This is an antitrust action brought pursuant to 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, by a class of several hundred cigarette wholesalers ("the class" or "the wholesalers") against Philip Morris, Inc. ("PM"), R.J. Reynolds Tobacco Co. ("RJR"), Brown & Williamson Tobacco Corp. ("B & W") and Lorillard Tobacco Co. ("Lorillard") (collectively "the manufacturers"). The class alleges that the manufacturers conspired between 1993 and 2000 to fix cigarette prices at unnaturally high levels, and that this collusion resulted in wholesale list price overcharges of nearly $12 billion. The district court ultimately entered summary judgment in favor of the manufacturers. It reasoned that the wholesalers had failed to demonstrate the existence of a "plus factor," as is necessary to create an inference of a price fixing conspiracy, and that even if the class had shown that a plus factor was present, the manufacturers were able to rebut fully the inference of collusion, as the economic realities of the 1990s cigarette market rendered the class's conspiracy theory untenable. Rather, the district court held that the manufacturers' pricing behavior evidenced nothing more than "conscious parallelism," a perfectly legal phenomenon commonly associated with oligopolistic industries.

On appeal, the wholesalers say that the district court misapplied the summary judgment standard, that they presented sufficient evidence to withstand the manufacturers' motions, and that the court erred by excluding portions of the testimony proffered by their primary expert witness. In the end, we conclude that none of the class's arguments are compelling and that the district court's treatment of the wealth of complicated issues in this case was nuanced, insightful and, ultimately, correct. Accordingly, we affirm the court's entry of final summary judgment in favor of PM, RJR, B & W and Lorillard.

* The modern American tobacco industry is a classic oligopoly. Between 1993 and 1999, appellees — the nation's four largest cigarette manufacturers — along with Liggett Group, Inc.,1 manufactured more than 97% of the cigarettes sold in the United States. Moreover, the composition of the industry has been remarkably stable over time, a condition that has resulted largely from the fact that during the twentieth century the major tobacco players engaged in minimal price competition. Because price fluctuations were relatively rare, smokers typically had no reason to change brands, brand loyalties were solidified and sizable market share shifts were uncommon.

During the early 1990s, however, a price gap widened between premium brands like Marlboro, Newport and Camel and discount and deep discount brands2 such as GPC, Basic and Doral. This price differential was the result of extremely competitive pricing of the non-premium brands, especially by B & W and RJR, which focused a large percentage of their competitive efforts on the discount and deep discount markets. This led some "premium smokers" to shift to one of the non-premium brands, and by 1993 these brands had captured over 40% of the United States market. At that time, there were 10 different wholesale list price points, i.e., cigarette price tiers.

Although this trend toward the discount and deep discount brands benefitted RJR and B & W, it was extremely undesirable from the perspective of premium-intensive manufacturers like PM and Lorillard.3 As such, PM — which at the time was (and remains) the market leader, with a market share that ranged from 42% to 50% during the period of the alleged conspiracy — sought in April, 1992 to raise the price of its lowest tier products by $4 per thousand. This effort was unsuccessful, however, because RJR, B & W and Lorillard did not follow suit, and PM was forced to rescind its price increase. Although PM again attempted to increase its deep discount prices in March, 1993, this effort similarly was rebuffed by its competitors.

Though temporarily unsuccessful, PM continued looking for ways to reverse the trend toward discount cigarettes, and roughly one year after its failed $4 per thousand price increase it found one. On April 2, 1993, PM decided to take what appellants refer to as "the single boldest commercial move in U.S. cigarette market history": it announced that it was cutting the retail price of Marlboro cigarettes — which were by far the single best selling brand in America, enjoying a 21% market share — by 40 cents per pack and foregoing price increases on other premium brands "for the foreseeable future." April 2, 1993 subsequently became widely known throughout the industry as "Marlboro Friday." This highly competitive pricing decision was extremely significant for several reasons. First, it left no doubt that PM was willing to take drastic competitive measures (indeed, to sacrifice profits) in order to protect the market share of its flagship brand. Second and quite importantly, it slimmed the price gap between premium and discount cigarettes. Because this price differential was constricted, consumers suddenly had less of an economic incentive to purchase discount cigarettes, and as a result premium brands like Marlboro regained some of market share they had lost prior to Marlboro Friday.

Finally, it set off a price war among appellees, as RJR, B & W and Lorillard were immediately confronted with a need to respond in some way to PM's bold action. In order to remain competitive, these manufacturers matched PM's retail price reductions. Although these pricing actions cut into the market share held by discount brands generally, and thus led to a reduction in the overall share held by RJR and B & W, which, as stated, were more heavily invested in these brands, the decision to match PM's price reduction meant that no manufacturer suffered unduly large market share losses.

However, this vast decrease in cigarette prices was disastrous for PM, RJR, B & W and Lorillard alike in terms of profits,4 and appellees were forced to rethink their profitability strategies. Indeed, appellants recognize that this economic landscape became especially difficult in light of increasing regulation of the industry and the surge of health-related litigation. To exacerbate its competitors' predicament, on July 20, 1993, PM announced that its Marlboro Friday price reduction would be made permanent and expanded to all of its premium brands, e.g., Parliament and Virginia Slims. Moreover, PM simultaneously lowered the wholesale price of its discount cigarettes and raised the wholesale price of its deep discount brands by 10 cents per pack, thereby consolidating the prices of these brand categories. This action reduced what had been 10 price points in the American cigarette market to four: 85 mm (regular length) premium cigarettes occupied one price tier, 100 mm (extra-long) premiums occupied another, and 85 mm and 100 mm discounts occupied the remaining two tiers. Again, PM's competitors promptly matched these newly announced prices. What's more, the very next day RJR announced that it would collapse the prices of its regular and 100 mm cigarettes, thereby reducing the price tiers to two: premium and discount. PM, B & W and Lorillard quickly followed suit.

Appellants contend that it is only at this point that PM, RJR, B & W and Lorillard began conspiring to fix and steadily increase prices to make up for the tremendous financial losses they suffered as a consequence of this price war. The class says that PM's actions on Marlboro Friday constructively informed its competitors that price discounting to gain market share would no longer be tolerated, and that only when such efforts were abandoned, and the premium/discount price gap narrowed, could prices again rise. By appellants' account, the manufacturers' conspiracy began in earnest when appellees began using trade press — i.e., tobacco industry financial analysts like Gary Black — to "signal" each other regarding their willingness to comply with PM's implicit demands so as to facilitate price increases. For example, the class points to a statement made by Martin Broughton, the CEO and Deputy Chairman of British American Tobacco ("BAT"), B & W's parent company, that "BAT may be one of those who started the price war in the U.S., but we have no wish to escalate it."5 Appellants argue that this statement somehow was a signal sent by B & W to its competitors that B & W was willing to reduce or end price competition. The class alleges that RJR conveyed a similar message on November 2, 1993 by publicly announcing that it would no longer sacrifice profitability for market share.

Appellants say that PM signaled its acceptance of its competitors' overtures by putting its distributors on "permanent allocation" — that is, limiting the quantities of product the distributors could order — on November 5, 1993. Historically, this practice had been implemented as a temporary measure prior to a price increase, with the purpose being to prevent the wholesalers from engaging in "trade loading," i.e., stocking up on cigarettes before the increase so as to deprive appellees of the benefit of their price adjustment. The class says that PM's placement of its wholesalers on permanent allocation actually was a signal that the price increase sought by RJR, B & W and Lorillard was coming.

Appellants suggest that RJR responded to PM's signal on November 8, 1993 with one last signal of its own. Specifically, the class argues that by announcing an increase of $2 per thousand cigarettes (4¢ per pack)6 in both the premium and discount categories, and thus maintaining the constricted discount/premium price gap, RJR indicated that it was acceding to PM's conditions for increasing prices. By November 22, 1993, PM, B & W and Lorillard had matched RJR's increase. The class argues that it was in the economic interests of RJR and B & W to attempt to widen the premium-discount price gap, and that their failure to do so is evidence of collusion.

This initial RJR-led price increase was followed by eleven more parallel increases between May 4, 1995 and January 14, 2000. See Holiday Wholesale Grocery, 231 F.Supp.2d at 1264 (detailing these increases). The class contends that not only is this synchronous movement evidence of a price fixing agreement, but that many of these increases were not in the economic interest of all four alleged co-conspirators. For example, appellants argue that on November 23, 1998, following the industry's settlement of health care litigation with numerous state attorneys general, PM raised its list prices by $22.50 per thousand (45¢ per pack), which was more than was needed to offset the settlement costs. Although RJR, B & W and Lorillard matched the increase, the class says that had PM behaved rationally it would have increased prices less, as it was in a stronger financial position than its competitors, who would have been forced to forfeit market share in order to remain competitive.

Appellants contend that signaling was also accomplished through the use of "credit memos" that accompanied the manufacturers' price increases. The class says that these memos granted wholesalers credits for the difference between the old (pre-increase) prices and the new prices for several weeks after the announcement of a price increase that purportedly was immediately effective. This allegedly was an opportunity for the competitors of the manufacturer that implemented the increase to match the increase before the implementing manufacturer began to profit from it.

The manufacturers respond to the class's signaling allegations by pointing to several facets of the record that undermine appellees' allegations of conspiracy. Preliminarily, PM, RJR, B & W and Lorillard observe that although wholesale price competition largely abated during the alleged conspiracy period, this resulted in significant part from a shift in their competitive efforts to the retail arena. Indeed, during the alleged conspiracy period, the manufacturers spent on retail competition more than twice the amount the wholesalers claim to have been overcharged. For example, from 1994 to 1999, PM's spending on retail promotions and incentive programs more than doubled. In 1999 alone, PM — the company that appellants say led the price fixing agreement — spent nearly $2.9 billion (60% of its operating income) on retail promotions, couponing and its retailer incentive program. Notably, during 1998, retail competition within the tobacco industry was so intense that RJR, B & W and Lorillard brought an antitrust and unfair competition suit against PM based on PM's unilateral adoption of a program that it called "Retail Leaders." See R.J. Reynolds Tobacco Co. v. Philip Morris, Inc., 199 F. Supp. 2d 362, 365 (M.D.N.C.2002).

This shift to retail competition was efficacious, RJR says, because (1) wholesalers cannot, by themselves, increase demand for appellees' products, as demand necessarily is generated by consumers; (2) retail promotions can be applied more precisely than wholesale incentives, for example, a "buy one pack get a second pack free" promotion can be targeted to convenience stores that sell cigarettes by the pack as opposed to retail outlets that sell them by the carton; (3) similarly, retail promotions serve as targeted defensive responses to the competitive initiatives of other manufacturers, and this targeting is more easily accomplished at the retail level; (4) manufacturers cannot guarantee that wholesale list price reductions will be passed through both the wholesalers and retailers to consumers, and that the only sure-fire way to guaranty retail savings is through retail-targeted measures; and (5) unlike wholesale incentives, retail price promotions help manufacturers compete for retail advertising and display space, which is of paramount importance from a marketing perspective.

Moreover, it is indisputable that from 1993 to 1996, the manufacturers increased wholesale list prices only four times for a total of 16 cents per pack, a sum less than half the price decrease that resulted from PM's actions on Marlboro Friday. Accordingly, during the first half of the alleged conspiracy period, prices were uniformly lower and rose at a substantially slower pace than during the competitive time prior to the conspiracy's alleged onset. Indeed, appellees note, it was only five years later (during 1998), as a result of the health care settlements, that cigarette prices surpassed their pre-Marlboro Friday levels. In fact, adjusted for settlement costs, federal excise tax increases and inflation, wholesale list prices did not reach these levels until August, 1999.

The evidence also establishes that significant market share shifts occurred within the tobacco industry during the alleged conspiracy period. In fact, from 1993 to 2000, PM's share of the United States market rose 20%, Lorillard's share rose nearly 50%, RJR's share fell approximately 25% and B & W's share declined roughly 19%. These changes in appellees' respective market shares were at least as great as those that occurred between 1980 and 1991, an unquestionably competitive period. For example, RJR lost twice as much market share from 1993 to 2000 as it did from 1982-1991.

The class also labels as part of the conspiracy appellees' adoption of permanent allocation programs despite the fact that they each had excess manufacturing capacities.7 According to the class, restricting supply in the face of consumer demand for more product was directly contrary to the manufacturers' economic interests and thus indicative of illegal collusion.

Appellants also argue PM, RJR, B & W and Lorillard furthered their collusive enterprise by exchanging sales data through a common consultant, Management Science Associates ("MSA"), which allegedly enabled appellees to ensure that all were adhering to their allocation programs and to detect and punish what plaintiffs' expert Franklin M. Fisher termed "defections from an industry understanding on price." The MSA system tracks shipments from the manufacturers to wholesalers and from the wholesalers to retailers and provides reports to each appellee regarding the shipments of its competitors. Appellants allege that although in 1994 PM began collecting sales data on RJR, B & W and Lorillard through MSA,8 and thereby incurred a great competitive advantage, in 1995 it inexplicably began sharing this system with its competitors. Moreover, the class posits, over time the MSA system has been modified to make the cigarette market more transparent, and all of these alterations have been implemented with the unanimous consent of PM, RJR, B & W and Lorillard.

Appellants also designate other activities purportedly undertaken by appellees as non-competitive. First, they say that PM, RJR and B & W (along with other manufacturers) conspired during the 1950s to refrain from seeking any competitive advantage by developing more health-conscious tobacco products or by competing on the basis of health generally. The class argues that this understanding remained in effect during the conspiracy period, and says that this is further evidence that appellees agreed not to compete on pricing grounds.

Second, the class contends that PM, RJR and B & W9 explicitly fixed prices in other countries — including Argentina, Canada, Costa Rica, France, Hungary, Saudi Arabia and Venezuela, among others — as well, and that they did so as a "trial run" for their price fixing activities in the United States. They say that appellees' "experience recognizing their competitors' signals on price coordination in foreign markets would facilitate collusion in the U.S. market."

Third, the class designates as evidence of a price fixing conspiracy the very structure and history of the tobacco industry. It specifically points to the facts that this industry features a concentration of sellers, inelastic demand at competitive prices, high barriers to entry, a fungible product, principal firms selling at the same level in the chain of distribution, prices that can be changed quickly, cooperative practices and a record of antitrust violations. Fourth, the class alleges that appellees conducted little analysis of whether to implement each of the eleven individual price increases between 1993 and 2000.

Fifth, appellants say that although wide market share shifts indisputably occurred during the alleged conspiracy period, collusion among the manufacturers is nevertheless evidenced by the dampening during this period of the even wider market share swings that pervaded between 1991 and 1993. Sixth, the class claims that appellees enjoyed numerous opportunities to conspire, and says that this is further evidence that they actually did collusively fix prices. Finally, appellants assert that the manufacturers' consolidation of control over pricing decisions in the hands of a select few high ranking corporate officers represents additional evidence of collusion.

The class contends that the efficacy of the conspiracy is established by its results. Between Marlboro Friday (April 2, 1993) and February, 2000, cigarette prices rose more than $1.17 per pack, and appellants say that they were overcharged — that is, charged more than they would have been in the absence of the conspiracy — a total of $11.9 billion. They further note that the manufacturers' profit margins increased during the period of the alleged conspiracy, whereas they fell during the early 1990s, and that market shares stabilized during this time.

Appellees, by contrast, vigorously dispute each of these contentions. They do so most powerfully by reiterating that (1) retail competition throughout the alleged conspiracy period was indisputably intense; (2) cigarette prices actually were lower during most of the alleged conspiracy period than they had been prior to Marlboro Friday; and (3) appellees' market shares fluctuated far more dramatically during this time than could possibly be explained in the event of an industry-wide agreement to fix prices (indeed, the market share shifts were at least as great as those that occurred between the competitive years of 1980-91). In fact, the only time the market share fluctuations were greater was during the period between 1991 and 1993, when the discount brands made significant inroads into the market.

In February, 2000, faced with this pattern of what they perceived to be illegitimate, collusive price increases, several wholesalers filed individual lawsuits in various judicial districts around the country. The Judicial Panel on Multidistrict Litigation consolidated these actions and transferred them to the United States District Court for the Northern District of Georgia, where they were assigned to Judge Forrester. Appellants subsequently filed an amended consolidated class action complaint alleging that PM, RJR, B & W and Lorillard conspired to fix prices in violation of 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. Appellants also alleged that the manufacturers had fraudulently concealed their activities, thereby tolling the four year statute of limitations that generally applies to federal antitrust actions. See 15 U.S.C. § 15b.

On appellees' motion, the district court subsequently dismissed without prejudice the fraudulent concealment allegations for failure to satisfy pleading requirements, and struck from the complaint as "evidence pleading" appellants' claims regarding foreign markets. On January 23, 2001, the court certified a class of "[a]ll persons in the United States ... that purchased cigarettes directly from one or more of the Defendants, or any parent, subsidiary or affiliate thereof, at any time from February 8, 1996 to February 8, 2000." Notably, the district court rejected the wholesalers' fraudulent concealment argument, and the class was temporally limited by the four year antitrust statute of limitations, 15 U.S.C. § 15b; claims that pertained to the period prior to February 8, 1996 were dismissed as untimely. On February 22, 2001, the district court granted the class leave to file a second amended complaint, featuring claims under the same sections of the Sherman and Clayton Acts. The manufacturers again moved to dismiss the allegations of fraudulent concealment and any claims for relief that concerned actions taken prior to February 8, 1996, and the court granted both of these requests.

Subsequently, on February 8, 2002, PM, RJR, B & W and Lorillard moved for final summary judgment on all claims against them. On July 11, 2002, in what can be fairly described as a comprehensive and well-reasoned opinion, the district court granted this motion, and it is the correctness of this ruling that is at issue on appeal. In brief, appellants argue that the district court applied an overly stringent legal standard in evaluating appellees' summary judgment motion. Specifically, they claim that the court improperly required them to exclude all non-conspiratorial explanations for the manufacturers' behavior and impermissibly weighed the evidence, thereby usurping the function of the jury. Appellants further contend that the district court erred by considering each piece of evidence they presented in isolation, as opposed to evaluating their proofs in concert. The wholesalers then identify 11 "plus factors," see infra, and argue that each plus factor creates a presumption that PM, RJR, B & W and Lorillard conspired to fix cigarette prices. They say that this presumption enables them to survive appellees' summary judgment motion. The class also argues that the district court's entry of final summary judgment was improper because it was predicated on the erroneous exclusion of portions of Fisher's expert testimony.

The manufacturers respond by defending the district court's opinion on its terms. They argue that the court applied precisely the correct summary judgment standard, that it did not improperly weigh appellants' proffered evidence or fail to consider that evidence in concert, that none of the court's evidentiary rulings were erroneous and that none of the class's alleged "plus factors" actually tends to indicate a price fixing conspiracy any more than it does rational, lawful, competitive oligopolistic behavior. Distilled to its essence, the argument espoused by PM, RJR, B & W and Lorillard is that in an oligopoly, reason and economic rationality often dictate parallel pricing behavior, and that the evidence presented by the class is perfectly consistent with such "conscious parallelism." The manufacturers further assert that various facts on which appellants do not focus, or that the class mischaracterizes, establish conclusively that appellees' behavior from 1993 through 2000 is dramatically inconsistent with any collusive behavior. Accordingly, they say, even if appellants had established the existence of a plus factor, and thus created an inference of conspiracy, that inference has been fully rebutted. For all of these reasons, the manufacturers assert that the district court properly entered final summary judgment in their favor.

We review de novo a summary judgment ruling, applying the same legal standard used by the district court. See Johnson v. Bd. of Regents, 263 F.3d 1234, 1242-43 (11th Cir.2001). In conducting this examination, we view the materials presented and all factual inferences in the light most favorable to the non-moving party. See Adickes v. S.H. Kress & Co., 398 U.S. 144, 157, 90 S. Ct. 1598, 1608, 26 L. Ed. 2d 142 (1970). Summary judgment is appropriate where "there is no genuine issue as to any material fact" and "the moving party is entitled to a judgment as a matter of law." Fed.R.Civ.P. 56(c). The burden of demonstrating the satisfaction of this standard lies with the movant, who must present "pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any" that establish the absence of any genuine, material factual dispute. Id.

We review the district court's exclusion of expert evidence for an abuse of discretion. See Maiz v. Virani, 253 F.3d 641, 662 (11th Cir.2001) ("We review a trial court's evidentiary rulings on the admission of expert witness testimony only for abuse of discretion.") (citing Toole v. Baxter Healthcare Corp., 235 F.3d 1307, 1312 (11th Cir.2000)).

A. The Summary Judgment Standard in the Price Fixing Context

In order to fully explain the summary judgment standard that we apply in price fixing cases, it is important to distinguish at the outset between collusive price fixing, i.e., a "meeting of the minds" to collusively control prices,10 which is prohibited under the Sherman and Clayton Acts, and "conscious parallelism," which is not.

As the Supreme Court has recognized, oligopolies — of which the modern tobacco industry is by all measures a prime example11 — often feature coordinated pricing and related behaviors. See Brooke Group, Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 227, 113 S. Ct. 2578, 2590, 125 L. Ed. 2d 168 (1993); United States v. Von's Grocery Co., 384 U.S. 270, 300, 86 S. Ct. 1478, 1495, 16 L. Ed. 2d 555 (1966) (Stewart, J., dissenting) (discussing "the sort of consciously interdependent pricing that is characteristic of a market turning the corner toward oligopoly"). Indeed, we have explicitly recognized that "`the distinctive characteristic of oligopoly is recognized interdependence among the leading firms: the profit-maximizing choice of price and output for one depends on the choices made by others.'" Bailey, 284 F.3d at 1251 (quoting IIA Areeda & Hovenkamp, Antitrust Law ¶ 404a).

When they are the product of a rational, independent calculus by each member of the oligopoly, as opposed to collusion, these types of synchronous actions have become known as "conscious parallelism." The Court has defined this phenomenon as "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 Group, 509 U.S. at 227, 113 S. Ct. at 2590 (citations omitted). "In other words," we have further explained, "conscious parallelism is the practice of interdependent pricing in an oligopolistic market by competitor firms that realize that attempts to cut prices usually reduce revenue without increasing any firm's market share, but that simple price leadership in such a market can readily increase all competitors' revenues." City of Tuscaloosa v. Harcros Chems., Inc., 158 F.3d 548, 570 (11th Cir.1998) (citations omitted). Thus, these behaviors typically result from firms' rational recognition that the market structure in which they operate will most easily yield profits by means other than price competition.

As numerous courts have recognized, it often is difficult to determine which of these situations — illegal price fixing or conscious parallelism — is present in a given case. This is largely attributable to the efforts typically made by those who fall on the wrong side of this line to disguise the illegal nature of their endeavors. In this vein, we have recognized that:

"[I]t is only in rare cases that a plaintiff can establish the existence of a conspiracy by showing an explicit agreement; most conspiracies are inferred from the behavior of the alleged conspirators," and from other circumstantial evidence (economic and otherwise), such as barriers to entry and other market conditions.

Harcros, 158 F.3d at 569 (quoting Seagood Trading Corp. v. Jerrico, Inc., 924 F.2d 1555, 1573 (11th Cir.1991)).

In the unusual case where the plaintiff is able to muster direct evidence of price fixing, summary judgment is categorically inappropriate. See In re Citric Acid Litig., 191 F.3d 1090, 1093 (9th Cir.1999) ("[Plaintiff] argues ... that it has produced direct evidence that [Defendant] entered into illegal price-fixing agreements with the admitted conspirators[,] in which case summary judgment would, of course, be inappropriate...." (citing McLaughlin v. Liu, 849 F.2d 1205, 1208 (9th Cir.1988) and T.W. Elec. Serv., Inc. v. Pacific Elec. Contractors Ass'n, 809 F.2d 626, 631-32 (9th Cir.1987))); In re Baby Food Antitrust Litig., 166 F.3d 112, 117-18 (3d Cir.1999).

More frequently, price fixing plaintiffs are relegated to relying on indirect means of proof. The problem with this reliance on circumstantial evidence, however, is that such evidence is by its nature ambiguous, and necessarily requires the drawing of one or more inferences in order to substantiate claims of illegal conspiracy. Over time, courts have become attuned to the economic costs associated with using circumstantial evidence to distinguish between altogether lawful, independent, consciously parallel decision-making within an oligopoly on the one hand, and illegal, collusive price fixing on the other. See, e.g., Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 594, 106 S. Ct. 1348, 1360, 89 L. Ed. 2d 538 (1986) ("[M]istaken inferences in cases such as this one are especially costly, because they chill the very conduct the antitrust laws are designed to protect. `[W]e must be concerned lest a rule or precedent that authorizes a search for a particular type of undesirable pricing behavior end up by discouraging legitimate price competition.'" (citing Monsanto Co. v. Spray-Rite Serv. Corp., 465 U.S. 752, 763-64, 104 S. Ct. 1464, 1470, 79 L. Ed. 2d 775 (1984) and quoting Barry Wright Corp. v. ITT Grinnell Corp., 724 F.2d 227, 234 (1st Cir.1983))).

This recognition is reflected in the summary judgment formulation that is employed in price fixing cases. See PepsiCo, Inc. v. Coca-Cola Co., 315 F.3d 101, 104 (2d Cir.2002) ("In the context of antitrust cases ... summary judgment is particularly favored because of the concern that protracted litigation will chill pro-competitive market forces.") (citation omitted). In order to ensure that only potentially meritorious claims survive summary judgment, the Supreme Court has required that inferences of a price fixing conspiracy drawn from circumstantial evidence be reasonable. See Matsushita, 475 U.S. at 588, 106 S. Ct. at 1356 ("Respondents ... must show that the inference of conspiracy is reasonable in light of the competing inferences of independent action or collusive action...."). Indeed, "antitrust law limits the range of permissible inferences from ambiguous evidence in a § 1 case." Id. (emphasis added).

In practice, this means that "[t]o survive a motion for summary judgment ... a plaintiff seeking damages for [collusive price fixing] ... must present evidence `that tends to exclude the possibility' that the alleged conspirators acted independently." Matsushita, 475 U.S. at 588, 106 S. Ct. at 1356 (quoting Monsanto, 465 U.S. at 764, 104 S. Ct. at 1471). Evidence that does not support the existence of a price fixing conspiracy any more strongly than it supports conscious parallelism is insufficient to survive a defendant's summary judgment motion. See id. ("[C]onduct as consistent with permissible competition as with illegal conspiracy does not, standing alone, support an inference of antitrust conspiracy." (citing Monsanto, 465 U.S. at 764, 104 S. Ct. at 1470)) (other citation omitted); Harcros, 158 F.3d at 569 ("[P]laintiffs' circumstantial documentary evidence is in equipoise, and in the absence of further evidence of collusion, summary judgment against the plaintiffs would be in order."); id. ("`A fact that can only be decided by a coin toss has not been proven by a preponderance of the evidence, and cannot be submitted to the jury.'" (quoting James v. Otis Elevator Co., 854 F.2d 429, 432 n. 3 (11th Cir.1988))).

In applying these principles, we have fashioned a test under which price fixing plaintiffs must demonstrate the existence of "plus factors" that remove their evidence from the realm of equipoise and render that evidence more probative of conspiracy than of conscious parallelism. See Harcros, 158 F.3d at 570-71 ("`[I]t is well settled in this circuit that evidence of conscious parallelism [alone] does not permit an inference of conspiracy unless the plaintiff [either] establishes that . . . each defendant engaging in the parallel action acted contrary to its economic self-interest' or offers other `plus factors' tending to establish that the defendants were not engaging merely in oligopolistic price maintenance or price leadership but rather in a collusive agreement to fix prices or otherwise restrain trade.") (quoting Todorov v. DCH Healthcare Auth., 921 F.2d 1438, 1456 n. 30 (11th Cir.1991)) (other citations omitted).

Although our caselaw has identified some specific plus factors, for example, "a showing that the defendants' behavior would not be reasonable or explicable (i.e. not in their legitimate economic self-interest) if they were not conspiring to fix prices or otherwise restrain trade," Harcros, 158 F.3d at 572, any showing by appellants that "tend[s] to exclude the possibility of independent action" can qualify as a "plus factor." Id. at 571 n. 35. In cases where the plaintiff establishes the existence of one or more "plus factors," these factors "only create a rebuttable presumption of a conspiracy which the defendant may defeat with his own evidence...." Todorov, 921 F.2d at 1456 n. 30.

In short, there are three steps to the summary judgment analysis in the price fixing context. First, the court must determine whether the plaintiff has established a pattern of parallel behavior. Second, it must decide whether the plaintiff has demonstrated the existence of one or more plus factors that "tends to exclude the possibility that the alleged conspirators acted independently." Matsushita, 475 U.S. at 588, 106 S. Ct. at 1356 (internal punctuation and citation omitted). The existence of such a plus factor generates an inference of illegal price fixing. Third, if the first two steps are satisfied, the defendants may rebut the inference of collusion by presenting evidence establishing that no reasonable factfinder could conclude that they entered into a price fixing conspiracy.

In undertaking this analysis, the district court is obligated to give the price fixing plaintiff(s) "`the full benefit of their proof without tightly compartmentalizing the various factual components and wiping the slate clean of scrutiny of each.'" Holiday Wholesale Grocery, 231 F.Supp.2d at 1268 (quoting Continental Ore Co. v. Union Carbide & Carbon Corp., 370 U.S. 690, 699, 82 S. Ct. 1404, 1410, 8 L. Ed. 2d 777 (1962)). However, as indicated by this discussion, it unquestionably is the duty of the district court to evaluate the evidence proffered by the plaintiffs not to ascertain its credibility, but instead to determine whether that evidence, if credited, "tends to" establish a conspiracy more than it indicates conscious parallelism. As the district court put this matter, "the import of the piece of circumstantial evidence [that appellants argue constitutes a plus factor] must actually be of the import that is ascribed to it by [appellants]. If, when considered in its entirety, it is totally ambiguous or to the opposite effect, it ... may not be relied upon by the jury." Id. at 1270. The district court then properly observed that a statement could not constitute a plus factor — i.e., could not be said by a reasonable factfinder to tend to exclude the possibility independent behavior or to establish a price fixing conspiracy — (1) if it required the jury to "engage in speculation and conjecture to such a degree as to render its finding `a guess or mere possibility,'" id. at 1271 (quoting Daniels v. Twin Oaks Nursing Home, 692 F.2d 1321, 1326 (11th Cir.1982)); or (2) if to infer conspiracy from the evidence the jury necessarily would engage in "fallacious reasoning." Id. Moreover, "if [appellants'] theory is economically senseless, no reasonable jury could find in its favor, and summary judgment should be granted." Eastman Kodak Co. v. Image Tech. Servs., Inc., 504 U.S. 451, 468-69, 112 S. Ct. 2072, 2083, 119 L. Ed. 2d 265 (1992).

In this case, the district court set forth this standard correctly and in detail, and none of appellants' criticisms of the court's articulation of the summary judgment standard are persuasive. The class says that the district court required it to exclude all "innocent" explanations for appellees' conduct in order to survive summary judgment. However, this plainly is incorrect. The court did not mandate that appellants exclude the possibility of conscious parallelism in order to survive summary judgment, but instead it required them to present some evidence that "tends to" exclude lawful, synchronous behavior. This is precisely in line with the method of proof established by the Supreme Court in Matsushita and refined by this court in Harcros. Nor, in a similar vein, did the district court insist that the existence of a conspiracy be the sole inference that a reasonable juror could draw from appellants' evidence. Again, the court required only that the class's evidence tend to exclude the possibility of conscious parallelism or, stated differently, to establish a price fixing conspiracy, nothing more, nothing less.

The class also is incorrect insofar as it suggests that the Supreme Court's decision in Eastman Kodak casts doubt on the correctness of the summary judgment standard applied by the district court. The Court said in Eastman Kodak that "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." 504 U.S. at 468, 112 S.Ct. at 2083 (citing Matsushita, 475 U.S. at 587-88, 106 S. Ct. at 1356). As the district court explicitly recognized, see Holiday Wholesale Grocery, 231 F.Supp.2d at 1270 n. 9, the Court further held in Eastman Kodak that its "requirement in Matsushita that the plaintiffs' claims make economic sense did not introduce a special burden on plaintiffs facing summary judgment in antitrust cases." 504 U.S. at 468, 112 S. Ct. at 2083.

The class argues that the district court's imposition of a "tends to exclude the possibility that the alleged conspirators acted independently" standard contradicted Eastman Kodak because it imposed a "special burden" on antitrust plaintiffs. However, this argument is unpersuasive for at least two distinct reasons. First, the "tends to exclude ..." standard itself was announced by the Supreme Court in Matsushita, and as such Eastman Kodak's reaffirmation of Matsushita reaffirms the correctness of this standard as well. See Eastman Kodak, 504 U.S. at 468, 112 S. Ct. at 2083 ("Matsushita demands only that the nonmoving party's inferences be reasonable in order to reach the jury...."). Moreover, as recently as 1998, in binding precedent, we applied this standard unambiguously in Harcros, observing that "[t]o survive a motion for summary judgment or for a directed verdict, a plaintiff seeking damages for a violation of § 1 must present evidence `that tends to exclude the possibility' that the alleged conspirators acted independently. [Such plaintiffs], in other words, must show that the inference of conspiracy is reasonable in light of the competing inferences of independent action or collusive action that could not have harmed respondents." 158 F.3d at 570 (quoting Matsushita, 475 U.S. at 575, 106 S. Ct. at 1350).

Second, it also is firmly established that a price fixing plaintiff's evidence must create a "reasonable" inference of conspiracy in order to withstand a summary judgment motion. See, e.g., PepsiCo, Inc., 315 F.3d at 105 ("Although all reasonable inferences will be drawn in favor of the non-movant, those inferences `must be reasonable in light of competing inferences of acceptable conduct.'" (quoting Tops Mkts., Inc. v. Quality Mkts., Inc., 142 F.3d 90, 95 (2d Cir.1998))). The "tends to exclude..." standard articulated in Matsushita and applied by the district court in this case simply represents an explication of this requirement; it does not represent a new hurdle. In other words, evidence creates the requisite reasonable inference of conspiracy if it "tends to exclude the possibility that the alleged conspirators acted independently." Thus, the summary judgment standard applied by the district court does not in any way run afoul of Eastman Kodak. We also note that Eastman Kodak, Matsushita and Harcros were accurately cited at length by the district court in this case in discussing the summary judgment standard to be applied in price fixing cases.

Similarly unpersuasive is appellants' argument that although they must demonstrate the existence of one or more plus factors, the evidence constituting such a factor must merely "tend to establish" a conspiracy, as opposed to "tend to exclude" independent action. In support of this proposition the class cites Harcros, where we said:

[I]t is well settled in this circuit that evidence of conscious parallelism [alone] does not permit an inference of conspiracy unless the plaintiff [either] establishes that, assuming there is no conspiracy, each defendant engaging in the parallel action acted contrary to its economic self-interest, or offers other "plus factors" tending to establish that the defendants were not engaging merely in oligopolistic price maintenance or price leadership but rather in a collusive agreement to fix prices or otherwise restrain trade.

158 F.3d at 570-71 (citations and internal punctuation omitted) (emphasis added). However, the distinction that appellants draw between this language and the "tends to exclude ..." standard announced in Matsushita is semantic only; it makes no substantive difference. Evidence that tends to exclude the possibility of independent action necessarily tends to establish collusion. The class's formulation is simply the linguistic flip side of the test announced in Matsushita. Indeed, in Harcros we used these formulations interchangeably. Compare 158 F.3d at 570 ("[A] plaintiff seeking damages for a violation of § 1 must present evidence that tends to exclude the possibility that the alleged conspirators acted independently") (internal punctuation and citations omitted) (emphasis added) with id. at 570-71 (discussing "`plus factors' tending to establish that the defendants were ... in a collusive agreement to fix prices or otherwise restrain trade") (emphasis added).

Appellants further argue that because of its "fundamental error" in conflating the "tends to exclude ..." and "tends to establish..." standards in the context of the plus factor analysis, the district court examined each piece of evidence separately to determine whether summary judgment was appropriate instead of considering the wholesalers' proofs in concert. As we observed above, however, this assertion is unfounded. See Holiday Wholesale Grocery, 231 F.Supp.2d at 1268 ("The court is... mindful that it must consider the evidence presented by Plaintiffs as a whole and that `plaintiffs should be given the full benefit of their proof without tightly compartmentalizing the various factual components and wiping the slate clean after scrutiny of each.'" (quoting Continental Ore Co., 370 U.S. at 699, 82 S. Ct. at 1410)); id. at 1270 ("[T]he import of the piece of circumstantial evidence must actually be of the import that is ascribed to it by Plaintiffs. If, when considered in its entirety, it is totally ambiguous or to the opposite effect, it is not relevant and may not be relied upon by the jury.") (emphasis added). The class is correct that the district court at one point indicated that it was analyzing "whether a piece of evidence can be admitted as relevant because it creates a reasonable inference." Id. However, in the face of the foregoing statements, the stray comment to which the class refers is vastly insufficient to establish that the court did not actually examine the class's evidence as a whole.

This is especially so given that the district court ultimately expressed its conclusion as being "that the evidence adduced by Plaintiffs does not singularly or in combination exclude the possibility that Defendants acted independently." Id. at 1328 (emphasis added). Instead, the court's statement indicates simply that the district court properly analyzed appellants' alleged plus factors both individually and in concert. Under Matsushita and Harcros, this certainly did not constitute legal error. Moreover, we have independently reviewed appellants' alleged plus factors, both individually and in combination, and we agree that none of these factors actually tends to exclude the possibility of independent behavior or tends to establish a price fixing agreement.

The class further argues that the district court improperly substituted its assessment of the evidence for that of the jury, i.e., that it acted as factfinder at the summary judgment stage. We think the court did no such thing. Instead, it merely ascertained, as it was required to do under Matsushita and Harcros, whether appellants' evidence tended to exclude the possibility of independent action. It made no finding as to whether such independent action did or did not exist. Although it did make determinations as to the reasonableness of the inferences that could be drawn from the evidence, these were threshold legal determinations that appropriately were made by the district court.

Finally, appellants argue that the district court failed to view the evidence in the light most favorable to them and that it also failed to distinguish between the existence of a conspiracy and its efficacy. However, they neither explain nor support these allegations, and as such we are unable to evaluate them.

B. The District Court's Application of the Antitrust Summary Judgment Standard in This Case

As a preliminary matter, no party contests the satisfaction of the first prong of our inquiry, viz., the existence of parallel pricing behavior. As evidenced by the repeated, synchronous pricing decisions that occurred within the tobacco industry between 1993 and 2000, appellees plainly priced their products in parallel. Similarly uncontroversial is appellants' lack of direct evidence of a price fixing conspiracy among PM, RJR, B & W and Lorillard. Although the class argued in the district court that it had presented such evidence, it has abandoned this contention on appeal. Accordingly, the key questions on appeal are (1) whether appellants have shown the existence of a plus factor so as to create an inference of conspiracy; and (2) if so, whether appellees are able to rebut that inference.

After articulating the summary judgment standard in an antitrust case, the district court delineated eleven distinct factors that appellants had denominated "plus factors." These are: "(1) signaling of intentions; (2) permanent allocations programs; (3) monitoring of sales; (4) actions taken contrary to economic self-interest, including (a) little analysis of whether to follow price increases, (b) B & W and RJR pulling away from the discount cigarette market, (c) the May 1995 price increase lead by RJR and followed by Philip Morris, (d) Philip Morris' agreement to base the initial [Management Science Associates]... payments on market capitalization rather than market share, and (e) `excessive' price increases after the MSA; (5) nature of the market; (6) strong motivation; (7) reduction in the number of price tiers; (8) opportunities to conspire; (9) pricing decisions made at high levels; (10) the smoking and health conspiracy; and (11) foreign conspiracies." Holiday Wholesale Grocery, 231 F.Supp.2d at 1274. Although the district court found that only the first four of these factors "merit[ed] serious consideration," id., it analyzed each of them in detail. The district court ultimately concluded that none of them actually tended to exclude the possibility of independent behavior, and the class contests the correctness of the court's conclusions as to each factor. Accordingly, we consider each of these alleged plus factors.

a. Signaling

First, appellants identify several "signals" that allegedly were transmitted among PM, RJR, B & W and Lorillard as to their pricing intentions. The class claims that these signals were the means by which appellees' price fixing conspiracy was carried out, as the manufacturers formulated and cemented their plans to collusively fix cigarette prices by indirectly communicating with each other through media outlets and other public announcements.

However, the district court rejected these communications as a plus factor, saying that it found:

[The class's] theory of signaling among the companies to be based on a combination of statements taken out of context, as well as ominous readings of typical industry reporting on strategy. To reach the inferences suggested would require the jury to engage in speculation and does not tend to exclude the possibility that [appellees] acted independently.... [Appellants] have done nothing more than show that in an oligopoly, each company is aware of the others' actions. This is the nature of the economic interdependence of the companies in an oligopoly.

Holiday Wholesale Grocery, 231 F.Supp.2d at 1275. The district court further noted that "in competitive markets, particularly oligopolies, companies will monitor each other's communications with the market in order to make their own strategic decisions[. As such], antitrust law permits such discussions even when they relate to pricing because the `dissemination of price information is not itself a per se violation of the Sherman Act.'" Id. at 1276 (quoting United States v. Citizens and S. Nat'l Bank, 422 U.S. 86, 113, 95 S. Ct. 2099, 2115, 45 L. Ed. 2d 41 (1975)). In this vein, lead counsel for the class conceded at oral argument before the district court on the manufacturers' summary judgment motions that the ultimate outcome of all of the statements and actions that appellants label signals was to return the tobacco industry from the pricing chaos that followed Marlboro Friday to a "traditional normal oligopoly," which, of course, is perfectly legal. Id. at 1316.

Nonetheless, the class posits that the signaling among appellees began when PM publicly announced on Marlboro Friday that its prospective strategy with respect to Marlboro was to "grow market share and longer term profits," and that it would "forgo any further price increases on premium brands for the foreseeable future." The translation of this announcement, the class says, is that PM would not engage in any more price cuts provided that its competitors did not attempt to alter the price gap between premium and discount or deep discount cigarettes.

Thus, appellants do not argue that the actions taken by PM on Marlboro Friday were in any way anti-competitive; in fact, they explicitly concede (as surely they must) that this was an exceptionally competitive move. See generally NCAA v. Bd. of Regents of Univ. of Okla., 468 U.S. 85, 104-106 & n. 27, 104 S. Ct. 2948, 2961-62 & n. 27, 82 L. Ed. 2d 70 (1984) (discussing generally the Sherman Act's goals of reducing prices and promoting competition, and observing specifically that "`[the Act] rests on the premise that the unrestrained interaction of competitive forces will yield the best allocation of our economic resources, the lowest prices, the highest quality and the greatest material progress'" (quoting Northern Pacific R. Co. v. United States, 356 U.S. 1, 4-5, 78 S. Ct. 514, 517-518, 2 L. Ed. 2d 545 (1958))); Monahan's Marine, Inc. v. Boston Whaler, Inc., 866 F.2d 525, 527 (1st Cir.1989) ("The Sherman Act's very purpose is to help consumers, in part by bringing about low, nonpredatory prices."). Indeed, as the Supreme Court explained in Brooke Group:

Even in an oligopolistic market, when a firm drops its prices to a competitive level to demonstrate to a maverick the unprofitability of straying from the group, it would be illogical to condemn the price cut: The antitrust laws then would be an obstacle to the chain of events most conducive to a breakdown of oligopoly pricing and the onset of competition. Even if the ultimate effect of the cut is to induce or reestablish supra-competitive pricing, discouraging a price cut and forcing firms to maintain supra-competitive prices, thus depriving consumers of the benefits of lower prices in the interim, does not constitute sound antitrust policy.

509 U.S. at 223-24, 113 S. Ct. at 2588.

Similarly, we believe that PM's announcement of its plan not to increase prices on premium brands (and thus relinquish any basis for gaining market share that it enjoyed as a result of Marlboro Friday) does not tend to exclude the possibility of independent behavior or to establish a price fixing conspiracy. Indeed, this announcement was a legitimate communication within the market on a day when an explanation for PM's unprecedented — and unquestionably competitive — behavior plainly was necessary. Contrary to the class's allegation that "[i]t seems unusual for a manufacturer of a consumer good to conduct a press conference to announce that it has decided to engage in retail competition," PM's explanation of the dramatic events of Marlboro Friday is by no means surprising. Thus, as the district court said, "the allegation of Marlboro Friday as `invitation to collude' is not supported by the materials cited by [appellants]." Holiday Wholesale Grocery, 231 F.Supp.2d at 1278. Moreover, we note that the Supreme Court's caveat in Brooke Group against discouraging competitive pricing behavior applies with equal force to PM's July, 1993 price reductions as it does to those instituted on Marlboro Friday.

To reiterate, PM reinforced its Marlboro Friday actions by announcing on July 20, 1993 that it would similarly reduce the prices of all of its premium brands and that it was consolidating its discount and deep-discount price tiers. In a commensurately competitive response to PM's actions, RJR, B & W and Lorillard — which already had reduced their premium prices in response to Marlboro Friday — followed suit and consolidated their discount and deep-discount prices. This had the effect of finally narrowing the premium/discount price gap to a level PM found acceptable. Subsequently, on November 8, 1993 RJR led a 4¢ per pack increase for both of these brand categories, which was promptly followed by PM, B & W and Lorillard. The class argues that RJR and B & W — which had been the primary beneficiaries of the broader, pre-Marlboro Friday gap — should have attempted to re-widen the premium-discount price gap. By instead acquiescing in the prices imposed on the market by PM, appellants say, RJR and B & W signaled their acceptance of the condition for a price increase established by PM, i.e., that any such increase could not result in a widening of the gap. Notably, however, appellants do not specify precisely how RJR and B & W should have tried to re-widen the price gap; that is, they do not identify the particular combination of premium price increases and/or discount price reductions that would have created a gap more favorable from the perspective of RJR and B & W.12

Moreover, RJR and B & W's decisions to accept the narrower premium-discount price gap preferred by PM do not tend to exclude the possibility of independent behavior or to establish a price fixing conspiracy. It is plain to us that RJR and B & W's actions are readily explained as economically rational, self-interested responses to Marlboro Friday. PM's premium price cuts strongly suggested to its competitors that any attempt to gain market share for discount brands through further price cuts would not be accepted. For example, had RJR or B & W tried to cut its discount prices while raising premium prices, as the class seems to argue they should have, PM would likely (and readily) have matched the lower discount prices while simultaneously lowering its premium prices, thereby underselling its competitors in the premium market and restoring the premium-discount price gap to the level it desired. Thus, any such wholesale price competition by RJR or B & W would only have decreased industry revenues across the board without increasing their market shares. The only viable way for RJR and B & W to increase their revenues was to raise prices in a manner that would not provoke a competitive response from PM, which is precisely the action that the class labels a signal of conspiracy. Put simply, this action is no more indicative of collusion than it is of lawful, rational pricing behavior.

The class denotes as still another signal the statement by Martin Broughton that "[BAT] may be one of those who started the price war in the U.S., but we have no wish to escalate it." The central problem with this argument, however, is that appellants have not fully represented what Broughton said. As the district court noted, the full statement reads: "BAT may be one of those who started the price war in the U.S., but we have no wish to escalate it. But we shall be ready to respond tactically where necessary." Holiday Wholesale Grocery, 231 F.Supp.2d at 1278 (emphasis added). As the court continued: "[T]he rest of the interview also notes that Mr. Broughton was `perplexed' by [PM's] tactics and that BAT `will wait and see what happens in the market before playing its own cards.'" Id. The district court concluded that when Broughton's entire statement is considered, "[n]o reasonable jury could infer that B & W agreed to any particular future course of conduct...." Id. We agree; Broughton's statement is at best ambiguous and, more realistically, cuts against a finding of collusion.13

As explained supra, it is clear that further price reductions of either the premium or discount brands would have reduced B & W's revenues without increasing its market share. Broughton's statement simply recognizes this economic fact, nothing more. Indeed, it would have been utterly irrational for B & W to have reacted to the events of Marlboro Friday in any way other than to have matched PM's drastic price reduction (so as to retain its market share) and then to attempt to raise prices again (albeit slowly and over time) so as to regain some of its lost revenue.

The class also says that B & W and RJR both "clarif[ied] their agreement not to compete on price" by publicly announcing their future pricing intentions at meetings with stock analysts. It posits that B & W was the first to signal, when on September 30, 1993, its CEO, Thomas Sandefur, told stock analysts that "our company fully intends to pursue options other than price for our V-F-M [value-for-money, an industry term for discount] brands." Notably, however, the class has not cited to the full text of this statement. Sandefur said, in full, that:

We believe that price cutting or discounting — while still a factor — will be less important than in the past. That's because gaps in the list price have narrowed between premium and V-F-M... brands. Premium brands, therefore, should regain some strength. Over time, smaller price increases coupled with state and federal tax increases, will bring premium prices up once again. This will make V-F-M brands more appealing. But in the near term, value-for-money brands will need to compete on some basis other than price. And our company fully intends to pursue options other than price for our V-F-M brands.

Holiday Wholesale Grocery, 231 F.Supp.2d at 1279 (emphasis in original).

Sandefur's complete statement simply does not tend to exclude the possibility that B & W's pricing decisions were reached independently (or to establish a price fixing conspiracy). Indeed, when considered in its entirety, Sandefur's statement clearly "explains that with Philip Morris narrowing the price gap and forcing a return to strength for premium brands, `in the near term,' discount brands would have to come up with another way to compete. This statement is anything but a commitment to abandon price competition, as [appellants] allege. It is, in fact, a declaration of an intent to do so if and when the prices of premium brands are increased." Holiday Wholesale Grocery, 231 F.Supp.2d at 1279.

The class next alleges that RJR similarly signaled its intent to refrain from price competition in a November 2, 1993 statement by Charles Harper, the CEO of RJR/Nabisco, RJR's parent company. However, Harper said no more than that RJR was going to focus on earnings growth and was "willing to accept modest market share losses as the cost of improving earnings." We cannot reasonably infer collusive price fixing from this brief comment. This general statement, like so many of those on which appellants focus, plainly is a rational response to the economic incentives created by PM on Marlboro Friday, and does not tend to exclude the possibility that RJR acted independently or to establish a price fixing conspiracy.

The class claims additionally that PM signaled the acceptability of its competitors' overtures at eliminating price competition by announcing its permanent allocation program, which typically had been associated with an impending price increase. Appellants say that this action actually was an acknowledgment by PM of its competitors' announcements that they had abandoned price competition. Like the other actions on which the class focuses, however, PM's adoption of a permanent allocation program does not tend to exclude the possibility that PM acted independently or to establish a price fixing conspiracy. This is so for at least two independent reasons: "First, [PM] had no idea that its competitors would react to a price increase and, to that end, had a competitive contingency plan." Holiday Wholesale Grocery, 231 F.Supp.2d at 1280. Indeed, as the district court noted, "[t]here is nothing to suggest that [PM was] responding to RJR's willingness to move to a profitability mode except the temporal proximity of the two announcements." Id. Second, as we will discuss in greater detail, infra, the allocation programs were viewed by appellees as necessary to combat trade loading by wholesalers, which resulted in excess product returns, disruptions in shipping patterns and stale product. When these factors are considered, the district court was correct to say that the mere fact that PM instituted a permanent allocation program cannot give rise to a reasonable inference of conspiracy.

Finally, the class posits that appellees' signaling reached a peak on November 8, 1993, when RJR increased wholesale prices in equal amounts for both the premium and discount price tiers. It contends that it was in RJR's economic interest to attempt to widen the price gap, yet it did not do so. However, RJR's actions do not tend to exclude the possibility of independent behavior or to establish a price fixing conspiracy. Indeed, appellants' allegation has stood economic reality on its head. As explained supra, it would have been in the interests of RJR and B & W to widen the gap between the premium and discount price points only if doing so would have generated market share or profitability. Yet because they knew as a result of Marlboro Friday that PM would not readily accept such a strategy, and that any effort to re-widen the price gap ultimately would have further depressed industry revenues without increasing their market shares, economic self-interest dictated that they raise premium and discount prices as much as they could. Again, we cannot say that these pricing decisions tend to exclude the possibility of independent action on the part of appellees or to establish that the manufacturers colluded to fix prices.

None of the actions taken by PM, RJR or B & W that appellants label "signals" tend to exclude the possibility that the primary players in the tobacco industry were engaged in rational, lawful, parallel pricing behavior that is typical of an oligopoly. Nor, conversely, do they tend to establish that a price fixing conspiracy was afoot. Notably, this conclusion does not change when we consider these actions cumulatively. In other words, in this case the whole of the manufacturers' actions is no greater than the sum of its parts. Because none of appellees' largely ambiguous statements and actions come close to meeting the mark, it is unhelpful to the class to consider those actions in concert.

b. Actions Against the Manufacturers' Economic Interests

It is firmly established that actions that are contrary to an actor's economic interest constitute a plus factor that is sufficient to satisfy a price fixing plaintiff's burden in opposing a summary judgment motion. See Harcros, 158 F.3d at 572 ("One prominent `plus factor,' to which antitrust plaintiffs often take recourse, is a showing that the defendants' behavior would not be reasonable or explicable (i.e. not in their legitimate economic self-interest) if they were not conspiring to fix prices or otherwise restrain trade — that is, that the defendants would not have acted as they did had they not been conspiring in restraint of trade." (citing Theatre Enters., Inc. v. Paramount Film Distrib. Corp., 346 U.S. 537, 540-42, 74 S. Ct. 257, 259-60, 98 L. Ed. 273 (1954))).

However, we must exercise prudence in labeling a given action as being contrary to the actor's economic interests, lest we be too quick to second-guess well-intentioned business judgments of all kinds. See In re Citric Acid Litig., 191 F.3d at 1101 ("Courts have recognized that firms must have broad discretion to make decisions based on their judgments of what is best for them and that business judgments should not be second-guessed even where the evidence concerning the rationality of the challenged activities might be subject to reasonable dispute."). Accordingly, appellants must show more than that a particular action did not ultimately work to a manufacturer's financial advantage. Instead, in the terms employed by Matsushita, the action must "tend[ ] to exclude the possibility of independent action." Thus, if a benign explanation for the action is equally or more plausible than a collusive explanation, the action cannot constitute a plus factor. Equipoise is not enough to take the case to the jury.

The district court identified 7 actions taken by PM, RJR, B & W and/or Lorillard that the class said were contrary to the actor's economic interest:

(1) non-initiating Defendants always followed price increases; (2) after Marlboro Friday, B & W and RJR turned away from discount cigarettes; (3) RJR led, and Philip Morris followed, a price increase despite planning documents which reflected they would not take increases; (4) each Defendant exchanged information through [MSA]; (5) each Defendant had permanent allocation programs; (6) Philip Morris based settlement payments on market capitalization and not market share; and (7) Defendants agreed to pay "excessive" settlement price increases.

Holiday Wholesale Grocery, 231 F.Supp.2d at 1296. With the exception of the sixth,14 appellants contest the district court's rejection of each action as a plus factor. Simply put, we find each of the class's arguments to be unpersuasive. None of the actions on which appellants focus tends to exclude the possibility of independent behavior (or tends to establish a price fixing conspiracy), and accordingly they do not constitute plus factors.

Preliminarily, we already have explained why the district court was correct to reject as a plus factor the decisions by RJR, B & W and, in some cases, PM to follow a price increase initiated by one of their competitors. Simply put, "[t]estimony and contemporaneous documents show that RJR, B & W, and Lorillard found their options extremely limited after [PM's] Marlboro Friday announcement." Holiday Wholesale Grocery, 231 F.Supp.2d at 1297. In the case of RJR and B & W, they could have (indeed, the class suggests they should have) continued to focus on widening the premium-discount price gap. However, given PM's unambiguous message on Marlboro Friday that it would act aggressively to attempt to maintain its desired price differential, this likely would have resulted in little if any market share gain. It simultaneously would have minimized profits, given that lower prices generate smaller revenues.

Moreover, lowering wholesale prices likely would have been an ineffective means of increasing market share, as price cuts are unlikely to be passed on by the wholesaler to the retailer or by the retailer to the consumer. Thus, appellees' only viable route back to profitability was to increase prices; that they did so in a parallel manner does not establish collusion. Because the decisions to follow price increases were in the economic interests of RJR and B & W, these actions are not plus factors.

Nor can the lack of analysis that preceded the decisions to match each such individual increase be construed as evidence of collusion. Once appellees perceived that raising prices was their wisest economic strategy, no extensive discussion or planning was needed to implement this strategy each time a competitor initiated a price increase. As the district court put it in discussing RJR: "Once RJR determined what its strategic response would be to Marlboro Friday, there was no need to re-complete that evaluation each time RJR needed to decide whether to respond to a price increase." Holiday Wholesale Grocery, 231 F.Supp.2d at 1298.

As for appellants' claims that RJR and B & W acted against their interests by deciding to turn away from the discount market that had been so profitable for them prior to Marlboro Friday, there are two simple answers. First, it is evident that RJR and B & W did not stop focusing on its discount brands, but simply changed tactics with respect to them. Specifically, instead of continuing with wholesale list price competition, RJR and B & W began competing at the retail level, where they could be certain that savings would accrue to the consumer, that is, would not be swallowed by the wholesalers or retailers. Second, to the extent that RJR or B & W did shift any focus away from their discount brands, doing so plainly was in their interests.

Although the class cites a statement by Gary Black for the proposition that "`analysts expected' and it was `well recognized' that RJR and B & W should continue to pursue the discount market," Holiday Wholesale Grocery, 231 F.Supp.2d at 1300, these "expectations" were not empirically supported. Marlboro Friday rendered the discount and deep discount markets less profitable. Both RJR and B & W learned that if they "continued to pursue discount prices downward, Philip Morris could respond in the same manner it did on Marlboro Friday." Id. at 1301. What's more, the profit margins on the premium brands made by RJR (Camel and Winston) and B & W (Barclay, Capri, Lucky Strike, etc.) were larger than those associated with these manufacturers' discount brands, see id. at 1301 n. 33 (discussing "the lower profit margins" associated with discount brands), and the post-Marlboro Friday narrowing of the premium/discount price gap rendered the discount brands less attractive to consumers. These factors cemented the wisdom of RJR and B & W's decisions to shift the focus away from gaining market share for their discount brands and toward increasing profitability by concentrating on, and increasing prices of, the premiums.

Similarly unpersuasive are the class's arguments concerning subsequent decisions by PM and RJR to initiate price increases despite internal analyses that such increases were not advisable. As the district court recounted in the context of RJR's decision to lead a May, 1995 increase, the internal doubts concerning the wisdom of these raises were undergirded by an uncertainty as to whether they would be followed, not whether they would be economically beneficial if followed. See Holiday Wholesale Grocery, 231 F.Supp.2d at 1303 ("[W]hile it is true that internal RJR staff recommended against raising prices, that recommendation was made because it was `doubtful' whether the other companies would follow RJR's price increase, a concern that belies [appellants'] allegation of price fixing. Nonetheless, RJR's CEO Jim Johnston decided in the spring of 1995 to increase prices in order to determine whether its competitors would do the same.").

Indeed, in 1995 the industry was still feeling the loss in revenues that resulted from the sharp price reductions spawned by Marlboro Friday, and as Professor Fisher explicitly recognized, "RJR was under considerable pressure in 1995 to improve its profits," as was PM. Id. Simply put, the increases that were in the interests of each manufacturer had to be initiated by someone, and it made good economic sense for the industry's two largest players, PM and RJR, to be the ones to have done so. Thus, the decisions to lead these increases are significantly more consistent with the economic interests of PM and RJR than they are with the conspiracy theory advanced by the class.15

Appellants next raise two arguments related to the exchange by each appellee of wholesale-to-retail sales information through Management Science Associates. Around 1993-94, PM developed with MSA a system for tracking wholesale to retail shipments of its products, which it believed would afford it a great competitive advantage. Subsequently, in 1995, PM permitted MSA to share this service with its competitors. In exchange for their acceptance of the service, RJR, B & W and Lorillard agreed to share their own sales information.

Appellants first argue that PM's decision to share this service cannot be seen as being in its economic interest, and instead must be viewed as a means of facilitating the monitoring of the conspiracy by all of the conspirators. PM responds that by sharing the MSA service it shifted the financial burden of gathering sales and market share data for its competitors to RJR, B & W and Lorillard, and that as a result it actually realized an annual savings of millions of dollars. Viewed in the light most favorable to appellants, both explanations are plausible, and thus, at the most, this action by PM stands in equipoise; that is, it is equally consistent with collusion as with lawful competition, and accordingly, under Matsushita and Harcros, it cannot represent a plus factor. Second, appellants argue that the participation by all appellees in the MSA data sharing system was contrary to their respective interests.

Preliminarily (and quite significantly), we note that the evidence establishes that appellees exchanged only sales, not pricing, information through MSA. Simply put, it is far less indicative of a price fixing conspiracy to exchange information relating to sales as opposed to prices. Moreover, it plainly was economically beneficial for each individual appellee to keep tabs on the commercial activities of its competitors, so the receipt of information concerning their sales does not tend to exclude the possibility of independent action or to establish anti-competitive collusion. Indeed, as RJR argues, "[e]ven if one assumes that, all else being equal, RJR would prefer that data about its products not be available to its rivals . . ., that does not tell us that RJR acts irrationally if it concludes that the competitive benefit of obtaining its rivals' data outweighs whatever preference it has against sharing its own data."

Thus, although the sharing of information can be seen as suggesting conspiracy, as appellants allege, it also can be seen equally as a necessary means to the receipt of its competitors' information. If a particular manufacturer ceased providing its own information, its entitlement to that of its competitors would similarly end. To draw an analogy, each company's willingness to give its own information can be viewed as the ante in a poker game. To ante is irrational only if there is no legitimate reason why one would be playing the game; yet here, the game is oligopolistic competition, which everyone concedes is lawful, and the ante is perfectly consonant with the desire to play. For both of these reasons, the delivery of wholesale-to-retail sales information to MSA does not tend to exclude the possibility of independent action (or tend to establish a price fixing conspiracy), and thus cannot constitute a plus factor.

Nor were the manufacturers' adoption of permanent allocation programs16 contrary to their economic interests. Thus, the adoption of these systems does not tend to exclude the possibility of lawful, independent behavior, does not tend to establish a price fixing conspiracy, and cannot constitute a plus factor. It is uncontroverted that during "the late 1980s and early 1990s, [the manufacturers'] direct customers engaged in `trade loading,' a practice whereby the direct customers would purchase an amount of cigarettes in excess of what they could sell either (1) in anticipation of expected list price increases or (2) in response to end-of-quarter or end-of-year manufacturer discounts and incentives. [The class] admits that [appellees] experienced increased costs due to trade loading because that practice would distort sales results and cause excess product returns, disruptions in shipping patterns, and stale product." Holiday Wholesale Grocery, 231 F.Supp.2d at 1284. By implementing the allocation programs, appellees were able to eliminate this costly practice, and, the district court found, contemporaneous documents demonstrate conclusively that this was the actual motivation for the programs. See id. at 1284-85. It is clear that the adoption by appellees of permanent allocation programs was in their respective interests.

In addition, the evidence establishes that allocations were increased in cases where a particular wholesaler demonstrated a genuine need for more product, for example where it took on a new retail customer. Although appellants attempt to point to specific instances in which wholesalers were unable to procure such increases, we agree with the district court that the refusal to increase a particular allocation generally occurred only where the demand for more product stemmed from the perception that a price increase was imminent, precisely the phenomenon that the systems were designed to eliminate. In summarizing its findings, the court found that "while it is clear the wholesalers did not like the allocation system, the testimony presented by [the class] is either ambiguous or shows the opposite of what [it] contend[s]. As such, no reasonable inference could be drawn from [the class's] evidence that [the manufacturers] used the allocation system to restrict output." Id. at 1287.

Finally, the class claims that the price increases that followed the settlement of several healthcare lawsuits with various state Attorneys General were excessive — the largest occurred on November 23, 1998, when PM raised its list prices by $22.50 per thousand — and contrary to appellees'17 (or at least PM's) economic interests. Specifically, it says that "PM knew that the large increases ... benefitted RJR and B & W more than PM. PM had the financial ability to take smaller price increases that only offset settlement costs, and had it done so, PM would have benefitted, and RJR and B & W would have lost market share." Appellants even question whether PM needed to increase its prices at all following the settlement. By contrast, they stress that regardless of PM's pricing decision, RJR and B & W were forced to raise prices to meet the financial obligations imposed by the settlement. Accordingly, they posit that by not raising prices following the settlement, or at least limiting the amount of its price increase, PM could have undersold its competitors, thereby gaining market share.

However, this argument fails because it unfairly constricts the range of PM's possible competitive responses to the healthcare settlement. PM could have attempted to increase its profits by either of two means: instituting a smaller price increase in an effort to gain market share from its competitors or, more directly, simply increasing prices to the maximum extent the market would bear. The class's argument is predicated on the notion that only the first was economically advantageous for PM. Nowhere does it explain why this should be so, and it disregards the fact that the second was likely a more direct means of accomplishing the same end.

Indeed, given that the need for a significant price increase was a product of a sudden and sizable cash outflow, PM can readily point to sound economic reasons for taking the most surefire, direct route to the recoupment of its losses. Instead of attempting to recover its losses through market share gains, which would have required a shift in consumers' brand loyalties — which, although not immutable, are not evanescent either — PM instead chose to directly replenish its coffers by increasing its prices. This plainly was not an uncompetitive decision. As such, the extent of PM's post-settlement price raises does not tend to exclude the possibility of independent behavior among appellees or to establish a collusive price fixing arrangement, and consequently is not a plus factor.

c. Monitoring of Sales through MSA

Previously, we indicated that the district court correctly concluded that appellees' provision of sales and pricing information to MSA was not a plus factor. However, the class raises a separate argument related to this monitoring. In particular, it says that "MSA solicited and received each [appellee's] approval for the modification of its services and the introduction of new services and specifically conditioned them on [appellees'] unanimous approval," and that this concerted behavior tends to exclude the possibility of independent action.

The district court rejected this argument, observing that appellants had "offer[ed] no support for their contention that each modification to the service provided by MSA ... was made with the `joint agreement of all [appellees].' Rather, the documents cited by Plaintiffs show that MSA ... discussed various proposals with each company individually and made modifications to their services based on ... MSA's ... views about what would be the most acceptable to their customers. There is no evidence in the documents cited by [appellants] that any of [appellees] negotiated together or reached any agreement together about what services they would request from MSA...." Holiday Wholesale Grocery, 231 F.Supp.2d at 1295 n. 29.

On appeal, appellants fail to demonstrate how this conclusion was erroneous. Moreover, even if the class could show that on some occasions PM, RJR, B & W and Lorillard each agreed to a modification to the service before that change was instituted, this does not tend to establish the existence of a conspiracy any more than it reflects that each individual subscriber to the joint data provision system made its decision based on an individualized calculus. To the extent that appellants argue that the very joint nature of this system gives rise to a reasonable inference of collusion, we already have discussed the practical and financial advantages that stem from the ability to monitor the sales of one's competitors and from the spreading of the responsibility for providing market share and pricing data to those competitors.

This leaves only the following alleged plus factors that the district court deemed unworthy of extended discussion.

d. Smoking and Health Conspiracy

Appellants say that beginning in 1953, PM, RJR and B & W conspired to restrict competition on the basis of health. They assert that this agreement remained effective throughout the alleged conspiracy and that it evidences appellees' agreement to fix prices. As the manufacturers note, however, Professor Fisher conceded that there is no evidence that these activities continued to transpire during the period in question, i.e., 1993-2000.

It is entirely possible that firms' decisions to limit health-based marketing in the past continued to be felt between 1993 and 2000. Such decisions may have checked companies' incentives to vigorously pursue health-related research. To the extent that health-related innovations were retarded as a result, cigarettes would have exhibited a higher degree of homogeneity between 1993 and 2000 than they would have otherwise. Similarly, firms limited their health-based advertising in the early to mid 1990s, which may have served to maintain perceived homogeneity. As explained above, greater homogeneity facilitates coordination on price. Thus to the extent that firms' decisions resulted in greater homogeneity between 1993 and 2000, it would have facilitated coordination on price in those years.

Holiday Wholesale Grocery, 231 F.Supp.2d at 1311 (emphasis in original).

In essence, for a jury to infer from this evidence that appellees collusively fixed prices during the 1990s, it would have to engage in propensity reasoning, i.e., "they did it 40 years before so they probably did it again," albeit now with Lorillard on board and with the object of the conspiracy being prices as opposed to health. In other words, "[i]n order to draw an inference of conspiracy [from] this testimony, a jury would have to engage in such speculation that its finding would be a `guess or mere possibility.'" Id. Moreover, Fisher's opinion itself is stated only in terms of theoretical possibility, hardly the stuff of proof that tends to exclude lawful, synchronous behavior.

The "smoking and health conspiracy" does not tend to exclude the possibility that appellees were engaged in lawful, competitive pricing behavior or to establish a price fixing conspiracy, and cannot constitute a plus factor. See Matsushita, 475 U.S. at 595, 106 S. Ct. at 1360 (dismissing arguments against granting summary judgment in favor of antitrust defendants where "where the evidence of conspiracy is speculative"); see also Eastman Kodak, 504 U.S. at 468-69, 112 S. Ct. at 2083 (reaffirming the Matsushita standard, and noting specifically that evidence of predatory pricing that is "`speculative,' and [is] not `reasonable'" is insufficient to withstand an antitrust defendant's summary judgment motion); SEC v. Alejandro Duclaud Gonzalez de Castilla, 184 F. Supp. 2d 365, 376 (S.D.N.Y.2002) ("While in a motion for summary judgment every inference based upon an established fact must be drawn against ... movants, as juries are regularly advised, speculation cannot substitute for proof.").

e. Foreign Conspiracies

The class also suggests that appellees conspired to fix prices in 10 different foreign countries during and prior to the alleged American conspiracy around which this lawsuit centers. The district court found that evidence of the manufacturers' actions in these countries was not admissible under Fed.R.Evid. 404(b) (other crimes, wrongs or acts) because appellants had failed to show that any of the actions allegedly undertaken by appellees overseas were illegal under the applicable foreign law. It summarized its disposition of the foreign conspiracy claim by saying that:

[Appellants] [ ]assert that they have "extensive evidence of anti-competitive activity" by [appellees] in Hungary, France, Argentina, Venezuela, and Panama. [Appellants] fail, however, to address the court's previous ruling that "price-fixing agreements were not at all times unlawful under the laws" of these countries. Furthermore, [appellants] baldly contend that [appellees], or their affiliates, undertook anti-competitive price fixing agreements in Canada, Costa Rica, El Salvador, Guatemala, and Saudi Arabia (Gulf States). Nowhere do [appellants] describe the legal landscape in these countries that would make any such activities unlawful, nor do [they] do any more than cite to documents in describing the nature of these alleged anti-competitive activities. This is simply too thin a reed upon which to allege that [the manufacturers'] "foreign conspiracies" are a "plus factor" in the instant litigation.

Holiday Wholesale Grocery, 231 F.Supp.2d at 1312.

On appeal, the class cites Professor Fisher's testimony for the proposition that regardless of the legality of the manufacturers' activities in these countries, the foreign agreements provided a mechanism to "establish and revise, and to monitor and enforce agreements to coordinate." Appellants say that a reasonable jury could infer the existence of such an American conspiracy from these allegedly anti-competitive activities abroad.

To begin with, we cannot say the district court abused its discretion in resolving this evidentiary question. In the absence of some palpable tie between these overseas activities and appellees' pricing actions in the United States, the foreign undertakings of PM, RJR, B & W and Lorillard do not tend to exclude the possibility of independent action in the setting of domestic cigarette prices any more than does the health conspiracy that allegedly began during the 1950s. Nor does the evidence readily establish a prior crime, wrong or act where appellants ultimately have failed to establish that the foreign conduct was a crime or wrong under the laws of the foreign sovereigns. See generally Fed.R.Evid. 404(b). Accordingly, appellants' alleged evidence of foreign agreements to collude does not rise to the level of a plus factor.

f. History and Structure of the Tobacco Industry 18

First, appellants say that the structure of the tobacco industry is conducive to price fixing agreements because of a concentration of sellers, inelastic demand at competitive prices, high barriers to entry, a fungible product, principal firms selling at the same level in the chain of distribution, prices that can be changed quickly, cooperative practices and a record of antitrust violations. In support of the contention that this industry structure constitutes a plus factor, the class cites In re High Fructose Corn Syrup Antitrust Litig., 295 F.3d 651, 655 (7th Cir.2002). The problem with this argument, however, is that the majority of the market characteristics on which the class focuses are simply indicia that the tobacco industry is an oligopoly, which is perfectly legal. See generally Harcros, 158 F.3d at 570-71. Moreover, as Judge Posner recognized in In re High Fructose Corn Syrup, "almost any market can be cartelized if the law permits sellers to establish formal, overt mechanisms for colluding, such as exclusive sales agencies." 295 F.3d at 655.

As for the industry's history of antitrust violations, the district court noted that the class had failed to direct it to any precedent for holding such to be indicative of a present antitrust violation. See Holiday Wholesale Grocery, 231 F.Supp.2d at 1305 ("The law generally disfavors use of such `historical' evidence.") (citations omitted). Appellants have failed to do so on appeal as well.

g. Dampening of Market Share Shifts

Appellants next say that the shifts in the market shares enjoyed by each appellee moderated during the period of the alleged conspiracy, thus indicating that the collusion had its intended effect, i.e., bringing stability to a most turbulent industry. They appear to claim that this is a plus factor. The class raises this argument in response to the district court's focus on the fact that there were significant market share shifts between 1993 and 2000, which it deemed a strong indication of competitive behavior. Appellants' contention is misplaced, however, because (1) as the district court noted, these shifts remained sizable between 1993 and 2000; and (2) it is beyond debate that no reasonable jury could infer from the fact that significant market share swings did not cease but instead were only somewhat smaller during the period in question than those that occurred between 1991 and 1993 that appellees conspired to fix prices.

Indeed, from 1993 to 2000 PM's share of the United States market rose 20%, Lorillard's share rose nearly 50%, RJR's share fell approximately 25% and B & W's share declined roughly 19%. These sharp market share shifts were at least as great as those that occurred between 1980 and 1991, an unquestionably competitive period — to reiterate, RJR lost twice as much market share from 1993 to 2000 as it did from 1982-1991 — and plainly are inconsistent with a price fixing conspiracy. See generally Brooke Group, 509 U.S. at 212-18, 113 S. Ct. at 2582-85 (describing the competitiveness and market share shifts that characterized the American cigarette market during the 1980s). Moreover, to the extent that the class argues merely that the degree of these shifts is probative of a conspiracy (but does not rise to the level of a plus factor), this contention is vastly less compelling than is its inverse, namely that the very fact that there were significant market share shifts during this period of alleged industry-wide collusion strongly undermines appellants' conspiracy allegations.

h. "Credit Memos" Made Price Increases Prospective

The class argues that appellees' "price increase announcements were accompanied by `credit memos' so that the price increases were effectively delayed for several weeks. This in effect allowed the [appellee] initiating the price increase to `offer' it to the other [appellees] before it became effective, to ensure that it met with all [appellees'] approval." The district court squarely rejected this argument because the class provided no evidence that the credit memos actually had this effect. It held: "The only testimony [appellants] cite for their contention that credit memos delayed price increases fails to support this assertion. [Appellants] assert that Carl Schoenbachler, former Chief Financial Officer of B & W, testified that the credit memos delayed the impact of the price increase for a week or more.... [H]owever, a more complete reading of Mr. Schoenbachler's testimony shows that [appellants'] meaning cannot be attributed to his testimony.19.. Furthermore, [appellees'] price announcements took effect immediately, within one to two days of the announcement date." Holiday Wholesale Grocery, 231 F.Supp.2d at 1274 n. 11. Because there is no evidence that these credit memos actually delayed the implementation of price increases, appellants' argument that they constitute a plus factor necessarily fails, as does any reliance they place on these memos as support for their allegation of conspiracy.

i. Opportunities to Conspire and the Restriction of Decisionmaking Authority to High-Ranking Corporate Officers

The class posits that PM, RJR, B & W and Lorillard enjoyed numerous opportunities to conspire, and that this supports their collusion claim. We unambiguously held in Todorov, however, that "the mere opportunity to conspire among antitrust defendants does not, standing alone, permit the inference of conspiracy." 921 F.2d at 1456 (citing Bolt v. Halifax Hosp. Med. Ctr., 891 F.2d 810, 827 (11th Cir.1990), overruled in part on other grounds by City of Columbia v. Omni Outdoor Adver., 499 U.S. 365, 111 S. Ct. 1344, 113 L. Ed. 2d 382 (1991)). Indeed, the opportunity to fix prices without any showing that appellees actually conspired does not tend to exclude the possibility that they did not avail themselves of such opportunity or, conversely, that they actually did conspire. Appellants may not rely on this proposition to support their allegations in this case.

Similarly, we held in Harcros that "memoranda showing that [an alleged conspirator] exercised strict control over chlorine pricing at Harcros support an inference that [that individual] was frustrated with his subordinates' failure to follow an innocuous allocation of responsibility as easily as they support an inference that Harcros centralized pricing authority in [the individual] in order to conceal the alleged conspiracy from other employees." 158 F.3d at 569. Such a showing cannot constitute a plus factor, as it does not tend to exclude the possibility of independent behavior or to establish a price fixing conspiracy. Firms routinely consolidate decision-making authority in high ranking officers for a multitude of wholly legitimate reasons. Standing alone, this tells us nothing about collusive price fixing.

As a final note, we observe that appellants have neglected Lorillard in nearly all of their allegations, and agree with the district court that this repeated omission also "casts doubt on [appellants'] theory." Holiday Wholesale Grocery, 231 F.Supp.2d at 1315.

Thus, after thorough review of this case, we are satisfied that none of the actions on which appellants' arguments are based rise to the level of plus factors. Nor do they constitute plus factors when considered in concert. Indeed, when all of appellees' actions are considered together, the class has established nothing more than that the tobacco industry is a classic oligopoly, replete with consciously parallel pricing behavior, and that its members act as such.

Because the class has not carried its burden of demonstrating the existence of a plus factor, it is unnecessary to reach the question of whether the manufacturers are able to rebut any inference that they conspired to fix prices. Nonetheless, we briefly discuss the arguments appellees raise in this connection, as we believe that these showings also powerfully establish that the wholesalers' allegations of collusive price fixing are economically untenable.

In general terms, PM, RJR, B & W and Lorillard argue that we must consider whether the conspiracy theory makes sense not as a general matter, but rather against the background of the economic realities of the 1990s cigarette market. When we do so, they continue, we should find that the class's conspiracy theory makes no economic sense. In particular, they say that as evidenced by a multitude of contemporaneous documents (1) cigarette prices were lower and actually rose more slowly between 1993 and 2000 than before and their combined operating income during the alleged conspiracy period never reached pre-1993 levels; (2) appellees' competitive spending at retail between 1993 and 2000 (approximately $25.256 billion) actually was much greater than the alleged list price overcharges to the wholesalers ($11.9 billion); and (3) market shares shifted dramatically during the alleged period of collusion, with clear "winners" and "losers" emerging. Appellees say that these considerations demonstrate that all of the business decisions they made during the period in question were motivated by strategic, competitive concerns and reflect uncertainty regarding various aspects of the market that is inconsistent with collusive behavior.

The district court analyzed the manufacturers' specific arguments in support of their contentions, agreed with each of them and concluded that even if the class had created an inference of conspiracy, appellees had fully rebutted it. See Holiday Wholesale Grocery, 231 F.Supp.2d at 1322-28. We agree.

As for the indisputable fact that cigarette prices were lower during most of the alleged period of collusion than they previously had been, the class says that this is a strawman, because these reduced prices were a direct product of Marlboro Friday, which induced the price fixing conspiracy. However, as the district court observed, appellees "demonstrated that prices rose at a lesser rate during the alleged conspiracy period than during the period from 1988 to 1993, which [the class] contend[s] is a period of competitive pricing. . . ." Holiday Wholesale Grocery, 231 F.Supp.2d at 1323. Had a price fixing arrangement actually been afoot — that is, in the absence of competition-induced pressure to keep prices low — the court reasoned, prices would have risen faster than if the manufacturers were competing with each other. Simply put, it does not require an advanced degree in economics to predict that prices will rise more quickly in the absence of competition than in a competitive environment. The district court did not err in concluding that this factor weighs meaningfully against the economic viability of appellants' theory.

Perhaps even more significantly, the class concedes that appellees "increased retail promotional payments during the conspiracy." Indeed, PM, RJR, B & W and Lorillard "spent a combined $25.256 billion on promotional allowances, coupons and retail value added" between 1993 and 2000, an amount that is more than double the sum allegedly overcharged. Holiday Wholesale Grocery, 231 F.Supp.2d at 1326. In 1999, for example, PM alone spent nearly $2.9 billion (60% of its operating income) on retail promotions, couponing and its retailer incentive program. Moreover, we reiterate that during 1998 competition within the industry was sufficiently pervasive that RJR, B&W and Lorillard brought an antitrust and unfair competition suit against PM based on PM's adoption of a program that it called "Retail Leaders." See R.J. Reynolds Tobacco Co., 199 F.Supp.2d at 365.

The district court recognized, simply and powerfully, that if prices are fixed, and market share consequently is unlikely to change very much, there is no rational reason to undertake extremely significant and expanding retail promotional expenditures, which are a paradigmatically competitive activity. Framed differently, it makes no sense for firms that are fixing wholesale prices to then undermine the economic benefit of their collusion by spending more than twice the amount of their allegedly illegal profits on retail competition. This seems beyond reasoned debate, and thus the manufacturers' sizable retail expenditures also undermines the economic viability of appellants' theory.

Finally, as for shifting market shares during the period in question, this also is contrary to the notion that appellees were conspiring to fix prices. Specifically, from 1993 to 2000 PM's share of the United States market rose 20%, Lorillard's share rose nearly 50%, RJR's share fell approximately 25% and B&W's share declined roughly 19%. Contrary to the class's contentions, these market shares are far from stable. Moreover, it makes little sense to posit that RJR and B&W would continue to participate in a conspiracy that cut their market shares so dramatically.

Accordingly, the economic circumstances under which the price fixing agreement allegedly transpired, when considered carefully, strongly undermine the allegation of anticompetitive behavior. Simply stated, several of the actions taken by PM, RJR, B&W and Lorillard are wholly inexplicable if we assume that a price fixing arrangement had been reached. In sum, we believe the district court correctly concluded that even had appellants demonstrated the existence of one or more plus factors — and on this record they have not — the manufacturers would have readily rebutted the resulting inference of collusion.

Besides arguing that the district court erred by rejecting their arguments concerning plus factors, the class further contends that the court erroneously excluded the expert testimony of Professor Fisher on some of these points. Specifically, appellants say that the district court improperly excluded Fisher's conclusion that the manufacturers engaged in activities beyond mere conscious parallelism,20 which he expressed in the context of several of the particular actions on which appellants base their claims.21

As we said in Harcros: Federal Rule of Evidence 702, as explained by the Supreme Court in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 589, 113 S. Ct. 2786, 2794-95, 125 L. Ed. 2d 469 (1993), and its progeny, controls determinations regarding the admissibility of expert testimony. Expert testimony may be admitted into evidence if: (1) the expert is qualified to testify competently regarding the matters he intends to address; (2) the methodology by which the expert reaches his conclusions is sufficiently reliable as determined by the sort of inquiry mandated in Daubert; and (3) the testimony assists the trier of fact, through the application of scientific, technical, or specialized expertise, to understand the evidence or to determine a fact in issue.

158 F.3d at 562; see also Quiet Tech. DC-8 v. Hurel-Dubois UK, Ltd., 326 F.3d 1333, 1340-41 (11th Cir.2003) (setting forth the standards governing the admission of expert testimony). As we have said, "we defer to the district court's evidentiary ruling unless that ruling is `manifestly erroneous.'" Id. (quoting General Elec. Co. v. Joiner, 522 U.S. 136, 142, 118 S. Ct. 512, 517, 139 L. Ed. 2d 508 (1997)). Furthermore, "`[w]e will not overturn an evidentiary ruling and order a new trial unless the objecting party has shown a substantial prejudicial effect from the ruling.'" Quiet Tech. DC-8, 326 F.3d at 1341-42 (quoting Maiz, 253 F.3d at 667).

In this case, the district court held that Fisher's opinions that collusive price fixing was afoot should be excluded because they were unhelpful and thus irrelevant, i.e., they did not tend to make it any more probable that appellees were (or were not) engaged in a price fixing conspiracy. See Holiday Wholesale Grocery, 231 F.Supp.2d at 1320-22. Additionally, the court held in at least one case that Fisher's testimony was unhelpful because he had misunderstood the evidence. See, e.g., Holiday Wholesale Grocery, 231 F.Supp.2d at 1288 (holding that Fisher's testimony regarding the institution of permanent allocation systems was "premised on an erroneous understanding of the evidence" and thus inadmissible). More specifically, the district court excluded Fisher's ultimate opinion that there was an illegal price fixing conspiracy as irrelevant because Fisher did not differentiate between lawful, conscious parallelism and collusive price fixing. Accordingly, the court held, these conclusions were of absolutely no use to a factfinder. See generally id. at 1321 ("The failure of Plaintiffs' expert to distinguish conscious parallelism from cartel behavior makes his subsequent opinion inadmissible as he finds inferences of collusion where the law finds none. Dr. Fisher is positing a new theory where certain aspects of conscious parallelism should be found to be anticompetitive. This is not the state of the law. This meshing of two terms is particularly troubling when the hallmark of a `plus factor' under the law is that it tends to exclude the possibility that the defendant was engaged in mere conscious parallelism.").

In Daubert, the Supreme Court observed that:

Rule 702 ... requires that the evidence or testimony "assist the trier of fact to understand the evidence or to determine a fact in issue." This condition goes primarily to relevance. Expert testimony which does not relate to any issue in the case is not relevant and, ergo, non-helpful. An additional consideration under Rule 702 — and another aspect of relevancy — is whether expert testimony proffered in the case is sufficiently tied to the facts of the case that it will aid the jury in resolving a factual dispute. The consideration has been aptly described... as one of "fit."

509 U.S. at 591, 113 S. Ct. at 2795-96 (internal punctuation and quotations omitted).

Simply stated, we can perceive no abuse of discretion in the district court's decision to exclude some of Fisher's testimony. Fisher's conclusion was that plaintiffs participated in an illegal price fixing conspiracy, and he expressed this opinion in the context of several aspects of appellants' evidence, e.g., by saying that the manufacturers' participation in the MSA information sharing system created an inference of collusion. However, Fisher defined "collusion" to include conscious parallelism. Put differently, he did not differentiate between legal and illegal pricing behavior, and instead simply grouped both of these phenomena under the umbrella of illegal, collusive price fixing. This testimony could not have aided a finder of fact to determine whether appellees' behavior was or was not legal, and the district court properly excluded it. See Fed.R.Evid. 702 (requiring that expert evidence "assist the trier of fact" in order to be admissible); Fed.R.Evid. 402 ("Evidence which is not relevant is not admissible."). Moreover, this was an equally proper basis at the summary judgment stage as it would have been in the context of a trial. See Kumho Tire Co. Ltd. v. Carmichael, 526 U.S. 137, 158, 119 S. Ct. 1167, 1179, 143 L. Ed. 2d 238 (1999) (reinstating the district court's grant of summary judgment where the expert testimony offered to defeat summary judgment did not meet the Daubert requirements).

We also note that the "addendum" to Fisher's report on which the class focuses specifically in its brief was not timely filed, and this alone was a sufficient basis for the district court to reject its admission. See id.; Fed.R.Civ.P. 26(a)(2)(C) (requiring that expert disclosures be filed "at the times and in the sequence directed by the court").

Finally, in every instance where the district court refused to admit Fisher's testimony, it did so only after considering it at length and concluding that it did not establish the existence of the particular plus factor in support of which it was offered. We too have reviewed and considered all of the expert testimony in question, and we are satisfied that it would not have assisted a factfinder in determining whether appellees engaged in lawful conscious parallel behavior or collusive price fixing.

Because appellants cannot demonstrate the existence of a plus factor, they cannot establish an inference of conspiracy, as they must to carry the burden imposed on them at the summary judgment stage of a collusive price fixing case. Moreover, appellees would have rebutted any inference that they conspired to fix prices by demonstrating that the class's conspiracy theory is utterly implausible. In addition, the court's exclusion of several portions of Fisher's testimony was in no sense an abuse of discretion, i.e., did not constitute manifest error. See Joiner, 522 U.S. at 142, 118 S. Ct. at 517. Accordingly, we affirm in all respects the district court's final summary judgment in favor of the manufacturers.

AFFIRMED.