Federal Trade Commission (FTC) appeals from the district court's denial of a preliminary injunction in an action instituted pursuant to § 13(b) of the Federal Trade Commission Act, 15 U.S.C. § 53(b) (Supp. IV 1974), to prevent consummation of a merger between Atlantic Richfield Company (Arco) and The Anaconda Company (Anaconda) during the pendency of administrative antitrust proceedings which are being conducted before FTC. Because we are persuaded that the district court correctly determined that a substantial likelihood has not been shown that FTC will prevail on the merits of the underlying antitrust proceeding, we affirm.

Arco, the fifteenth largest publicly-held corporation in terms of sales and revenues, and the thirteenth largest in terms of assets, is principally engaged in the production and sale of petroleum, petroleum products and natural gas. Arco has also engaged in exploration for uranium and, through a joint venture, the production and sale of uranium oxide. It has, however, divested itself of this aspect of its operations, to which fuller reference will be made later. Arco has never been engaged in the copper business.

Anaconda is principally engaged in the mining and processing of copper and aluminum and the manufacture of copper, copper alloy and aluminum products. It is also engaged in exploration for uranium and the production of uranium oxide. It is a fully integrated copper company. It ranks 188th in terms of sales and revenues and seventy-first in assets, notwithstanding that in recent years many of its foreign assets have been appropriated by a foreign government with less than full compensation. Anaconda ranks third in the market for copper ore and concentrates and fourth in the market for refined copper, respectively. The two leaders in the market for copper ore and concentrates control approximately thirty percent of the market, and the three leaders in the market for refined copper control approximately sixty percent of the market. Anaconda has a market share of 8.27 percent in the market for copper ore and concentrates and 9.78 percent in the market for refined copper.

In March, 1976, Arco made a successful cash tender offer to shareholders of Anaconda as a result of which it acquired twenty-seven percent of Anaconda's common stock. In July, 1976, Arco and Anaconda agreed to merge, Anaconda to become a wholly-owned subsidiary of Arco. Anaconda's shareholders have approved the merger.

FTC commenced administrative proceedings challenging the merger on the ground that it would violate § 7 of the Clayton Act, 15 U.S.C. § 18 (1970), or § 5 of the Federal Trade Commission Act, 15 U.S.C. § 45 (1970) as amended (Supp. IV 1974),1 because of its anticompetitive effect in the markets for copper ore and concentrates, the production and sale of refined copper and the production and sale of uranium oxide. It then sued in the district court under § 13(b) of the Federal Trade Commission Act to restrain consummation of the merger which, by agreement of the parties, must occur by March, 1977, if at all. After a full hearing, the district court, in an opinion from the bench, denied relief. It made numerous findings, the pertinent ones of which will be separately addressed.

Section 13(b) of the Federal Trade Commission Act, the statutory basis for this litigation,2 requires a judicial consideration of two factors in determining whether to grant a preliminary injunction while the administrative determination of whether the antitrust laws have been or would be violated is being made: "the equities" and "the Commission's likelihood of ultimate success." See FTC v. Food Town Stores, Inc., 539 F.2d 1339 (4 Cir. 1976) (Winter, J., sitting as a single circuit judge). The district court found overall that FTC had not shown a reasonable likelihood that it would prevail in establishing that the merger would violate the antitrust laws. We think this conclusion essentially correct. We are led to this result from our analysis of FTC's grounds of attack and the evidence of record.

In the administrative proceedings, FTC alleged that the effects of the proposed merger may be substantially to lessen (a) potential competition between Arco and producers of copper in the market for copper ore and concentrates and in the production and sale of refined copper, (b) actual competition between Arco and Anaconda in the production and sale of uranium oxide, and (c) potential competition between Arco and producers of uranium oxide.3 We will consider separately the copper and uranium oxide claims.

With regard to copper ore concentrates and refined copper, it is undisputed that Arco is not now engaged in such activities and hence is not an actual competitor in the copper markets. FTC's theory that the merger will substantially lessen competition is therefore predicated upon the claim that the evidence shows that but for the merger, Arco could reasonably be expected to engage in such activities and to become an entrant into these markets and areas of competition. Specifically, FTC contends that it has established that, if the merger is not allowed to be consummated, there is nonetheless a "reasonable probability" that Arco would enter the copper markets by original entry, by joint venture, by acquisition of an ore body, or by toehold acquisition.4 Presumably, FTC would not claim that entry by joint venture or by toehold acquisition would violate the antitrust laws;5 obviously original entry, or modified original entry by acquisition of a known ore body rather than exploration to discover an ore body, would not violate the antitrust laws.

At the outset, it must be recognized that the doctrine of actual potential entry as a basis for finding a violation of the antitrust laws is an evolving doctrine and not one with regard to which a body of controlling authority has yet been developed.6 The question of whether elimination of competition by an actual potential entrant by reason of an acquisition, without more, may violate § 7 was reserved in United States v. Falstaff Brewing Corp.,410 U.S. 526, 537, 93 S. Ct. 1096, 35 L. Ed. 2d 475 (1973), and United States v. Marine Bancorporation, 418 U.S. 602, 625-26 & n. 28, 94 S. Ct. 2856, 41 L. Ed. 2d 978 (1974). The Supreme Court seems to have sanctioned the theory of actual potential entry in the context of a joint venture in United States v. Penn-Olin Chemical Co., 378 U.S. 158, 84 S. Ct. 1710, 12 L. Ed. 2d 775 (1964), but it did not address the question where only actual potential entry was claimed. There are a number of lower court decisions to the effect that the elimination of a probable future entrant is the type of "effect" which "may be substantially to lessen competition" within the meaning of § 7.7 One of the best discussions of the doctrine, endorsing it, is found in Turner, Conglomerate Mergers and Section 7 of the Clayton Act, 78 Harv.L.Rev. 1313, 1362-86 (1965). Turner notes that the doctrine is not uniformly embraced by text writers who reason that § 7 should not be applied to a merger which neither lessens actual competition nor harms future competition by raising barriers to entry, but simply prevents an increase in competition that would otherwise take place. See Rahl, Applicability of the Clayton Act to Potential Competition, 12 A.B.A. Antitrust Section 128, 143 (1958). See also Robinson, Recent Developments: 1974, 75 Colum.L.Rev. 243 (1975); J. Markham, Conglomerate Enterprise and Public Policy 176-78 (1973); Posner, Antitrust Policy and the Supreme Court: An Analysis of the Restricted Distribution Horizontal Merger and Potential Competition Decisions, 75 Colum. 282, 323-24 (1975) (criticizing in various ways the theory).

The novelty of the doctrine and the absence of definitive authority sanctioning it and defining its parameters could well serve as a basis for denial of a preliminary injunction under § 13(b), since it is difficult, if not impossible, to determine FTC's chances of ultimate success when the law is so uncertain and the parameters of the doctrine obscure. But we prefer to rest our decision on another basis, namely, that the district court correctly concluded that a showing of ultimate success under the legal standards which have evolved has not been made. Stated in terms of the proper scope of appellate review of the district court's denial of a preliminary injunction, we cannot say that the district court was so patently in error that its denial of the injunction should be disturbed. See Caldwell v. HUD, 522 F.2d 4 (4 Cir. 1975); Conservation Council of N. C. v. Costanzo, 505 F.2d 498 (4 Cir. 1974); West Virginia Highlands Conservancy v. Island Creek Coal Co., 441 F.2d 232 (4 Cir. 1971).

Turner, in his careful analysis of the actual potential entrant theory, speaks to the threshold question of the quantum of proof required to show a violation of § 7 by loss of a potential new entrant in a market like the copper market which is controlled by a tight oligopoly:

I therefore conclude that when the only (emphasis in original) alleged anticompetitive consequence of a merger is the elimination of what would have been a new entrant in a tight oligopoly, there must, in order to support prohibition, be clear proof that the firm would in fact have entered an admittedly rare case, and one bound to become even less frequent if this rule were adopted. (Emphasis added.) 78 Harv.L.Rev. at 1384.

When the sole (emphasis in original) alleged anticompetitive consequence of a merger is the loss of what would have been a new entrant, prohibition is appropriate only when the market is a tight oligopoly and when entry by internal expansion would appear to have been certain. (Emphasis added.) 78 Harv.L.Rev. at 1386.8

The Turner concept of the quantum of proof to sustain a claim of actual potential competition finds peripheral support in both Falstaff Brewing Co. and Marine Bancorporation. In Marine Bancorporation, the Supreme Court explained that the principal focus of the potential competition doctrine has been on perceived potential competition rather than actual potential competition because "(u)nequivocal proof that an acquiring firm actually would have entered de novo but for a merger is rarely available," 418 U.S. at 624, 94 S.Ct. at 2871 (emphasis added), thereby implying that the standard is one of "unequivocal proof" in a case where only actual potential competition is claimed. The Supreme Court also admonished that "it is to be remembered that § 7 deals in 'probabilities,' not 'ephemeral possibilities'," 418 U.S. at 622-23, 94 S.Ct. at 2870.

In Falstaff, the Supreme Court did not disturb the district court's finding that Falstaff was not an actual potential entrant while at the same time indicating that the district court relied almost solely on management's post-acquisition statements that Falstaff would not enter de novo. See 410 U.S. at 532, 93 S. Ct. 1096. Given the questionable value of such statements, it is clear that little evidence is required to prove that there would not be de novo entry. See also Case Comment, Mergers-Acquisition of Leading Coal Company By Potential Entrant Into Coal Industry Violated Section 7 of Clayton Act, 86 Harv.L.Rev. 772, 777 n. 26 (1973) ("Section 7 requires a showing of reasonable probability that a merger will have anticompetitive effects . . .. It is doubtful that the mere showing of (the company's) ability to enter by toehold or independent expansion, without a showing that it intended or had any reason to do so, satisfies the reasonable probability requirement.")

There is more reason to require a higher burden of proof to show anticompetitive effect of a merger like that in the instant case than in a traditional vertical or horizontal merger. The reasons are well articulated by Turner, see 78 Harv.L.Rev. at 1320-22, and will not be repeated here. Moreover, the conglomerate merger in the instant case involves no product or market extension; it is a merger purely for purposes of diversification. Arco is not poised on the fringe of the copper markets; it has no technological skills readily transferrable to the copper markets; it has no channels of distribution which may be utilized to distribute copper. In short, since it is not probable that an anticompetitive effect in copper should necessarily result from the merger, strict proof of any anticompetitive effect should be required.

When we examine the record, we find that entry into the copper markets is more burdensome than entry into many other possible markets. FTC's copper consultant testified that de novo entry into copper is "extremely difficult" and the entry barriers are "severe." Estimates in the record of the cost of entry into mining and refining range from $200 to $450 million. Entry time is estimated from ten to nineteen years. The starting point is, of course, discovery of a commercially exploitable deposit of ore. This alone may take ten years and involve an expenditure of up to $50 million. An entrant must also construct smelting facilities. In order to assure successful entry, the entrant must attain a certain level of technical expertise. Presently, seventy-five companies are currently engaged in exploration for copper, the first stage of entry, but the obvious deposits have already been discovered. It is clear that entry into the copper market is not easily accomplished.

Unquestionably, FTC has shown that Arco has strong economic incentives to seek diversification inasmuch as oil resources in the United States continue to dwindle and may be depleted in the 1980's. As well, copper offers Arco an investment in an industry which historically has been as profitable as the oil business. Arco is accumulating revenues which it desires to utilize, and it is willing to make large expenditures for a successful investment. In sum, Arco does possess the financial resources to make a de novo entry into the copper markets although the minimum capital costs of successful entry are substantial.

On the other hand, the difficulty of entry into a market to which Arco is neither committed not impelled, cf. United States v. Phillips Petroleum Co., 367 F. Supp. 1226 (C.D.Cal.1973), aff'd mem., 418 U.S. 906, 94 S. Ct. 3199, 41 L. Ed. 2d 1154 (1974) (Phillips on edge of lucrative California market in which it had made some investments), certainly indicates that de novo entry or entry by acquisition of other than a large company was not entirely economically suitable. See United States v. Marine Bancorporation, 418 U.S. 602, 642, 94 S. Ct. 2856, 41 L. Ed. 2d 978 (1974); Missouri Portland Cement Co. v. Cargill, Inc., 498 F.2d 851, 857, 864 (2 Cir.), cert. denied, 419 U.S. 833, 95 S. Ct. 150, 42 L. Ed. 2d 123 (1974); United States v. Falstaff Brewing Corp., 332 F. Supp. 970, 972 (D.R.I.1971), rev'd on other grounds, 410 U.S. 526, 93 S. Ct. 1096, 35 L. Ed. 2d 475 (1973). This is particularly so considering the long lead time, see Missouri Portland Cement Co. v. Cargill, Inc., 498 F.2d 851, 857 (2 Cir.), cert. denied, 419 U.S. 833, 95 S. Ct. 150, 42 L. Ed. 2d 123 (1974); United States v. Black & Decker Mfg. Co., 5 Trade Reg.Rep. P 61,033, at 69,589 n. 64 (D.Md.1976), and the need to acquire more complete technical expertise than Arco presently possesses. See Missouri Portland Cement Co. v. Cargill, Inc., 498 F.2d 851, 857, 863 (2 Cir.), cert. denied, 419 U.S. 833, 95 S. Ct. 150, 42 L. Ed. 2d 123 (1974); United States v. Black & Decker Mfg. Co., 5 Trade Reg. Rep. P 61,033, at 69,586, 69,592 (D.Md.1976). Objectively, then, Arco might reasonably be shown to be ripe for diversification and financially capable of entering the copper markets, but there are as well good reasons why copper would not be an appropriate area for grass roots entry. Indeed, objectively, Arco could have been found to be clearly not a prospective entrant.

There is, however, evidence that Arco, at its various levels of management, has not been unmindful of possible entry into the copper markets, and FTC argues that this interest is indicative of an intent or desire to enter these particular markets. As a forward-looking corporation, progressively and aggressively managed, one having a substantial cash flow and one whose management is keenly aware that it is engaged in exploiting a diminishing natural resource, Arco, not surprisingly, has long given consideration to diversification into other forms of business activity, particularly into the extraction, processing and sale of minerals other than petroleum and natural gas. The record shows that since the late 1960's, Arco had expressed various degrees of interest in entry into the copper industry and had begun to investigate various methods of entry. Since the late 1960's, some members of Arco's senior management have favored diversification into copper by acquisition and were committed to diversification into copper only if diversification could be accomplished by acquisition.

Pursuant to this interest, Arco made inquiries concerning the possible acquisition of other copper producers. As well, Arco undertook studies regarding the copper industry and its future importance in the domestic economy. The studies concluded that copper was the most appropriate non-ferrous metal into which Arco could diversify and Arco's Executive Group (consisting of senior management), in late 1974 or early 1975, approved the conclusion that copper was the most attractive non-ferrous metal into which to diversify. Through its approval of Arco's long-range plan for future goals, Arco's Board of Directors formally approved diversification into copper by September, 1975, without any decision, however, as to the mode, time or method of market entry.

Other groups within Arco's lower management, more particularly the Synthetic Crude and Minerals Division (SCMD), began in the early 1970's to evaluate entry by exploration, by acquisition of a known copper ore deposit, by joint venture with a company already active in copper exploration or developing a known ore deposit, or by acquisition of a copper producing company. A joint task force from Corporate Planning and SCMD recommended grass roots exploration for non-energy minerals including copper, lead, zinc and their associated metals as the most appropriate diversification vehicle at that time, and it recommended financial studies and a thorough analysis of the existing copper-lead-zinc industry, the companies in the mining industry, and how other energy companies have diversified into minerals. In March and August, 1975, senior management directed further study of copper exploration and the possible acquisition of copper reserves, as well as accumulation of technological assessment expertise in copper.

While the proof thus shows a continuing interest on the part of Arco in the copper industry and continuing studies as to the best means of entry, it fails to show a significant commitment at the decisional level that Arco was seriously considering original entry into the copper markets or entry by toehold acquisition.9 Indeed, the proof is equally consistent with an attitude of gathering information and watchful waiting for a future determination of the means of entry to be employed if diversification into copper was to be undertaken. Thus, the decision to diversify cannot be divorced from the means to be employed.

FTC sought to complete its proof by the affidavit of Dr. George F. Leaming, a professional economist who is an expert in the copper industry and who has had widespread experience as an economic consultant both for industry and various governmental groups. Dr. Leaming found Arco to be a "leading potential entrant" and it was his expert opinion that from Arco's history of making overtures to copper companies regarding possible acquisition, Arco's investigation of possible purchase of copper deposits, Arco's study of copper exploration techniques, Arco's avowed corporate interest in diversification into the copper industry and Arco's large annual income, and Arco's capital investment capabilities, that Arco was a "likely potential entrant into the United States copper industry by means other than the acquisition of an existing large copper producer." While certainly relevant and entitled to consideration, we cannot say that Dr. Leaming's expert opinion is a substitute for a specific commitment at Arco's top managerial level to enter the copper markets by original entry or by toehold acquisition, particularly when Arco's top management indicated a conviction against de novo entry while approving large acquisition entry. Nor do we think that Dr. Leaming's opinion demonstrates the existence of such factors as would reasonably impel Arco to such a decision if the merger is not permitted. As we have earlier stated, we think that the quantum of proof necessary to show Arco to be an actual potential entrant by original entry or by toehold acquisition is greater than that provided by FTC.10

Our conclusion is reinforced by subjective evidence on the part of Arco's top management that Arco would never seriously consider original entry or entry by toehold acquisition. Admittedly, this proof is subjective and mainly consists of self-serving statements made by Arco officials subsequent to the commencement of FTC's administrative antitrust proceeding or subsequent to the institution of suit by FTC in the district court. The matter of subjective evidence, particularly subjective statements post litem moten, was the subject of extended discussion by Mr. Justice Marshall concurring in the result in Falstaff Brewing Corp. Mr. Justice Marshall pointed out the obvious, that such statements are inherently unreliable and must be used with great care because of the likelihood that they are designed to fend off regulatory attack rather than to reflect accurately a policy decision made when bias and self-interest are not operative facts. Inter alia, Mr. Justice Marshall stated, "subjective statements of management are probative in cases where the acquiring firm is alleged to be an actual potential entrant," 410 U.S. at 564, 93 S.Ct. at 1117, but he later qualified this statement by adding, "subjective evidence has, at best, only a marginal role to play in actual potential-entry cases," 410 U.S. at 566, 93 S.Ct. at 1117. The capsule of his views appears from the following:

To summarize, then, I would not hold that subjective evidence may never be considered in the context of an actual potential-entry case. Such evidence should always be admissible as expert, although biased, commentary on the nature of the objective evidence. And in a rare case, the subjective evidence may serve as a counterweight to weak or inconclusive objective data. But when the district court can point to no compelling reason why the subjective testimony should be believed or when the objective evidence strongly points to the feasibility of entry de novo, I would hold that it is error for the court to rely in any way upon management's subjective statements as to its own future intent.

Basically, resort to subjective evidence raises problems of credibility made acute by the self-serving nature of statements made after litigation has ensued. In the instant case, however, consistent with the views of Mr. Justice Marshall, we do not think that FTC's objective evidence strongly points to the feasibility of entry de novo and therefore Arco's subjective evidence is entitled to some weight. Particularly is this true when the evidence here shows that entry de novo would be tremendously expensive, time-consuming, and unusually difficult, so that it may be fairly concluded that entry de novo is not readily feasible even for a company possessing the economic and technological resources of Arco.

Because we are thus persuaded that the district court did not improperly conclude that there was not a substantial likelihood that FTC would be able to establish a violation of § 7 under the actual potential entry doctrine,12 we think that the district court correctly denied a preliminary injunction with respect to the effect of the merger on Arco's entry into the copper markets.

In the administrative proceeding and in the initial stages of this litigation, FTC claimed, with respect to uranium oxide, that the merger would substantially lessen actual competition. FTC asserted that the merger was in effect a horizontal merger inasmuch as Arco had uranium reserves and was engaged in the sale of uranium oxide which it extracted by a special in situ leaching process (which is not apparently not usable on most uranium found in the United States), and Anaconda was also engaged in the production and sale of uranium oxide. That claim is now moot because on November 1, 1976, Arco agreed to sell its only uranium oxide production and selling operation to United States Steel Corporation.13 The closing under the agreement of sale was held on December 6, 1976.

It is, of course, true that the agreement provides that Arco will continue to operate the facility (presumably to train personnel for the buyer and to provide an orderly transition of operation), but Arco has exercised its right to discontinue operating the facility by giving notice that it will withdraw no later than December 7, 1977; and United States Steel has evidenced an intent to take over the operation by July 1, 1977, if technologically feasible. Arco will continue to receive an overriding royalty on Dalco's twenty-five percent interest in the operation and Arco retains the right to use its patents and the technological processes it has developed for in situ leaching. Arco also retains a property interest in any technological advances made at the facility and it has a non-exclusive royalty-free license to use any patents, processes or technological know-how that Dalco or United States Steel may develop in the future with respect to the operation. The significant element, however, is that after December 7, 1977, Arco will not be responsible for preparing programs and budgets; it will not conduct uranium exploration operations; and it will not act as sales agent for the uranium production.

Divestiture prior to merger is an acceptable technique to avoid an antitrust violation. See United States v. Atlantic Richfield Co., 297 F. Supp. 1061, 1068-69 (S.D.N.Y.1969). See also FTC v. Food Town Stores, 539 F.2d at 1345 (dictum). We think that divestiture renders inapposite the holding in Stanley Works v. FTC, 469 F.2d 498 (2 Cir. 1972), cert. denied,412 U.S. 928, 93 S. Ct. 2750, 37 L. Ed. 2d 155 (1973) (one percent holding by Stanley to be sufficient to be merger between actual competitors), on which FTC places great reliance. Therefore, we do not find that actual competition would be affected.

FTC nonetheless contends that the merger will violate the antitrust laws because Arco is a "most likely potential entrant" into the uranium oxide market either by original entry or by toehold acquisition. For many of the same reasons we have expressed with regard to copper, we think that the record provides substantial support for the district court's contrary holding. Of course, Arco possesses the financial resources to enter the uranium market; it has demonstrated a past presence in that market. And Arco expended and authorized the expenditure of funds for uranium exploration prior to its decision to withdraw from the market. But what Arco has done and what it is capable of doing are not enough to show Arco to be an actual potential entrant, the removal of which will substantially lessen competition. It bears repeating that § 7 speaks of probabilities and not mere possibilities. While the FTC concedes that to constitute a potential entrant Arco must have "interests and incentives to enter," proof of them is lacking here.

Objectively, full-fledged entry by Arco will require substantial capital outlays over an extended period of time. Return on investment will be slow in an area of investment which is very risky. Arco does not presently have expertise in the type of uranium extraction normally used in an area where technological expertise is necessary. Arco has no ore reserves which can be easily utilized nor has it discovered any commercially exploitable ore bodies in the last five years of exploration. There are then significant factors militating against full-fledged entry by Arco into uranium.

While FTC's expert found Arco to be a most likely potential entrant, this opinion is overborne by the fact that Arco has just withdrawn from the uranium market, albeit a very special segment of the market in situ leaching under an agreement of sale which contains no provision for recapture whether or not the merger is consummated, and has in the past decreased its percentage of holdings in that same facility from one-hundred percent to fifty percent and has withdrawn in past years from participation in fuel enrichment programs and nuclear waste disposal.14 Once again, a subjective post litem moten statement reinforces this objective fact as Arco's top executive indicates that Arco is planning withdrawal from uranium. Additionally, there are eighty-six companies presently exploring for uranium in the United States; and in the past ten years, seven firms have entered uranium production by internal expansion, despite the barriers. It is, therefore, unlikely that even were Arco to be a potential entrant, its loss would have a significant anticompetitive effect. See FTC v. Proctor & Gamble Co., 386 U.S. 568, 581, 87 S. Ct. 1224, 18 L. Ed. 2d 303 (1967); United States v. Hughes Tool Co., 415 F. Supp. 637, 646 (C.D.Cal.1976); United Nuclear Corp. v. Combustion Engineering, Inc., 302 F. Supp. 539, 557 (E.D.Pa.1969); Posner, Conglomerate Mergers and Antitrust Policy: An Introduction, 44 St. John's L.Rev. 529, 531 Spec. ed. (1970).

While we do not foreclose the possibility that one may withdraw from a market and still be a realistic potential entrant, clear proof that Arco, notwithstanding withdrawal, would probably reenter the market if the merger fails and the loss of Arco as such an entrant would significantly lessen an opportunity for increased competition, is necessary to show a violation of § 7. There was ample support for the district court's conclusion that such proof was not forthcoming and that an injunction should be denied.

AFFIRMED.