In this appeal we are called upon to review a grant of summary judgment in a complex antitrust conspiracy case. This litigation arises out of the circumstances surrounding the settling of February 1982 futures contracts for No. 2 home heating oil. Simply stated, plaintiff-appellant Apex Oil Company ("Apex") alleges that defendants-appellees, which include competing oil companies, a broker, the New York Mercantile Exchange (the "Exchange"), and an officer of the Exchange named Julian Raber, conspired to force Apex to deliver much of the oil it had agreed to deliver in February 1982 in the first three delivery days of that month at one delivery site. Apex alleged that this behavior made it impractical or impossible for Apex to deliver on time and forced Apex to pay premium prices to other oil companies, including defendants, to obtain oil for delivery to defendants in order to avoid a default.

Since this action centers upon a series of events surrounding the futures market for No. 2 heating oil for February 1982, it is helpful to outline the workings of the commodities futures market.

In brief, a commodity futures contract entitles the buyer of the contract (the "long") to purchase from the seller of the contract (the "short") a set quantity of some commodity at a set date in the future. Typically, all of the terms of the contract, except price, are standardized under the rules of an exchange. The price is negotiated on the floor of the exchange.

This case concerns futures contracts for No. 2 heating oil sold on the Exchange. On the Exchange, one futures contract for No. 2 oil consists of a contract to deliver 1000 barrels (42,000 gallons) of oil in New York Harbor at a set price during some future month.

Most contracts, however, never result in actual delivery since the market is used by many investors for speculation. Speculation is made possible by allowing holders of long or short "positions" to "liquidate" those positions by purchasing or selling the necessary contracts. For example, a holder of short contracts who does not wish to deliver oil may buy a number of long contracts equal to the number of short contracts he or she holds. Similarly, a long who does not wish to take delivery of the oil may enter into short contracts equal in number to the number of long contracts held. The speculator makes a profit or suffers a loss based on the differences in the prices paid for the long and short contracts.

Many actors take part in the market for No. 2 heating oil futures. A futures commission merchant or broker executes the desires of the actual buyers or sellers of contracts by requesting a member of the Exchange ("clearing member") to enter into the transaction through a floor broker on the floor of the Exchange. The Exchange functions as a "clearing house," acting as a buyer to all sellers and a seller to all buyers thereby enabling an investor who wishes to buy or sell a contract to find a partner.

When trading is closed for contracts for a particular month, the Exchange determines which persons hold long positions and which persons hold short positions by offsetting each trader's long contracts and short contracts. The Exchange then matches the longs with the shorts for the purpose of delivery. Under rules discussed below, the matched parties work out the time, place and method of delivery. Delivery may still be avoided by means of an "exchange futures for product" transaction ("EFP"). In an EFP, the parties agree to supersede the futures contract with a new contract to deliver oil on the cash ("wet") market. If the holder of a short position defaults, i.e., fails to deliver timely, the Exchange rules provide for fines and other sanctions.

In light of the complexity and mass of material presented by the parties herein, we briefly summarize what appear to us to be undisputed facts concerning the subject February 1982 No. 2 heating oil contracts.

The rules of the Exchange provided that any February contracts not liquidated by Friday, January 29, 1982--the close of trading for such contracts--would be matched, long with short, for the purpose of delivery (or EFP). At the close of trading, Apex was short 4,378 contracts. In other words, Apex had agreed to deliver 4,378 contracts (183,876,000 gallons) of oil in February 1982. The total number of outstanding contracts designated for delivery by all persons holding short positions, including Apex, was 4,906.

Of the 4,906 contracts open at the close of trading, 1,945 were held by a number of oil companies and their affiliates who were named as defendants herein1 (the "long defendants"):

As required by Exchange rules, on February 1, 1982, Apex selected a delivery site through its broker and clearing member Goldman Sachs & Co. ("Goldman"). For 4,281 of its 4,378 contracts, it selected GATX terminal in Carteret, New Jersey, as the delivery site. Apex designated B.P. terminal at Tremley Point, New Jersey, as the delivery site for the remaining contracts. On the same day, also pursuant to Exchange rules, each long defendant submitted "Notice of Intention to Accept Petroleum Products Futures." The clearing house then matched up longs and shorts for delivery. All of the open long positions held by the long defendants, except for fifteen contracts to be delivered to Warren, were designated by the Exchange for delivery by Apex.

The Exchange rules then required each long to nominate a delivery date and method by 4:30 p.m. on February 5. Deliveries could be requested by pipeline, intra-facility transfer, inter-facility transfer, inventory (also known as book) transfer, tanker, barge or truck. Inventory transfer entails transfer of title of oil without actually moving or delivering it.

The rules then in effect stated that "[t]he buyer shall give written delivery or loading instructions," that "[a]ll deliveries must be completed after the fifth business day and before the last business day of the delivery month," and that in the case of barge or tanker delivery, "[t]he seller must supply the product as soon as the barge or tanker reports readiness to load," or in the case of on-shore delivery, "as soon as the buyer reports the transfer facility or truck is ready to accept the product." Herein, the first day for which the long defendants could demand delivery was February 6.

Prior to February 1982, the nominations by holders of long positions typically were spread throughout the month. Moreover, it was not unusual for long and short parties to be somewhat flexible in the actual timing and location of delivery. For example, in December 1981 and January 1982, the nominations of the long defendants were spread fairly evenly throughout each month. The actual delivery dates differed from the nominated dates for roughly two-thirds of the nominations. The actual delivery method was changed in two-thirds of the nominations. The actual location for delivery differed in nearly one-third of the nominations. Thus, although the rules apparently allowed longs to rigidly demand delivery at a particular time and place, in practice the longs were not always inflexible.

According to Apex, this pattern was broken in February 1982. Apex alleges that the long defendants, aided by DiMauro, a broker, and by the Exchange and Raber, conspired to nominate an inordinate number of their contracts for the first three delivery days of February. There is no dispute that the defendants actually did nominate a large number of contracts for the first few days. Indeed, in over 75% of the long defendants' 1,945 nominations, delivery was demanded for one of the first three delivery days. Significantly, fewer than 25% of long positions held by non-defendants were nominated for the first three days. Further, our calculations indicate that while over half of the contracts held by defendants were nominated for the first day of delivery, only about 10% of contracts held by non-defendants were nominated for the first day. Apex alleges, and the defendants deny, that a conspiracy among the defendants was responsible for this nomination pattern.

There is evidence that on February 3, a broker representing several long interests passed along a rumor on the Exchange floor that Apex would not be able to make delivery. That afternoon, the Exchange Control Committee convened for the purpose of resolving any market problem. Apparently, the broker repeated his statement regarding the possibility of an Apex default and also stated that "one or two" of his customers were "out to get Apex." The record indicates that similar rumors circulated among a number of other brokers and oil companies.

On the next day, February 4, Apex's broker, Goldman, represented to the Exchange Control Committee that Apex would not default, claiming that it expected to receive 750,000 barrels from Amoco and that it expected to have another 400,000 barrels available at GATX. However, on February 5, Goldman told the committee that Apex no longer expected to have the oil available at GATX for deliveries nominated for February 6, 7 or 8. Apex apparently only had available approximately 30,000 barrels of No. 2 oil at GATX on February 5.

Meanwhile, on February 4 and 5 the Exchange Control Committee asked several clearing members to suggest that their customers holding long positions agree to change delivery dates and sites. Apex alleges that its own requests to change delivery dates, methods or sites were either ignored or rejected by the long defendants. The long defendants do not appear to contest this allegation. Apex further alleges that this inflexibility was the result of a conspiracy among the defendants to "squeeze" Apex by forcing it to deliver an enormous amount of oil on short notice. As a result of this inflexibility, Apex claims that it was forced to purchase oil from the long defendants at inflated prices in order to meet its delivery obligations. Indeed, there is evidence that while the closing price for February contracts was 89.7cents per gallon, Apex paid prices as high as 97.5cents per gallon to several long defendants for oil that was immediately delivered back to the long defendants to satisfy Apex's delivery obligations. These purchases of oil enabled Apex to avoid a default.

The gravamen of Apex's claims is that the defendants conspired to "squeeze" Apex: the long defendants, DiMauro, TIC, and Triad did so by manipulating the market; Raber and the Exchange by assisting or failing to prevent the other defendants from manipulating the market. The amended complaint asserts eight "claims for relief," which can be placed into four categories.

1. Antitrust Conspiracy. Apex alleges four Sherman Act claims; treble damages are requested. Three claims allege violations of section 1 of the Act and one alleges a violation of section 2. See 15 U.S.C. Secs. 1, 2.

The first claim asserts that the long defendants and the exchange defendants conspired to nominate for early delivery in violation of section 1 of the Act. The early nominations allegedly enabled the long defendants to "manipulate, raise, maintain, stabilize and/or fix the price of No. 2 Heating Oil" by creating an "artificial shortage." Amended Complaint paragraphs 46-52; see 15 U.S.C. Sec. 1.

The second claim alleges that the long defendants conspired to refuse to deal with Apex by refusing to sell No. 2 oil to Apex, thereby weakening Apex's position as a competitor and creating an artificially high price for the oil. Amended Complaint paragraphs 53-59; see 15 U.S.C. Sec. 1.

The third claim asserts that the long defendants conspired to prevent other sources of No. 2 oil from supplying it to Apex, thereby artificially raising the price for the oil. Amended Complaint paragraphs 60-64; see 15 U.S.C. Sec. 1.

The fourth claim alleges that the long defendants conspired to monopolize the wet market for No. 2 oil available for delivery in New York Harbor during early February 1982, thereby causing Apex to pay artificially high prices for such oil. Complaint paragraphs 65-69; see 15 U.S.C. Sec. 2.

2. Commodities Manipulation. The fifth claim in the amended complaint alleges that the long defendants violated sections 9(b) and 13(a) of the Commodity Exchange Act ("CEA") by manipulating the price of oil and that all the defendants violated the CEA by spreading false reports about Apex. Amended Complaint paragraphs 70-79; see 7 U.S.C. Secs. 13(b), 13c(a). This claim apparently has been abandoned with respect to the Exchange defendants.

3. Failure to Regulate. The seventh claim in the amended complaint asserts that the exchange defendants failed to regulate properly the activities of brokers, traders and others with respect to the trading and delivery of oil during the period in question, thus violating the CEA and regulations thereunder. Amended Complaint paragraphs 86-91; see 7 U.S.C. Secs. 7(d), 7a(8).

4. Fraud. The sixth claim alleges fraud under the Commodity Exchange Act, see 7 U.S.C. Secs. 6b, 13c(a), and the eighth claim alleges common law fraud by the long defendants. Amended Complaint paragraphs 80-85, 92-97.

Discovery on all the claims having been completed, all the defendants moved for summary judgment dismissing all the claims. The district court granted the motions and entered judgment dismissing the amended complaint. See Apex Oil Co. v. DiMauro, 641 F. Supp. 1246 (S.D.N.Y.1986).

We affirm in part, reverse in part and remand.

The principal issue in this appeal is whether Apex has produced enough evidence in support of the claimed conspiracy to survive defendants' motions for summary judgment. These motions with respect to the antitrust claims were aimed only at this issue and the district court's grant of the motions was founded upon the district court's perception of a lack of evidence of a conspiracy. While we agree with the district court that, despite the great mass of evidence presented by Apex, the evidence of conspiracy is insufficient as to most of the defendants, we conclude that sufficient evidence has been presented for the case to proceed as to one defendant.

A motion for summary judgment shall be granted if "there is no genuine issue as to any material fact and ... the moving party is entitled to a judgment as a matter of law." Fed.R.Civ.P. 56(c). Herein, we focus primarily on the first part of the test: whether there is an issue of fact as to the existence of a conspiracy.

The movant bears the initial burden of demonstrating "the absence of a genuine issue as to any material fact." See Adickes v. S.H. Kress & Co., 398 U.S. 144, 157, 90 S. Ct. 1598, 1608, 26 L. Ed. 2d 142 (1970). This may consist simply of pointing out that the plaintiff has failed to present any evidence to establish a necessary element of the cause of action. See Celotex Corp. v. Catrett, 477 U.S. 317, 106 S. Ct. 2548, 2552-53, 91 L. Ed. 2d 265 (1986). In response, the non-moving party must then "set forth specific facts showing that there is a genuine issue for trial." Fed.R.Civ.P. 56(e); see Celotex Corp., 106 S. Ct. at 2552 & n. 3; Adickes, 398 U.S. at 159-61, 90 S.Ct. at 1609-10. While the non-moving party must show more than a "metaphysical doubt" as to material facts, Matsushita Electric Industrial Co. v. Zenith Radio Corp., 475 U.S. 574, 106 S. Ct. 1348, 1356, 89 L. Ed. 2d 538 (1986), all inferences must be drawn in favor of the non-moving party. Id. at 1356-57. However, the Supreme Court has stated that "antitrust law limits the range of permissible inferences from ambiguous evidence in a Sec. 1 case." Id., 106 S.Ct. at 1357.

Specifically, "conduct as consistent with permissible conduct as with [an] illegal conspiracy does not, standing alone, support an inference of antitrust conspiracy." Id.; see Monsanto Co. v. Spray-Rite Service Corp., 465 U.S. 752, 764, 104 S. Ct. 1464, 1470, 79 L. Ed. 2d 775 (1984). There must be some evidence that "tends to exclude the possibility that the [defendants] were acting independently." Id. at 764, 104 S.Ct. at 1471. Thus, evidence of parallel conduct alone cannot suffice to prove an antitrust conspiracy. A plaintiff "should present direct or circumstantial evidence that reasonably tends to prove that the [defendants] 'had a conscious commitment to a common scheme designed to achieve an unlawful objective.' " Id. (quoting Edward J. Sweeney & Sons v. Texaco, Inc., 637 F.2d 105, 111 (3d Cir.1980), cert. denied, 451 U.S. 911, 101 S. Ct. 1981, 68 L. Ed. 2d 300 (1981)); see also H.L. Moore Drug Exchange v. Eli Lilly & Co., 662 F.2d 935, 941 (2d Cir.1981), cert. denied, 459 U.S. 880, 103 S. Ct. 176, 74 L. Ed. 2d 144 (1982). As stated by this court, "[a]t a minimum, ... 'the circumstances [must be] such as to warrant a jury in finding that the conspirators had a unity of purpose or a common design and understanding, or a meeting of minds in an unlawful arrangement.' " International Distribution Centers, Inc. v. Walsh Trucking Co., 812 F.2d 786, 793 (2d Cir.1987) (quoting Michelman v. Clark-Schwebel Fiber Glass Corp., 534 F.2d 1036, 1043 (2d Cir.), cert. denied, 429 U.S. 885, 97 S. Ct. 236, 50 L. Ed. 2d 166 (1976)).

We must note, however, that while summary judgment is a valuable means for avoiding unnecessary trials, and recent Supreme Court as well as Second Circuit cases have tended to encourage its use in complex cases such as this one, see, e.g., Celotex Corp., 106 S. Ct. at 2555; Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 106 S. Ct. 2505, 91 L. Ed. 2d 202 (1986); Matsushita, supra; Knight v. U.S. Fire Ins. Co., 804 F.2d 9, 12 (2d Cir.1986), cert. denied, --- U.S. ----, 107 S. Ct. 1570, 94 L. Ed. 2d 762 (1987); Eastway Construction Corp. v. City of New York, 762 F.2d 243, 250-51 (2d Cir.1985), it should not be regarded as a substitute for trial. Thus, while the Supreme Court has indicated that trial courts should draw only reasonable inferences in favor of the non-moving party viewing the evidence as a whole, see Matsushita, Id. at ---, 106 S.Ct. at 1357; see also International Distribution Centers, 812 F.2d at 793-94, and while some assessing of the evidence is necessary in order to determine rationally what inferences are reasonable and therefore permissible, it is evident that the question of what weight should be assigned to competing permissible inferences remains within the province of the fact-finder at a trial.

In Matsushita the plaintiffs had alleged that various Japanese television manufacturers were engaged in a predatory pricing scheme, financed by the charging of higher than competitive prices in Japan, designed to put American companies out of business. In that case, the Supreme Court stated that summary judgment dismissing the action was appropriate since predatory pricing schemes are inherently unlikely and since there was no sign of the alleged goal of such a scheme ever coming within reach after many years of its alleged operation. Thus, since the scheme was not "plausible," the Court stated that the jury should not be permitted to infer the existence of a conspiracy from evidence of higher than competitive prices in Japan since to do so could deter pro-competitive conduct, i.e., the setting of low prices in the United States. Id. at ---, 106 S.Ct. at 1360; see also Monsanto, 465 U.S. at 762-64, 104 S.Ct. at 1470. If the scheme alleged is implausible, a conspiracy must be proved by strong direct or strong circumstantial evidence, and the implausibility of a scheme will reduce the range of inferences that may permissibly be drawn from ambiguous evidence. See Matsushita, Id. at ---, 106 S.Ct. at 1360-62.

It is worth noting that the conspiracy alleged herein seems plausible, as evidenced by the apparent ability of some defendants to sell oil to Apex at prices significantly over market. Our task herein is to discern whether on the evidence presented it could reasonably be concluded that the higher prices came about as a result of a conspiracy, rather than through independent action of the defendants.

We also think it worth noting that in the Matsushita and Monsanto cases, the types of conspiracies alleged were more susceptible of direct proof since they concerned long-term complex relationships among competitors. See Posner, Oligopoly and the Antitrust Laws: A Suggested Approach, 21 Stan.L.Rev. 1562, 1569-75 (1969) (discussing difficulty of maintaining cartel of many firms and probable need for explicit agreement); see also Gainesville Utilities Department v. Florida Power & Light Co., 573 F.2d 292, 303 (5th Cir.), cert. denied, 439 U.S. 966, 99 S. Ct. 454, 58 L. Ed. 2d 424 (1978); F. Sherer, Industrial Market Structure and Economic Performance 199-200 (2d ed. 1980). In a case where the alleged scheme is short-term and relatively simple in operation, we must as a practical matter accept that the best proof available most often will only tend to show the existence of an informal, perhaps even tacit, rather than explicit, agreement. Therefore, we carefully examine the conduct and communications of the defendants in order to determine whether it evidences an agreement. See Gainesville Utilities Department, supra (citing Posner, supra, at 1562).

Much of the evidence in this case concerns the alleged parallel conduct of the defendants. Parallel conduct can be probative evidence bearing on the issue of whether there is an antitrust conspiracy. However, parallel conduct alone will not suffice as evidence of such a conspiracy, even if the defendants "knew the other defendant companies were doing likewise." Modern Home Institute, Inc. v. Hartford Accident & Indemnity Co., 513 F.2d 102, 110 (2d Cir.1975). More must be shown.

Since mere parallel behavior can be consistent with independent conduct, courts have held that a plaintiff must show the existence of additional circumstances, often referred to as "plus" factors, which, when viewed in conjunction with the parallel acts, can serve to allow a fact-finder to infer a conspiracy. See id.; In re Plywood Antitrust Litigation, 655 F.2d 627, 633-37 (5th Cir.1981), cert. dismissed, 462 U.S. 1125, 103 S. Ct. 3100, 77 L. Ed. 2d 1358 (1983). Such circumstances might include a common motive to conspire or a high level of interfirm communications. See, e.g., In re Plywood Antitrust Litigation, 655 F.2d at 633 (interfirm communications); Ambook Enterprises v. Time Inc., 612 F.2d 604, 616 (2d Cir.1979) (common motive to conspire), cert. dismissed, 448 U.S. 914, 101 S. Ct. 35, 65 L. Ed. 2d 1179 (1980). See generally P. Areeda, Antitrust Analysis 371-82 (3d ed. 1981). However, such plus factors may not necessarily lead to an inference of conspiracy. For example, such factors in a particular case could lead to an equally plausible inference of mere interdependent behavior, i.e., actions taken by market actors who are aware of and anticipate similar actions taken by competitors, but which fall short of a tacit agreement. See id. at 374-75. In such a case, a court might find it difficult to hold that the parallel acts "tend to exclude the possibility" of independent action.

As discussed below, see Parts IV.A.2 to IV.A.5 infra, while Apex points to evidence indicating that the long defendants may have been motivated by a desire to make inordinate profits, to to "get" Apex, this evidence of motivation creates an equally valid inference of independent action. Apex also points to evidence of a number of communications among the defendants. However, aside from the particular communications discussed below, see Part IV.A.2 infra, for the reasons stated by the district court, we believe that inferring a conspiracy from the existence of these communications would amount to speculation under the circumstances herein. See 641 F. Supp. at 1266-69. Accordingly, we believe that in this case Apex must show more.

Apex presents evidence which it believes shows that the parallel acts were "against the apparent individual economic self-interest of the alleged conspirators." Modern Home Institute, 513 F.2d at 111. Such a showing, if successful, might tend to exclude the possibility of independent parallel behavior. See Bogosian v. Gulf Oil Corp., 561 F.2d 434, 446 (3d Cir.1977), cert. denied, 434 U.S. 1086, 98 S. Ct. 1280, 55 L. Ed. 2d 791 (1978); see also Matsushita, supra. See generally 6 P. Areeda,Antitrust Law paragraphs 1415, 1434, at 84-93, 213-21 (1986).

Apex argues that it made no economic sense for defendants individually to nominate early. However, the most we can say in support of this position is that the evidence concerning the rationality of nominating early is ambiguous. For example, it is not really controverted that the price of oil was falling in early February 1982. The defendants suggest that this provided them with a good reason to take delivery early; the oil would be more valuable earlier in the month than later. Apex counters that this makes sense only if there was an immediate market for the oil since it costs money to store oil, since the money used to buy the oil early instead could be invested, and since the defendants allegedly already had inventories of oil on hand. However, these costs which might be incurred as a consequence of early delivery were very likely small in comparison with the potential decrease in value of the oil by the end of the month. Apex provides no probative evidence tending to show that defendants would have experienced difficulty in selling the oil after delivery or even that any defendant believed that late delivery would be harmful. Indeed, it was an officer of Apex who testified that in a falling market early delivery was indicated. While Apex presented the testimony of an expert who asserted that the defendants did not need the oil, the expert did not state that taking the oil early would be to the defendants' disadvantage. Since the evidence as to the rationality of nominating early is ambiguous, it appears that we do not have a case where a fact-finder could be allowed to infer from the parallel conduct the existence of a conspiracy in the absence of additional evidence. See id. p 1415b, at 86-87 (question is whether no rational person would engage in the alleged parallel behavior individually) (discussing Modern Home Institute, supra ).

This does not mean that evidence of the long defendants' parallel conduct becomes irrelevant, however. A court deciding whether to grant summary judgment should not view each piece of evidence in a vacuum. Seemingly innocent or ambiguous behavior can give rise to a reasonable inference of conspiracy in light of the background against which the behavior takes place. Evidence can take on added meaning when viewed in context with all the circumstances surrounding a dispute. Thus, while we must carefully determine what inferences reasonably may be drawn from each piece of evidence, we must make this determination in light of all of the evidence proffered by Apex.

Perhaps the strongest evidence of a conspiracy offered by Apex is contained in a notebook kept by Dennis Denihan of Northeast memorializing a telephone conversation on February 3 between Fred Slifka, an officer of Global, and Don Oliver at Northeast. The entry in the notebook read:

(1) All nominations in writing

(2) If not accepted, NPC [Northeast] to demand $100/contract/day penalty for 5 days

At end of five days window NPC to demand 10 pct penalty + oil plus go to court

(3) NPC to go to newspaper and expose exchange

(4) All loadings must be GATX

[Portion obliterated].

Under the circumstances herein, we believe a reasonable inference could be drawn that this notebook entry evidences an agreement between Slifka of Global and Oliver of Northeast to take the actions set forth in it. The district court suggested that since Apex itself picked GATX as the place of delivery, "it is more likely" that the conversation was a natural one between traders faced with the possibility of an Apex default on deliveries; that it was merely a discussion of shared concerns. See 641 F. Supp. at 1267. In other words, since those in the market were aware that Apex might well default, it was reasonable for competitors to discuss this possibility and also possible responses to be taken in order to ensure delivery of the oil.

However, while Apex itself initially chose GATX as the delivery location, there was some evidence indicating that Apex had not in the past been held to its initial selection of a delivery site. Moreover, it seems that if these two parties truly were concerned primarily with early delivery of the oil, a better way to ensure delivery might have been to show some flexibility as to the mode and place of delivery. Instead, the notebook entry seems to suggest that Northeast and Global may have agreed to insist on delivery of the oil at GATX. The long defendants cite evidence suggesting they feared loss of priority if they showed flexibility. However, this merely represents evidence disputing the inference which can be drawn from the notebook entry.

Moreover, the district court seems to have accepted the long defendants' assertion that since the price of No. 2 oil was falling at the time the nominations were made, no issue of fact could exist as to the motivation behind the early nominations. As this argument runs, it was sensible to accept the oil early so that a higher price could be charged on resale. This would tend to show that the notebook entries merely evidence an innocent conversation about how to achieve that end.

We cannot agree with this characterization of the evidence. First, assuming prices were falling, it was not inevitable that early nomination would be advantageous. For example, it would not have made sense to incur storage costs if fairly rapid resale was not feasible. A table compiled by the long defendants indicates that for all long positions in February 1982 No. 2 oil contracts, defendant and non-defendant, only about 30% of all nominations specified the first two days of the delivery period. Our calculations, based upon apparently undisputed compilations, submitted by Apex, of the number of nominations for each day made by each long defendant, indicate that over 60% of the long defendants' nominations were for the first two days (over 75% for the first three days); and approximately 65% of Global's nominations were for the first two days (100% in the first three days). While only about 10% of Northeast's nominations were for the first two days, over 55% were for the first three days. The non-defendant longs nominated only 15% of their contracts for delivery for the first two days (19% in the first three days). In fact, it appears that the non-defendant longs nominated at most 23% of their Apex contracts for the first three days. It therefore seems clear that not all holders of long positions in oil contracts in a climate of falling prices saw fit to take delivery early, even with respect to those contracts to be delivered by Apex. Indeed, some evidence indicates that in the two months prior to February 1982, No. 2 oil prices may have been dropping, albeit not as rapidly. In those months, very few contracts were nominated early by holders of long positions.

We see no conflict between our view here that evidence of falling prices does not necessarily indicate that early delivery was in the interest of every holder of a long position, and our discussion above in which we stated that demanding early delivery was not automatically suspect. We believe that the demand of early delivery, by itself, creates no dominant inference.

However, in light of the incriminating nature of the notes regarding the telephone conversation, since we must draw reasonable inferences in favor of Apex, we conclude that the finder of fact should decide whether the denial of conspiracy by the defendants should be accepted. We do not reverse as to Global and Northeast since the claims against them have been settled. See note 1, supra. However, this discussion is relevant to interpreting additional communications among certain defendants which are discussed below.

We next consider several entries made in the notebook of William Berberich of Belcher on February 1, 1982, prior to the nominations by the long defendants. In that notebook, the name of Don Oliver, a trader at Northeast, appears. An arrow is drawn from the name to the notation, "Losses." Under these notations there are additional notations that appear to relate to prices of No. 2 oil. Then there appears the following: "Demand delivery " (emphasis in original), "Look for Default" and "Merc--EFP." Several pages later, in another notation for February 1, the notebook reads, "Sat. demand 12:01 Sat. [February 6] AM"; and in notations for February 2 appear the words, "promote crunch" and "press for default."

Apex asks us to infer from this that Berberich spoke to Oliver on February 1 and agreed to "demand delivery " and thereby "press for default" on the part of Apex. Apex offers no other evidence indicating that the conversation ever took place. Looking at this evidence alone, we agree with the defendants and the district court that inferring that a conversation took place would amount to speculation. Proof of Belcher acting independently in an attempt to create a short squeeze would not show a violation of the Sherman Act as alleged in the amended complaint.

Apex states that the suggested inference is a proper one because Belcher held a long position that was relatively small in comparison to the size of Apex. It is posited that Belcher alone could not have hoped to create a "crunch" and that therefore Belcher must have conspired to create the "crunch."

This reasoning ignores the undisputed evidence that Belcher was aware that Apex had relatively small amounts of No. 2 oil at GATX. On February 1, Apex actually had a deficit balance of oil at GATX. It also was widely known that Apex had a huge short position. Thus, it would not be unreasonable for officials at Belcher to assume that other holders of long positions might nominate some contracts early. One might also note that the 315 contracts held by Belcher amount to 12,600,000 gallons, not an insignificant amount of oil.

At oral argument, counsel for Apex stated that it could have made delivery on all of its obligations were it not for the insistence on the use of GATX. Accordingly, we are asked by Apex to conclude that Belcher could not have reasonably expected to "squeeze" Apex with only 315 contracts. First, even if we assume this were so, it says nothing of the knowledge or beliefs of Belcher officials as to the matter. Apex in its brief cites a passage from a deposition of Paul Novelly, a principal of Apex. However, in that deposition Novelly merely stated that Apex had enough oil in its system "to deliver against the February contract[s]" "for the month." He did not testify that Apex could complete delivery the first three days of the delivery period. Thus, even if we were to assume that Belcher knew or believed that Apex was capable of delivering the oil over the course of the month, it could have harbored a reasonable expectation that it could "squeeze" Apex by nominating its 315 contracts for the first minute of the delivery period. We therefore cannot draw an inference of conspiracy from this notebook entry alone.

However, Apex also points to another conversation, between Slifka of Global and Berberich of Belcher, which Apex argues supports an inference that Belcher took part in a conspiracy. On February 3, Slifka spoke to Berberich. Berberich admitted that the conversation took place, but stated that it was limited to whether Slifka thought there would be a default and to the fact that Warren held a long position. In his notebook for that day, Berberich entered Slifka's name and telephone number, followed by, "1) G.E. Warren .150. 2) Sent Wires and Want Immediate Answer. 3) Threatened WS [illegible]." Two pages later in the notebook, on the same date, an entry reads:

--Penalty to start Fri.--

--10% after 5 days--

--oil at 5 days--

--Bring suit against exchange

--Restraining order to prevent any further trades on Merc.

Apex points out that the above entries in Berberich's notebook closely resemble the entries in Denihan's notes, discussed above, regarding a conversation between Slifka (Global) and Oliver (Northeast) on the same day. Apex also points out that " 'once a conspiracy is shown, only slight evidence is needed to link another defendant with it.' " United States v. Wilkinson, 754 F.2d 1427, 1436 (2d Cir.) (Mansfield, J.) (quoting United States v. Marrapese, 486 F.2d 918, 921 (2d Cir.1973), cert. denied, 415 U.S. 994, 94 S. Ct. 1597, 39 L. Ed. 2d 891 (1974)), cert. denied, 472 U.S. 1019, 105 S. Ct. 3482, 87 L. Ed. 2d 617 (1985).

In our view, the similarity between the notes seems striking. If a finder of fact were to infer a conspiracy between Global and Northeast, a possibility we have concluded would be reasonable in light of Denihan's notebook entries, see supra, it could reasonably conclude in light of Berberich's notebook entries regarding the Slifka conversation that Belcher also participated, especially if we view the Slifka conversation in light of Berberich's notations made two days earlier (e.g., "promote crunch" and "press for default"). The long defendants point out that the conversation between Berberich and Slifka took place after Belcher had made its nominations, and argue that the entries merely refer to a discussion of whether there was going to be a default. However, the notes could more easily be interpreted as outlining the method of continuing the pressure that already had been placed on Apex. Thus, it reasonably could be inferred that the conversation consisted of more than mere speculation concerning Apex's ability to deliver. Accordingly, in light of all of the circumstances, including communications among the defendants and their parallel conduct, we must conclude that sufficient evidence has been produced from which a factfinder could infer that Belcher, Global, and Northeast " 'had a unity of purpose or a common design and understanding, or a meeting of the minds in an unlawful arrangement.' " International Distribution Centers, 812 F.2d at 793 (quoting Michelman, 534 F.2d at 1043). We therefore reverse the dismissal of the antitrust claims as to Belcher.

Apex points to evidence of a series of additional conversations said to have occurred among the long defendants. Bearing in mind that evidence of the mere exchange of information by competitors cannot establish a conspiracy, see Kreuzer v. American Academy of Periodontology, 735 F.2d 1479, 1487 (D.C.Cir.1984), we have carefully considered the arguments of Apex with respect to these conversations. For essentially the reasons stated by the district court, see 641 F. Supp. at 1266-69, we conclude that inferring the existence of a conspiracy from the remaining conversations would amount to mere speculation.

Apex presents a mass of other material that Apex characterizes as evidence of a conspiracy. We have carefully reviewed this evidence and, while we reverse and remand as to the antitrust claims against Belcher for the reasons stated above, and remain aware that " 'once a conspiracy is shown, only slight evidence is needed to link another defendant with it,' " Wilkinson, supra, since we do not find such a link, we conclude that affirmance of the grant of summary judgment is appropriate as to the remaining defendants.

a. The Activities of Sal DiMauro

Apex devotes much of its argument to discussing a number of activities engaged in by Sal DiMauro of Triad Petroleum, Inc. and TIC Commodities, Inc. We can find no permissible inference of conspiracy which can be drawn from any of this evidence.

Apex first points to an offer on February 1 by Sal DiMauro on behalf of Stinnes to sell Apex 300,000 barrels of oil for approximately 87cents per gallon. This offer was withdrawn ten minutes later after Joseph DiMauro, also of Triad and TIC, spoke to the president of Stinnes. On February 2, Sal DiMauro offered Apex 300,000 barrels from Stinnes at 97 cents. And on February 3, Joseph DiMauro offered Apex 1.5 million barrels at 93.5 cents. Apex wishes us to infer a conspiracy from this. However, Apex offers no evidence tending to show that this sequence was part of any agreement beyond one between Stinnes and its broker, DiMauro. Moreover, the fact that Stinnes used its broker to make the offers does not amount to an illegal conspiracy. A more likely inference from this evidence is that once the market became aware of Apex's delivery obligations, Stinnes and others, as rational sellers, raised the price of oil in order to take advantage of Apex's plight.

Apex also argues that Sal DiMauro invited one Jerry Schultz, a trader at Tipco Products, Inc. ("Tipco"), not a defendant, to participate in a short squeeze. Apex finds support for this proposition in the deposition of Schultz. However, a fair reading of the Schultz deposition leads to no such conclusion. Schultz had heard rumors of an impending short squeeze and discussed this with DiMauro. DiMauro asked Schultz "if [Tipco was] going to take delivery early." Schultz replied that Tipco had not yet decided and expressed concern about a lack of storage space. DiMauro suggested that delivery could be worked out. Schultz also asked how to go about nominating early. In response, DiMauro had a representative of TIC explain the nomination process to Schultz. Still unsure of the process, Schultz requested a telex explaining the necessary steps. TIC instead sent a telex to ContiCommodity Inc., Tipco's clearing member at the Exchange, which stated that Tipco planned to take delivery of all of its contracts on February 6. It also stated that "final delivery instructions [would] be communicated no later than February 5, 1982." Schultz testified that Tipco never authorized the sending of this telex.

One possible explanation for this activity is that DiMauro was soliciting Tipco to join a conspiracy. The problem with this approach is that it assumes that there was a conspiracy of which DiMauro was a member. Apex suggests that Schultz testified that he took the call to be an invitation to participate in a short squeeze along with certain other defendants. However, nothing in the transcript suggests that DiMauro was shown to be part of a conspiracy. Nor does it suggest with which long defendants DiMauro is supposed to have been conspiring. So, even if we assume that DiMauro had invited Tipco to take part in a squeeze, there is no indication that he did so pursuant to an agreement with any particular long defendants. We are therefore unable to conclude that the Schultz deposition "tends to exclude" the possibility that DiMauro was simply trying to solicit business for his brokerage firm. We conclude that this evidence does not give rise to any permissible inference of a conspiracy under the circumstances herein.

Next, Apex points to a conversation between Joseph DiMauro and Slifka of Global on February 4. True, the conversation was followed by the submission by TIC of a nomination on behalf of Stinnes of 300 contracts by inventory transfer at GATX on February 6. However, DiMauro was the broker for a number of the defendants, and it is a normal function for a broker to make transactions on behalf of a client. Further, given that there is no evidence of what was said during the conversation, we cannot conclude, in light of all of the circumstances, that a factfinder should be allowed to infer that the conversation between DiMauro and Slifka concerned an agreement to squeeze Apex.

b. Statements of Certain Market Actors

Apex offers evidence of statements by certain persons that it claims suggest evidence of a conspiracy. This evidence adds little to what we already have discussed. The district court found that most of this evidence consists of inadmissible hearsay and that in any event none of the evidence tended to prove the existence of a conspiracy. We need not reach the hearsay question since the proffered evidence does not give rise to an inference of conspiracy.

For example, Apex argues that Tony Densieski, a trader for Texas Energy (not a defendant) informed the Vice-President of the Exchange that certain longs had invited Densieski to take advantage of Apex's position. However, Densieski never specified which longs had made this offer. We clearly cannot infer the participation on the part of any defendant based on such a generalized accusation.

The testimony of Jay Bernstein, an employee of Northville Caribbean is similarly unhelpful. Bernstein merely testified as to his understanding that there was a conspiracy among "other longs." These longs were unspecified, and Bernstein testified that he had no personal knowledge of any actions taken by the longs.

Similarly, while Julian Raber allegedly told Patricia Lyons of Apex that he was aware of "a conspiracy to 'squeeze the shorts,' " and referred to the actions of the longs as a "sting," the testimony concerning these statements lacked any indication of what facts Raber based his opinion on or even whether his opinion was based on anything at all other than rumor or speculation.

Upon examination of these and other alleged statements by market participants proffered by Apex in support of its conspiracy theory, we must conclude that taken either as a whole or individually, none of them creates an inference of a conspiracy among any additional named defendants.

c. Evidence of the Long Defendants' Motivation

Apex has presented evidence that it believes tends to show that the defendants were motivated by greed and by a strong dislike of Apex. Apex argues that such motivation tends to exclude the possibility of independent action because the long defendants had nothing to gain unless they acted in concert. As discussed above, however, it is not at all clear that one long defendant could not have forced Apex into some concessions. See Part IV.A.2, supra. Moreover, as each defendant became aware of the early nominations by the others, the advantages of inflexibility as to delivery date and location became evident. Since the early nominations became public knowledge, it became evident to anyone in the business that a profit could be made by forcing Apex to deliver. Thus, any motivation to "get" Apex gives rise to an inference of individual action which is equally as plausible as an inference of conspiracy. We therefore conclude that the long defendants' alleged motivation cannot succeed in establishing a conspiracy on the evidence herein.

Little of the evidence discussed thus far relates to the Exchange defendants. The strongest evidence presented by Apex to show that the Exchange defendants participated in a conspiracy consists of evidence of a number of communications between the Exchange defendants and the long defendants. However, the content of these conversations, even as characterized by Apex, amounts to "jawboning" in an attempt to avert an impending default. We note that the Exchange also communicated with Apex as part of this effort. Moreover, Apex has offered no plausible motivation for the Exchange defendants to have participated in the alleged scheme. We therefore affirm the grant of summary judgment as to the Exchange and Raber, its vice-chairman.

The district court held alternatively that in addition to failing to prove participation by the Exchange defendants in a conspiracy, Apex failed to show bad faith, which the district court believes is necessary in an antitrust claim asserted against a self-regulating market such as the Exchange when the claim relates to the duty of the Exchange to maintain a fair and orderly market. See 641 F. Supp. at 1282; see also P.J. Taggares Co. v. New York Mercantile Exchange, 476 F. Supp. 72, 78-79 (S.D.N.Y.1979). Since we have concluded that Apex has failed to present evidence linking the Exchange defendants to a conspiracy, we need not reach the issue of whether bad faith is a necessary element of the antitrust claims against the Exchange.

Apex has presented massive additional materials that it claims amount to proof of a conspiracy when viewed in light of all of the circumstances herein. Some of it tends to indicate that several of the defendants disliked Apex. Some of it seems to indicate that some of the long defendants may have used the circumstances as they existed in February 1982 to "get" Apex. However, aside from the evidence related to Global, Belcher and Northeast, discussed above, none of it gives rise to a permissible inference of a conspiracy among the defendants. Consequently, for the reasons discussed principally in parts IV.A.1 through IV.A.4, we affirm the dismissal of the antitrust claims as to all the other defendants.

The fifth claim in the amended complaint alleges that the defendants manipulated the commodities market in violation of the CEA, 7 U.S.C. Sec. 13(b), which prohibits, inter alia, (1) any "attempt to manipulate the price of any commodity ... for future delivery" and (2) delivery of "false or misleading or knowingly inaccurate reports concerning ... market information or conditions that affect or tend to affect the price of any commodity in interstate commerce."

The amended complaint alleges first that the long defendants, with the assistance of the other defendants, "attempted to and did manipulate the price of No. 2 Heating Oil" by "cornering" or "squeezing" the market for No. 2 oil. Amended Complaint paragraphs 72-73. It also alleges that the long defendants, with the assistance of the other defendants, disseminated false reports in violation of the CEA. The district court dismissed both aspects of this claim. On appeal, Apex only presents arguments for reinstatement of the manipulation aspect of the claim. We therefore leave undisturbed the district court's dismissal of the false reports aspect of the claim.

The district court dismissed the manipulative scheme portion of the claim because while Apex alleged that the long defendants together engineered a squeeze, it failed to present evidence of a conspiracy. 641 F. Supp. at 1275. Having reversed this determination with respect to Belcher in the context of the antitrust claims, we also do so in the context of the commodities manipulation claim.

Apex urges us to reverse as to all the long defendants on this claim despite the lack of evidence of a conspiracy, suggesting that each defendant alone could have manipulated the market. However, the amended complaint did not allege that each long defendant manipulated the market. Rather, it alleges that the "long defendants" did so. Since the complaint alleges collective action, this suit cannot be allowed to proceed as if it alleges individual action.

We do, however, reverse the dismissal of the manipulation allegation against Belcher. The dismissal of the fifth claim is in all other respects affirmed.

Apex alleges as its seventh claim that the Exchange defendants failed to remedy the alleged market manipulation by the long defendants in violation of the CEA, 7 U.S.C. Secs. 7(d), 7a(8). The district court dismissed this claim because it found that Apex failed to present evidence tending to show that the Exchange acted or failed to act as a result of self-interest or bad faith as interpreted in a number of this court's decisions. See, e.g., Brawer v. Options Clearing Corp., 807 F.2d 297 (2d Cir.1986); Sam Wong & Son v. New York Mercantile Exchange, 735 F.2d 653 (2d Cir.1984). Apex on appeal argues that a number of statements by persons connected with the Exchange, most notably Julian Raber, could lead a reasonable juror to infer the existence of an improper motive in light of all of the circumstances herein. Having carefully reviewed the allegations and the proffered evidence, we affirm the district court's dismissal of the claim of failure to regulate for the reasons stated in the district court opinion. 641 F. Supp. at 1276-81.

The only remaining claims are those alleging (1) common law fraud by the long defendants and (2) a violation of section 4(b) of the CEA by the long defendants. Apex does not request reversal of the dismissal of these claims in its brief and offers no arguments suggesting why reversal would be appropriate. We therefore do not consider these claims.

Accordingly, for the reasons stated above, we reverse the dismissal of the antitrust and commodities market manipulation claims as to Belcher. These claims are remanded for further proceedings consistent with this opinion. As noted in note 1, supra, this appeal has been dismissed as to Northeast. The judgment of the district court in all other respects is affirmed.