concurring and dissenting:
To understand this case one must begin with a description of the market as established by the evidence at trial.
THE MARKET
Geography. A given bakery can provide fresh bread for only a relatively small area, covering a radius of no more than 200 miles. See In the Matter of International Tel. & Tel. Corp., 104 F.T.C. 280, 409 (1984). The San Francisco area, the Lake Tahoe area, and the Sacramento area were *1346the principal loci of competition of the bakers involved here. Northern California is the relevant geographic area.
The Product. The product is bread— bread eventually to be consumed by the eaters of bread. Arguably all breads are interchangeable, and variety and heath-baked breads such as whole wheat, rye, pumpernickel, and raisin are part of the same market as white pan bread. However, this case was tried on the basis that the white pan loaf was the product, and this single kind of bread must be the focus of decision.
The white pan loaf was produced by bakeries in three ways: (1) advertised bread, as to which there was a brand name that was publicly advertised; (2) private label bread, bread that was not advertised under a brand name but sold by individual retailers as their bread, although it had been purchased from wholesalers; and (3) bread from the “captive bakeries,” that is, bakeries maintained by major grocery chains.
Inglis has maintained that the market is further limited to the white pan loaf sold by wholesalers to retailers as “private label.” Inglis’ expert, Professor Wheeler, testified that such a submarket existed. His testimony is contrary to the accepted legal definition of a submarket. A sub-market can be determined by examining “industry or public recognition of the sub-market as a separate economic entity, the product’s peculiar characteristics and uses, unique production facilities, distinct customers, distinct prices, sensitivity to price changes, and specialized vendors.” Brown Shoe Co. Inc. v. United States, 370 U.S. 294, 325, 82 S.Ct. 1502, 1524, 8 L.Ed.2d 510 (1962). The private label market is not recognized as a separate economic entity. The bread sold in it does not have any peculiar characteristics and uses. Unique production facilities are not employed to produce the bread. The customers are not distinct and there are no specialized vendors. Price changes are sensitive not only to the price for private label but to the cost of captive bakery bread. In short, there is only a single characteristic that is peculiar to private label bread: distinct prices. That is not enough to establish a submark-et.
“Private label” bread is no different in kind or quality from advertised or “captive bakery bread.” In each of these categories, there is simply white pan bread. Chemically and physically, it is the same thing. The only difference is the price at which the bread is sold. Where the product is the same, a difference in its pricing does not create a different product or a different market. A market is “composed of products that have reasonable interchangeability for the purposes for which they are produced.” United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377, 404, 76 S.Ct. 994, 1012, 100 L.Ed. 1264 (1956). This definition is long-established and still current. State of California v. American Stores Co., 872 F.2d 837, 841 (9th Cir.1989) rev’d on other grounds, 495 U.S. 271, 110 S.Ct. 1853, 109 L.Ed.2d 240 (1990). One pound white loaves of bread are produced to eat. The three kinds of bread were interchangeable. The market here was the market for these breads in Northern California.
The Competitors. As of 1970 when this case began, the following bakeries were competitors in this market:
Campbell-Taggart, a national company, which sold advertised “Kilpatrick” and “Rainbo” brands and private label bread. In the period 1970 to 1975 it operated six bakeries in the relevant area and made from 41 to 52 percent of the wholesalers’ sales.
American, which operated the largest bakery in the market and sold advertised bread chiefly under the brand name “Langendorf” and private label bread. Until 1973 it was the second largest wholesaler in the market.
Continental, the largest national wholesale bakery in the United States, but only the third largest wholesaler in the relevant market. It sold advertised bread under the label “Wonder” and private label bread. From 1970 through 1975 it made 22 percent of the wholesalers’ sales.
Inglis, the plaintiff in this case. It operated one bakery, located in Stockton, and sold advertised bread under the Sun*1347beam label and private label bread. In 1969, prior to the events in this lawsuit, it declared bankruptcy, but apparently emerged to continue business through 1975. Beginning in 1965 until 1975, In-glis lost money every year except 1972 when, for part of the period, Campbell-Taggart and Continental were closed by a strike. Since 1966 Inglis did not reinvest enough capital to cover depreciation. Because of its financial difficulties Inglis had effectively stopped advertising its Sunbeam brand by the mid-1960s. In 1975, because of irregularities in Inglis’ books, the company financing it withdrew its support. Inglis made one last attempt to secure financing from the Teamsters Union and then folded. This suit, now in its twentieth year, appears to be its chief asset.
Safeway, Lucky and Alpha Beta. These three large grocery chains baked and sold about half the bread in the white pan bread market.
Interstate Bread Company, a national bakery selling advertised bread under the Blue Seal label.
Welsh’s Baking Co., a regional company selling advertised bread under the Sheepherder label.
Other wholesalers: Cottage, Home Craft, Modern, Nielson.
In summary, there were five national and five regional wholesale bakeries and three large grocery chains in the Northern California market. It is undisputed that these bakers had excess capacity for the production of bread. There was ample evidence that the price of the bread was dominated by the three chains which produced their own bread and could produce more of it if the price went up beyond the price they preferred.
THE SHERMAN ACT CLAIM OF ATTEMPTED MONOPOLIZATION
Inglis failed to prove this claim, and Continental is entitled to judgment notwithstanding the verdict for three reasons:
First. Although Professor Wheeler testified that there was such a thing as a market in private label white pan bread, the evidence adduced at trial was overwhelmingly to the contrary, as the above description of the market shows. It is the plaintiffs burden to establish the relevant market. R.C. Dick Geothermal Corp. v. Thermogenics, Inc., 890 F.2d 139, 143 (9th Cir.1989) (en banc). The proper market is a question of fact and a jury finding will be reversed if clearly erroneous. Twin City Sportservice, Inc. v. Charles O. Finley & Company, 676 F.2d 1291, 1299 (9th Cir.), cert. denied 459 U.S. 1009, 103 S.Ct. 364, 74 L.Ed.2d 400 (1982).
Inglis’ failure to prove a special market in private label white bread ends its case. It has offered no proof of predation in the sale of advertised bread. Yet sales of private label and advertised bread must be aggregated if a case in the relevant market is to be made. See In the Matter of International Tel. & Tel. Corp., 104 F.T.C. at 435. The “ultimate standard” is whether the alleged predation created “a market structure enabling the seller to recoup his losses.” William Inglis & Sons Baking Co. v. I.T.T. Continental Baking Co., Inc., 668 F.2d 1014, 1035 (9th Cir.1981) (Inglis I), cert. denied, 459 U.S. 825, 103 S.Ct. 57, 74 L.Ed.2d 61 (1982). Nothing has been proved by Inglis as to a market structure that includes advertised white bread.
Second. As a distinct and independent ground for decision, it may be accepted arguendo that the market contended for by Inglis was the private label white pan bread sold by the wholesalers, but the claim that Continental had a predatory strategy is implausible.
Four characteristics of this market are of capital importance for this case: (1) Continental’s sales in the relevant period were no more than half that of the largest wholesale baker in the market, Campbell-Taggart. (2) Inglis was a small and weak competitor that had declared bankruptcy before the conduct complained of began. (3) The grocery chains played a key role in the pricing of private label bread because if the price rose the chains could use their captives. (4) All of the wholesalers had substantial excess baking capacity.
As evidence of a predatory scheme, In-glis points to the testimony of Professor Wheeler that in “an oligopolistic or duopo-*1348listic market with barriers to entry” a firm would raise prices because the higher prices would “be beneficial to all participants in the market.” This testimony is not helpful to Inglis’ case. The fact that all participants would benefit would mean that Inglis would benefit too; the market strategy would not have a negative impact on Inglis.
Inglis also offered Professor Wheeler’s testimony that a rational businessman might engage in predatory pricing as part of “a long-term market strategy to achieve market penetration and eliminate competitors.” Such a possibility theoretically exists. But to testify that the possibility exists is not to testify that a rational businessman in Continental’s shoes would so have behaved in the Northern California white pan bread market. This piece of testimony proves nothing concretely.
Entry barrier is irrelevant where excess capacity exists. Competition cannot be eliminated or market discipline imposed where major consumers possess their own alternative ways of getting the product. Even if Professor Wheeler’s testimony is believed that there were entry barriers for new bakers, a predatory strategy would have been unproductive for Continental because it would have opened the field to its larger competitor, Campbell-Taggart, to use its excess capacity, and because the chains would have continued to hold the key to price.
“Absent some assurance that the hoped-for monopoly will materialize, and that it can be sustained for a significant period of time,” a predator would embark on a highly speculative and economically irrational scheme. Matsushita Elec. Indust. Co. v. Zenith Radio Corp., 475 U.S. 574, 589, 106 S.Ct. 1348, 1357, 89 L.Ed.2d 538 (1986) (emphasis in original). It is for that reason that “predatory pricing schemes are rarely tried, and even more rarely successful.” Id. In the highly competitive white pan bread market, Continental had no hope of achieving monopoly power and no prospect of sustaining it.
The ultimate purpose of the antitrust laws, federal and state, is to benefit the consumer. Such is “the modern conception.” American Academic Suppliers, Inc. v. Beckley-Cardy, Inc., 922 F.2d 1317, 1319 (7th Cir.1991). “[Mistaken inferences” about predation “are especially costly, because they chill the very conduct the antitrust laws are designed to protect.” Matsushita, 475 U.S. at 594, 106 S.Ct. at 1360. Here a case that has gone on twenty years has ended in a multimillion dollar verdict imposed on a baker because its price for bread was too low! To sustain the verdict would violate federal antitrust law as it is now authoritatively interpreted.
Third. As Judge Canby’s opinion demonstrates, Inglis completely failed to show that Continental acted with anticompetitive intent, that is, as he puts it, that it “specifically intended to control prices or destroy competition.” In Judge Canby’s words, “Inglis’ principal evidence of intent was also the centerpiece of its entire case, a study prepared by the accounting firm of Ernst & Whinney.” The “methodology of the E & W study” was, as he says, “to include expenses that Continental may have incurred regardless of its private label bread production.” As a consequence, the study “failed to identify the unique cost of producing private label bread, and consequently failed to show Continental’s prices were below that cost.”
INGLIS’ ROBINSON-PATMAN ACT CLAIM
As Judge Canby has concluded, the failure of the E & W study to show intent to monopolize is also failure to show evidence of intent to discriminate in prices in order to lessen competition or create a monopoly. As Judge Canby has also concluded, Inglis’ market analysis “failed to show the substantial possibility of injury to competition.” Again Continental is entitled to judgment notwithstanding the verdict.
THE CALIFORNIA UNFAIR TRADE PRACTICES CLAIM
The Law of the Case. Inglis I established there was a single difference be*1349tween the Sherman Act and California antitrust law:
The only real difference between the California statute and federal law is in the allocation of evidentiary burdens. Assuming proof of injury to a competitor has been made, California law allows plaintiffs to establish a prima facie case with proof of prices below average total cost. The defendant then has the burden of negating the inference of illegal intent or of establishing an affirmative defense. Under federal law, a plaintiff who relies on prices below average total cost, but above average variable cost, has not created a presumption of illegal intent.
Inglis I, 668 F.2d at 1049.
Unless there is “manifest injustice,” the law of the case binds us. See United States v. Miller, 822 F.2d 828, 832 (9th Cir.1987). No manifest injustice has been shown in adhering to what we earlier laid down as the governing law. To the contrary, injustice is done when a case which was tried twice, and considered here once as a federal case, should in the end be decided by an unsupported theory of what state law might be. That this case is still conceived of by the majority as a federal case is indicated by its citation of federal authority on damages said to be awardable under California law.
In particular, in a way that is fatal to Inglis’ case, we are committed to the following standard as to what constitutes predatory pricing:
Predation exists when the justification of these prices is based, not on their effectiveness in minimizing losses, but on their tendency to eliminate rivals and create a market structure enabling the seller to recoup his losses. This is the ultimate standard, and not rigid adherence to a particular cost-based rule, that must govern our analysis of alleged predatory pricing.
Guided by these principles, we hold that to establish predatory pricing a plaintiff must prove that the anticipated benefits of defendant’s price depended on its tendency to discipline or eliminate competition and thereby enhance the firm’s long-term ability to reap the benefits of monopoly power.
Inglis I, 668 F.2d at 1035.
This standard, enunciated in relation to the Sherman Act, has, under the law of the case, equal application to the California antitrust law. It is clearly a standard different from older decisions that treated predation as something that could be proved apart from a relevant market, e.g., Blanton v. Mobil Oil Corp., 721 F.2d 1207, 1214 (9th Cir.1983), cert. denied, 471 U.S. 1007, 105 S.Ct. 1874, 85 L.Ed.2d 166 (1985). The change is consistent with modern Supreme Court precedent, e.g., Cargill, Inc. v. Monfort of Colorado, Inc., 479 U.S. 104, 119-20 n. 15, 107 S.Ct. 484, n. 15, 93 L.Ed.2d 427 (1986).
As Inglis I makes explicit, predatory pricing can be proved only by considering its effect on “a market structure.” Predatory pricing does not take place in a void. It is predatory only if it is designed to affect the market structure. To determine whether Continental’s pricing here was predatory, we must first determine what the market was that it is charged with affecting.
As the earlier description of the market demonstrated, Inglis attempted to show an impact on a nonexistent private label white pan bread market. No impact was shown on the actual market in white pan bread. Inglis’ failure of proof in this respect is a failure to prove predatory pricing.
The California Law. The foregoing has proceeded on the assumption that we are bound by the law of the case set out in Inglis I. But there is no inconsistency between California law and the law of the case. Inglis I merely took up a question which California had not explicitly addressed. Already, the language of Inglis I on the difference between the two statutes consisting in the burden of proof is being cited as authoritative in the state courts. Turnbull & Turnbull v. ARA Transp., Inc., 219 Cal.App.3d 811, 825, 268 Cal.Rptr. 856, 864 (1990).
Contrary to the wooden interpretation of a single section of the statute that the majority has adopted here, California has been careful to explain that the statute has *1350a purpose as a whole: “Throughout the Act the Legislature has manifested its intent to discourage practices which injure the seller’s competitors (§§ 17040, 17043, 17045, 17071) and thereby tend to create the monopolies condemned by section 17001 (ante, fn. 3). Equally apparent is the Legislature’s concern to allow the seller to meet in good faith the prices of his competitors (§§ 17040, 17050), thereby fostering the competition promoted by section 17001.” Harris v. Capitol Records Distributing Corp., 64 Cal.2d 454, 50 Cal.Rptr. 539, 544, 413 P.2d 139 (1966). The purpose is to prevent the creation of monopolies. It is for that reason that the statute does not put into play the presumption on which the majority relies simply by proof of sales below cost but only when that proof is joined “with proof of the injurious effect of such acts.” Cal.Bus. & Prof.Code § 17071 (West 1987). There has been no proof of the injurious effects of the acts attributed to Continental.
For the same reason, the legislature made it easy to defeat the presumption, and the Supreme Court of California has made it even easier. The presumption does not apply to any sale made “[i]n an endeav- or made in good faith to meet the legal prices of a competitor selling the same article or product, in the same locality or trade area and in the ordinary channels of trade.” Id. at § 17050(d).
The standard set for how a defendant may rebut the presumption created by the statute is as follows:
The obvious and only effect of this provision is to require the defendants to go forward with such proof as would bring them within one of the exceptions or which would negative the prima facie showing of wrongful intent. They may present facts showing that they were within the express exceptions regardless of actual intent; or they may introduce evidence of another necessity not expressly included to show that sales were made in good faith and not for the purpose of injuring competitors or destroying competition.
People v. Pay Less Drug Store, 25 Cal.2d 108, 114, 153 P.2d 9 (1944).
The subsequent application of Pay Less has established that testimony by the seller as to his intention to meet competition and to maintain and hopefully expand his sales by selling below cost is sufficient to rebut the statutory presumption of an intent to injure competitors or to destroy competition. Dooley’s Hardware Mart v. Food Giant Markets, Inc., 21 Cal.App.3d 513, 98 Cal.Rptr. 543 (1971); Ellis v. Dallas, 113 Cal.App.2d 234, 248 P.2d 63 (1952); Sandler v. Gordon, 94 Cal.App.2d 254, 210 P.2d 314 (1949). The simple denial of the seller of his intent to injure competition has sufficed as evidence to rebut the presumption. Tri-Q, Inc. v. Sta-Hi Corp., 63 Cal.2d 199, 45 Cal.Rptr. 878, 883, 404 P.2d 486, 491 (1965).
The Supreme Court of California has not decided the precise issue of recoupment. Nonetheless, the liberal decisions of the supreme and appellate courts of California permitting easy refutation of the presumption are data not to be disregarded unless we are convinced by other persuasive data that the California Supreme Court would decide otherwise. Dimidowich v. Bell and Howell, 803 F.2d 1473, 1482 (9th Cir.1986) modified on other grounds, 810 F.2d 1517 (1987). The data we have all point in the same direction.
Our basic task “when we decide a claim that involves a novel question of state law” is to “try to predict how the highest state court would decide the issue.” Hal Roach Studios, Inc. v. Richard Feiner & Co., Inc., 896 F.2d 1542, 1548 (9th Cir.1990). In making such a prediction we “may also look to ‘well reasoned-decisions from other jurisdictions.’ ” Id. (citation omitted). It seems reasonably clear that an informed prediction as to how the present Supreme Court of California would decide this case would recognize that the California Supreme Court would take into account modern federal law and decide this case as United States Supreme Court precedent dictates federal law would decide it.
In the present case Inglis’ own expert testified that because of the excess capaci*1351ty in the market Continental had to sell below its average cost. Counsel for Inglis conceded the point. Once it was established that Continental had to sell below average cost in order to sell at all, the statutory presumption disappeared. Once the statutory presumption disappeared, In-glis had no evidence that Continental was selling below cost for the purpose of injuring it.
Judge Canby’s opinion says that Continental could have been sold at a price that was below cost but still higher than it did sell. This contention appears to be speculation.
There is a final defect with Inglis’ claim. The presumption on which it relies does not come into effect simply by proof of sales below cost. The presumption comes into play only if sales below cost are joined “with proof of the injurious effect of such acts.” Cal. Bus. & Prof.Code § 17071 (West 1987 & 1991 Supp.). In a market in which there was excess capacity no seller was going to be able to sell their bread except below cost. There has been no showing that the price Continental chose to sell at caused any injury to Inglis.
DAMAGES
On the foregoing analysis, the question of damages is not reached. However, as it has been reached by the court, the damages awarded Inglis must be considered. On this issue alone, Continental is entitled to judgment notwithstanding the verdict.
The Lost Profits. The E & W study, on which Inglis relied, showed that in the second quarter of 1970 Continental’s price for private label bread was 3.1 cents above costs, averaged over a time span of up to seven years, that were not uniquely related to private label bread. As the time span that was used increased so did the number of costs that were thrown into the computation. There is no basis in the evidence for believing that the cost figures used by E & W correctly measured Continental’s costs under California law.
Assuming, however, that in the second quarter of 1970 there was a 3.1 cent spread between these averaged costs and the price, E & W calculated what Continental would have charged for its private label bread if it had maintained the same spread from this period until early 1976. E & W then projected the profit that Inglis would have made had it too been able to sell bread at this hypothetical price. The result was a lost profits figure of $2 million. This methodology made all of Inglis’ lost profits turn on a hypothetical price at which Continental could have sold its private label bread.
On appeal, Continental does not contend that Inglis, to prevail on this issue, must show that all of its losses were attributed to Continental’s failure to sell at the hypothetical price. Continental’s position is that for Inglis to recover any damages, it must show that these damages were specifically caused by Continental’s alleged conduct. Inglis has failed to do so.
Among the causes of Inglis’ repeated losses are the following:
1. Inglis’ poor competitive position, emerging from a recent declaration of bankruptcy, failing to advertise, maintaining poor quality control of its product and holding a position in the industry so weak that the Quality Bakers Association, of which Inglis was a member, questioned in 1973 whether Inglis was “even needed in the market.”-
2. The dominant pricing role played by the captives of the grocery chains. All wholesalers had to meet the competition of the captives.
3. The competitive pricing of Campbell-Taggart, which was the largest wholesaler in the market.
4. The competitive pricing of American which, according to Inglis in 1972, engaged in “extreme predatory pricing.”
In this highly competitive market it is simply nonsense to suppose that if Continental had maintained a certain price, In-glis would have been able to have maintained it too. Inglis’ losses were clearly the result of a confluence of factors. Continental cannot be made responsible for damages which are not attributable to con*1352duct which violated the antitrust laws. It is basic that the plaintiff “have suffered damages flowing directly from the violation of the Act.” Weissensee v. Chronicle Publishing Co., 59 Cal.App.3d 723, 728, 129 Cal.Rptr. 188, 191 (1976). Inglis has failed to prove its lost profits were due to conduct of Continental.
The Going Concern Value. Inglis’ expert testified that Inglis’ going concern value was $3.6 million, a figure 31 times its hypothetical pre-tax earnings of $116,000 and 62 times Inglis’ hypothetical after-tax earnings. On an after-tax basis a purchaser at this price would have had a return of 1.2 percent. To suppose that there would have been any purchasers at this price of this shaky company in a highly competitive market is contrary to the weight of the evidence and could not be a rational basis for the jury’s verdict.
In addition, Inglis’ expert’s estimate made the assumption that Inglis would have increased its total sales of all bread products by 3.12 percent. This assumption flew in the face of Inglis’ contention that the relevant market was the one pound private label and not the total white bread market. There is no reason to permit In-glis to prevail on one theory as to the market and to be awarded damages on a totally contradictory theory.
Finally, Judge Canby acknowledges that the expert picked a number out of the air in supposing that Inglis could have achieved a profit of 1.7 percent on sales. Judge Can-by attempts to salvage this part of the damages claim using the more modest figure of .8 percent. But this court has no basis for substituting its own judgment for that of the expert. The expert picked an incredible figure. It should not be our task to- salvage his credibility.
In sum, the expert’s testimony as to valuation was against the weight of all the other evidence in the case and was based on an assumption as to the relevant market which, if consistently applied, would have undermined Inglis’ basic case. There was, therefore, no evidence before the jury as to which it could rationally put a value on Inglis as a going concern. Inglis has proved no damages.
CONCLUSION
This case has dragged on for 20 years. It is apparently the only asset of a company that went out of existence 15 years ago — went out of existence because the quality of its products, the character of its marketing, and the management of its business made it unprofitable. In desperation it has tried to. get something out of its bigger and more successful competitor. If we were to act on the theory that winners in an economic competition have to share the wealth with the losers, or that any defendant with a deep pocket should be made to disgorge for the benefit of an unhappy competitor, we could sustain the verdict which a jury, remarkably generous with other people’s money, has awarded Inglis. The verdict is against the weight of the evidence on both federal claims; on the California claim; and as to the damages awarded. I would direct the entry of judgment for Continental.