Rebel Oil Co. v. Atlantic Richfield Co.

BOOCHEVER, Circuit Judge.

Rebel Oil Company appeals from the district court’s grant of summary judgment in favor of Atlantic Richfield Corporation, after this court remanded Rebel’s antitrust action against ARCO. We affirm.

FACTS

The following statement of facts is taken in part from this court’s decision in Rebel Oil Co. v. Atlantic Richfield Co., 51 F.3d 1421 (9th Cir.), cert. denied, 516 U.S. 987, 116 S.Ct. 515, 133 L.Ed.2d 424 (1995), and from the district court opinion that is the subject of this appeal, Rebel Oil Co. v. Atlantic Richfield Co., 957 F.Supp. 1184 (D.Nev.1997).

Gasoline sold in the Las Vegas, Nevada area is first produced from crude oil in refineries in Los Angeles, California. Wholesale marketers then pump the gasoline to Las Vegas over a pipeline operated by the Cal-Nev Pipeline Co., which transports ninety-five percent of Las Vegas’ gasoline. Wholesale marketers then sell the gasoline to Las Vegas retail marketers, who then sell the gasoline to consumers.

Rebel Oil Company, Inc. and Auto Flite Oil Company, Inc. (collectively “Rebel”) are retail marketers of gasoline in Las Vegas, selling only self-serve, cash-only gasoline. Rebel has 16 retail stations. In addition to its retail sales, Rebel is also a wholesaler who ships gasoline refined in Los Angeles over the Cal-Nev pipeline, to be sold to retail marketers in Las Vegas. Atlantic Richfield Corporation (“ARCO”) is both a wholesale and a retail marketer of gasoline in Las Vegas, and also drills and refines crude oil into gasoline. ARCO supplies gasoline to 53 retail stations in Las Vegas under the “ARCO” name, all of which sell only self-. serve, cash-only gasoline. A subsidiary of ARCO, Prestige Stations (“PSI”), operates 15 of those 53 stations. The remaining 38 stations operating under the ARCO name are owned and operated by independent dealers.

Rebel and other discount marketers of gasoline saw their business grow in the 1970s, as cost-conscious motorists driving low-mileage vehicles saved money by buying self-serve, cash-only gasoline rather than visiting full-service stations. In 1982, ARCO adopted a nationwide strategy to respond to this change in the gasoline market. The strategy called for the elimination of full-serve and credit-card sales and offered dealers incentives for matching the prices of the discount independent stations. ARCO’s strategy increased its sales and market share nationwide.

PROCEDURE

In 1990, Rebel filed an antitrust suit for damages against ARCO under Section 4 of the Clayton Act. Rebel alleged that between 1985 and 1989, ARCO charged predatory prices in an attempt to take away market shares from its competitors, with the goal of monopolizing the gasoline market in Las Vegas. Rebel claimed that the predatory pricing scheme forced competitors out of the Las Vegas market, and that Rebel lost sales as a result and had to concentrate on the wholesale market. Rebel also claimed that when ARCO’s predation ended, ARCO had fifty-four percent of the market for self-serve, cash-only gasoline, and that ARCO then engaged in price gouging, charging prices in Las Vegas that were higher than Los Ange-les prices to recoup its losses resulting from the previous low predatory prices.

The district court limited Rebel’s discovery solely to the “more discrete” issues of ARCO’s market share and the entry barriers in the relevant market, which would determine whether ARCO had sufficient market power to charge prices above competitive levels (“supracompetitive prices”). Rebel Oil Co., Inc. v. Atlantic Richfield Co., 133 F.R.D. 41, 45 (D.Nev.1990). The court reserved discovery on the issue whether ARCO had engaged in predatory pricing until after Rebel made an adequate showing of market power, because “[pjroving that a defendant has engaged in pricing below cost entails extensive discovery regarding virtually all aspects of a company’s business.” Id.

In the fall of 1992, both parties filed motions for summary judgment and the district court granted summary judgment for ARCO on all three antitrust claims. Rebel Oil Co. v. Atlantic Richfield Co., 808 F.Supp. 1464 (D.Nev.1992). The court concluded that ARCO’s market share of the retail gasoline market in Las Vegas was insufficient as a matter of law to establish market power; that Rebel failed to show that barriers to entry prevented other retailers from entering the retail market; and that as a result of the lack of market share and entry barriers, ARCO lacked the power to charge supracom-petitive prices to recoup any losses from predatory pricing.

Rebel appealed, and this court affirmed in part and reversed in part in Rebel Oil Co., Inc. v. Atlantic Richfield Co., 51 F.3d 1421 (9th Cir.), cert. denied, 516 U.S. 987, 116 S.Ct. 515, 133 L.Ed.2d 424 (1995) (“Rebel I”). Rebel I affirmed the district court’s summary judgment in favor of ARCO on the claims of attempted monopolization, because Rebel had not established the existence of a genuine issue of material fact whether ARCO had the market power required to support an attempted monopolization claim. Id. at 1448.

Rebel I reversed the grant of summary judgment dismissing Rebel’s claim that ARCO had practiced “primary line” price discrimination, by “chargftng] predatory, below-cost prices in one [geographical] market in an attempt to eliminate competitors there, and charging] supracompetitive prices in another market.” Id. at 1445. Because a lesser showing of market power was required for a claim of price discrimination than for a claim of attempted monopolization, id. at 1447, the court concluded that Rebel had produced sufficient evidence to create a disputed question of material fact whether ARCO had enough market power to enforce supracompetitive pricing. Id. at 1448. Although Rebel had not shown that ARCO had sufficient market power for an attempted monopolization claim, for price discrimination Rebel needed only to prove that ARCO possessed a level of market power to threaten oligopolization. An oligopoly is “[a] market condition in which sellers are so few that the actions of any one of them will materially affect price and hence have a measurable impact upon competitors.” American Heritage Dictionary of the English Language 916 (1970).

The economic forces at work in an oligopoly are very different than in a monopoly. A predator is able to establish and maintain supracompetitive prices in an oligopoly by making it too painful for its existing competitors to challenge its prices, and thus, “disciplining” them.... Read in the most favorable light, Rebel’s evidence tends to indicate that no new competition can enter the market to challenge ARCO, and that the existing competition, while it may be able to challenge ARCO, lacks the will to do so.

Rebel I, 51 F.3d at 1448. The court remanded to the district court for further proceedings on the price discrimination claim. Id. at 1448.

On remand, the parties conducted discovery on the issue of predatory pricing. ARCO then moved for summary judgment. ARCO simultaneously filed a motion in limine to exclude Rebel’s expert testimony under Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), which the district court granted in part. [ER p. 281]

The district court also granted ARCO’s motion for summary judgment. Rebel Oil Co. v. Atlantic Richfield Co., 957 F.Supp. 1184, 1196-1204 (D.Nev.1997) (“Rebel II”). The court rejected Rebel’s contention that, under the law of the case doctrine, our decision in Rebel I compelled a finding that Rebel had offered sufficient evidence to preclude summary judgment on its price discrimination claim. Id. at 1194-95. The court acknowledged that Rebel I determined that material questions of fact existed regarding both ARCO’s ability to recoup any losses from predatory pricing and Rebel’s damages. The court concluded, however, that none of Rebel’s measures of ARCO’s costs was legally adequate to create an issue of material fact regarding whether ARCO priced its gasoline below its costs. Id. at 1200-03. The court also determined that Rebel had failed to present clear and convincing evidence of ARCO’s predatory intent. Id. at 1204.

ANALYSIS

This court reviews a grant of summary judgment de novo, evaluating the evidence in the light most favorable to the nonmoving party. Rebel I, 51 F.3d at 1432. Summary judgment is appropriate only when the pleadings, affidavits, and other materials in evidence present no genuine issue of material fact. Id.

Section 2(a) of the Clayton Act, as amended by the Robinson-Patman Act, prohibits price discrimination:

It shall be unlawful for any person engaged in commerce, in the course of such commerce, either directly or indirectly, to discriminate in price between different purchasers of commodities of like grade and quality ... where the effect of such discrimination may be substantially to lessen competition or tend to create a monopoly. in any line of commerce, or to injure, destroy, or prevent competition with any person who either grants or knowingly receives the benefit of such discrimination, or with customers of either of them....

15 U.S.C. § 13(a). The statute prohibits “primary-line” price discrimination, when the seller charges predatory, below-cost prices in one geographical market to eliminate competitors there, but charges supracompetitive prices in another market. Rebel I, 51 F.3d at 1445. Rebel alleges primary-line price discrimination, contending that between 1985 and 1989, ARCO charged below-cost prices in the Las Vegas market to eliminate competition there, while charging higher prices in the Los Angeles market.

To prove primary-line price discrimination, Rebel must prove two things: “that the prices complained of are below an appropriate measure of its rival’s costs .... [and] that the competitor had a reasonable prospect ... of recouping its investment in below-cost prices.” Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 222-24, 113 S.Ct. 2578, 125 L.Ed.2d 168 (1993); USA Petroleum Co. v. Atlantic Richfield Co., 13 F.3d 1276, 1277 (9th Cir.1994); Rebel I, 51 F.3d at 1445.

I. Below-Cost Pricing

Rebel presented expert testimony of below-cost pricing, using three different measures of ARCO’s cost in Las Vegas: an exchange agreement between ARCO and Tosco Corporation; the Los Angeles “rack price” for gasoline, adjusted for transportation costs; and profit-and-loss statements from Prestige Stations, Inc. (“PSI”), a wholly-owned subsidiary of ARCO. The parties’ central dispute on appeal is whether Rebel’s evidence established a question of fact whether ARCO priced its Las Vegas gasoline below its costs.

[0]nly below-cost prices should suffice [for recovery for price discrimination], and we have rejected ... the notion that above-cost prices that are below general market levels or the costs of a firm’s competitors inflict injury to competition cognizable under the antitrust laws.... As a general rule, the exclusionary effect of prices above a relevant measure of cost either reflects the lower cost structure of the alleged predator, and so represents competition on the merits, or is beyond the practical ability of a judicial tribunal to control without courting intolerable risks of chilling legitimate price-cutting.

Brooke Group, 509 U.S. at 223, 113 S.Ct. 2578; see USA Petroleum, 13 F.3d at 1285 (“above cost below-market level pricing even when coupled with a structural showing such as recoupment, cannot constitute predatory pricing.”).

Rebel and ARCO agree that prices are not predatory unless they are below a “relevant measure of cost,” but they disagree on whether, as a matter of law, Rebel presented a relevant measure of cost.

Neither the Supreme Court nor this circuit has concluded what would be the appropriate measure of cost in a predatory pricing case. See Brooke Group, 509 U.S. at 222 n. 1, 113 S.Ct. 2578 (“Because the parties in this case agree that the relevant measure of cost is average variable cost, however, we again decline to resolve the conflict among the lower courts over the appropriate measure of cost.”); see Multistate Legal Studies, Inc. v. Harcourt Brace Jovanovich Legal and Profl Publications, Inc., 63 F.3d 1540, 1549 n. 5 (10th Cir.1995) (“Unfortunately for litigants, neither the Supreme Court nor we have taken a position on which of various cost measures is the definitive one, although we have spoken of marginal and average variable costs as being relevant.”), cert. denied, 516 U.S. 1044, 116 S.Ct. 702, 133 L.Ed.2d 659 (1996). It is generally agreed that “the relation between the cost of producing a product and the price charged for it is the criterion for determining whether the price is predatory,” Transamerica Computer Co., Inc. v. International Bus. Machs. Corp., 698 F.2d 1377, 1384 (9th Cir.), cert. denied, 464 U.S. 955, 104 S.Ct. 370, 78 L.Ed.2d 329 (1983).

Economists, however, measure a firm’s costs in a number of ways.... Costs are divided into fixed costs (those that do not vary with changes in output) and variable costs (which do so vary). Total cost is the sum of fixed and variable costs. Marginal cost is the increment to total cost that results from producing an additional unit of output. Average cost, or average total cost, is obtained by dividing total cost by output. Likewise, average variable cost is the sum of all variable costs divided by output. Average cost is thus higher than average variable cost for all output levels.

Id. at 1384-85; see Vollrath Co. v. Sammi Corp., 9 F.3d 1455, 1461 (9th Cir.1993) (plaintiff used average variable cost to argue that prices were predatory), cert. denied, 511 U.S. 1142, 114 S.Ct. 2163, 128 L.Ed.2d 886 (1994); 3 P. Areeda & H. Hovenkamp, Antitrust Law §§ 739-40 (1996) (“marginal-cost pricing is the economically sound division between acceptable competitive behavior and unlawful, ‘below-cost’ predation,” but because marginal costs are usually difficult to measure, average variable cost may be a substitute test).

We must determine whether Rebel presented evidence of any of these cost measures that creates a question of material fact on its predatory pricing claim.

A. Law of the Case

Rebel’s first argument is that this panel does not even need to decide the cost issue, because it was decided by the panel in Rebel I, and the law of the case governs all further proceedings. Rebel I described expert testimony offered by Rebel: “The expert concluded that ARCO’s retail prices in Las Vegas were consistently below the wholesale prices of all other wholesale suppliers in Las Vegas, and at times were 10 cents or more per gallon below ARCO’s marginal cost.” Rebel I, 51 F.3d at 1431. Rebel points to footnote 1 of Rebel I, which immediately follows the foregoing quotation, and states:

Marginal cost is the cost that a firm incurs in the production of one additional unit of output. Herbert Hovenkamp, Economics and Federal Antitrust Law 10 (1985). Rebel measured ARCO’s marginal cost based upon an agreement that ARCO signed with Tosco Corp. to purchase incremental supplies of gasoline. Rebel claims that ARCO bought the additional supplies of gasoline from Tosco because it'was expanding output but was unable to obtain the additional output from its own refineries. The cost of this “last barrel” of gasoline represents ARCO’s marginal cost for gasoline.

Id. at 1431 n. 1 (emphasis added). Rebel claims that this footnote and the highlighted sentence conclusively establish that the Tosco agreement presents a material issue of fact regarding below-cost pricing. Rebel also points to the following statement in Rebel I as establishing that another potential measure of marginal cost (the “rack price”) presented a material question of fact:

Rebel’s claim alleged primary-line discrimination. As proof, Rebel’s expert relied on pricing data showing that for four years, between 1985 and 1989, ARCO charged predatory prices in Las Vegas that, when adjusted for transportation costs, were 6 to 14 cents below those charged in Los Ange-les.

Id. at 1445.

Under the doctrine of “law of the case,”

a court is generally precluded from reconsidering an issue that has already been decided by the same court, or a higher court in the identical ease. The doctrine is not a limitation of a tribunal’s power, but rather a guide to discretion. A court may have discretion to depart from the law of the case where ... the evidence on remand is substantially different_ Failure to apply the doctrine of the law of the case absent one of the requisite conditions constitutes an abuse of discretion.

United States v. Alexander, 106 F.3d 874, 876 (9th Cir.1997) (citations and quotations omitted). “For the doctrine to apply, the issue in question must have been decided explicitly or by necessary implication in [the] previous disposition. A significant corollary to the doctrine is that dicta have no preclu-sive effect.” Milgard Tempering, Inc. v. Selas Corp., 902 F.2d 703, 715 (9th Cir.1990) (citations and quotations omitted).

We conclude that the two sections of Rebel I cited as establishing the law of the case were not explicit decisions on below-cost pricing, and therefore do not constitute a basis for the application of the law of the case doctrine. First, as the district court noted in Rebel II, “[b]elow cost pricing was not an issue before the Ninth Circuit Court of Appeals in Rebel I.” Rebel II, 957 F.Supp. at 1196 n. 7. The district court had bifurcated discovery, and “[t]here was no discovery on predatory pricing, intent and collusion.” Rebel I, 51 F.3d at 1432. Rebel I reviewed a district court decision regarding market power. That decision did not evaluate Rebel’s proposed measures of ARCO’s costs, which related not to market power but to whether ARCO’s prices were predatory. Second, as discussed below, the issue of the appropriate measure of cost is a thorny one, and the two brief references in Rebel I to the Tosco agreement and to “rack price” in Los Ange-les hardly “decided explicitly” that either measure was sufficient to withstand ARCO’s motion for summary judgment. Both statements are better read as descriptions rather than dispositions of Rebel’s claims. “As the issue of below cost pricing was not before the court in Rebel I, any dicta addressing cost measurements is not law of the case or binding upon this Court.” Rebel II, 957 F.Supp. at 1201.¹

B. The Tosco Agreement

As evidence to show that “the prices complained of [were] below an appropriate measure of [ARCO’s] costs,” Brooke Group, 509 U.S. at 210, 113 S.Ct. 2578, Rebel presented a 1986 “Exchange Agreement” between ARCO and Tosco Corporation. [SER pp. 555-84] Because “ARCO is willing to exchange ANS [Aaska North Slope Crude Oil] for gasoline; and ... Tosco is willing to exchange gasoline for ANS,” the parties entered into a “buy/sell exchange of ANS for gasoline,” [SER p. 555], providing that ARCO would deliver ANS to Tosco, and Tosco would deliver gasoline to ARCO. The quantities to be exchanged were established by a formula set forth in the agreement. The agreement also provided for cash payments by ARCO to Tosco and by Tosco to ARCO when crude oil or gasoline was delivered, and “the dollar amounts paid by each party will equal those paid by the other.” [SER p. 567] The cash payments were tied to the market price of ANS. [SER pp. 563, 584]

Rebel argues there is at least a question of material fact whether the agreement is not an exchange but an agreement to purchase, and thus is a direct measure of ARCO’s marginal cost of producing an additional unit of gasoline. This circuit has concluded that a swap of like products is not a “sale” for the purposes of the Robinson-Patman Act. Airweld, Inc. v. Airco, Inc., 742 F.2d 1184, 1191 (9th Cir.1984), cert. denied, 469 U.S. 1213, 105 S.Ct. 1184, 84 L.Ed.2d 381 (1985); see American Oil Co. v. McMullin, 508 F.2d 1345, 1353 (10th Cir.1975) (where oil company supplies petroleum products from its refinery in Salt Lake City to other oil companies, in return for petroleum products from the other companies’ refineries in other areas, “[t]he record does not show that these transactions are actually sales.”). Because crude oil and gasoline, although related products, are not “fungible,” the Tosco Agreement does not document an “exchange of the same product” or “like types of gasoline.” Airweld, 742 F.2d at 1192. The agreement is therefore not a “swap,” and we must examine ARCO’s marginal cost.

ARCO submitted a declaration stating:

The Exchange Agreement provided for cash payments by Tosco to ARCO, when crude was delivered, and by ARCO to Tosco, when gasoline was accepted, but those payments were security arrangements, and were in equal amounts. In substance, the Agreement was an in-kind transaction in which crude oil was exchanged for gasoline.

[SER p. 455 n. 7] ARCO’s declaration also stated: “ARCO’s cost for the gasoline delivered by Tosco was equal to ARCO’s cost to produce and deliver the crude oil to Tosco since there were no net cash payments.” [SER p. 455]

Rebel’s expert witness, Dr. Keith Leffler, testified that

In my Report I state that “[t]he gasoline received by ARCO from Tosco represents an increment to the supply of ARCO gasoline. The effective economic costs of the gasoline to ARCO thus indicates the marginal cost of ARCO gasoline.” The calculation of the “economic cost” in'what is effectively a barter situation of course requires determination of the value of what ARCO gives up. Regardless of whether ARCO bought the crude oil, produced the crude oil or was given the crude oil, the marginal costs to ARCO of obtaining the additional gasoline from Tosco is given by the value of that crude oil that they gave up.

[ER p. 194]. Leffler’s statement echoes ARCO’s labelling of the agreement as an in-kind transaction, but his cost measurement is different. ARCO states that the cost to it was the cost of producing and delivering the crude oil; Leffler states that the cost is instead the market price of the crude oil, i.e. what ARCO would have received for the crude oil if ARCO had sold it rather than traded it to Tosco.

The measure of cost proposed by Rebel-the market price of the crude oil-is problematic for two reasons. First, the market price for crude oil is not what ARCO spent in obtaining the crude oil: ARCO produced the crude oil, and did not purchase the crude oil on the market. Second, the market price represents only what ARCO would have received had ARCO sold the crude oil in the market rather than trading it to Tosco for gasoline. The measure of marginal cost proposed by Rebel is thus really the opportunity cost to ARCO of choosing to enter into the exchange agreement rather than selling the crude oil elsewhere. Opportunity costs are vastly different from ARCO’s marginal or variable costs, and we agree that “the use of the concept of opportunity costs [to show predatory pricing] must be held improper as a matter of law.” In re IBM Peripheral EDP Devices Antitrust Litig., 459 F.Supp. 626, 631 (N.D.Cal.1978); Continental Airlines, Inc. v. American Airlines, Inc., 824 F.Supp. 689, 701 (S.D.Tex.1993) (opportunity cost is not one of the costs of implementing a particular business choice, and cannot be included in considering average variable price).

Rebel’s expert, Dr. Leffler, admitted that “it’s accepted that ARCO was a low cost producer of crude in this era so, if anything, I would expect it to have lower cost than what was reflected in the market value of that barrel.” [SER p. 202] Aside from this admission that ARCO’s cost of producing the crude oil was low, Rebel presented no evidence of the actual costs of production. Nevertheless, Leffler argued for the use of market price as a measure of ARCO’s cost, taking strong issue with the “courts’ language in rejecting ... the economic concept of opportunity costs for measuring below cost sales in predatory pricing cases.” [ER p. 193 n. 5] Because Rebel offered “evidence” (Leffler’s expert opinion) in its opposition to ARCO’s summary judgment motion, and ARCO did not offer conflicting expert opinion, Rebel argues that “the district court acted as a finder of fact,” impermissibly weighing Rebel’s evidence and effectively conducting a bench trial. But where, as here, the expert offers an opinion that courts have rejected as a matter of law, that opinion is not enough to create an issue of material fact that survives summary judgment. “In the context of antitrust law, if there are undisputed facts about the structure of the market that render the inference [to be drawn from expert affidavits] economically unreasonable, the expert opinion is insufficient to support a jury verdict.” Rebel I, 51 F.3d at 1435-36 (citing Eastman Kodak Co. v. Image Technical Serv., Inc., 504 U.S. 451, 468-69, 112 S.Ct. 2072, 119 L.Ed.2d 265 (1992)). “Expert testimony is useful as a guide to interpreting market facts, but it is not a substitute for them.” Brooke Group, 509 U.S. at 242, 113 S.Ct. 2578; Wallace v. Bank of Bartlett, 55 F.3d 1166, 1170 (6th Cir.1995), cert. denied, 516 U.S. 1047, 116 S.Ct. 709, 133 L.Ed.2d 664 (1996).

We conclude that using the Tosco agreement as a measure of ARCO’s cost in the way urged by Rebel would require equating ARCO’s cost of the crude oil to market price, which is counter to legal definitions of cost. Therefore, the agreement cannot provide a legally sufficient measure of ARCO’s cost.

C. The Los Angeles “Rack Price”

Rebel alleges that ARCO’s wholesale prices in Las Vegas were below the wholesale “rack price” in Los Angeles, and were therefore predatory.

“Rack prices” are the prices charged by marketers for gasoline deliveries “at the rack,” sales in small tanker-truck quantities at loading-rack distribution points. Leffler acknowledged that “ARCO doesn’t post a rack price or make sales at a rack price,” [SER p. 219] but stated that the cost of buying gasoline at the Los Angeles rack plus transportation costs (over the Cal-Nev pipeline) equalled the cost an efficient seller would face in Las Vegas. Rebel contends that this rack price is an adequate proxy for ARCO’s marginal cost, and because ARCO sold wholesale gasoline in Las Vegas below that price/cost, ARCO’s price was predatory.

The district court concluded that Rebel’s comparison of wholesale prices in Las Vegas and Los Angeles showed only that ARCO charged a lower price in Las Vegas. Because “Rebel has made no showing of ARCO’s actual costs of producing gasoline,” however, Rebel still had not created a factual issue regarding whether the Las Vegas price was below ARCO’s costs. Rebel II, 957 F.Supp. at 1203-04.

Using the rack price as a measure of cost would substitute a market price for ARCO’s actual costs. We affirm the district court’s conclusion that the evidence of rack price was insufficient to create a material question of fact whether ARCO’s Las Vegas prices were predatory.

D. Losses at PSI Outlets

Another Rebel witness, Harry Hol-combe, reviewed the income statements of ARCO’s PSI stations and concluded that almost all the stations lost money during the alleged predatory period. Because ARCO sold PSI wholesale gas and PSI stations lost money, Holcombe stated that this demonstrated that ARCO was willing to lose money by pricing gas at its PSI outlets below the outlets’ cost of obtaining the gas.

There are a number of reasons why this does not present a triable issue. First, the district court excluded Holcombe’s testimony regarding “economic measurements and concepts such as the measurement of below marginal or average variable cost selling, recoupment, and data supporting price discrimination.” [ER p. 274] His testimony was restricted to general industry practices. Nevertheless, the district court in Rebel II chose to address his testimony so as to address the PSI method of determining cost. Rebel II, 957 F.Supp. at 1203 n. 20.

Second, PSI is a wholly-owned subsidiary of ARCO, and Holcombe stated that PSI had no independent existence. Under these circumstances, the price paid by PSI to ARCO is a transfer price, not a sale. Brown v. Hansen Publications, Inc., 556 F.2d 969, 971 (9th Cir.1977). If PSI and ARCO are a combined unit, the costs of ARCO and PSI together must be considered, not simply PSI’s costs versus its sales. Vollrath, 9 F.3d at 1460-61 (if entities are considered a combined unit, “then the appropriate consideration is the costs of the combined group to determine if the combined group sold below its costs”); Caribe BMW, Inc. v. Bayerische Motaren Werke Aktiengesellschaft, 19 F.3d 745, 749 (1st Cir.1994) (100% ownership makes firm and subsidiary single person under Robinson-Patman Act). Finally, income statements such as those analyzed by Holcombe may show that PSI stations lost money, but such statements alone are not sufficient to show that below-cost sales were made. Janich Bros., Inc. v. American Distilling Co., 570 F.2d 848, 858 (9th Cir.1977) (as amended), cert. denied, 439 U.S. 829, 99 S.Ct. 103, 58 L.Ed.2d 122 (1978) (although conventional business records do provide information to use in calculating costs, the inclusion of other nonvaria-ble costs in conventional records, such as depreciation, taxes, and licenses, means that “a firm’s marginal costs cannot be ascertained from conventional business records”).

II. Daubert Ruling

In its ruling on Rebel’s expert witness testimony, a week before its summary judgment ruling, the district court held that Leffler’s opinions were admissible under Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579,'113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). Rebel contends that the later grant of summary judgment in ARCO’s favor and the district court’s rejection of Leffler’s conclusions is irreconcilable with the admission of Leffler’s testimony.

“The test [for the admission of expert testimony] under Daubert is not the correctness of the expert’s conclusions but the soundness of his methodology.” Daubert v. Merrell Dow Pharms., Inc., 43 F.3d 1311, 1318 (9th Cir.), cert. denied, 516 U.S. 869, 116 S.Ct. 189, 133 L.Ed.2d 126 (1995) (on remand from Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993)). As the district court’s Daubert order explicitly stated, “Whether Leffler’s cost analysis is an economically valid method is [a] different question from whether it is legally sufficient to support a price discrimination claim.” [ER p. 275] The district court’s conclusion at summary judgment that Leffler’s testimony was not legally sufficient evidence to create a question of material fact regarding whether ARCO priced its gasoline below its costs is not inconsistent with its conclusion following the Daubert healing that Leffler’s methodology was sound. An expert witness may be qualified to testify even though the expert’s conclusions are legally incorrect. See General Elec. Co. v. Joiner, — U.S. -, -, 118 S.Ct. 512, 519, 139 L.Ed.2d 508 (1997) (district courts may reject expert testimony that is based on sound methodology when “there is simply too great an analytical gap between the data and the opinion proffered”).

The Daubert ruling did not require the district court to deny ARCO’s motion for summary judgment.

III. Predatory Intent

Rebel complains that the district court sua sponte raised the issue whether ARCO had predatory intent and then decided that Rebel had not produced evidence that would allow a jury to find “clear and convincing evidence” of that intent. See Rebel II, 957 F.Supp. at 1203-04. Rebel claims this was improper because it had no notice that intent was in issue because ARCO did not address intent in its motion, and Rebel therefore had no opportunity to address the issue. Because we conclude that a showing of predatory intent is not sufficient in the absence of below-cost pricing to establish an antitrust violation, we do not decide this issue.

In Brooke Group, the Supreme Court made it clear that predatory intent alone, no matter how clear and convincing the evidence, cannot substitute for a plaintiffs failure to demonstrate a reasonable prospect of recoupment. See 509 U.S. at 222-23 (below-cost pricing and reasonable prospect of re-coupment required). In reviewing the sufficiency of the evidence, the Court stated:

the record contains sufficient evidence from which a reasonable jury could conclude that Brown & Williamson envisioned or intended this anticompetitive course of events. There is also sufficient evidence in the record from which a reasonable jury could conclude that for a period of approximately 18 months, Brown & Williamson’s prices on its generic cigarettes were below its costs, and that this below-cost pricing imposed losses on Liggett.... Liggett has failed to demonstrate ... that in pursuing this scheme, Brown & Williamson had a reasonable prospect of recovering its losses from below-cost pricing through slowing the growth of generics....

No inference of recoupment is sustainable on this record, because no evidence suggests that Brown & Williamson-w/jaiewer its intent in introducing black-and-whites [generics] may have been-was likely to obtain the power to raise the prices for generic cigarettes above a competitive level.

Id. at 231-32, 113 S.Ct. 2578 (citations and quotations omitted) (emphasis added).

If intent cannot substitute for the required element of recoupment, it likewise cannot establish an antitrust violation in the absence of evidence of below-cost pricing. Rebel did not establish a question of material fact regarding ARCO’s below-cost pricing. The district court’s discussion of intent was therefore unnecessary, and we need not discuss whether it was erroneous.²

AFFIRMED.

1 . In Areeda and Hovenkamp's, Antitrust Law, supra, § 740, the authors describe footnote 1 of Rebel I (which stated "The cost of this ‘last barrel’ of gasoline [in the Tosco agreement, discussed supra] represents ARCO's marginal cost for gasoline.”) as follows:

In the Ninth Circuit's Rebel Oil case the plaintiff alleged that the defendant was unable to supply gasoline at the predatory price from its own refineries, so it contracted to purchase more from an independent supplier, and paid more than it charged upon resale. If so, the resulting price was clearly below marginal cost. Further, as the court noted, that would be true of all sales made at that price, not only for the sales of the gasoline bought at a high price. "Marginal” cost does not look at the average cost, but the incremental cost of the last unit produced, which in this case was the higher cost units procured from the independent supplier.

Id. at p. 378 (emphases added) (footnotes omitted). Rebel claims that this shows that footnote 1 established the law of the case. We disagree. Rebel's selection from Antitrust Law states only what Rebel alleged: that the Tosco agreement represented a "purchase" for which ARCO "paid,” and that "if so,” the agreement would serve as a measure of marginal cost.

2 . Because we affirm the grant of summary judgment on below-cost pricing, we do not address ARCO's challenge to the district court’s decision on recoupment, nor do we address Rebel’s argument regarding the measure of its damages.