want to retain its employees and JELD-WEN would have to shift more production to its remaining plants. JELD-WEN also surmised that it would have to pay contractual penalties for failing to meet other customers’ supply needs, but the district court found this unlikely, noting that those contracts contained force majeure clauses that encompass the divestiture order.

The court’s dismissal of such conjecture did not constitute impermissible burden-

shifting. It was proper for JELD-WEN to bear the burden of proving its own hardships, as such facts were “peculiarly in [its] knowledge.” See Smith v. United States, 568 U.S. 106, 112 (2013) (cleaned up).

c.

The fourth factor required Steves to show “that the public interest would not be disserved by a permanent injunction.” eBay, 547 U.S. at 391. The district court found divestiture to be in the public interest because it would add a third supplier to the doorskin market, thereby promoting competition. JELD-WEN asserts (1) that the court wasn’t equipped to assess this factor without knowing who would buy Towanda (or at least knowing more about the likely bidders); (2) that Towanda won’t compete effectively after it’s divested, so the remedy will backfire; and (3) that letting Steves buy Towanda (which it has expressed interest in doing) won’t increase competition.

JELD-WEN’s first argument requires some unpacking. In this case, the district court undertook a two-step process of ordering divestiture first, and then (if affirmed on appeal) holding an auction with a special master’s help. This approach is commonly taken in suits filed by the Federal Trade Commission. The Supreme Court recognized its virtues

50

in Brown Shoe, 370 U.S. at 309–310, and since then, our sister circuits have often affirmed divestiture orders in government suits where important details—like who would buy the divested entity—had not been sorted out yet. See Saint Alphonsus Med. Ctr.-Nampa Inc. v. St. Luke’s Health Sys., Ltd., Nos. 1:12-CV-560 and 1:13-CV-116, 2014 WL 407446 (D. Idaho Jan. 24, 2014), at *26 (ordering a company “to fully divest itself” of assets acquired in a merger and “take any further action needed to unwind the [a]cquisition,” without addressing who would buy the assets), aff’d, 778 F.3d at 792–793; In re ProMedica Health Sys., Inc., No. 9346, 2012 WL 2450574, at *6 (F.T.C. June 25, 2012) (ordering a company to fully divest merged assets “to an [a]cquirer that receives the prior approval of the [Federal Trade] Commission”), aff’d sub nom. ProMedica, 749 F.3d at 573.

JELD-WEN contends that a two-step process is inappropriate in private suits like this one. We disagree. In both government and private suits, a court may order divestiture if it’s needed to “restore competition,” i.e., to “protect the public interest.” United States v. E.I. du Pont de Nemours & Co., 366 U.S. 316, 326 (1961). If a court can properly assess the public interest in a government suit without having found a buyer, it can also do so in a private suit. Brown Shoe’s rationale for approving of a two-step process applies with equal force here: Potential buyers may hesitate to place bids while a lengthy appeal looms. See 370 U.S. at 309. Indeed, over two years have passed between the divestiture order and our decision, and more time will pass if JELD-WEN presses its appeal further. Delaying the auction until this appeal wraps up will attract more buyers and thus serve the public interest.

51

The facts of this case also support the use of a two-step process. The district court reasonably found that Towanda will likely attract a buyer capable of competing with JELD- WEN and Masonite. Several serious bidders emerged in 2012, when both CMI and the molded-door market as a whole were doing poorly. Since then, demand for doorskins has increased dramatically and the Towanda plant has undergone capital improvements, making it an even more appealing investment. In other cases, where a buyer is less likely to emerge, a two-step process may be inadvisable because the appeal is likely to be a waste of time that prejudices the defendant. But on this record, the district court acted within its discretion by ordering such a process.

JELD-WEN also faults the district court for not requiring Steves to submit more proof of other bidders’ interest. There’s some merit to this critique. Perhaps there were ways, short of an auction, to confirm with greater certainty that multiple bidders would emerge. The court could have, for instance, sought input from the parties that placed bids for CMI in 2012, or asked the Justice Department (or another expert) to share its expectations of how the auction process might unfold. But the district court’s failure to take such steps here wasn’t an abuse of discretion, as no case law required them, and ample evidence supported a finding that buyers would likely emerge.

We turn now to JELD-WEN’s second public-interest argument. As our sister circuits and the Justice Department have recognized, a key factor in divestiture analysis is whether the divested entity will be “a willing, independent competitor capable of effective production in the [relevant] market.” White Consol. Indus., Inc. v. Whirlpool Corp., 781 F.2d 1224, 1228 (6th Cir. 1986); see also J.A. 970 (listing the factors that the Justice

52

Department considers before pursuing divestiture). JELD-WEN disputes the district court’s finding that a divested Towanda can compete effectively. As this is a factual finding, we review it for clear error. See Equinor USA Onshore Props. Inc. v. Pine Res., LLC, 917 F.3d 807, 813 (4th Cir. 2019).

We conclude that the district court’s finding wasn’t clear error because substantial evidence supports it. Specifically, CMI had 16% market share before the merger, making it a significant player in the market. It thrived until the housing bubble burst, and even afterward, its doorskin business remained profitable until the merger. There was ample interest in buying CMI in 2012, suggesting that investors believed in its ability to compete even when the doorskin market wasn’t doing well. And since then, JELD-WEN has made capital improvements to Towanda, helping the plant generate substantial profits. Demand for doorskins has also increased, creating more room for a third supplier. Further, trial evidence showed that, since 2012, JELD-WEN and Masonite’s prices have increased, while their product quality has decreased. So, even if a divested Towanda’s market share doesn’t exceed the 16% enjoyed by CMI pre-merger, its presence should still stoke competition and thereby serve the public interest.

JELD-WEN posits that Towanda won’t be profitable without relying on JELD-

WEN’s sales, administration, and technology services. But the eventual buyer of Towanda may offer similar or better services. While this is a factor that the special master should consider when it holds an auction, it’s not a reason to reject divestiture outright.

Lastly, we consider the possibility that Steves may buy Towanda. So far, Steves is the only entity to have expressed interest in doing so. As JELD-WEN points out, there’s a

53

tension between Steves’s desire to pay as little as it can for Towanda and the public’s interest in finding the best purchaser. Indeed, Steves’s preference for divestiture in lieu of a $139.4 million damages award may signify that it hopes to buy Towanda at a bargain price. But Steves won’t run the auction process. The special master will be charged with soliciting bids. So, we need not worry about any conflict of interest.

If Steves does buy Towanda, each of the three doorskin suppliers would be vertically integrated. That’s not ideal for promoting competition, as the three suppliers would share a collective incentive not to sell to the Independents. But three is better than two. And reducing market concentration generally promotes competition because, “where rivals are few, firms will be able to coordinate their behavior, either by overt collusion or implicit understanding, in order to restrict output and achieve profits above competitive levels.” Heinz, 246 F.3d at 715 (cleaned up). Indeed, a “duopoly . . . is presumptively unlawful in and of itself.” ProMedica, No. 3:11-CV-47, 2011 WL 1219281, at *56.

The facts of this case bear out those principles, as trial evidence shows that there was more competition in the doorskin market pre-merger, even though all three suppliers at the time were vertically integrated. Thus, the chance that Steves may buy Towanda doesn’t render the divestiture order an abuse of discretion.

***

In sum, the record supports the district court’s findings as to each of the equitable factors. Of course, the divestiture process is far from over. If the special master can’t locate a satisfactory buyer, the district court may have to revisit its ruling. And, when a

54

buyer is selected, JELD-WEN may challenge whether a sale to that particular buyer will serve the public interest.

But as it stands, this case is a poster child for divestiture. A merger has resulted in a duopoly. Each doorskin supplier is vertically integrated. Evidence indicates that they’ve used their market power to threaten the Independents’ survival. And it’s reasonable to expect that a third supplier—even one that’s vertically integrated—will promote competition, as CMI did before the 2012 merger. Thus, the district court acted within its discretion by ordering divestiture.

E.

JELD-WEN also challenges the $139.4 million damages award for future lost profits, which would only kick in if divestiture doesn’t occur. Unlike with its claim for equitable relief, Steves had to “show actual injury” to obtain damages, not merely threatened injury. See Cargill, 479 U.S. at 111. We conclude that, because Steves has not yet suffered the injury on which its claim for future lost profits rests, this claim wasn’t ripe for adjudication and thus should have been dismissed without prejudice.

“The doctrine of ripeness,” which is rooted in both Article III and prudential doctrines, “prevents judicial consideration of issues until a controversy is presented in clean-cut and concrete form.” Lansdowne on the Potomac Homeowners Ass’n, Inc. v. OpenBand at Lansdowne, LLC, 713 F.3d 187, 198 (4th Cir. 2013) (cleaned up). As ripeness is a jurisdictional issue, Sansotta v. Town of Nags Head, 724 F.3d 533, 548 (4th Cir. 2013), we must consider it whenever it draws our attention, including after the plaintiff has been awarded judgment. See Lansdowne at 198–99 (reviewing ripeness argument after

55

the plaintiff had been granted summary judgment). “To determine if a case is ripe, we balance the fitness of the issues for judicial decision with the hardship to the parties of withholding court consideration.” Id. at 198 (cleaned up).

We begin with fitness. As relevant here, a claim is fit for adjudication when it’s “not dependent on future uncertainties.” See id.; see also Thomas v. Union Carbide Agric. Prods. Co., 473 U.S. 568, 580–81 (1985) (indicating that a claim is unripe if it “involve[s] contingent future events that may not occur as anticipated, or indeed may not occur at all” (cleaned up)).

Steves’s claim for future lost profits depends on a future uncertainty: whether it will have access to doorskins after September 2021. And this uncertainty is especially problematic because it’s entirely within JELD-WEN’s control. See SureShot Golf Ventures, Inc. v. Topgolf Int’l, Inc., 754 F. App’x 235, 240–41 (5th Cir. 2018) (dismissing an antitrust damages claim as unripe because it was based on an allegation that the defendant, who had just acquired the plaintiff’s supplier, would refuse to deal with the plaintiff in the future); Volvo N. Am. Corp. v. Men’s Int’l Prof’l Tennis Council, 857 F.2d 55, 64–65 (2d Cir. 1988) (holding that an antitrust challenge to rules that the defendant had proposed (but not yet adopted) was ripe only where the possibility of the rules’ adoption had “a present anti-competitive effect”); cf. Clapper v. Amnesty Int’l USA, 568 U.S. 398, 413 (2013) (stating that the Supreme Court is “reluctant to endorse standing theories that require guesswork as to how independent decisionmakers will exercise their judgment”).

Specifically, JELD-WEN could offer to supply Steves after September 2021 on terms agreeable to Steves, which would prevent the injury on which this claim is premised.

56

If JELD-WEN were to take that route, the damages award would be superfluous. Of course, the jury found that JELD-WEN planned not to do that, but JELD-WEN might well change its tune if the alternative is to pay Steves almost $140 million. And Steves’s duty to mitigate damages may oblige it to accept a reasonable offer. Cf. Golf City, Inc. v. Wilson Sporting Goods, Co., Inc., 555 F.2d 426, 436 (5th Cir. 1977) (“An antitrust plaintiff has a duty to mitigate damages.”).

It’s true that any future-lost-profits claim involves uncertainty as to the extent of those lost profits. But “[p]roof of injury (whether or not an injury occurred at all) must be distinguished from calculation of damages (which determines the actual value of the injury).” Newton, 259 F.3d at 188. The problem here is that Steves’s potential injury—its loss of access to doorskins after September 2021—hasn’t “occurred at all.” See id.

In rejecting JELD-WEN’s ripeness argument, the district court found that Steves’s future damages will flow from JELD-WEN’s notice of termination in September 2014. But that’s beside the point. JELD-WEN had every right to give such notice; doing so didn’t inflict an actual injury (as opposed to a threat of future injury) upon Steves. Indeed, JELD- WEN and Steves terminated a previous contract in 2010 before negotiating the Supply Agreement. It will take an intervening event—JELD-WEN’s anticipated refusal to sell to Steves on any reasonable terms after September 2021—to block Steves’s access to doorskins, which is the injury on which its claim is premised. Whether that event will occur is uncertain and entirely up to JELD-WEN.

Because of this uncertainty, Steves’s future-lost-profits claim wasn’t fit for judicial decision. And by the same token, withholding adjudication of this claim should not cause

57

Steves any hardship. For one thing, if divestiture occurs, Steves’s threatened injury will be avoided and it won’t get its future damages anyway. If divestiture doesn’t pan out, and JELD-WEN refuses to sell doorskins to Steves, Steves could then seek a conduct remedy or damages, as it would have suffered an actual injury by that point. And, of course, JELD- WEN would be estopped from relitigating issues that have already been decided, such as whether the merger was illegal.

We therefore vacate that portion of the district court’s judgment awarding future lost profits as an alternative to divestiture. 15 F.

We turn now to the trade-secrets trial. JELD-WEN challenges two of the district court’s jury instructions, which we review for abuse of discretion, United States ex rel. Oberg v. Pa. Higher Educ. Assistance Agency, 912 F.3d 731, 736 (4th Cir. 2019).

1.

JELD-WEN first assails the court’s instruction that, “as to trade secret 23, and only trade secret 23[,] . . . . a trade secret may exist if some, or even all, of its individual elements are public, provided that the trade secret as a whole remains confidential,” J.A. 2991. We understand JELD-WEN to argue that, by limiting this instruction to trade secret 23 (the only combination trade secret in the case), the court suggested that the bar for proving the existence of a non-combination trade secret is higher than it really is. That is, this

15 As we conclude that Steves’s future-lost-profits claim wasn’t ripe, we need not address JELD-WEN’s alternative argument that the damages award didn’t account for benefits that Steves gained from the CMI merger.

58

instruction implied that JELD-WEN had to prove that its sixty-six other alleged trade secrets were wholly confidential, which (JELD-WEN claims) isn’t the right legal standard.

This argument misses the mark. The court was right to limit its instruction to trade secret 23. By definition, a combination trade secret has multiple elements, not all of which must be confidential so long as the fact of their “unified process, design, and operation” are confidential. AirFacts, 909 F.3d at 96. In contrast, a non-combination trade secret by definition has only one element. That element “must be secret”; otherwise, it couldn’t be a trade secret. Hoechst Diafoil Co. v. Nan Ya Plastics Corp., 174 F.3d 411, 418 (4th Cir. 1999) (quoting Kewanee Oil Co. v. Bicron Corp., 416 U.S. 470, 475 (1974)).

JELD-WEN conflates two concepts: combination trade secrets and the level of secrecy required of trade secrets generally. While it’s true that an item can be a trade secret even if it’s publicly available—for instance, if it appears in an unsealed court filing, see id.—it still must be relatively secret, i.e., it “must not be of public knowledge or of a general knowledge in the trade or business,” id. (quoting Kewanee, 416 U.S. at 475). After all, that’s the point of a trade secret. JELD-WEN’s position—i.e., that a non-combination trade secret need not be secret—strains common sense.

2.

Next, JELD-WEN protests the district court’s instruction that malicious trade-secret appropriation, under federal and Texas law, requires an “intent to cause injury or harm” to the plaintiff. According to JELD-WEN, the proper standard is whether the defendant showed “conscious disregard for the rights of another.” Appellant’s Br. at 73. JELD-WEN

59

insists that this error requires a retrial as to the eight secrets that the jury deemed protectable and misappropriated.

JELD-WEN’s federal and state claims warrant separate analysis.

a.

Congress enacted the Defend Trade Secrets Act to create a federal private cause of action for trade-secret misappropriation. See Defend Trade Secrets Act of 2016, Pub. L. 114-153 § 2, 130 Stat. 376, 376–82 (May 11, 2016). The statute provides, inter alia, that plaintiffs are entitled to exemplary damages and attorney’s fees if a trade secret is “willfully and maliciously misappropriated.” 18 U.S.C. § 1836(b)(3)(C), (D). But it doesn’t define “malicious,” and we have found no cases that define that term in the context of this statute.

Nor have we found a clear definition of the term in states’ versions of the Uniform Trade Secrets Act—on which the federal statute is modeled. See Kuryakyn Holdings, LLC v. Ciro, LLC, 242 F. Supp. 3d 789, 797 (W.D. Wis. 2017) (looking to state statutes with respect to another Defend Trade Secrets Act issue). To the contrary, cases interpreting those statutes show that they offer competing definitions, each based on how “malice” is defined in other contexts under the relevant state’s laws. In some of these cases, courts have adopted the “intent to cause injury or harm” standard that was used here (or something similar). See Contour Design, Inc. v. Chance Mold Steel Co., Ltd., No. 09-CV-451, 2011 WL 6300622, at *12 (D.N.H. Dec. 16, 2011) (looking to New Hampshire law), aff’d in part and rev’d in part, 693 F.3d 102 (1st Cir. 2012); Wellogix, Inc. v. Accenture, LLP, 823 F. Supp. 2d 555, 570 (S.D. Tex. 2011) (looking to Texas law). The leading trade-secrets treatise also supports that approach. See 1 Milgrim on Trade Secrets § 1.01[2][c][iv][C],

60

[5][d][ii]. Other courts, however, have used JELD-WEN’s favored standard. See Macquarie Bank Ltd. v. Knickel, 793 F.3d 926, 940 (8th Cir. 2015) (looking to North Dakota law); Learning Curve Toys, Inc. v. PlayWood Toys, Inc., 342 F.3d 714, 730 (7th Cir. 2003) (looking to Illinois law).

Next, we consider legislative history and how Congress and courts have defined “malicious” in analogous contexts. But JELD-WEN offers no help on these points. It devotes less than a page of briefing to this issue, and the only authorities it cites involve the Illinois and North Carolina versions of the Uniform Trade Secrets Act. It fails to recognize that states have differing views on this issue, and that we must look to other evidence to assess what Congress intended.

“[I]t is not the obligation of this court to research and construct legal arguments open to parties, especially when they are represented by counsel,” and “perfunctory and undeveloped arguments . . . are waived.” Judge v. Quinn, 612 F.3d 537, 557 (7th Cir. 2010) (cleaned up). Thus, in light of JELD-WEN’s perfunctory and undeveloped argument, we have no cause to reverse the district court on this issue.

b.

We turn now to the Texas-law claim. The current version of the Texas Uniform Trade Secrets Act includes JELD-WEN’s favored definition of “willful and malicious.” See Tex. Civ. Prac. & Rem. Code § 134A.002(7). But that’s due to a statutory amendment that took effect on September 1, 2017, and doesn’t apply to actions that commenced before that date. Texas Unif. Trade Secrets Act, 2017 Tex. Sess. Law Serv. Ch. 37, §§ 6, 7 (H.B. 1995) (“Chapter 134A, Civil Practice and Remedies Code, as amended by this Act, applies

61

only to an action that commences on or after the effective date of this Act . . . . This Act takes effect September 1, 2017.”). Thus, it doesn’t apply to JELD-WEN’s misappropriation claims, because JELD-WEN first sought leave to file these claims in March 2017 and the district court granted leave that May.

JELD-WEN’s claims are instead governed by pre-September 2017 Texas law. See id. § 6 (“An action that commences before the effective date of this Act is governed by the law applicable to the action immediately before the effective date of this Act . . . .”). Before the 2017 amendment, Texas courts defined “malice” in misappropriation suits as “a specific intent by the defendant to cause substantial injury or harm to the claimant,” in accordance with the Texas civil code’s general definition of the term. Horizon Health Corp. v. Acadia Healthcare Co., Inc., 520 S.W.3d 848, 866 (Tex. 2017) (quoting Tex. Civ. Prac. & Rem. Code § 41.001(7)); accord Eagle Oil & Gas Co. v. Shale Exploration, LLC, 549 S.W.3d 256, 283 (Tex. Ct. App. 2018). The jury instruction in this case matches that standard, so no retrial is necessary.

G.

JELD-WEN also asserts that the district court lacked authority to enter judgment for the Intervenors in the trade-secrets case because JELD-WEN brought no claims against them. As this is an issue of law, we review it de novo. See Equinor, 917 F.3d at 813.

We agree with JELD-WEN. Judgment as a matter of law may be granted only “on a claim or defense.” Fed. R. Civ. P. 50(a)(1)(B). To our knowledge, no federal court has ever recognized an exception to this principle. Nor would such an exception make sense, as the civil plaintiff is the “master of his complaint” and “determines the claims . . . to

62

bring.” See United States ex rel. Bunk v. Gosselin World Wide Moving, N.V., 741 F.3d 390, 405–06 (4th Cir. 2013) (cleaned up); see also Hudson v. Air Line Pilots Ass’n Int’l, 415 B.R. 653, 660 (N.D. Ill. 2009) (denying summary judgment to an intervenor because the plaintiffs had not brought any claims against it).

The district court nevertheless granted judgment to the Intervenors for two reasons:

(1) the claims against Steves centered on the Intervenors’ conduct, and (2) by intervening, they gained the right to participate in the case and exposed themselves to potential liability if JELD-WEN pursued a claim against them. Neither reason passes muster. Suits against corporations often focus on certain individuals’ conduct; that doesn’t make such individuals defendants. And an intervenor’s right to participate in a case—by, e.g., filing motions, participating in discovery, and appealing judgments—doesn’t give them a right to judgment on claims that don’t exist. Further, the mere act of intervening didn’t expose the Intervenors to liability. That would have required JELD-WEN to pursue relief against them in this action, which it didn’t do.

The cases cited by Steves and the district court are inapposite. For example, in Schneider v. Dumbarton Devs., Inc., the plaintiff successfully pursued a claim against the intervenor. See 767 F.2d 1007, 1017 (D.C. Cir. 1985) (stating that the plaintiff “asked for,” and the court granted after a bench trial, judgment specifically against the intervenor). And United States v. Oregon was about whether intervention constituted a waiver of tribal immunity. See 657 F.2d 1009, 1013–14 (9th Cir. 1981). As for Alvarado v. J.C. Penney Co., Inc., the intervenors in that case were granted judgment on their own claims. See 997 F.2d 803, 805 (10th Cir. 1993) (noting that the intervenors made “their claims known” by

63

“requesting a declaratory judgment of sorts” against the parties). None of these cases suggest that a party can be awarded judgment on a claim or defense that was never raised.

H.

Finally, JELD-WEN asks that we reassign this case on remand, alleging that the district judge “made repeated and critical errors that fundamentally skewed the proceedings against JELD-WEN.” Appellant’s Br. at 77. We decline to do so.

“Absent a claim of bias,” which JELD-WEN doesn’t make here, “reassignment is appropriate in unusual circumstances where both for the judge’s sake and the appearance of justice an assignment to a different judge is salutary and in the public interest, especially as it minimizes even a suspicion of partiality.” United States v. North Carolina, 180 F.3d 574, 582–83 (4th Cir. 1999) (cleaned up). “In determining whether such circumstances exist, a court should consider:

(1) whether the original judge would reasonably be expected upon remand to have substantial difficulty in putting out of his or her mind previously expressed views or findings determined to be erroneous or based on evidence that must be rejected,

(2) whether reassignment is advisable to preserve the appearance of justice, and

(3) whether reassignment would entail waste and duplication out of proportion to any gain in preserving the appearance of fairness.

Id. at 583.

Overall, the district judge presided admirably over this exceedingly complex case.

While we disagree with some of the judge’s rulings, they nonetheless reflect thoughtful analysis, and we’re confident that the judge can put his previous rulings out of his mind.

64

“And, in light of the lengthy history of this case, reassignment would entail a waste of judicial resources.” Id. at 583. We therefore decline to reassign this case.

IV.

For the reasons given, we vacate the district court’s judgment as to Steves’s future-

lost-profits claim and the Intervenors. We otherwise affirm the district court’s judgment and remand for further proceedings consistent with this opinion.

AFFIRMED IN PART, VACATED IN PART, AND REMANDED

65

RUSHING, Circuit Judge, concurring:

I join the Court’s opinion in full. As the Court observes, we have not previously had occasion to speak on the issue of divestiture sought by a private plaintiff under Section 16 of the Clayton Act, 15 U.S.C. § 26. Supra at 19. But other courts have considered such requests, and none has yet encountered a case in which divestiture was an appropriate award. In particular, courts have been reluctant to order divestiture at the behest of a private plaintiff after consummation of the allegedly anticompetitive merger.

Divestiture is “simple, relatively easy to administer, and sure,” California v. Am.

Stores Co., 495 U.S. 271, 281 (1990) (quoting United States v. E.I. du Pont de Nemours & Co., 366 U.S. 316, 331 (1961)), only “before the transaction is consummated, or if stock or discrete tangible assets are all that later need be divested,” Ginsburg v. InBev NV/SA, 623 F.3d 1229, 1234 (8th Cir. 2010). After a merger closes and the two entities combine their assets and operations into a single corporate unit, divestiture becomes decidedly more complex. The passage of time exacerbates those complexities, not only for the combined entity but also for nonparties who will be affected by a court order dividing the company. See, e.g., id. at 1235–1236 (noting “obvious hardship” divestiture would work on employees and distributors of the acquired company and potential damage to competition and consumers); Blue Cross & Blue Shield United of Wis. v. Marshfield Clinic, 883 F. Supp. 1247, 1264 (W.D. Wis. 1995) (observing that divestiture “would have a large impact on third parties . . . that have not been before this [c]ourt to protect their interests”), aff’d in part, rev’d in part on other grounds, 65 F.3d 1406 (7th Cir. 1995).

66

Threatened antitrust injury is often foreseeable from the anticompetitive combination itself. Courts accordingly have measured the reasonableness of a private plaintiff’s delay in suing for divestiture relative to the announcement of the transaction and its subsequent consummation rather than analogizing to the four-year limitations period applicable to suits for damages. See, e.g., Ginsburg, 623 F.3d at 1235 (finding delay “inexcusable” when plaintiffs “waited nearly two months” after the merger announcement to sue and moved for a preliminary injunction nine days before the contemplated closing date); Taleff v. Sw. Airlines Co., 828 F. Supp. 2d 1118, 1124 (N.D. Cal. 2011) (holding divestiture unavailable because plaintiffs “delayed in filing their suit until after [d]efendants’ merger had already been consummated”); Antoine L. Garabet, M.D., Inc. v. Autonomous Techs. Corp., 116 F. Supp. 2d 1159, 1173 (C.D. Cal. 2000) (concluding that divestiture was not an available remedy because plaintiffs waited until the day of the merger’s consummation to sue); see also Am. Stores, 495 U.S. at 297–298 (Kennedy, J., concurring) (reasoning that plaintiff’s delay in seeking divestiture for several months after notice of the intended merger and after completion of the stock sale but before the entities combined their business operations “should bear upon the ultimate disposition of the case,” including “the bar of laches”). But see supra at 40 n.13.

Here, Steves alleged an antitrust injury that could not have been foreseen until two years after the merger closed, namely the loss of its doorskin supply and resulting threat to its survival. As the Court explains, it was reasonable for Steves to wait to pursue equitable relief until 2014, when it learned of this threat. Supra at 41–43. Although “problems of proof and causation” may attend claims alleging that a merger caused an antitrust injury to

67

occur years later, Midwestern Machinery, Inc. v. Nw. Airlines, Inc., 167 F.3d 439, 442 n.3 (8th Cir. 1999), here the district court and jury found causation established. Steves was not on notice of the threatened injury that forms the basis for its divestiture claim until 2014, therefore its delay in filing suit must be evaluated from that point rather than from the time the merger was announced or consummated, as might otherwise be the case.