dissenting.
I respectfully dissent.
The essential question here is whether thirty-one (31) parallel price increase announcements by a small group of suppliers over a ten (10) year period were mere coincidence (lawful and, in fact, expected in the world' of oligopolies) or evidence of an agreement to fix prices (unlawful even among oligopolists). I think there- are enough factual issues in this case that the question whether it was a lawful, coincidence or an unlawful agreement should be decided by a jury.
The majority’s ruling creates an unworkable burden, not supported by our precedent, for plaintiffs seeking to prove a Sherman Act price-fixing case with circumstantial evidence. Second, it affirms a decision where á district judge weighed and compartmentalized evidence, a task better suited for juries—not judges.
An antitrust plaintiff may avoid summary judgment based upon circumstantial evidence alone. That concept is almost a legal axiom, yet it finds no home in the majority opinion. We have long held that a “plaintiff in a section 1 case does not have to submit direct evidence, i.e., the so-called smoking gun, but can rely solely on circumstantial evidence and the reasonable inferences drawn from such evidence.” Petruzzi’s IGA Supermarkets, Inc. v. Darling-Delaware Co., 998 F.2d 1224, 1230 (3d Cir. 1993); see, e.g., In re Baby Food Antitrust Litig., 166 F.3d 112, 124 (3d Cir. 1999) (“A plaintiff in a Section 1 conspiracy can establish a case solely on circumstantial evidence and the reasonable inferences to be drawn therefrom”). In other words, an antitrust plaintiffs burden at summary judgment “is no different than in any other case.” Id. (quoting Big Apple BMW, Inc. v. BMW of N. Am., Inc., 974 F.2d 1358, 1363 (3d Cir. 1992)); In re Chocolate Confectionary Antitrust Litig., 801 F.3d 383, 396 (3d Cir. 2015). Today’s opinion all but explicitly states that, now, “the so-called smoking gun,” Petruzzi's, 998 F.2d at 1230, is required.
As a general principle, “antitrust law limits the range of permissible inferences from ambiguous evidence in a § 1 case,” Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 588, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986), and conduct that is “as consistent with permissible competition as with illegal conspiracy” cannot, on its own, support an inference of an antitrust conspiracy, id. (citing Monsanto Co. v. Spray-Rite Serv. Corp., 465 U.S. 752, 104 S.Ct. 1464, 79 L.Ed.2d 775 (1984)).
This Court favors a sliding scale approach to determine the types of inferences allowed to be drawn from circumstantial evidence in antitrust cases. According to our Circuit’s precedent, and that of the Supreme Court’s, the range of inferences that may be drawn from circumstantial evidence depends upon “the plausibility of the plaintiffs’ theory and the dangers associated with such inferences.” Chocolate, 801 F.3d at 396 (quoting In re Flat Glass Antitrust Litig., 385 F.3d 350, 357 (3d Cir. 2004)). In cases where an antitrust plaintiffs economic theory of liability “makes no economic sense,” and drawing inferences in the plaintiffs favor would deter healthy competition, the plaintiff must produce “more persuasive evidence” to bolster its claim. Id. On the other hand, when a plaintiffs theory makes economic sense, courts draw more liberal inferences in favor of the plaintiff. Id. at n.8.1 Valspar presented an economic theory that makes perfect economic sense, yet the District Court and majority did not draw any inferences in Valspar’s favor.
The majority performs a thorough analysis of the evidence of parallel conduct and the “plus factors.” Viewing all this evidence as a whole, I believe it clear that summary judgment was not proper in this case.
A. Parallel Conduct
It is true that conscious parallelism alone cannot create an inference of conspiracy. The majority has taken this to mean that any evidence of conscious parallelism is therefore incapable of raising an inference of conspiracy. This is incorrect. Parallel pricing is a necessary requirement of any § 1 price-fixing claim, and simply because parallel pricing alone cannot preclude summary judgment does not mean that courts ignore evidence of it. Indeed, our precedent has repeatedly warned against overlooking this important factor in these types of cases, especially where the plaintiffs economic theory—as it does here—makes perfect economic sense.2
The sheer number of parallel price increase announcements in this case—31 to be exact—is unprecedented. Cf. Flat Glass, 385 F.3d at 369 (reversing summary judgment in case involving 7 parallel price increases in 5 years); Chocolate, 801 F.3d at 410 (3 parallel price increase announcements insufficient to withstand summary judgment, in part, because there was no “abrupt” or “radical” shift in pre-conspira-cy conduct). While this sheer number, in itself, cannot carry the day for Valspar, the other evidence viewed in conjunction with these parallel price increase announcements can.
This amount of parallel price increase announcements, in a relatively short time period, commands attention. In Flat Glass, we considered the temporal proximity between the companies’ respective price increase announcements as evidence of an agreement to conspire. 385 F.3d at 364-67. Here, there is evidence that many of the manufacturers’ price increase announcements were made within hours, days, or weeks of each other. For example, in one instance, DuPont announced a price increase at 11:00 a.m., Tronox matched it seven hours later, and Kronos matched it eight hours later. The next day, Millennium and Huntsman announced identical price increases. In another instance, all five Ti02 manufacturers made the same price increase announcement within a four-day period. This close timing creates a strong inference of conspiracy.
The unprecedented amount of parallel price increase announcements, while not dispositive, would undoubtedly raise red flags to any reasonable fact finder. Vals-par’s theory of liability makes “perfect economic sense.” Chocolate, 801 F.3d at 396. Accordingly, “more liberal inferences from the evidence,” which necessarily includes the 31 parallel price increase announcements, should be drawn. Id. The majority’s unwillingness to allow more liberal inferences in the plaintiffs favor seems to me to be a mistake. Instead, the majority gave little weight to the amount of parallel price increase announcements simply because parallel conduct itself is insufficient to create an inference of conspiracy. This approach sees the trees, not the forest. See Flat Glass, 385 F.3d at 357 (mandating courts analyze the evidence “as a whole” to determine whether “it supports an inference of concerted action”).
Of course, “[f]or parallel pricing to go beyond mere interdependence, it must be so unusual that in the absence of an advance agreement, no reasonable firm would have engaged in it.” Baby Food, 166 F.3d at 135. With that said, it is undoubtedly a question of fact as to whether the parallel pricing in this case was sufficiently “unusual.” Id. No case to ever reach us has contained even half the amount of parallel price increase announcements present here—not to mention that many of them here were separated by mere days or hours. This raises an obvious and serious question after today’s decision: What will it now take for a plaintiff relying on circumstantial evidence to move the ball across the goal line?
The sheer amount of parallel conduct in this case, coupled with the plausibility of Valspar’s economic theory, should inform our analysis of the plus factors. Flat Glass, 385 F.3d at 358 (“[A]n agreement among oligopolists to fix prices at a supracompeti-tive level ... makes perfect economic sense” and therefore “more liberal inferences from the evidence should be permitted than in Matsushita because the attendant dangers from drawing inferences recognized in Matsushita are - not present”); id. at n.8 (“Matsushita itself said little about proof requirements in a case where underlying structural evidence indicates that the offense is quite plausible” (quoting Herbert Hovenkamp, The Rationalization of Antitrust, 116 Hahv. L. Rev. 917, 925-26 (2003))).3 It did not. The majority paid very little mind to these distinctions—especially the plausibility of Vals-par’s economic theory.
■The majority’s formulation of the summary judgment standard in this case, coupled with its dismissive treatment of unprecedented parallel-conduct evidence, creates too high a hurdle for plaintiffs attempting to prove a price-fixing conspiracy using circumstantial evidence. The limitations in antitrust cases announced in Matsushita, and that we followed in Chocolate, were never meant to require something more than circumstantial evidence of ans agreement to preclude summary judgment. Nor did they impose some “special” burd|en. Eastman Kodak Co. v. Image Tech. Servs., Inc., 504 U.S. 451, 467, 112 S.Ct. 2072, 119 L.Ed.2d 265 (1992); see Flat Glass, 385 F.3d at 359 n.9 (“[Ujnfortunately, many courts have read Matsushita as requiring a certain quantum evidence of verbal agreement before summary judgment can be. avoided,” (quoting Herbert Hovenkamp, The Rationalization of Antitrust, 116 Harv. L. Rev. 917, 925 (2003))); see also Petruzzi’s, 998 F.2d at 1230 (to create a genuine issue of material fact, the plaintiff' “need not match, item for item, each piece of evidence proffered by the movant, but simply must exceed the ‘scintilla’ standard.”).
The U.S. Supreme Court has gone so far as to caution against this kind of misapplication of Matsushita. In Eastman Kodak Co. v. Image Technical Services, Inc., it emphasized:
The Court’s requirement in Matsushita that the plaintiffs’ claims make economic sense did not introduce a: special burden on ’ plaintiffs facing summary judgment in antitrust cases. The Court did not hold that if the moving party enunciates any economic theory supporting its behavior, regardless of its accuracy in reflecting the actual market, it is entitled to summary judgment. 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. If the plaintiffs theory is economically senseless, no reasonable jury could And in its favor, and summary judgment should be granted.
504 U.S. at 468-69, 112 S.Ct. 2072 (footnote omitted). The Court in Eastman Kodak also expressed a preference to “resolve antitrust claims on a case-by-case basis, focusing on the particular facts disclosed by the record.” Id. at 467, 112 S.Ct. 2072. This principle is particularly poignant here, where the “facts disclosed by the record” (ie., 31 parallel price increase announcements) are strongly suggestive of an agreement to fix prices.
B. The Plus Factors
Although the majority recognizes there is no exhaustive list of- plus -factors, Flat Glass, 385 F.3d at 360, it considers only a few select plus factors and fails to consider others. There is no one plus factor that is “strictly necessary.” Id. at 361 n.12 (quoting In re High Fructose Corn Syrup Antitrust Litig., 295 F.3d 651, 655 (7th Cir. 2002) (Posner, J.)). The presence of certain plus factors does not .automatically preclude summary judgment. Id.; see Petruzzi’s, 998 F.2d at 1242 (recognizing that a “wide range of circumstantial evidence can be used to establish needed plus factor”).
While we often rely on the “big 3” plus factors (motive, actions contrary to interest, and traditional conspiracy), the plus-factor inquiry is not intended to be rigid or formulaic. Flat Glass, 385 F.3d at 361 n.12; Petruzzi’s, 998 F.2d at 1242. There is a slew of other viable plus factors, including, among others: (i) fixed relative market shares; (ii) exchanges of price information; and (iii) price, output, and capacity changes at the formation of the cartel. See William E. Kovacic et al., Plus Factors and Agreement In Antitrust Law, 110 Mich. L. Rev. 393, 415 (2011) (listing Pos-ner’s “fourteen plus factors”); see also id. at 423 (recognizing that a company’s redis-tributions of gains and losses—or “true-ups”—are circumstantial evidence of a conspiracy); Richard A. Posner, Antitrust Law 87 (2d ed. 2001) (recognizing “signaling” as a plus factor, especially when the announcement occurs before the actual implementation of the price increase).
The majority is correct that evidence of the first two plus factors may not aiways nudge the ball over the goal line for a plaintiff at summary judgment because they “often restate interdependence.” Flat Glass, 385 F.3d at 361. But this is not always the case. Because our approach to these cases is fluid, and we must not compartmentalize evidence, there are some cases where these two factors may not simply restate interdependence. Id: at n.12. For instance, certain acts' against self-interest (e.g., non-price acts against self-interest) “cannot simply be explained as a result of oligopolistic interdependence.” Id. • •
1. Motive to Enter Into a Conspiracy
Motive is “important to a court’s analysis, because [its] existence tends to eliminate the possibility of mistaking the workings of a competitive market ... with interdependent, supracompetitive pricing.” Flat Glass, 385 F.3d at 361.
The majority mentioned only' that evidence of motive often restates interdependence and thus does not create an inference of concerted action. However, in Chocolate, cited often by the majority, we simply recognized that “evidence, of motive without more does not create a reasonable inference of concerted action.” 801 F.3d at 298 (emphasis added). Here, there is much “more.”
2. Actions Against Self-Interest
In the District Court, Valspar pointed to substantial evidence that DuPont and the other manufacturers acted contrary to their self-interest. First, Valspar noted that Ti02 prices rose despite no change in the Ti02 market. Second, Valspar argued that the market shares of the Ti02 manufacturers remained relatively stable from 2002 to 2013. Third, Valspar relied on the fact that DuPont and the other manufacturers made intercompany sales to each other at below-market value.
The District Court seemed not to heed Flat Glass's pronouncement that “certain types of ‘actions against self interest’ may do more than restate economic intérdepen-dence.” 385 F.3d at 361 n.12.4 The explicit examples cited in Flat Glass were “unilateral exchanges of confidential price information,” Id. Unilateral exchanges of confidential price information, like other non-price actions against self-interest, “cannot simply be explained as a result of oligopo-listic interdependence.” Id. Valspar presented a triable issue of fact on this point below. There is evidence that the GSP allowed DuPont and the other manufacturers to gain confidential information about each other regarding supply, demand, inventory, and market share.
In a 2002 email, Paul Bradley at Huntsman noted that it would be possible to “derive” each individual manufacturer’s production of Ti02 from the GSP- data. Valspar also produced evidence suggesting that the manufacturers collectively used industry consultant Jim Fisher as a conduit to share confidential information. Fisher attended an industry conference which the manufacturers—including DuPont—attended. According to Fisher, at this conference the Ti02 manufacturers “discussed the need to take advantage of tight market conditions to improve pricing.”5
The same actions contrary to self-interest that led the Seventh Circuit to reverse summary judgment in High Fructose are also present here. Areeda’s treatise recognizes this in identifying the type of oligopoly that may nonetheless be collusive: “there were numerous oral and some written statements by employees of the defendant to the effect that they had an understanding that they would not undercut one another’s prices and that they were involved in an organization (i.e. cartel) controlling prices.” Phillip E. Areeda <& Herbert Hovenkamp, Antitrust Law ¶ 1431b, at 232 (3d ed. 2010) (citing High Fructose, 295 F.3d at 662). As in High Fructose, here there are numerous statements from the manufacturers’ employees (including DuPont) expressing an understanding that they “would not undercut one another’s prices” and that they were involved in an organization to control prices. Id. In one email, ironically, the DuPont author parrots Areeda’s above “undercutting]” language verbatim in advising others to modify pricing “[o]nly if you are not undercutting a Kronos price increase!!!!!” Cf. High Fructose, 295 F.3d at 662 (statement from defendant that “[w]e have an understanding within the industry not to undercut each other’s prices” served as evidence of “an explicit agreement to fix prices” (Posner, J.)). Another email from a different competitor recognized that “all are still acting in a disciplined manner, respecting each other’s market positions and share and holding price.”
There is no shortage of these emails, all of which support an inference that the Ti02 manufacturers were working together. pursuant to an agreement to maintain price. There are further emails indicating that all the Ti02 suppliers were “on the bus,” that DuPont was “training” others on price, and that some of the suppliers were planning price increase announcements in order to allow other suppliers to “get on their horses.” In 2006, a DuPont executive went so far as to comment about another’s price increase: “the timing may be no coincidence - their reading of the CEFIC info like ours should give them confidence that NA [North American] price increases can be prosecuted despite the flat market.” This email is particularly probative of an agreement, given the DuPont executive’s recognition that they could continue to hike up prices even though demand was decreasing. See Flat Glass, 385 F.3d at 358 (finding oligopolists “raising prices” indicative of a conspiracy when the price increases were made “absent increases in marginal cost or demand”).
The repetitive pattern of public price increase announcements is also a garden variety example of action against self-interest. When there is evidence that “the publication of wholesale price increases was intended to make, and has the effect of ... ensuring competitors could quickly learn of, and respond to” these price increases, an inference of an agreement to fix prices arises. See Petroleum Prods. Antitrust Litig., 906 F.2d 432, 446-47 (9th Cir. 1990). A “price announcement given in the hope that rivals will follow” evinces an agreement if “repetition creates an expectation of such behavior.” Areeda & Hovenkamp, Antitrust Law ¶ 1422b, at 171 (3d ed. 2010); see id. (“Although ... mere proof of interdependent pricing, standing alone, may not serve as proof of an antitrust violation, we believe that the evidence concerning the purpose and effect of price announcements, when considered together with the evidence concerning the parallel pattern of price restorations, is sufficient to support a reasonable and permissible inference of an agreement, whether express or tacit, to raise or stabilize prices.” (quoting Petroleum Prods., 906 F.2d at 446-47)).6
The majority relies heavily on the fact that not all of the co-conspirators’ parallel price increase announcements resulted in a sale at the actual announced price. Drawing a distinction between price increase announcements and actual increases has been criticized by judges and scholars alike. Judge.Posner has emphasized that “[i]n deciding whether there is enough evidence of price fixing to create a jury issue, a court asked to dismiss a price-fixing suit on summary judgment must be careful to avoid three traps.” High Fructose, 295 F.3d at 661.
One of Judge Posner’s traps is “to distinguish between the existence of a conspiracy and its efficacy.” Id. at 656. In other words, arguing that although there was an agreement to fix prices, the goods were not actually sold at that price. Id, The majority does just that by relying on the fact that, often times, DuPont and the manufacturers sold Ti02 at a lower price than reflected in their initial parallel price increase announcements. As explained by Areeda (who the majority cites frequently and whose opinions permeate Third Circuit antitrust jurisprudence), a “price announcement given in the hope that rivals will follow” evinces an agreement if “repetition creates an expectation of such behavior.” Areeda & Hovenkamp, Antitrust Law ¶ 1422b, at 171 (3d ed. 2010) (emphasis added).
Our own precedent is at odds with such an analysis. See Flat Glass, 385 F.3d at 362 (rejecting the argument that “regardless of the ... list prices, ... the prices at which flat glass producers actually Sold their product to customers [ ] declined during the period of the alleged conspiracy” because “[a]n agreement to fix prices is ..: a per se violation of the Sherman Act even if most or for that matter all transactions occurred at lower prices” (citing High Fructose, 295 F.3d at 656) (emphasis added)).
The majority fell into this trap. There is plenty of evidence, in the form of emails, that DuPont and the other Ti02 manufacturers made price increase announcements “in the hope that rivals will follow.” Areeda & Hovenkamp, Antitrust Law ¶ 1422b, at 171. While this is not conclusive evidence of an agreement, the scale tips in favor of finding an agreement when there is “repetition” of such price increase announcements. Id. If nothing else, this case involves repetitive price announcements. This repetition clearly gave the suppliers “an expectation,” id., which is further.illuminated by the drastic increase in parallel price increase announcements from 2002 to 2013. These emails, in conjunction with the pattern, frequency, and. effect of the price announcements, tend to exclude the possibility that the Ti02 manufacturers were acting independently.
Valspar also presented evidence that DuPont- and the other suppliers consciously maintained static market shares.-Market share stability is a well-recognized symptom of collusive and concerted action in antitrust cases. See William E. Kovacic et al., Plus Factors and Agreement in Antitrust Law, 110 MICH. L. REV. 393, 415, 422 (2011).7 The logic is not difficult to understand-: a- company acting in a healthy, competitive, and self-interested manner would seek to expand—not maintain or decrease—its own market share.8 The-District Court acknowledged this evidence. Instead of submitting it to the jury as a disputed question of fact, however, the District Court summarily concluded that this evidence did not- show collusion because the Ti02 market is an oligopoly. Valspar Corp. v. E.I. Du Pont De Nemours, 152 F,Supp.3d 234, 242-43 (D. Del. 2016). The majority took the same approach. Just like the District Court, the majority weighed the expert evidence on this issue and made a finding: that the evidence of (likely unilateral) market share stability was insufficient in this case to show'concerted action or ’ agreement. It seems to me that if the court is “weighing evidence” or “making findings” it' should be at trial, on a full record and done by a fact finder, i.e;, a jury or judge sitting without a jury.
Dovetailing with this evidence of static market shares is evidence that the Ti02 manufacturers made intercompany sales of Ti02, meaning, they sold Ti02 to one another. This evidence might 'indicate pure competition but for-the. fact that .the manufacturers frequently sold the Ti02 to their competitors, at below-market prices. For example, when DuPont would sell Kronos Ti02, Kronos paid an average of 16% less for the Ti02 than DuPont’s own customers did. DuPont also sold Ti02 to Millennium at below-market prices.9 One of Valspar’s experts, Dr. Williams, was able to identify years of below-market sales between the Ti02 manufacturers.
The majority downplays this evidence of below-market intercompany sales. It apparently considered these sales just as consistent with non-collusive activity as with conspiracy because: (1) DuPont used some Ti02 in its own production in 2005 and 2006; (2) a DuPont plant in Mississippi shut down for five months; and (3) DuPont’s sales to Kronos were governed by a patent settlement agreement from 2006 to 2008. The District Court, and the majority, found the volume of intercompa-ny sales insufficient to constitute a “true-up.”10
The majority notes that the intercompa-ny sales were low in number and unlikely to show an agreement. According to this logic, there is no evidence of a “trué-up” because the intercompany sales were fairly low in quantity. Yet the very purpose of a “true-up” is for the companies within a cartel to -maintain their market share. Therefore,-it might not necessarily make sense for a company to make a large cross-sale, or a large number, of cross-sales, in order to maintain its relative market share.11
Obviously, intercompany sales “could” be redistributions of gains or losses, Valspar, 152 F.Supp.3d at 244, and “might” be explained by the closed DuPont plant, id. However, these “coúlds” and “mights” cast doubt on—hot support—DuPont’s argument for summary judgment. Where there are reasonable inferences that there was more likely than not a conspiracy to fix prices, summary judgment is not proper.12 The majority, like the District Court, accepted each of DuPont’s explanations of possibly conspiratorial conduct and adopted each without much explanation. This approach should be unacceptable at the summary judgment stage. See Flat Glass, 385 F.3d at 368 (explaining we should not “consider each individual piece of evidence and disregard it if we could feasibly interpret it as consistent with the absence of an agreement to raise prices”); Petruzzi’s, 998 F.2d at 1230 (to create a genuine issue of material fact, the plaintiff “need not match, item for item, each piece of evidence proffered by the movant, but simply must exceed the ‘scintilla’ standard.”); Areeda & Hovenkamp, Antitrust Law § 14.03b, at 14-25 (4th ed. 2011) (plaintiffs need not “disprove all noncon-spiratorial explanations for the defendants’ conduct” to prevail at summary judgment).
3. Traditional Conspiracy Evidence
Traditional conspiracy evidence is often the most important “plus factor” in a case like this one. Chocolate, 801 F.3d at 401. Traditional conspiracy evidence is evidence that “the defendants got together and exchanged assurances of common action or otherwise adopted a common plan even though no meetings, conversations, or exchanged documents are shown.” E.g., id. at 398. Even though “no meetings, conversations, or exchanged documents” are required as direct evidence of a conspiracy, DuPont urged us to require, in essence, exactly that. .
This Court has explicitly and repeatedly held that “traditional conspiracy” evidence may exist “even though no meetings, conversations, or exchanged documents are shown.” Chocolate, 801 F.3d at 398 (emphasis added); Superior Offshore Internat’l, Inc. v. Bristow Grp., 490 Fed.Appx. 492, 499 (3d Cir. 2012); Burtch v. Milberg Factors, Inc., 662 F.3d 212, 227 (3d Cir. 2011); In re Ins. Brokerage Antitrust Litig., 618 F.3d 300, 322 (3d Cir. 2010); Flat Glass, 385 F.3d at 361. Yet the majority seems to require Valspar to present evidence of direct meetings and conversations. The majority reasoned that since “there is no evidence that there was any discussion of prices during these meetings and certainly no evidence of an agreement,” Valspar’s argument “falls short.” Maj. Op. at 197.
In reality, Valspar presented various forms of traditional conspiracy evidence. For example, Valspar presented a Millennium email stating “we have competition on board for the Oct 1 price increase announcement.”13 Having “competition on board” for a price increase announcement certainly conveys that the suppliers somehow got together and exchanged assurances of “common action,” i.e., to announce the same prices. Id. The same goes for the suppliers’ emails about the “collective needs” of the industry14 and getting everyone “on the bus” or, put another way, “on their horses.”
Today’s decision could easily be read to require direct evidence of an agreement in an oligopoly/antitrust case despite the fact that neither our prior jurisprudence (nor the Supreme Court’s) has ever required such evidence. What’s more, it is not even correct to state that no meetings or conversations between competitors took place. In 2004, the CEO of Millennium met with the President of Huntsman. The very next day, an internal Millennium email stated that they had “competition on board for the Oct 1 price increase announcement.” A few years earlier in 2002, DuPont announced a price increase a few days after Jim Fisher met with Kronos and then with DuPont. And if that wasn’t enough, according to Jim Fisher, the suppliers, at a meeting together, “discussed the need to take advantage of tight market conditions to improve pricing.”15
I am not sure how this circumstantial evidence could be stronger. It unequivocally shows that one alleged conspirator’s (Millenium’s) CEO met with another alleged conspirator’s (Huntsman’s) President days before a parallel price increase announcement. This meeting occurred at the same time an email was written stating that Ti02 “competition” was “on board” with a particular price increase announcement. Even more persuasive, there is evidence that all the Ti02 suppliers discussed “improv[ing] pricing” at an industry conference in 2005 and that in 2002 DuPont and Kronos announced an identical price increase just days after Jim Fisher met with these two “competitors.”
A jury should be allowed to determine whether Fisher’s meetings with both Kro-nos and DuPont—days before a parallel price increase announcement—were suspect. A jury should be allowed to determine whether an email that “competition” is “on board” for a price increase announcement was concerted action, particularly when this email was written one day after Huntsman’s President personally met with Millennium’s CEO. A jury should be been permitted to decide whether a meeting with the Ti02 manufacturers, in which they explicitly discussed “improving] pricing,” supports an inference of concerted action. This is the exact sort of powerful evidentiary synergy the majority implies is absent from Valspar’s case.16 This approach misses by a mile an essential truth of actual courtroom litigation: that circumstantial evidence is competent, valid, and vital evidence in almost every conspiracy trial, civil or criminal. The courtroom litigation process, though sometimes messy and unpredictable, is the preferred method for the resolution of factual questions under our Seventh Amendment. And, with all its quirks, the civil trial is a far better method of evaluating evidence, direct or circumstantial, than an academic parsing of a printed record developed in discovery.
The inculpatory flavor of these emails is enhanced by the fact that the suppliers were all a part of the TDMA, which gave them access to highly confidential information via the GSP. The majority attempts to analogize the GSP to In re Citric Acid Litigation, 191 F.3d 1090 (9th Cir. 1990). In Citric Acid, the U.S. Court of Appeals for the Ninth Circuit rejected a theory that the alleged conspirators used their membership in a trade association as a front for conducting conspiratorial activities. 191 F.3d at 1097-98. It aptly pointed out that if courts “allowed conspiracy to be inferred ■ from such activities alone, we would have to allow an inference of conspiracy whenever a trade association took almost any actión.” Citric Acid, 191 F.3d at 1098 (emphasis added).
I agree with the majority that membership in a trade association,- in itself, cannot serve as traditional evidence of a conspiracy. Nonetheless, our task is not to view the Ti02 suppliers’ membership in the TDMA in a vacuum. When viewed in conjunction with the other evidence, the membership can be seen in a much different light than Citric Acid. For starters, there is evidence here ■ (absent in Citric Acid, Chocolate, or any other case) of 31 parallel price increase announcements. Nearly all of these announcements- came within thirty days of a TDMA meeting.17 Also absent from Chocolate or-Citric Acid is the presence of an industry consultant (Jim Fisher) who was simultaneously retained by multiple “competitors” to gather pricing information. Obviously, no court could say with certainty that DuPont agreed to fix prices with the other suppliers at the TDMA meetings. But making a judicial determination with certainty is not our job at summary judgment. The point is that, the above evidence, viewed in a light , most favorable to Valspar—whose theory makes perfect economic sense—creates an inference of concerted action sufficient to reach ajury.
(The majority downplays the role of this, key “industry consultant” Jim Fisher. There is evidence suggesting Fisher was used as a vehicle to carry out the suppliers’ collusive agreement. On June 11, 2002, DuPont publicly announced a price increase.18 Four days prior, Jim Fisher had called DuPont’s Competitive Intelligence Manager, Connie Hubbard. In that conversation, Fisher conveyed confidential pricing information about one of DuPont’s competitors: Huntsman. This was the first Hubbard .had heard of this Huntsman increase because it had not been announced publicly. This interaction, in itself, provides traditional conspiracy evidence in that it suggests Huntsman and DuPont may have used Fisher to implement a “common' plan” to fix prices “even though no meetings, conversations, or exchanged documents are shown” between Huntsman and DuPont. Chocolate, 801 F.3d at 398.19
There .is plentiful evidence of price signaling (another plus factor) in this case. In Petroleum Products Antitrust Litigation, the U.S. Court of Appeals for the Ninth Circuit confronted nearly identical circumstantial evidence of price signaling. 906 F.2d 432, 446 (9th Cir. 1990). Like here, in Petroleum Products, there was evidence that the competitors’ “price increases were occasionally announced in advance of their effective date.” 906 F.2d at 446 n.11. This type of preemptive announcement, the Ninth Circuit recognized, is “effective in allowing the price leader to communicate its intention and to receive reactions without having to incur substantial risk.” Petroleum Prods., 906 F.2d at 446 n.11.
, DuPont maintains that it made price announcements publicly because it was required to do so per its contracts with customers. While it is true that DuPont was required to notify its customers when it changed its price, there is no evidence that DuPont was required to do so publicly, In fact, the record evidence suggests the opposite. Notably, DuPont never publicly announced price decreases. This supports an inference that DuPont’s public price increase announcements were for the purpose of collusion rather than customer notification.
The Ti02 manufacturers even admitted they did not use public price increase announcements to notify customers. When asked about the purpose of public announcements, Kronos’s Jay Becker agreed he would “never” rely on a public price increase to provide Kronos customers with notice of a price increase. Gary Cianfíchi, from Millennium, similarly stated that he did not believe any customer contract required public notice.. Larry Rogers, from Kronos, testified he “really, couldn’t say why” price increases were announced publicly given that the customers were notified privately in writing. This evidence also supports an inference that the, manufacturers used public price increase announcements as a concerted method of fixing prices market-wide.
The traditional conspiracy evidence in this case is much different than it was in Chocolate. We recognized in Chocolate that a company’s departure from pre-conspira-cy conduct can serve as traditional conspiracy evidence, which is the “most important plus factor.” Chocolate, 801 F.3d at 401, 410. The caveat is that the change from pre- to post-conspiracy conduct must be “radical” or “abrupt.” Id. at 410. In Chocolate, there was no radical or abrupt change. This case differs from the departure of pre-conspiracy conduct in Chocolate in three significant respects.
First, and most basic; Chocolate involved comparing 2 pre-conspiracy price increases with 3 post-conspiracy price increases. Id. Here, by stark contrast, we must compare 3 pre-conspiracy increases with 31 post-conspiracy increases. In other words, in Chocolate, the pre-post ratio was just above 1:1 whereas here the ratio is 10:1. It would be difficult to claim that such a change in conduct is not “abrupt” or “radical.” In a case with unprecedented (31) parallel price increase announcements, such a finding creates an unwarranted burden for plaintiffs relying on circumstantial evidence to prove a price-fixing conspiracy. The majority attempts to explain this radical and abrupt shift, implying it was “just an uptick in frequency.”
Second, the nature of the communications between competitors in Chocolate is different from the communications here. In Chocolate, we found the communications unpersuasive in part because “unlike in Flat Glass,” the communications did “not reveal pricing plans dependent on others following.” Id. at 408. Here there is evidence, as in Flat Glass, that could give rise to an inference that the Ti02 manufacturers’ pricing, decisions were dependent upon the decisions of others. Like Flat Glass and unlike Chocolate, the communications here were made between high-level rather than low-level employees. The evidence that Fisher communicated contemporaneously with people from Kronos, Millennium, Huntsman, and DuPont suggests the individual suppliers’ pricing plans were “dependent on others following.” Id. The DuPont email advising to modify pricing “[o]nly if it meant a Kronos price increase would not be “undercut” similarly suggests pricing plans “dependent on others following.”' Id. (emphasis added). As does Fisher’s direct testimony that, at an industry meeting, the Ti02 suppliers “discussed the need to take advantage of tight market conditions to improve pricing.”20 Finally, evidence that certain executives asked Jim Fisher to confirm others’ planned price increase announcements further indicates the suppliers were making decisions not on their own—as the evidence showed in Chocolate—but rather, as in Flat Glass, based on other suppliers’ price decisions.
Third, the pre-conspiracy prices in Chocolate related to “different products” than the post-conspiracy price increases. 801 F.3d at 410. Here, the pre-conspiracy and post-conspiracy price increase announcements related to the same fungible product: Ti02. This argument, unlike the argument of the appellants in Chocolate, is an exact “apples-to-apples” comparison. Chocolate, 801 F.3d at 410.
One final point has been overlooked in comparing this case to Flat Glass and Chocolate: neither Flat Glass nor Chocolate involved nearly as many parallel price increase announcements as we have here. To be clear, again, these parallel price increase announcements, viewed alone, are not enough to defeat summary judgment. However, we are not to “consider each individual piece of evidence and disregard it if we could feasibly interpret it as consistent with the absence of an agreement to raise prices.” Flat Glass, 385 F.3d at 368. When viewed alongside all the other evidence in this case, the unprecedented parallel price increase announcements—many of which were made hours or days within each other—create an inference that the suppliers’ conduct was collusive.
These principles are especially important given that the majority continually relies on the proposition that “[cjonduct as consistent with permissible competition as with illegal conspiracy does not, standing alone, support an inference of conspiracy sufficient to survive summary judgment.” Matsushita, 475 U.S. at 588, 106 S.Ct. 1348; see Valspar, 152 F.Supp.3d at 240, 244, 246, 249, 252-53 (relying on this principle with respect to individual pieces of evidence). But see Flat Glass, 385 F.3d at 368 (courts shall not “consider each individual piece of evidence and disregard it if we could feasibly interpret it as consistent with the absence of an agreement to raise prices”). Matsushita seems wrongly applied when used to discredit each separate piece of proffered evidence an antitrust plaintiff brings forth. Flat Glass, 385 F.3d at 358 n.8, 359 n.9. It is also wrongly applied in a case like this, where the plaintiffs’ theory—oligopolists conspired to fix prices—makes perfect economic sense. Id. at 358; Petruzzi’s, 998 F.2d at 1231-33 (rendering the Matsushita presumption against liberal inferences “unnecessary” because the plaintiffs’ “theory is not implausible”).21
It would not be too difficult to view the 31 parallel price increase announcements, standing alone, as consistent with interdependence. This, of course, would ignore the comparatively miniscule amount of pre-conspiracy price increase announcements: 3. It would also not be too difficult to view the relative market share stability of the Ti02 suppliers, standing alone, as consistent with interdependence. This, of course, would ignore the simultaneous intercompa-ny sales at below-market value. The same goes for the manufacturers’ TDMA and CEFIC membership, meetings, and the GSP. There is nothing inherently collusive about trade associations, industry meetings, or aggregated statistics. This too, of coprse, would ignore the role Jim Fisher played as a communicator of confidential information between the Ti02 suppliers and that the co-conspirators’ executives had meetings together days and hours before they announced parallel price increases.
The majority seems to discount the plausibility of Valspar’s economic theory. This factor has been a focal point in our antitrust jurisprudence for decades. Matsushita, 475 U.S. at 588-91, 106 S.Ct. 1348; Chocolate, 801 F.3d at 396; Flat Glass, 385 F.3d at 358; Petruzzi’s, 998 F.2d at 1232. The majority substitutes this distinct factor with the more general theory of interdependence. See Maj. Op. at 191 n.l. According to the majority, in Flat Glass, this Court refused to draw liberal inferences in favor of the plaintiff because of the theory of interdependence. In fact, this Court did draw liberal inferences in Flat Glass, and reversed summary judgment, partly because the plaintiffs economic theory made perfect sense. It simply stated, in passing, that courts must be “cautious in accepting inferences from circumstantial evidence” in these types of cases—not that they do not do so. Flat Glass, 385 F.3d at 358.
There is no disagreement here that courts should take a “cautious” approach to accepting inferences from circumstantial evidence in price-fixing cases involving oligopolies. See Chocolate, 801 F.3d at 412 (explaining this approach); Flat Glass, 385 F.3d at 358-59 (same). I agree that this cautious approach is consistent with our Circuit’s law. I disagree with the majority’s transformation of this general “cautious” approach into a new approach that appears to shut the door on a district court’s ability to accept reasonable inferences in any case involving oligopolists. Such a black-and-white approach is not resonant of the type of “caution” discussed in Flat Glass and Chocolate, but rather acts to usurp the jury’s role in deciding cases loaded with circumstantial evidence of an actual agreement to fix prices.
In Chocolate, this Court confirmed that “[ujnder Matsushita, the range of acceptable inferences that may be drawn from ambiguous or circumstantial evidence varies with the plausibility of the plaintiffs’ theory.” 801 F.3d at 396; see also id. at n.8 (comparing cases where this Court has drawn liberal inferences when the plaintiffs theory made sense with cases where this Court refused to do so because the theory did not make sense). In Chocolate, this Court did not view the theory of interdependence as a complete roadblock to drawing liberal inferences. Id. at 396-97. The only time this should happen is if the plaintiff is relying on “ambiguous evidence alone.” Id. at 396. Up until today, the general theory of interdependence never supplanted a court’s consideration of the plaintiffs economic theory.22
C. Conclusion
I am, certainly mindful of the theory of interdependence and the presence of an oligopoly. With that said, from the very start, Valspar presented a theory that makes perfect economic sense. It supported this theory with strong evidence of parallel conduct in the form of 31 (an unprecedented amount) of parallel price increase announcements. Recognizing conscious parallelism to be insufficient on its own to survive summary judgment, Vals-par also presented viable evidence in support of the plus factors: (i) price.signaling, (ii) exchanges of confidential information, (iii) relatively static market shares, (iv) intercompany sales of Ti02 at below market price, (v) abrupt, departure from pre-conspiracy conduct, and (vi) a market susceptible to conspiracy. Although the Ti02 market is an oligopoly, Valspar also presented evidence that did not simply restate interdependence: non-price acts against self-interest. -Finally, it presented traditional conspiracy evidence. Viewed together, and not compartmentalized, all this evidence was more than sufficient to preclude summary judgment. : .
For these reasons, I respectfully dissent.
. Based on this sliding scale approach—first articulated in Matsushita—courts have taken varying approaches to cases depending on the strength of the plaintiffs theory. Compare Matsushita, 475 U.S. at 588-91, 106 S.Ct. 1348 (refusing to draw liberal inferences from plaintiffs’ "predatory pricing” theory, which posited that multiple companies conspired to lower prices, because a conspiracy to lower prices makes no economic sense), with Petruzzi's, 998 F.2d at 1232 (drawing liberal inferences and reversing summary judgment because the plaintiffs theory, that companies conspired not to compete with each other, made “perfect economic sense”), and Flat Glass, 385 F.3d at 358 (drawing liberal inferences in reversing summary judgment given that "an agreement among oligopolists to fix prices ... makes perfect economic sense”).
. We discuss the plausibility of Valspar’s economic theory in greater detail infra. For now, it is enough to say we have previously held that Valspar's exact economic theory (parallel price fixing among oligopolists) makes perfect economic sense: "an agreement among oligo-polists to fix prices at a supracompetitive level ... makes perfect economic sense.” Flat Glass, 385 F.3d at 358.
. • I share the concern, of the amicus—namely, that it would be an absurd result if, "in situations 'in which the danger of [parallel pricing] is most serious,” liability would actually’ be "less likely,” Amicus Br. at 15 (quoting Louis Kaplow, Competition Policy & Price Fixing 126 (2015)). A plain reading of our case law reveals this Court never intended to ramp up a price-fixing plaintiff’s burden of proof, especially when the plaintiff's economic theory makes perfect economic sense. Chocolate, 801 F.3d at 396-97; Flat Glass, 385 F.3d at 358. In fact, courts must do the opposite in such a scenario- by drawing liberal inferences in favor of the plaintiff. Id.
. The majority simply states, without consideration of the actual evidence, that "Valspar has not shown” that the first two plus factors "do more than restate the theory of interdependence.” Majority at 197 n.9. The actual evidence, discussed supra and infra, shows otherwise.
. In re Titanium Dioxide Antitrust Litig., 959 F.Supp.2d 799, 812-13 (D. Md. 2013).
. Citing to one single DuPont email, the majority reasons that "[h]ad the competitors gotjten] together and exchanged assurances of common action or otherwise adopted a common plan, ... there would have been no need for DuPont to resort to public announcements to 'test' whether its competitors were 'receiving/understanding [its] price increase messages." Majority at 201 n.14. While it may be possible to reach this conclusion by reading one email in isolation, the evidence as a whole creates a reasonable inference that there was more likely than not an agreement to fix prices. For example, evidence of the meetings between Fisher and all the alleged conspirators, as well as the actual meetings of top executives from Kronos and Huntsman (followed almost immediately by a parallel price increase announcement), as well as the conspirators all discussing "improving pricing” does create an inference of "a common plan.”
. The majority criticizes Valspar’s reliance .on legal scholarship as opposed to case law. This is interesting given that (1) this Court has long., turned to legal scholarship to inform their decisions in antitrust cases involving oligopolists, and (2) the majority itself cites to legal scholarship—including multiple law review articles—seven times.
.' For example, say Company A has a 30%, market share in a particular industry and Company B has a 40% market share in that industry. Obviously, Company A and B, assuming they are competitive, would want to acquire as much market-share for themselves as possible. Therefore, it would defy all logic and notions of procompetitive behavior for Company A (who has a lower market share) to take affirmative actions to stay at 30% rather than grow beyond a 30% market share.
. Titanium Dioxide, 959 F.Supp.2d at 814.
. A "true-up” occurs when companies in a conspiracy redistribute their individual gains and losses in order to comply with their conspiratorial agreement, Kovacic et al., Plus Factors, at 423. Such a transaction "leads to a strong inference of collusion” since there is "no reasonable noncollusive explanation” for intercompany sales at "nonmarket prices” between companies that are supposed to be competing with one another. Id.
. To use another example, say Company A enjoys 35% of the market share, while Company B has 50% of the market share. Assume A and B are colluding and, thus, they want to raise prices and maintain market shares per their agreement. Then assume that Company A’s share drops to 33%. Company B may sell a very small amount o'f product to Company A simply to allow Company A to maintain its market share. Mistaking this as an insignificant sale, merely because of its size or-lack of frequency, would be an oversight.
. Federal Rule of Civil Procedure 56 states: "The court shall grant summary judgment if the movant shows that there is no genuine dispute as to any material fact and the mov-ant is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a). Simply because courts must exercise caution in these types of cases does not .do away with Rule 56’s proposition that genuine disputes of material fact preclude summary judgment: "Generally, the movant’s burden on a summary judgment motion in an antitrust case ‘is no different than in any other case.’ ” Intervest, 340 F.3d at 159. In these cases, courts still must deny summary judgment if there is a genuine dispute of material fact and "view the facts and any reasonable inferences drawn therefrom in the light most favorable to the party opposing summary judgment.” Id. (emphasis added) (citing Eastman Kodak, 504 U.S. at 456, 112 S.Ct. 2072), The majority did not do this. Also contrary to the majority's entire analysis, the "special consideration” and "caution” we ap-' ply in these types of cases informs—not supplants—the general guidelines found in Rule 56.
. Titanium Dioxide, 959 F.Supp.2d at 829.
. The District Court read this to mean the "collective needs" of Millennium alone. However, read in the context of the entire email, a reasonable jury could certainly conclude the opposite: that the author was referring to the collective ■ needs of the Ti02 industry members.
. In re Titanium Dioxide, 959 F.Supp.2d at 812-13 (quoting Fisher) (emphasis added).
. The majority does not discuss this particularly damning evidence, but states generally that "the record does not show the existence of an actual agreement.” Majority at 198 n.l 1. Contrary to the majority’s insistence, an actual agreement can be shown in the exact way that Valspar has set out to do so in this case. See Areeda & Hovenkamp, Antitrust Law ¶ 1422b, at 171 (3d ed. 2010) ("Although ... mere proof of interdependent pricing, standing alone, may not serve as proof of an antitrust violation, we believe that the evidence concerning the purpose and effect of price announcements, when considered together with the evidence concerning the parallel pattern of price restorations, is sufficient to support a reasonable and permissible inference of an agreement, whether express or tacit, to raise or stabilize prices.”) (emphasis added).
. DuPont's insistence that it did not begin attending TDMA meetings until 2010 is moot given that it attended CEFIC meetings—long before 2010—that were, concurrent with the TDMA meetings. Valspar, 152 F.Supp.3d at 246 n.5.
. Titanium Dioxide, 959 F.Supp.2d at 811.
.Obviously, based on this evidence, and other evidence of Fisher’s cross-company communications, a reasonable jury could infer that no direct conversations between the Ti02 manufacturers were needed if Fisher acted as their mouthpiece. ■
, In re Titanium Dioxide, 959 F.Supp.2d at 812-13.
. DuPont relied on the requirement that, to survive summary judgment, Valspar must present evidence that "tends to exclude the possibility” that the alleged conspirators acted independently. Matsushita, 475 U.S. at 588, 106 S.Ct. 1348. All of the evidence in this case, viewed in its totality, tends to exclude that the Ti02 manufacturers acted independently. Contrary to DuPont’s interpretation, "tends to exclude the possibility” does not mean "unequivocally excludes the possibility.” Such a standard would defy our basic summary judgment jurisprudence, which views the evidence in a light most favorable to the nonmoving party.
. In an attempt to assuage concerns about its analysis, the majority tries to justify its heightened standard through Chocolate. According to the majority, this Circuit’s precedent has long required a plaintiff in this type of case to prove that it is “more likely than not” true that there was a price-fixing conspiracy in order to survive summary judgment. Majority at 192 n.1, 194, 194 n.4, 201-02, 201 n.14. The majority makes too much of this dicta. This was not, as the majority claims, some profound announcement of a new legal standard or rule. Indeed, this purportedly axiomatic language has never once been used in any other price-fixing case involving oligopolies. See generally Matsushita, 475 U.S. 574, 106 S.Ct. 1348 (not using this language anywhere); Monsanto, 465 U.S. 752, 104 S.Ct. 1464 (same); Flat Glass, 385 F.3d 350 (same); Baby Food, 166 F.3d 112 (same). The use of that phrase in Chocolate simply reflected the Court's conclusion that the plaintiffs, in that particular case, had not .been able, to point to any reasonable infer-enees of a conspiracy. Even assuming arguen-do this alleged "standard” were actually the measuring stick, the plaintiffs in this case have certainly met it.