HILL, Circuit Judge.

This is an antitrust action brought by Pacific Engineering Production Company of Nevada (PE) against Kerr-McGee Chemical Corporation (successor to American Potash and Chemical Co., which shall be referred to as AMPOT) and its parent corporation, Kerr-McGee Corporation. The trial court gave PE judgment for treble damages totaling $4,590,594 and for attorneys' fees of $528,000 based on its finding that AMPOT was guilty of monopolizing and attempting to monopolize in violation of § 2 of the Sherman Act, 15 U.S.C. § 2, and of price discrimination in violation of § 2(a) of the Robinson-Patman Act, 15 U.S.C. § 13(a). The court also dismissed AMPOT's counterclaims against PE.

The trial court's findings are commendably thorough and exact as to the basic facts. We disagree, however, with the application of the antitrust laws to those facts and reverse the judgment for treble damages. The judgment is affirmed as to the dismissal of AMPOT's counterclaims. We will first consider the Sherman Act issues.

Section 2 of the Sherman Act proscribes both monopolizing and attempting to monopolize any part of trade or commerce. The elements of a monopolization case are (1) the possession of monopoly power in the relevant market and (2) the willful acquisition or maintenance of that power as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident. United States v. Grinnell Corp., 384 U.S. 563, 570-71, 86 S.Ct. 1698, 16 L.Ed.2d 778 (1966). A case of attempted monopolization requires proof that defendant's conduct was motivated by specific intent to monopolize and that a dangerous probability of monopoly existed. E. J. Delaney Corp. v. Bonne Bell, Inc., 525 F.2d 296 (10th Cir. 1975), cert. denied, 425 U.S. 907, 96 S.Ct. 1501, 47 L.Ed.2d 758. The distinctions between the two offenses are not relevant to our decision, because the same conduct was relied upon to establish both the attempt and the willful maintenance of monopoly. The fundamental issue is whether AMPOT engaged in predatory price cutting.

PE and AMPOT manufacture ammonium perchlorate (A/P), a chemical used almost exclusively as an oxidizer in solid rocket fuel. The ultimate consumer of A/P is the federal government. Demand is fixed by government procurement policies and thus is not responsive to price. The direct customers of PE and AMPOT are missile manufacturers for NASA and the Department of Defense.

The first A/P producer was Western Electrochemical Company, which became AMPOT in 1954. Kerr-McGee acquired AMPOT in 1967 and reorganized it into the Kerr-McGee Chemical Corporation in 1970. PE was organized in 1955 by a former Western officer and began A/P production in 1958. The same year, two large industrial chemical manufacturers, Hooker Chemical Co. and Pennwalt Chemical Corp., also entered the market.

In the period of 1954-1957, when AMPOT was the only A/P producer, the price was above 40 cents a pound. From 1958, when the other three firms entered the market, until 1961, the price ranged from 32 to 35 cents a pound. During this time the industry outlook was bright; the mood was characterized by one witness as euphoric. Industry capacity was increased in response.

By December 1963, this optimism had vanished. The anticipated demand did not materialize, primarily because of NASA's conversion from solid to liquid rocket fuel and the diversion of defense appropriations from missile programs to the Vietnam war effort. Prices continuously declined as the companies competed frantically for the work that existed. In 1964, prices first went below 20 cents a pound when PE bid a major order at 19.25 cents. In 1965, AMPOT bid 16.40 cents on another major procurement. The low point came early in 1966 at 14.92 cents, bid by PE. Pennwalt closed its plant in 1965, and Hooker followed suit in 1966. Both continued to sell from inventory for about a year. The trial court found AMPOT's antitrust violations began in 1966 and continued through 1970. During this time, prices were between 15 and 20 cents a pound with the majority of sales being made at prices under 18.25 cents.

The extent of the price cutting must be attributed in some part to the practices of the three major consumers of A/P which were prime contractors on such projects as Minuteman, Poseidon, and Titan missiles. The so-called Big Three accounted for the bulk of total A/P sales. Their practice was to obtain all their A/P needs for a period of months or years in one contract. The winner of the contract secured a substantial share of the total market for the immediate future. In letting the contract, a buyer would first send out a request for quotes and bid the prime contract on the basis of the quotes received. These quotes in effect locked the A/P producers into a ceiling price. After receiving the prime contract, the buyer would again request quotes on which the A/P contract would supposedly be let. Instead of notifying the low bidder, however, the buyer would single out one or two low bidders and tell them to "sharpen their pencils" if they wanted the contract. Often a buyer would succeed in getting the low bidder to underbid itself. The trial court noted that these "whipsaw" tactics were "unethical if not unlawful."

The trial court recognized that until 1966 the decline in the A/P price was due to market forces beyond the control of AMPOT. The court held that after that time AMPOT began violating the antitrust laws because it "sought to take advantage of a troublesome situation and force PE out of the market." The primary factor in the trial court's decision was AMPOT's below-cost pricing. It is uncontested that AMPOT's prices of 15 to 20 cents were well below its average total cost which, although not specifically set out in the findings, was probably in the range of 20 to 26 cents. AMPOT's prices were above its "out of pocket cost," that is, above average variable cost. At this price, it was more profitable to produce than to shut down. The cost of producing A/P is highly volume-responsive and decreases consistently with increased production until full plant capacity is reached. Both AMPOT and PE possessed plant capacity sufficient to supply substantially the entire market for A/P.

There was ample evidence that AMPOT knew PE could not survive at the low price level. PE was a one-product company which was vastly undercapitalized, while AMPOT was larger and more diversified. In 1965, a memorandum on the situation was prepared for AMPOT by a Mr. Cable, its government contracts administrator and special financial analyst. Cable based his memorandum on the following assumptions which proved to be essentially correct:

1. The costs of producing A/P are approximately equal within the industry.

2. A/P consumers are interested in maintaining two sources.

3. Pennwalt will not actively compete for business at the present price levels.

4. Present price levels for A/P will not give PE a profit and probably will not cover all its cost.

Cable observed that if PE were getting the same price as AMPOT, it could not survive and its plant would be of interest only to a consumer of A/P; if PE were getting a higher price, AMPOT could raise its prices without affecting its market share. Cable recommended prices be established with the expectation that the market must be shared with PE either as a going concern or as a subsidiary of a consumer. He recommended against pricing on the expectation of forcing PE to shut down. In the face of this knowledge, AMPOT continued its low prices. In 1968, AMPOT was bidding two to three cents a pound below PE. AMPOT's market share is indicative of the degree of its success in foreclosing PE from the market. Year Percentage

The surveillance of PE was conducted by AMPOT's plant manager at Henderson, Nevada, where PE's plant also was located. It consisted of (1) a monthly check of the number of cars in PE's parking lot; (2) checking freight office records concerning the movement of freight cars in and out of PE's plant; (3) taking aerial photos of PE's plant on two or three occasions; (4) clipping articles in the local papers concerning PE's activities; (5) inquiring at the Colorado River Commission concerning PE's power consumption; and (6) encouraging the shipping foreman to inquire of truck drivers concerning shipments to PE.

AMPOT also engaged in intensive scrutiny of PE's relationship with American Cyanamid which had lent PE over $2,400,000 with a trust deed as security. When suit was filed to foreclose, AMPOT retained two Nevada attorneys to monitor the court files and keep it advised. In 1967, AMPOT learned American Cyanamid had agreed to subordinate its debt to allow PE to obtain a $500,000 loan and continue operating.

The "dumping" incident occurred in 1968 after PE had won a major contract from United Technology Center (UTC). J. C. Schumacher, AMPOT's vice president in charge of manufacturing electrochemicals, visited UTC's president and asked if he would be interested in purchasing surplus A/P at nine or ten cents a pound. There was in fact no surplus. AMPOT's evidence indicated its sales department knew nothing of Schumacher's offer and repudiated it when inquiries were made. The trial court found this should be considered some evidence of the attitude of AMPOT in view of Schumacher's position with the company. Moreover, another AMPOT officer, in a subsequent memo to his file, relied on the incident as having placed a "seed of doubt" that a better price was available.

AMPOT's opposition to PE's reclassification as a small business for purposes of obtaining financial assistance from the Small Business Administration was found to be entirely legal. It was considered "symbolic" of AMPOT's intent to drive PE from the market.

The trial court attached major significance to AMPOT's forward pricing. On different occasions, the large A/P purchasers would request submission of tentative bids for future years. Generally, AMPOT's forward price schedules were high at times when it appeared PE's demise was imminent and were replaced by lower prices when it appeared PE would be able to continue. For example, in 1966 AMPOT learned that American Cyanamid had taken a summary judgment against PE and that an involuntary petition in bankruptcy had been filed against PE. AMPOT then submitted a higher forward price schedule. In February 1967, AMPOT learned that American Cyanamid had agreed to subordinate its debt and that PE had obtained a loan. The next October, AMPOT submitted substantially lower forward prices.

AMPOT's witnesses testified that the forward prices were attempts to raise the A/P price and were withdrawn only under pressure from the purchasers. Shortly after the high price schedule was submitted in 1966, AMPOT met with representatives of two of the Big Three A/P consumers. AMPOT was told that its high prices were unacceptable. The consumers said they had been willing to let AMPOT become the sole supplier at the lower prices, but they would now attempt to keep PE alive. Two of the largest consumers did give PE "stay alive" orders at prices above those paid to AMPOT. The trial court rejected AMPOT's contention that it was forced to withdraw the higher prices and concluded instead that these events indicated AMPOT's intention to do whatever was necessary to drive out PE.

AMPOT also introduced evidence of its other attempts to rescue the A/P industry. In late 1965, AMPOT representatives met with the Deputy Assistant Secretary of Defense in charge of procurement to complain about conditions in the industry. They pointed out that procurement practices and excess capacity were endangering the industry and that two former suppliers had already "mothballed" their facilities. They said AMPOT desired a system providing both good prices for the government and a fair profit for suppliers. The government refused to take action and reminded AMPOT that it was bound by a "National Security Clause" in the contract under which it had purchased its plant from the Navy. The clause required that until March 1972 AMPOT maintain its plant so that it could be operable within 120 days. In light of this situation, AMPOT contended it had no choice but to produce at 15 cents a pound instead of shutting down.

By 1971, AMPOT felt it was close enough to the expiration of the National Security Clause that it could take affirmative action. It notified the Defense Department that it was considering permanent suspension of production. Subsequently, a solution known as the "SAMSO Agreement" was reached. The agreement imposed a division of the market between AMPOT and PE at a profitable price.

The trial court rejected AMPOT's contention that this evidence showed an absence of predatory intent. Another clause in AMPOT's government contract allowed it to petition for release from the National Security Clause in return for whatever consideration the government might request. Inquiries were made concerning whether AMPOT was interested in a release in return for a lump sum payment. AMPOT's failure to avail itself of this relief was considered evidence that it was not seriously interested in exerting pressure to raise prices during the 1966-1970 period.

The trial court found AMPOT must have known it could raise prices without affecting its market share. The prices the court found could have been charged, however, were on the average higher than the forward prices which had provoked strenuous objections from the powerful A/P buyers. During the period of alleged predation, prices actually became somewhat higher than they had been even though the industry continued to operate at less than 50 percent of capacity. Under the trial court's holding, the only way AMPOT could have avoided violating the antitrust laws was to raise prices to a noncompetitive level in order to save its smaller, undercapitalized rival. AMPOT's decision not to raise prices was held to be predatory conduct in violation of § 2 of the Sherman Act. Considering AMPOT's prices and its other practices together, the court found AMPOT's prices were predatory because they were set with the specific intent to drive PE from the market.

We believe there are fundamental flaws in the trial court's application of the Sherman Act. The court's holding that AMPOT's conduct was predatory is based on its effect as a competitor rather than its effect on competition. As we said in Atlas Building Products Co. v. Diamond Block Gravel Co., 269 F.2d 950, 954 (10th Cir. 1959), cert. denied, 363 U.S. 843, 80 S.Ct. 1608, 4 L.Ed.2d 1727: "Antitrust legislation is concerned primarily with the health of the competitive process, not with the individual competitor who must sink or swim in competitive enterprise." In an industry plagued by falling demand and excess capacity, the sinking of a competitor may be an indication of a healthy competitive process.

In a two-firm industry, the exclusion of one firm necessarily results in a monopoly. That does not mean the survivor necessarily violates the antitrust laws. In Union Leader Corp. v. Newspapers of New England, Inc., 180 F. Supp. 125, 140, modified, 284 F.2d 582 (1st Cir. 1960), cert. denied, 365 U.S. 833, 81 S.Ct. 747, 5 L.Ed.2d 744, the district court correctly stated:

[A] person does not necessarily have an exclusionary intent merely because he foresees that a market is only large enough to permit one successful enterprise, and intends that his enterprise shall be that one and that all other enterprises shall fail. . . To prove that a person has that type of exclusionary intent which is condemned in anti-trust cases there must be evidence that the person who foresees a fight to the death intends to use or actually does use unfair weapons. Putting the same idea in another way, we may say that there is no sharp distinction between (a) the existence of an intent to exclude and (b) the use of unfair means. . . .

Predatory pricing is the essential unfair means alleged in this case. The other practices employed by AMPOT inflicted no independent harm. They were merely auxiliaries to its general pricing policy and indicate nothing more than an intent that its pricing succeed in excluding PE from the market. "But intending the natural consequences of acts which are in all respects lawful, does not constitute the `exclusionary intent' that is a prerequisite for finding a violation of section 2 [of the Sherman Act]." Union Leader Corp. v. Newspapers of New England, Inc., supra at 584. If the trial court correctly found AMPOT's pricing was not predatory in itself, we do not believe AMPOT can be found guilty of violating the Sherman Act.

The use of the term "predatory" to describe conduct violative of the antitrust laws has left much to be desired. This court has noted, "The term probably does not have a well-defined meaning in the context it was used, but it certainly bears a sinister connotation." Telex Corp. v. IBM Corp., 510 F.2d 894, 927 (10th Cir. 1975), cert. dismissed, 423 U.S. 802, 96 S.Ct. 8, 46 L.Ed.2d 244. The Supreme Court has indicated that any price below cost may be considered predatory. Utah Pie Co. v. Continental Baking Co., 386 U.S. 685, 696 n. 12, 87 S.Ct. 1326, 18 L.Ed.2d 406 (1967). We do not believe, however, that the Court has established a definitive standard. The Court was speaking of the "inferential value" of below cost sales. In some instances there is no inference to be taken from selling below total cost. It may even be desirable and certainly could not be considered "sinister."

The common circumstance in which sales below total cost may be expected is that of excess capacity. Professor Samuelson has noted in this circumstance the distinction between the interest of individual competitors and the ideals of competition.

Competition, which the businessman regards as destructive, cutthroat, and ruinous, may actually be the only way to get the redundant plant capacity into operation or to discourage its maintenance. . . . Losses or subnormal profits is the free enterprise way of discouraging excess capacity.

P. Samuelson, Economics 496 (8th ed. 1970).

Excess capacity was precisely the problem facing AMPOT and PE. Each had plant capacity sufficient to supply substantially the entire demand for A/P. Individually, AMPOT and PE faced demand curves sharply responsive to price. With the nearly identical products, a small price cut by either of them could result in the acquisition of virtually all the business. Each of them also faced consistently decreasing marginal costs. Additional units of A/P could be produced at less cost than previous units. In this situation, each competitor would be tempted to lower price and expand output to reach a lower point on its marginal cost curve. This in turn would drive the other competitor back up its marginal cost curve and increase its losses at the new, lower price.

The textbook response to this situation is either a price war resulting in the destruction of all but one firm or the establishment of some form of imperfectly competitive price leadership oligopoly. See Samuelson, supra at 452. AMPOT, with knowledge of the consequences, chose to engage in price competition and was found guilty of violating the Sherman Act. The trial court found that a rational, non-predatory duopolist would have employed price leadership to raise the price to a profitable level for both competitors.

Price leadership is not desirable competitive conduct. Any conduct tending to fix prices has been condemned in the strongest language. United States v. Container Corp. of America, 393 U.S. 333, 89 S.Ct. 510, 21 L.Ed.2d 526 (1969); United States v. Socony-Vacuum Oil Co., 310 U.S. 150, 60 S.Ct. 811, 84 L.Ed. 1129 (1940). Conscious parallel pricing is circumstantial evidence of price fixing and may involve many of the same vices. See Theatre Enterprises, Inc. v. Paramount Film Distributing Corp., 346 U.S. 537, 74 S.Ct. 257, 98 L.Ed. 273 (1954); Cackling Acres, Inc. v. Olson Farms, Inc., 541 F.2d 242 (10th Cir. 1976).

In this case, price leadership would invoke virtually all the disadvantages of monopoly. Both PE and AMPOT would have had to restrict output and to incur higher production costs. The selling price of the product would then be forced higher than it would be if one producer left the market. In addition, society would be saddled with the deadweight welfare loss resulting from idle and misallocated plant capacity. We do not believe the antitrust laws should be interpreted to encourage this result.

We conclude that in the circumstances of this case prices below total cost would be more consistent with the competitive goals of the antitrust laws. In the case of Telex Corp. v. IBM Corp., supra at 927, we specifically noted that § 2 of the Sherman Act should not be construed "to prohibit price changes which are within a `reasonable' range, up or down." Although this statement points to the correct result in this case, it still leaves something to be desired. There is no indication of when downward price changes cease to be "reasonable." Even in the situation presented here, the price could be set so low that it should be considered predatory.

The proper identification of predatory pricing has recently been a subject of controversy. The discussion has focused on the employment of marginal cost as the floor below which prices should be considered predatory. In Predatory Pricing and Related Practices under Section 2 of the Sherman Act, 88 Harv.L.Rev. 697 (1975), Professors Areeda and Turner argue that any price above a firm's marginal cost, or in the alternative its average variable cost, should be conclusively presumed lawful. Any price below that level should be conclusively presumed unlawful. However, in Scherer, Predatory Pricing and the Sherman Act: A Comment, 89 Harv.L.Rev. 869 (1976), the author takes issue with the short-run cost approach. Scherer contends Areeda and Turner's proposed rule would permit anti-competitive pricing inconsistent with long-run welfare maximization. He recommends that courts continue to inquire into the monopolist's subjective intent and that several long-run variables be considered in determining whether the monopolist's conduct is anticompetitive.

We believe that evidence of marginal cost or average variable cost is extremely beneficial in establishing a case of monopolization through predatory pricing. The relationship of a firm's price and its marginal cost is regarded as the best indicator of the competitive health of an industry. Samuelson, supra at 438; C. McConnell, Economics, 446-448, 464 (3d ed. 1966). As an indicator of predatory pricing, marginal cost is extremely valuable because: "There is no reason consistent with an interest in efficiency for selling a unit at a price lower than the cost that the seller incurs by the sale." Posner, Exclusionary Practices and the Antitrust Laws, 41 U.Chi.L.Rev. 506, 519 (1974).

In the present case, the trial court specifically found that AMPOT's sales were always at prices above average variable cost and "contributed to the company's cash flow." These prices also were above AMPOT's marginal cost. AMPOT never reached its shut-down point — the point at which it would incur greater losses by operating than by shutting down. Under the circumstances, selling at these prices was rational, competitive behavior.

Although we do not intend to adopt a solely cost-based test, there are no other relevant factors indicating AMPOT's conduct was anticompetitive even in the long run. The low prices for A/P persisted into the long run (a period of nearly seven years), and the industry's problems were resolved only by the imposition of external controls in the interest of national defense. The evidence shows this is not a case of short-run price cutting designed to secure monopoly profits in the long run. See International Air Indus., Inc. v. American Excelsior Co., 517 F.2d 714 (5th Cir. 1975), cert. denied, 424 U.S. 943, 96 S.Ct. 1411, 47 L.Ed.2d 349. There is nothing in the record to warrant a finding that AMPOT engaged in predatory pricing.

We turn now to the trial court's finding that AMPOT also violated the price discrimination prohibition of the Robinson-Patman Act, 15 U.S.C. § 13(a). The discriminatorily low price was found in the sales to the large-volume buyers which we have been discussing up to this point. There also existed a number of small-volume purchasers constituting a relatively small percentage of the total A/P market. To these purchasers, both AMPOT and PE sold on the basis of published price schedules rather than secret bids. These prices were higher than the competitive bid prices. In 1968, the only year for which the trial court made specific findings, AMPOT's schedule prices varied from 18.62 cents a pound on orders of 100,000 pounds or more to 32.5 cents for orders under 2,000 pounds. In the same year, AMPOT's bid prices to the major purchasers were from 16 to 18 cents a pound. The price difference could not be cost-justified because AMPOT produced all its A/P in a continuous process and made all sales from accumulated inventory. AMPOT dominated the small-sale market primarily because PE made only nominal efforts to compete for this business.

Based on its finding that AMPOT's competitive bid prices were predatory, the trial court found the price discrimination gave AMPOT "an illegal `edge' in its ability to undercut PE" which supplied the necessary element of injury to competition. This edge was found to be "slight" but "not insubstantial." We disagree with the court's conclusion that the evidence established injury to competition.

Some cases suggest discriminatorily low prices might be considered a violation of the Robinson-Patman Act on facts which would not establish predatory pricing in violation of § 2 of the Sherman Act. See Areeda Turner, supra, 88 Harv.L.Rev. at 726. As it applies to the primary-line injury in this case, the Robinson-Patman Act should be interpreted no differently from the Sherman Act. The Supreme Court has admonished that the Robinson-Patman Act should be reconciled "with the broader antitrust policies that have been laid down by Congress." Automatic Canteen Co. v. FTC, 346 U.S. 61, 74, 73 S.Ct. 1017, 1025, 97 L.Ed. 1454 (1953). However, we perceive no substantial conflict. The primary-line decisions have consistently emphasized the element of predation. Utah Pie Co. v. Continental Baking Co., supra, 386 U.S. at 696 n. 12, 87 S.Ct. 1326, citing FTC v. Anheuser-Busch, Inc., 363 U.S. 536, 548, 80 S.Ct. 1267, 4 L.Ed.2d 1385 (1960).

The reasoning employed by the trial court in finding injury to competition has been referred to as the "double inference test." International Air Indus., Inc. v. American Excelsior Co., supra at 723. From below-cost pricing, predatory intent is inferred; from the finding of predation, injury to competition is inferred. See Continental Baking Co. v. Old Homestead Bread Co., 476 F.2d 97, 104 (10th Cir. 1973), cert. denied, 414 U.S. 975, 94 S.Ct. 290, 38 L.Ed.2d 218. It follows that if the first inference is lacking, there is no basis for the second.

We have held that AMPOT was competing on the merits; its prices were not predatory. The fact that AMPOT was charging higher prices in another submarket does not change its legitimate competition into an injury to competition. Areeda Turner, supra, 88 Harv.L.Rev. at 727; see Telex Corp. v. IBM Corp., supra. Although PE may have been at a disadvantage because it was not competing effectively in the high-price submarket, that does not establish the injury to the health of the competitive process with which the Robinson-Patman Act is concerned. Atlas Building Products Co. v. Diamond Block Gravel Co., supra; Anheuser-Busch v. FTC, 289 F.2d 835 (7th Cir. 1961).

We realize that our ruling leads to the somewhat untoward result that the larger of two competitors will survive while the smaller may expire. Bigness, however, is not a disqualification to compete. The Seventh Circuit made the point rhetorically in Anheuser-Busch v. FTC, supra at 843 n. 13:

If bigness be a disqualification for a firm to compete, interesting collateral questions arise, such as, how small does a company have to be before it has the right to enter into price competition with its competitors? And how large does it have to be before it must stop competing in price?

The Robinson-Patman Act was intended to provide small businesses with protection from abuses by large, powerful business, but legitimate price competition is not such an abuse. "[N]either the Act nor any social value compels the sheltering of an individual competitor, at the expense of the public interest, from the competitive process." International Air Indus., Inc. v. American Excelsior Co., supra at 721, accord, Atlas Building Products Co. v. Diamond Block Gravel Co., supra at 956.

Finally, we must consider AMPOT's appeal from the dismissal of its counterclaims. The only point urged on appeal is that an alleged agreement between PE and two major A/P purchasers to give PE "stay alive" orders constituted a group boycott illegal per se under Klor's, Inc. v. Broadway-Hale Stores, Inc., 359 U.S. 207, 79 S.Ct. 705, 3 L.Ed.2d 741 (1959). We need only note that the findings of fact which defeat the group boycott claim are not clearly erroneous. F.R.Civ.P. 52(a).

The judgment is reversed insofar as it holds AMPOT liable for violations of the Sherman and Robinson-Patman Acts. The judgment is affirmed as to the dismissal of AMPOT's counterclaims.