The plaintiffs in this antitrust suit, seventeen manufacturers of tortillas, appeal a take-nothing judgment entered in favor of Gruma Corporation (“Gruma”), a manufacturer of tortillas and related food products, two of its corporate divisions, and a related entity. Because the plaintiffs failed to offer evidence of damages and causation sufficient to sustain a rational judgment in their favor, we affirm.

I.

The nature of this suit and the conduct alleged to be anticompetitive are set forth in the district court’s opinion, 1 so we only briefly summarize them here. Plaintiffs challenge Gruma’s conduct downstream— its efforts to obtain shelf and display space in retail outlets and to induce retailers to promote and advertise its tortillas. Specifically, plaintiffs challenge Gruma’s use of marketing agreements with retailers whereby Gruma pays up-front fees to retailers—so-called “slotting fees”—or provides other price reductions or financial incentives to obtain (and in part manage) shelf space, advertising, and product promotion, as well as Gruma’s conduct as it acts as a “Category Captain,” a designa *452 tion given a particular product manufacturer by a retailer enabling the manufacturer to assist the retailer in display and promotional operations. 2 Plaintiffs charge Gruma with exclusive dealing in violation of section 1 of the Sherman Act and section 3 of the Clayton Act, monopolization and attempted monopolization in violation of section 2 of the Sherman Act, price discrimination in violation of the Robinson-Patman Act, and violations of state antitrust laws.

Before trial, Gruma moved for summary judgment, contending that plaintiffs could not, as a matter of law, establish any antitrust violations. Gruma also moved to exclude, on Daubert grounds, 3 the plaintiffs’ designated expert witnesses on damages and causation. The district court did not rule on Gruma’s summary judgment motion before trial; it entered an order indicating the motion would be carried with trial. The court summarily denied the motions to exclude but reserved the right to reconsider the question on a renewed objection at trial.

The case proceeded to trial before a jury. After plaintiffs presented their fact witnesses, they offered their damages expert, Kenneth McCoin, to opine on the profits allegedly lost as a consequence of Gruma’s challenged conduct. Gruma renewed its Daubert objection, and after a complete proffer and extended voir dire examination, the court sustained the objection and excluded McCoin from testifying. Plaintiffs then called their expert on causation and antitrust injury, Gregory Gundlach, to opine on the causal link between Gruma’s challenged conduct and the damages claimed; Gruma renewed its objection to his testimony, and after another proffer and examination, the court excluded Gundlach from testifying.

The plaintiffs having no admissible evidence of antitrust damages or causation on which a verdict could be based, the court dismissed the jury. Gruma moved for judgment as matter of law and moved the court to consider its pending motion for summary judgment. Thereafter, in a published opinion, the court articulated its reasons for excluding the experts and granted a take-nothing judgment in favor of Gruma, concluding that plaintiffs’ claims for money damages and injunctive relief fail as a matter of law.

II.

Because the district court granted Gruma’s motion for summary judgment and its motion for judgment as a matter of law, our review is de novo, and we may affirm on any basis supported by the record. 4 Private antitrust liability under § 4 of the Clayton Act requires a plaintiff to show (1) a violation of the antitrust laws, (2) the fact of damage, and (3) some indication of the amount of damage. E.g., Nichols v. Mobile Bd. of Realtors, Inc., 675 F.2d 671, 675-76 (5th Cir. 1982). The fact of damage requirement is one of causation; the plaintiff must show that the defendant’s unlawful conduct was a material cause of injury to its business. If the requisite causal link is proven, “a more relaxed burden of proof obtains for the amount of damages than would justify an award in other civil *453 cases.” Eleven Line, Inc. v. N. Tex. State Soccer Ass’n, 213 F.3d 198, 207 (5th Cir. 2000). Though relaxed, the standard for proving the quantum of damages is not without bounds, for antitrust damages may not be determined by guesswork or speculation; “we must at least insist upon a ‘just and reasonable estimate of the damage based on relevant data.’ ” Lehrman v. Gulf Oil Corp., 464 F.2d 26, 46 (5th Cir. 1972) (quoting Bigelow v. RKO Radio Pictures, 327 U.S. 251, 264, 66 S.Ct. 574, 90 L.Ed. 652 (1946)). 5

III.

We do not pause to consider the district court’s conclusion that the plaintiffs failed, as a matter of law, to establish any harm to competition rather than competitors in any properly defined market and thus any violation of the antitrust laws. Instead, affirmance is compelled on more narrow grounds: the failure of proof on damages and causation.

A.

To prove actual damages, plaintiffs chiefly relied on McCoin’s damages model. He used a yardstick measure of lost profits whereby he compared plaintiffs’ sales history with sales data and growth projections from trade association studies of national tortilla markets; he applied a uniform gross margin to each of the plaintiffs’ firms—the major assumption being that, absent Gruma’s illegal conduct, each of the plaintiffs would have performed to the rate of the market as a whole. See El-Aguila, 301 F.Supp.2d at 624 n. 14. The district court found this model wholly unreliable insofar as it attributed all of the measured lost profits to the alleged antitrust injury and irrelevant insofar as it was not in any respect anchored to the specific agreements or marketing practices challenged by plaintiffs. See id. at 624-26.

A district court has broad discretion in deciding to admit or exclude expert testimony, 6 and excluding McCoin’s testimony was anything but an abuse of discretion. Indeed, McCoin made no effort to demonstrate the reasonable similarity of the plaintiffs’ firms and the businesses whose earnings data he relied on as a benchmark. 7 Similarly, McCoin did not consider whether the plaintiffs’ firms were even capable of handling the excess capacity the projected rates of return necessarily entail. Moreover, by characterizing all variances between the trade association earnings data and plaintiffs’ respective earnings as “lost profits,” no allowance was made for losses caused by any other factor—including, for example, reductions in shelf space attributable to other dominant firms or new entrants into the relevant markets, the plaintiffs’ own failure to compete for shelf space on the terms sought by retailers, and their lack of capacity or efficiency relative to other firms.

In any event, even if McCoin’s testimony had been admitted, it would not have pro *454 vided a sufficient basis on which the jury could have arrived at a reasonable and just estimate of actual damages. 8 And the plaintiffs did not present additional evidence sufficient to prove damages. Because plaintiffs failed to offer substantial evidence on which a principled award of money damages could be based, the district court did not err in granting judgment in favor of Gruma on the claims for money damages.

B.

Nor did plaintiffs provide sufficient evidence to demonstrate that Gruma’s conduct was a material cause of actual or threatened damage (much less of the sort the antitrust laws were designed to prevent). 9 To prove causation, plaintiffs relied primarily on Gundlach, whose view was that Gruma’s marketing agreements and practice of paying slotting fees resulted in exclusivity and permitted preferential shelf-space and display positions in a manner inconsistent with sales and thus restricted competitors from the market.

The district court did not question Grundlach’s qualifications but excluded his testimony because his opinions amounted to abstract conclusions not adequately grounded in the facts of the case. See El-Aguila, 301 F.Supp.2d at 620-24. We cannot say this was an abuse of discretion, for the record indicates that Gundlach did not examine sales data from retailers in the relevant markets to determine whether space allocation among the various brands was disproportionate to their sales, nor did he attempt to tie space allocation to retailers with which Gruma had marketing agreements or paid slotting fees, or to quantify either the extent of exclusivity Gruma allegedly obtained or the extent of market foreclosure allegedly caused by Gruma’s challenged conduct. 10 Moreover, Grundlach failed adequately to account for alternative causes of plaintiffs’ reduction in shelf space, most notably their failure to compete for shelf space by offering similar incentives to reduce the net price paid for tortillas by retailers as well as the growing success of retailers’ own private lines of tortillas.

Beyond their designated expert, plaintiffs point to circumstantial evidence and argue that the jury could have inferred causation. Though jury inferences of causation are in some instances permissible, “the required causal link must be proved as a matter of fact and with a fair degree of certainty.” Alabama v. Blue Bird Body Co., 573 F.2d 309, 317 (5th Cir. 1978). And this is especially so where plaintiffs admittedly lost shelf space (and thus sales) because of salient factors distinct from the challenged conduct such as increasing competition in the tortilla category and their refusal even to seek shelf space in some retail outlets where Gruma’s products were sold. 11

*455 In sum, because plaintiffs failed to present substantial evidence demonstrating that Gruma’s conduct was a material cause of its actual or threatened injury as well as evidence on which a jury could base a reasonable award of money damages, their claims fail as a matter of law. Accordingly, the judgment in favor of defendants is AFFIRMED.