SUSAN L. CARNEY, Circuit Judge:

Plaintiffs appeal from an award of summary judgment to defendants in an antitrust class action alleging a conspiracy to fix prices in violation of section 1 of the Sherman Act, 15 U.S.C. § 1. Defendants Stora Enso North America Corporation ("SENA") and Stora Enso Oyj ("SEO") and former defendants UPM-Kymmene Corporation and UPM-Kymmene, Inc. (together, "UPM") are manufacturers and sellers of "publication paper," a type of paper used in preparing printed material of various types. Plaintiffs — a certified class consisting of direct purchasers of defendants' paper products — contend that, in the relevant period, they paid higher prices for publication paper than they would have in the absence of the alleged price-fixing agreement.

Plaintiffs' theory of conspiracy is relatively simple. In August and November 2002, and again in February 2003, SENA and UPM — as well as several other publication paper manufacturers — raised their list prices for certain types of publication paper. The price hikes mirrored each other in amount and occurred in close succession. Plaintiffs maintain that SENA and UPM instituted these three price increases pursuant to an agreement, rather than independently. Plaintiffs also contend that, in the same time frame, SENA and UPM coordinated the closure of paper mills in order to reduce the supply of publication paper, and that SEO played a material role in the overall price-fixing scheme. The U.S. District Court for the District of Connecticut (Stefan R. Underhill, J.) concluded that defendants were entitled to summary judgment on all counts because, in its view, and in light of the standards articulated by the Supreme Court in Matsushita Electric Industrial Co. v. Zenith Radio Corp., 475 U.S. 574, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986), plaintiffs "failed to offer sufficient evidence to dispel the possibility that SENA and UPM acted independently." In re Publication Paper Antitrust Litig., No. 3:04-md-1631, 2010 WL 5253364, at *13 (D.Conn. Dec. 14, 2010).

We hold that the district court erred in granting summary judgment to SENA because a jury could reasonably find that SENA and UPM entered into an agreement to raise the price of publication paper, and that, as implemented, this agreement damaged plaintiffs. SENA was thus not entitled to judgment as a matter of law. As to SEO, however, we conclude that the district court properly awarded it summary judgment because plaintiffs failed to offer sufficient evidence from which a jury could reasonably conclude that SEO had any direct involvement in decisions regarding SENA's marketing, sale, or pricing of publication paper in the United States. Accordingly, we vacate the district court's judgment in part, affirm it in part, and remand the case for further proceedings consistent with this opinion.

I. BACKGROUND

Unless otherwise noted, the following facts are not disputed by the parties.

A. The Parties

SEO is a pulp and paper manufacturer headquartered in Helsinki, Finland. In August 2000, SEO acquired Consolidated Papers, Inc., an American paper manufacturer located in Wisconsin. Thereafter, Consolidated Papers — renamed SENA — operated as a subsidiary of SEO. From August 2000 until May 2003, Kai Korhonen was President of SENA. Under Korhonen's leadership, SENA's sales and marketing team would recommend to him whether SENA should initiate a price increase or follow a price increase previously announced by a competitor. As President, Korhonen had the "final say" over SENA's pricing strategy and decisions. Trial Tr. at 262.1

UPM is also a pulp and paper manufacturer headquartered in Helsinki. In January 2002, Markku Tynkkynen became President of UPM's Magazine Paper Division. In this role, Tynkkynen had responsibility for, and "final authority" over, UPM's pricing of publication paper in the United States. Dep. of Markku Tynkkynen, Jan. 22, 2009, at 18.

B. The Publication Paper Industry

"Commercial paper" is a type of publication paper that is used for printing and advertising. Manufactured in several different grades, some of which are numbered 1 through 5, its quality ranges from "fine" (the most expensive type) to "supercalendered" (the least expensive). Fine paper — categorized in the industry as including grades 1, 2, and 3 — is typically used for high-end publications. Magazine paper — comprised of grades 4 and 5 — is generally used in ordinary magazines and store catalogues. Supercalendered paper — which is not assigned grade numbers — is often used to print the advertising inserts found in newspapers.

The present action involves "publication paper," which, for purposes of this litigation, is defined as grades 3, 4, and 5 of "coated" commercial paper.2 The publication paper that plaintiffs purchased thus includes the coated products of some fine, and all magazine, paper. Publication paper is commonly considered a commodity product — i.e., a product that is uniform rather than differentiated across sellers. The publication paper market in North America is an oligopoly.

In 2002 and 2003, when the price-fixing agreement is alleged to have been made and implemented, International Paper ("IP"), SENA, and UPM ranked first, second, and third, respectively, in volume of production of publication paper in North America. Each had a market share of between 17% and 21%; their combined market shares totaled approximately 60%.3 During the same two-year period, publication paper was selling at historically low prices, and the industry suffered from excess capacity and low demand.4 Despite these unfavorable conditions, in August 2002, IP and MeadWestvaco ("Mead"), another publication paper manufacturer, both announced price increases; in November 2002, Mead announced a price increase; and in February 2003, IP announced a price increase. Both in the timing of the announcements and the amount of the increases, SENA and UPM's price increases in August and November 2002 and in February 2003 closely followed those of their competitors.

C. The Alleged Conspiracy

The conspiracy alleged in this action centers around certain private meetings and phone calls between Korhonen and Tynkkynen that undisputedly occurred soon before each of these three price increase announcements.

The two men originally met in Finland in the late 1970s, when both were working as production engineers for Ahlström Company. During the approximately five years that they worked together, they were also neighbors, and became close friends. Tynkkynen left Ahlström Company in 1982; Korhonen, however, remained at the firm, which was later acquired by a predecessor company to SEO. Over the twenty years following Tynkkynen's departure, despite their earlier close friendship, Korhonen and Tynkkynen saw each other only on occasion, their interactions typically limited to social settings and meetings of the Finnish Paper Engineers Association, of which they were both members.

1. The August 2002 Price Increase

In August 2002, soon after he became President of UPM's Magazine Paper Division and twenty years after they last worked as colleagues, Tynkkynen invited Korhonen, who by then had become President of SENA, to lunch. As he later testified at SENA's criminal trial, Tynkkynen wanted Korhonen to provide him with insight into operating in the North American market, a market with which Tynkkynen was unfamiliar. Tynkkynen also sought to determine whether SENA had sufficient market power to lead a price increase in the United States.

In response to Tynkkynen's invitation, Tynkkynen and Korhonen met for approximately 90 minutes on August 8 in a private executive lunchroom at UPM's headquarters in Helsinki. During the meeting, Korhonen expressed his opinion that, contrary to Tynkkynen's view, SENA was not a market leader for coated grade 5 or supercalendered paper. Tynkkynen and Korhonen estimated that together SENA and UPM likely would have a combined 40% market share for coated grade 5 paper. Tynkkynen advised Korhonen, "UPM has been a follower and will be a follower" — meaning that UPM would follow price increases announced by any major competitor and would also try to implement those increases consistently with its customers.5 Trial Tr. at 177, 104. No one else was present during the meeting, and Korhonen did not disclose to anyone at SENA that he met with Tynkkynen or mention the substance of their discussion.

The following day, on August 9, IP and Mead announced a $2 per hundredweight ("cwt.") price increase for publication paper grades 2 through 4 effective October 1, 2002. Four days later, on August 13, SENA issued an identical announcement as to price, paper grades, and effective date. On August 21, UPM announced its $2 per cwt. price increase for publication paper grades 3 and 4, on the same schedule. (According to plaintiffs, UPM sold little or no grade 2 publication paper in the United States at that time.)

2. The November 2002 Price Increase

On November 6, again at Tynkkynen's invitation, Korhonen and Tynkkynen met at the Chicago O'Hare airport and had lunch together in a private dining room at the O'Hare Hilton Hotel. During their meeting, Korhonen and Tynkkynen discussed the poor economic conditions in the industry: prices were eroding because of an excess supply. Notwithstanding the August price increase, which had become effective approximately five weeks earlier, they agreed that the industry was in need of a price increase. Tynkkynen again advised Korhonen that UPM would follow if a competitor announced a price increase. Apart from his personal assistant, who arranged his itinerary, Korhonen did not disclose to anyone at SENA that he met with Tynkkynen.

The following week, on November 15, Mead announced a $2 per cwt. price increase for publication paper grades 1 through 4 effective January 1, 2003. On November 19, after UPM had decided to follow Mead's price increase but before any public announcement was made, Tynkkynen called Korhonen and left a voicemail stating, "Mead is out, we are following." Id. at 118. Korhonen did not return the call. The following day, on November 20, both SENA and IP announced a $2 per cwt. price increase for publication paper grades 2 through 4, also effective January 1, 2003. Two days later, on November 22, UPM issued a written announcement declaring a $2 per cwt. price increase for publication paper grades 2 through 4, also effective as of the turn of the year.

3. The February 2003 Price Increase

On February 3, 2003, executives at IP decided at an internal meeting to increase the price of their grade 5 publication paper, subject to final approval by senior IP officials. Three days later, on February 6, Korhonen called Tynkkynen. Each man asserts that he does not remember exactly what was discussed during the call, which lasted approximately five minutes.6

On February 10, IP publicly announced a $3 per cwt. price increase for grade 5 publication paper effective April 1, 2003. The following morning, on February 11, Korhonen called Tynkkynen and left a message requesting that Tynkkynen return his call. That afternoon, Tynkkynen called Korhonen, and the two men spoke for four minutes. During the call, Korhonen told Tynkkynen, "IP's out and [SENA] will match and follow." Id. at 84. Tynkkynen understood "match and follow" to mean that SENA would follow IP's price increase and also would "implement [the price increase] seriously" at the customer end — in other words, that SENA would hold a hard line in negotiating with customers. Id. Tynkkynen informed Korhonen that UPM had notified some of its customers of its intention to match IP's price increase. Approximately one half-hour after the men's phone call, SENA e-mailed and faxed a brief letter to its customers announcing a $3 per cwt. price increase for grade 5 paper, coated grade 4 paper, and supercalendered products. That same afternoon, UPM mailed a bulletin to its customers in which it, too, announced a $3 per cwt. price increase for grade 5 publication paper. Both price increases were to take effect April 1.

After Korhonen and Tynkkynen's conversation on February 11, Tynkkynen advised Kevin Lyden, President of UPM's North American subsidiary, that he thought SENA was "going to be tough on the price increase" it had announced. Dep. of Peter Littley, Feb. 18, 2010, at 14. Lyden understood Tynkkynen's statement to mean that SENA would stand firm on its price increase when negotiating contracts with its customers. According to Lyden, Tynkkynen's assessment of how aggressively SENA would implement the price increase was based, in part, on his February 11 phone call with Korhonen.

D. The Department of Justice Investigation and SENA's Criminal Trial

In 2004, the United States Department of Justice ("DOJ") launched a criminal antitrust investigation into the publication paper industry. In July 2004, after learning that government investigators sought to interview him in relation to the investigation, Tynkkynen called Korhonen from a public telephone booth. Tynkkynen concedes that he used a public telephone in order to ensure that he "ha[d] a clean line" — i.e., a phone line that was not being "followed or taped" by law enforcement authorities. Dep. of Markku Tynkkynen, Jan. 22, 2009, at 136. During their conversation, Tynkkynen proposed that he and Korhonen develop a "joint story" concerning what they had discussed during their private phone calls and meetings. Trial Tr. at 133. Korhonen concurred that developing a "joint story" was "a good idea." Id. They decided that if interviewed, they would tell law enforcement officials that in their conversations in 2002 and 2003 they discussed railroad fees, timber swaps, and the Finnish Paper Engineers Association. Pricing was not among those topics that they agreed to identify as having been discussed. (Later, in an August 2004 phone call, Tynkkynen told Korhonen that he intended to tell the government "everything" that he remembered about their meetings, rather than the "joint story" they had discussed in July. Id. at 135-36.)

In February 2006, the DOJ granted conditional full immunity to UPM in return for the company's cooperation in the price-fixing investigation. In December 2006, a grand jury issued a one-count indictment against SENA charging that from August 2002 through June 2003, SENA and others entered into a conspiracy to fix the price at which the companies sold publication paper to customers in the United States, in violation of section 1 of the Sherman Act. The case against SENA was tried to a jury in the U.S. District Court for the District of Connecticut (Christopher F. Droney, J.) in July 2007. United States v. Stora Enso N. Am. Corp., No. 3:06-cr-323.

During the criminal trial, Tynkkynen testified, inter alia, that during their phone conversation on February 11, 2003, he and Korhonen reached an "agreement" that UPM and SENA would "match[]" IP's February 2003 price increase for coated grade 5 paper and that both firms would "seriously implement" that price increase with their customers. Trial Tr. at 88, 247-48. For his part, Korhonen testified that at the time of the November 6, 2002 meeting, he understood Tynkkynen to be trying to "initiate something to get the prices improved." Id. at 281. Korhonen further testified that after the November 6 meeting, "some phone calls" occurred between him and Tynkkynen during which "intentions were exchanged about [publication] paper price increases," and that, during their February 11, 2003 phone conversation in particular, Korhonen told Tynkkynen what SENA was "going to do," and, likewise, Tynkkynen told Korhonen what UPM was "going to do." Id. at 282, 284-85. After both parties presented their evidence, SENA unsuccessfully moved for acquittal. On July 19, 2007, in a general verdict, the jury acquitted SENA of criminal antitrust violations.

E. Procedural History

After the DOJ's investigation became public, individual plaintiffs brought nine separate actions against various manufacturers of publication paper in federal district courts across the country, alleging price fixing in the U.S. market for publication paper. In November 2004, the Judicial Panel on Multidistrict Litigation ("MDL") consolidated and transferred these actions to the U.S. District Court for the District of Connecticut, thereby creating MDL 1631, In re Publication Paper Antitrust Litigation. Additional suits were filed over the course of the following year. They, too, were consolidated with MDL 1631.

In their fifth and final amended complaint, filed in March 2008, plaintiffs brought claims under section 1 of the Sherman Act, 15 U.S.C. § 1, and sections 4 and 16 of the Clayton Act, 15 U.S.C. §§ 15 and 26, seeking treble damages and alleging that from October 1, 2002, to May 31, 2004, SENA, SEO, and UPM conspired to fix the prices of coated paper grades 3, 4, and 5. In April 2008, the district court approved plaintiffs' settlement with UPM, leaving only SENA and SEO as defendants in the MDL proceedings. In May 2009, the district court certified the plaintiff class as "[a]ll persons ... who directly purchased Publication Paper for delivery in the United States from [SENA or SEO]... at any time during the period from October 1, 2002 to and including September 30, 2003" (the "class period"). Order Granting Class Certification, D. Ct. Dkt. No. 470. After the close of discovery, defendants moved for summary judgment, and the district court granted defendants' motion. Publication Paper, 2010 WL 5253364, at *1.

In its decision, the district court rejected plaintiffs' argument that Tynkkynen's testimony that he and Korhonen reached an "agreement" on pricing in February 2003 constituted sufficient "direct evidence" of price fixing to withstand defendants' motion for summary judgment. Id. at *8. As to plaintiffs' other evidence, the court acknowledged that SENA and UPM's three price increases constituted parallel conduct; that the publication paper industry was conducive to collusion; and that the communications between Tynkkynen and Korhonen supported the existence of a conspiracy. Nevertheless, the court determined that plaintiffs' evidence did not sufficiently "exclude the possibility that SENA acted independently in raising the prices charged for publication paper" and that, in any event, "even if Tynkkynen and Korhonen agreed that UPM and SENA would follow rivals' future price increases," plaintiffs had not demonstrated that such an agreement "had some effect on future price increases announced by SENA." Id. at *13-14. The district court also concluded that plaintiffs failed to show that defendants conspired to reduce the supply of publication paper by closing certain of their paper mills, and ruled that plaintiffs offered "virtually no evidence" demonstrating that SEO played a role in the alleged price-fixing scheme. Id. at *17.

This appeal followed.

On appeal, plaintiffs contend that the district court made several factual and legal errors that require us to reverse the judgment entered for defendants. In particular, they maintain that the district court improperly discounted direct evidence of a price-fixing agreement; improperly failed to draw reasonable inferences in their favor; and erred in concluding as a matter of law that the alleged agreement between Korhonen and Tynkkynen had no effect on SENA's pricing decisions.7 Our task is to determine whether, on the evidence presented, a jury could reasonably conclude that defendants reached an agreement "for the purpose and with the effect of raising ... the price of" publication paper. United States v. Socony-Vacuum Oil Co., 310 U.S. 150, 223, 60 S.Ct. 811, 84 L.Ed. 1129 (1940); accord Major League Baseball Props., Inc. v. Salvino, Inc., 542 F.3d 290, 336 (2d Cir.2008). This requires evidence sufficient to permit a preponderance finding that "higher prices came about as a result of [the agreement], rather than through independent action of the defendants." Apex Oil Co. v. DiMauro, 822 F.2d 246, 253 (2d Cir.1987).

II. DISCUSSION

A. Standard of Review

We review de novo a district court's grant of summary judgment. MBIA Inc. v. Fed. Ins. Co., 652 F.3d 152, 158 (2d Cir.2011). We will affirm only if, after construing the evidence in the light most favorable to the non-moving party and drawing all reasonable inferences in its favor, id., "there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law," Fed.R.Civ.P. 56(a).

Our Court has observed that "[b]y avoiding wasteful trials and preventing lengthy litigation that may have a chilling effect on pro-competitive market forces, summary judgment serves a vital function in the area of antitrust law." Tops Mkts., Inc. v. Quality Mkts., Inc., 142 F.3d 90, 95 (2d Cir.1998). Nevertheless, summary judgment is not a substitute for trial. Apex, 822 F.2d at 252. Thus, when the evidence admits of competing permissible inferences with regard to whether a plaintiff is entitled to relief, "the question of what weight should be assigned to [those] inferences remains within the province of the fact-finder at a trial." Id. at 253.

B. Price-Fixing Claims

Section 1 of the Sherman Act provides that "[e]very contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce... is ... illegal." 15 U.S.C. § 1. Notwithstanding its broad language, this provision prohibits "only unreasonable restraints of trade." Bus. Elecs. Corp. v. Sharp Elecs. Corp., 485 U.S. 717, 723, 108 S.Ct. 1515, 99 L.Ed.2d 808 (1988). Per se liability attaches, however, to "plainly anticompetitive" agreements. Texaco Inc. v. Dagher, 547 U.S. 1, 5, 126 S.Ct. 1276, 164 L.Ed.2d 1 (2006). An agreement between competitors to fix prices, known as a horizontal price-fixing agreement, categorically constitutes an unreasonable restraint, and, accordingly, is unlawful per se. Id.

To prevail on a claim of horizontal price fixing, a plaintiff must demonstrate that the defendants entered into a conspiracy "formed for the purpose and with the effect of raising ... price[s]." Socony-Vacuum, 310 U.S. at 223, 60 S.Ct. 811. In other words, the evidence must show that: (1) the defendants conspired to raise prices, and (2) this conspiracy caused injury to the plaintiff in the form of artificially inflated prices. See Apex, 822 F.2d at 253; In re Flat Glass Antitrust Litig., 385 F.3d 350, 356 (3d Cir.2004).

Parallel pricing among competitors, also known as "conscious parallelism," is often proffered as evidence of a price-fixing agreement. "Conscious parallelism" describes "the process, not in itself unlawful, by which firms in a concentrated market might in effect share monopoly power, setting their prices at a profit-maximizing, supracompetitive level by recognizing their shared economic interests and their interdependence with respect to price and output decisions." Brooke Grp. Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 227, 113 S.Ct. 2578, 125 L.Ed.2d 168 (1993).

Conscious parallelism alone, however, does not establish an antitrust violation. Such behavior is consistent with both unlawful conspiracy and lawful independent conduct. See Apex, 822 F.2d at 253. Accordingly, when the defendants' parallel pricing forms the basis for a price-fixing claim, a plaintiff must show additional circumstances — often referred to as "plus factors" — which, when viewed in conjunction with the parallel conduct, would permit a fact-finder to infer a conspiracy. Id. Such "plus factors" may include, for example: a common motive to conspire; evidence that the parallel acts were against the apparent economic self-interest of the individual alleged conspirators; or evidence of "a high level of interfirm communications." Id. at 254.

Section 1 of the Sherman Act does not itself provide a private right of action. That right is established by section 4 of the Clayton Act, which authorizes private suits by "any person who shall be injured in his business or property by reason of anything forbidden in the antitrust laws," 15 U.S.C. § 15, and affords prevailing plaintiffs a claim to three times their actual damages.

C. Evidence of an Agreement

1. The Scope of Matsushita

In concluding that summary judgment was appropriate in this case, the district court cited extensively to the Supreme Court's decision in Matsushita. Acknowledging that "the evidence as a whole arguably could support an inference of illegal collusive behavior," the district court nonetheless determined that plaintiffs "failed to offer sufficient evidence to dispel the possibility" that SENA acted independently. Publication Paper, 2010 WL 5253364, at *9, *13. The parties dispute the extent to which the "tends to exclude" formulation articulated in Matsushita, 475 U.S. at 588, 106 S.Ct. 1348 ("To survive a motion for summary judgment ... a plaintiff seeking damages for a violation of § 1 must present evidence that `tends to exclude the possibility' that the alleged conspirators acted independently."), applies under the circumstances presented here. We take this opportunity to clarify the application of the standards established in that case.

In Matsushita, the plaintiffs alleged that Japanese television manufacturers had engaged in a longstanding and wide-ranging conspiracy to fix prices in the United States below the level that their costs and market conditions made reasonable. According to the plaintiffs' theory, the defendants' predatory prices would eventually force competitors to leave the market, at which point the defendants would raise prices to monopoly levels. The Court observed that, as a general matter, "predatory pricing schemes are rarely tried, and even more rarely successful." Id. at 589, 106 S.Ct. 1348. Indeed, the conspiracy alleged in Matsushita had spanned 20 years without achieving the putatively desired monopoly. Because it found both that the plaintiffs' theory of conspiracy was implausible and that the defendants' price-cutting conduct could as readily be explained as legitimate competitive behavior, the Court required the plaintiffs to "come forward with more persuasive evidence to support their claim than would otherwise be necessary" to withstand summary judgment. Id. at 587, 106 S.Ct. 1348.

Matsushita, then, stands for the proposition that substantive "antitrust law limits the range of permissible inferences" that may be drawn from ambiguous evidence. Id. at 588, 106 S.Ct. 1348; accord In re High Fructose Corn Syrup Antitrust Litig., 295 F.3d 651, 654, 661-62 (7th Cir. 2002) (Posner, J.). It further holds that the range of inferences that may be draw from such evidence depends on the plausibility of the plaintiff's theory. See Matsushita, 475 U.S. at 588, 106 S.Ct. 1348; Monsanto Co. v. Spray-Rite Serv. Corp., 465 U.S. 752, 764, 104 S.Ct. 1464, 79 L.Ed.2d 775 (1984); Apex, 822 F.2d at 253. Thus, where a plaintiff's theory of recovery is implausible, it takes "strong direct or circumstantial evidence" to satisfy Matsushita's "tends to exclude" standard. Apex, 822 F.2d at 253. By contrast, broader inferences are permitted, and the "tends to exclude" standard is more easily satisfied, when the conspiracy is economically sensible for the alleged conspirators to undertake and "the challenged activities could not reasonably be perceived as procompetitive." Flat Glass, 385 F.3d at 358; cf. Eastman Kodak Co. v. Image Technical Servs., Inc., 504 U.S. 451, 468, 112 S.Ct. 2072, 119 L.Ed.2d 265 (1992) ("Matsushita demands only that the nonmoving party's inferences be reasonable in order to reach the jury, a requirement that was not invented, but merely articulated, in that decision.").

At several points in its opinion, intermingled with its use of the "tends to exclude" standard, the district court stated that plaintiffs were required to "exclude" or "dispel" the possibility that defendants acted independently. Publication Paper, 2010 WL 5253364, at *10-11, *13-14. Requiring a plaintiff to "exclude" or "dispel" the possibility of independent action places too heavy a burden on the plaintiff. Rather, if a plaintiff relies on ambiguous evidence to prove its claim, the existence of a conspiracy must be a reasonable inference that the jury could draw from that evidence; it need not be the sole inference. As Professors Areeda and Hovenkamp have advised:

Phillip E. Areeda and Herbert Hovenkamp, Fundamentals of Antitrust Law, § 14.03(b), at 14-25 (4th ed.2011) (footnotes omitted).

Of course, the standards established in Matsushita do not apply at all when a plaintiff has produced unambiguous evidence of an agreement to fix prices. As such, "an admission by the defendants that they agreed to fix their prices is all the proof a plaintiff needs." Corn Syrup, 295 F.3d at 654. Thus, at least three of our sister circuits have held that summary judgment is generally not appropriate where a plaintiff has produced direct, as opposed to circumstantial, evidence of an agreement to fix prices. See Williamson Oil Co., Inc. v. Philip Morris USA, 346 F.3d 1287, 1300 (11th Cir.2003); Petruzzi's IGA Supermarkets, Inc. v. Darling-Delaware Co., Inc., 998 F.2d 1224, 1233 (3d Cir.1993); In re Coordinated Pretrial Proceedings in Petroleum Prods. Antitrust Litig., 906 F.2d 432, 441 (9th Cir.1990); see also R.B. Ventures, Ltd. v. Shane, 112 F.3d 54, 58 (2d Cir.1997) (quoting, in dicta, Ninth Circuit decision drawing distinction between direct and circumstantial antitrust evidence). Some courts have reached this conclusion by reasoning that "direct evidence... requires no inferences to establish the proposition or conclusion being asserted." In re Baby Food Antitrust Litig., 166 F.3d 112, 118 (3d Cir.1999).

Neither party in this case questions this evidentiary distinction. Nevertheless, we do not think it necessary to draw bright lines to decide this case. See Corn Syrup, 295 F.3d at 661-62. All evidence, including direct evidence, can sometimes require a factfinder to draw inferences to reach a particular conclusion, though "[p]erhaps on average circumstantial evidence requires a longer chain of inferences." Sylvester v. SOS Children's Vills. Ill., Inc., 453 F.3d 900 (7th Cir.2006) (Posner, J.). Here, the totality of the evidence, viewed in the light most favorable to plaintiffs and with proper regard for the Matsushita standards, could support a reasonable inference of illegal collusive behavior.

2. Plaintiffs' Evidence of an Agreement to Fix Prices

In this case, the district court concluded that "[n]othing in the record concerning the February 2003 price increase suggests that Tynkkynen or Korhonen had an agreement to follow the IP increase." Publication Paper, 2010 WL 5253364, at *9. We disagree. As previously discussed, Tynkkynen testified during SENA's criminal trial that he and Korhonen reached an "agreement" that UPM and SENA would follow IP's February 2003 price increase and that they would implement that price increase on the customer end "seriously," i.e., to the fullest extent possible. Whether or not this testimony — a co-conspirator's acknowledgment that he understood his numerous communications with Korhonen to reflect a price-fixing agreement — admits any ambiguity as to Korhonen's parallel understanding of the same communications, the testimony is surely strong evidence of a collusive scheme between UPM and SENA. That would be sufficient to satisfy Matsushita's "tends to exclude" standard even if plaintiffs' theory were implausible, which it is not. See Apex, 822 F.2d at 253.8

The district court considered Tynkkynen's testimony to be of limited value, in part, because, as the court explained and is not disputed, "English is not the native language of Tynkkynen and Korhonen." Publication Paper, 2010 WL 5253364, at *8. Any notion that Tynkkynen did not understand the meaning of "agreement" in English or the import of his testimony, however, is belied by the record. Tynkkynen repeatedly explained in court that he and Korhonen had a "common understanding," and that the substance of that "understanding" was that "both companies are matching ... [IP's] coated # 5 price increase and will seriously implement the price increase. So it was a common understanding, and, thus, [an] agreement." Trial Tr. at 247-48. Moreover, defendants conceded at oral argument before this Court that translation and language were not an issue. Oral Arg. Tr. at 23. Although an "understanding" might in some contexts be intended to signal an "expectation" rather than "agreement," the substance of Tynkkynen's testimony and the context of the events he describes fully support the conclusion that when he said "understanding" — as when he said "agreement" — he meant "agreement."

Plaintiffs have provided additional evidence to support a reasonable finding that SENA and UPM engaged in price fixing. The district court found, and defendants do not contest, that the publication paper industry is conducive to collusion: publication paper is a commodity product with few substitutes; the market is controlled by a limited number of sellers (principally, IP, SENA, and UPM); and high capital investment costs limit the entry of new market players. See Corn Syrup, 295 F.3d at 656-57 (describing characteristics that make a market susceptible to collusion). Furthermore, during the class period, the publication paper industry suffered from excess capacity and historically low prices, conditions that make "price competition more than usually risky and collusion more than usually attractive." Id. at 657. Finally, there is ample evidence of conspiratorial behavior. Most notably, it is undisputed that in private phone calls and meetings — for which no social or personal purpose has been persuasively identified — Tynkkynen shared UPM's pricing strategies with Korhonen and both men disclosed to each other their companies' intentions to increase prices before those decisions had been publicly announced. Tynkkynen and Korhonen also developed a "joint story" to conceal from the government the true nature of their communications — behavior from which a jury could infer that both men were aware that their communications and related actions regarding publication paper pricing violated the law.

We recognize that despite the strength of plaintiffs' evidence of an agreement, the totality of the evidence admits of alternative interpretations — namely, that SENA as a corporation acted independently of UPM in deciding to increase its prices and that the price increases each company announced were the product of certain characteristics of the industry and not any agreement between Korhonen and Tynkkynen. But it is the province of the jury to determine how much weight to accord Tynkkynen's testimony and the other relevant evidence. We believe that, on the basis of Tynkkynen's testimony alone, a jury could reasonably find that SENA and UPM made an agreement to fix the prices at which their companies sold publication paper in the latter part of 2002 and early 2003. Plaintiffs' other evidence provides additional support for the theory that, as early as August 2002, Korhonen and Tynkkynen reached an agreement to follow price increases announced by their competitors. In sum, unlike in Matsushita, the record in this case presents strong, if not irrefutable, evidence of a conspiracy in a context where the conspiracy's goals were aligned with the conspirators' economic interests.

D. Causation

1. Applicable Law

Having determined that a jury could reasonably find that SENA and UPM unlawfully agreed to raise prices, we now consider whether there was enough evidence for a jury to find that such an agreement actually caused the price increases that occurred.9 This Court has had little occasion to explore at length the issue of causation in the antitrust context,10 although certain principles have emerged from the Supreme Court's opinions and our precedents. The Supreme Court has noted, for instance, that "[i]t is enough that the illegality is shown to be a material cause of the [antitrust] injury; a plaintiff need not exhaust all possible alternative sources of injury in fulfilling [its] burden of proving compensable injury." Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U.S. 100, 114 n. 9, 89 S.Ct. 1562, 23 L.Ed.2d 129 (1969). Relying on Zenith Radio, our Court has also observed that an antitrust defendant's unlawful conduct need not be the sole cause of the plaintiffs' alleged injuries; to prove a "causal connection" between the defendant's unlawful conduct and the plaintiff's injury, the plaintiff need only "demonstrate that [the defendant's] conduct was a substantial or materially contributing factor" in producing that injury. Litton Sys., Inc. v. AT & T Co., 700 F.2d 785, 823 n. 49 (2d Cir. 1983); see also U.S. Football League v. Nat'l Football League, 842 F.2d 1335, 1377 (2d Cir.1988). We have held further that to prevail on an antitrust claim, a plaintiff must establish that "the injuries alleged would not have occurred but for [the defendant's] antitrust violation," Argus Inc. v. Eastman Kodak Co., 801 F.2d 38, 41 (2d Cir.1986), adding necessity to the materiality requirement of our antitrust causation analysis.

Applying these basic concepts here, we consider whether there is sufficient evidence from which a jury could find that Korhonen and Tynkkynen's agreement, if proven, was both a material and but-for cause of the price increases. Our analysis is enriched and refined by study of causation principles as developed in the tort law context. See generally Associated Gen. Contractors of Cal., Inc. v. Cal. St. Council of Carpenters, 459 U.S. 519, 103 S.Ct. 897, 74 L.Ed.2d 723 (1983) (applying common law principles of causation derived from tort and contract law to determine whether the plaintiff had sufficiently alleged an antitrust injury under section 4 of the Clayton Act).

"There is a causal link between an act or activity and an injury when we conclude on the basis of the available evidence that the recurrence of that act or activity will increase the chances that the injury will also occur." Guido Calabresi, Concerning Cause and the Law of Torts: An Essay for Harry Kalven, Jr., 43 U. Chi. L.Rev. 69, 71 (1975). In other words, if an act is deemed wrongful because it is believed significantly to increase the risk of a particular injury, we are entitled — in the tort context at least — to presume that such an injury, if it occurred, was caused by the act. See, e.g., Liriano v. Hobart Corp., 170 F.3d 264, 271 (2d Cir.1999) ("When a defendant's negligent act is deemed wrongful precisely because it has a strong propensity to cause the type of injury that ensued, that very causal tendency is evidence enough to establish a prima facie case of cause-in-fact."); see also BCS Servs., Inc. v. Heartwood 88, LLC, 637 F.3d 750, 758 (7th Cir.2011) ("Once a plaintiff presents evidence that he suffered the sort of injury that would be the expected consequence of the defendant's wrongful conduct, he has done enough to withstand summary judgment on the ground of absence of causation."). Under this framework, the burden then shifts to the defendant "to bring in evidence tending to rebut the strong inference, arising from the [injury], that the [act] was in fact a but-for cause of the plaintiff's injury." Liriano, 170 F.3d at 271.

2. The Strength of the Evidence

As applied here, we consider the causal link between an agreement to raise prices and a subsequent price increase. Price-fixing agreements are "conclusively presumed to be unreasonable and therefore illegal" precisely because of their "pernicious effect on competition" and lack of any redeeming virtue, N. Pac. Ry. Co. v. United States, 356 U.S. 1, 5, 78 S.Ct. 514, 2 L.Ed.2d 545 (1958) — i.e., because such agreements are so likely to result in artificially higher prices being charged to consumers without accomplishing any legitimate business purpose. Therefore, the demonstrable existence of an agreement between Korhonen and Tynkkynen to follow price increases announced by competitors, if proven, constitutes strong evidence that the alleged agreement caused at least some element of the subsequent price increases (e.g., amount or effective date), or, at a minimum, the inability of plaintiffs to negotiate below the list price. Furthermore, the causal link is presumed to be particularly strong when, as alleged here, the agreement is between executives at rival companies, each of whom has final pricing authority. Compare Flat Glass, 385 F.3d at 368-69 (evidence of "upper level executives hav[ing] secret conversations about price" has more probative weight than "price discussion[s] among low level sales people"), with Baby Food, 166 F.3d at 125 ("Evidence of sporadic exchanges of shop talk among field sales representatives who lack pricing authority is insufficient to survive summary judgment.").

Defendants contend, and the district court concluded, that even if Korhonen and Tynkkynen entered into an illegal agreement, the record provides an inadequate basis on which a jury could find that the agreement caused plaintiffs' complained-of injuries. They so reason from the record's reflection that employees in SENA's sales and marketing department initiated, recommended, and implemented pricing decisions during the class period. Defendants maintain that these lower-level employees — not Korhonen — independently decided that SENA should follow price increases announced by competitors because doing so made economic sense and was consistent with SENA's pricing philosophy to be a market follower rather than a market leader. Defendants further assert that Korhonen never altered or vetoed these employees' pricing recommendations. Accordingly, they urge that the district court properly concluded as a matter of law that even if Korhonen and Tynkkynen reached a price-fixing agreement, it had no effect on SENA's pricing decisions.

Although defendants' argument has some force, and might well persuade a jury, we are not convinced that it so conclusively rebuts plaintiffs' strong evidence that the alleged agreement was both a material and but-for cause of the price increases as to permit an award of summary judgment for defendants. Even if other SENA employees had day-to-day responsibility for pricing strategy and implementation, it is undisputed that Korhonen, as President of SENA, had the final say on all pricing decisions. Defendants emphasize that Korhonen never told other SENA employees about his communications with Tynkkynen. But there was no reason for him to do so, and indeed there were strong reasons for him to remain silent on this score. Korhonen's decisions on pricing were final and, as Korhonen himself has admitted, he knew, at a minimum, that his communications with Tynkkynen violated SENA's internal antitrust policy, which prohibited "sett[ing] prices or verify[ing] the activities of [SENA's] competitors by calling competitors," Defs.' Resp. to Pls.' Local Rule 56(a)(2) Statement ¶ 499. Moreover, one component of the alleged pricing agreement was holding firm with customers as to price — an arrangement that Korhonen was in a position to enforce strictly or variably.

In addition, there is scant evidence in the record to support defendants' assertion that SENA has historically been a market follower rather than a market leader, and that any agreement was therefore of no effect. Defendants cite only Korhonen's qualified statement, made in a declaration filed in the course of this litigation, that "SENA usually did not initiate a price increase on its own, but rather would respond to (and normally match) price increases initiated by other paper manufacturers." Decl. of Kai Korhonen in Supp. of Defs. SEO and SENA's Mot. for Summ. J. ¶ 15 (emphasis added). We are aware of no specific information regarding SENA's pricing decisions outside of the class period in the record on summary judgment. Even during the class period, SENA did not uniformly follow the pricing decisions announced by other publication paper manufacturers: for example, SENA did not match Mead's price increase on all products in November 2002 and led price increases in February 2003 for coated grade 4 paper and supercalendered paper — products that were not covered by IP's price announcement. In any event, the mere fact that following price increases announced by competitors may have been consistent with SENA's overall pricing strategy does not immunize SENA from liability if it had an illegal agreement with UPM to adhere to that strategy. Defendants conceded as much at oral argument before this Court. Oral Arg. Tr. at 21-22.

Finally, the alleged agreement between Korhonen and Tynkkynen would have been valuable to SENA because it significantly reduced SENA's risk: by coordinating with UPM in advance, SENA could follow competitors' price increases secure in the knowledge that UPM, one of its biggest competitors, would not undercut SENA. Without an advance agreement, SENA ran the risk that UPM would hold its prices steady, thereby spiriting away SENA's customers and eating into SENA's market share. For these reasons, a jury could reasonably conclude that even if other SENA employees, based on their own independent business judgments, recommended matching competitors' price increases, Korhonen would neither have approved those blanket price increases nor enforced them strictly, absent assurances from Tynkkynen that UPM would do likewise.11

Defendants also argue that any agreement reached by Korhonen and Tynkkynen during their phone conversation on February 11, 2003, had no effect on pricing because, by that point, SENA had already decided to follow IP's price increase. But defendants do not dispute that SENA's decision to follow IP's price increase was not actually reached until February 11, the day of the phone call. Moreover, when the call was made, neither UPM nor SENA had publicly announced or implemented any price increase. Indeed, SENA sent its price increase announcement to customers approximately one half-hour after Korhonen and Tynkkynen's conversation, and UPM had not finished compiling its list of customer contacts or mailed out its price announcement letter until the afternoon of February 11. It therefore would be reasonable to infer that Korhonen required assurance from Tynkkynen as to UPM's plans before SENA would go forward as it did with its price increase and with its efforts to implement that price increase with customers to the fullest extent possible.12 See Flat Glass, 385 F.3d at 369 (concluding that exchanges of pricing information that are "tightly linked" with subsequent parallel price increases permit the inference that "the exchanges of information had an impact on pricing decisions").

Based on the foregoing, we conclude that the evidence regarding causation presents a genuine dispute of material fact: that is, whether Korhonen and Tynkkynen's agreement, if proven, was both a material and but-for cause of the price increases. The question of causation is therefore a jury question.13

E. SEO

Finally, plaintiffs contend that the district court erred in granting summary judgment to SEO. Plaintiffs have offered evidence that on October 30, 2002, Tynkkynen met with Bernd Rettig, an SEO executive, and that during that meeting Tynkkynen and Rettig agreed that UPM would lead a price increase for publication paper in Europe, and that SEO would match that price increase. Plaintiffs assert that the success of SEO's price-fixing efforts in Europe depended on its ability to ensure that SENA and its rivals fixed the price of publication paper in the United States. Plaintiffs have failed, however, to offer any concrete evidence in support of their theory. For that reason, and because the record is devoid of evidence that SEO had any direct involvement in decisions regarding the marketing, sale, or pricing of publication paper in the United States, we affirm the district court's grant of summary judgment to SEO.

III. CONCLUSION

For the foregoing reasons, we conclude that plaintiffs' Sherman Act section 1 claim against SENA is sufficiently supported to require a trial, but their claim against SEO is not. Accordingly, we VACATE the district court's judgment in part, AFFIRM it in part, and REMAND the case for further proceedings consistent with this opinion.