Fujifilm Manufacturing U.S.A., Inc. v. Goldman Sachs & Co.

UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF NEW YORK IN RE ALUMINUM WAREHOUSING ANTITRUST LITIGATION This Document Relates To: In re Aluminum Warehousing Antitrust Litigation (Direct Purchaser Plaintiffs), No. 14 Civ. 3116 (PAE) (S.D.N.Y.) Agfa Corporation and AGFA Graphics, 13 MD 2481 (PAE) NV v. The Goldman Sachs Group, Inc., 14 Civ. 3116 (PAE) No. 14 Civ. 211 (PAE) (S.D.N.Y.) 14 Civ. 211 (PAE) 14 Civ. 217 (PAE) Mag Instrument, Inc. v. The Goldman Sachs 14 Civ. 6849 (PAE) Group, Inc., 15 Civ. 8307 (PAE) No. 14 Civ. 217 (PAE) (S.D.N.Y.) OPINION & Eastman Kodak Company v. The Goldman ORDER Sachs Group, No. 14 Civ. 6849 (PAE) (S.D.N.Y.) Fujifilm Manufacturing U.S.A., Inc. v. Goldman Sachs & Co., No. 15 Civ. 8307 (PAE) (S.D.N.Y.) PAUL A. ENGELMAYER, District Judge: Plaintiffs in these actions1 contend that defendants have violated section 1 of the Sherman Act, 15 U.S.C. § 1, by conspiring to inflate prices in the primary-aluminum market.2 Specifically, they allege that defendants’ anticompetitive conduct increased regional premiums associated with primary-aluminum sales. This, plaintiffs contend, had the effect of inflating the prices they paid for such aluminum during the relevant period. The present motion does not concern whether plaintiffs’ theory is right, but whether they are the right parties to bring this lawsuit. For the most part, plaintiffs did not buy aluminum from defendants or their co-conspirators. Instead, they overwhelmingly bought aluminum from unrelated third parties: generally, smelters of aluminum. Defendants argue that this fact, viewed in light of the operation of the aluminum market as revealed by the extensive discovery record, is fatal to the claims of such indirect-purchaser plaintiffs. Specifically, given the causal distance between defendants’ conduct and plaintiffs’ harms, defendants contend that such plaintiffs are not efficient enforcers of the antitrust laws, and therefore lack antitrust standing. Plaintiffs counter that their aluminum purchases, even if made at some remove from any defendant, unavoidably incorporated the price component that defendants conspired to inflate. This debate draws upon 1 The pending motions relate to the claims brought by the four plaintiffs in 14 Civ. 3116 (“First Level Purchasers” or “FLPs”) and the four plaintiffs bringing individual actions against various defendants: (1) Agfa Corporation and AGFA Graphics, NV (together, “Agfa”), 14 Civ. 211; (2) Mag Instrument Inc. (“Mag”), 14 Civ. 217; (3) Eastman Kodak Company (“Kodak”), 14 Civ. 6849; and (4) Fujifilm Manufacturing U.S.A., Inc. (“Fujifilm,” and together with Agfa, Mag, and Kodak, the “Individual Purchasers” or “IPs”), 15 Civ. 8307. 2 “The term ‘primary aluminum’ is used in the industry to describe aluminum in the form produced at a smelter or primary aluminum plant, by original producers, as distinguished from ‘secondary aluminum,’ which is reconstituted aluminum scrap.” Eastman Kodak Co. v. Henry Bath LLC, 936 F.3d 86, 88 (2d Cir. 2019) (“Aluminum VI”). a growing line of cases in this Circuit, in which courts have largely dismissed, for want of antitrust standing, plaintiffs’ claims that defendants with whom they did not directly do business manipulated benchmark prices. Plaintiffs here seek to distinguish this line of authority. They also maintain that prior decisions in this case, by the judge to whom this case was previously assigned and by the Second Circuit, have already resolved that they have antitrust standing. With fact discovery complete as to the claims at issue, defendants move for summary judgment on those of plaintiffs’ claims that arise from purchases of primary aluminum from parties other than defendants. For the reasons below, the Court grants defendants’ motion. I. Background3 A. The Parties The First Level Purchaser plaintiffs are Ampal, Inc. (“Ampal”); Custom Aluminum Products, Inc. (“Custom”); Claridge Products and Equipment, Inc. (“Claridge”); and Extruded Aluminum Corporation (“Extruded”). TAC ¶¶ 44, 53, 60, 67. 3 The Court assumes familiarity with the facts and lengthy procedural history of this litigation. See, e.g., In re Aluminum Warehousing Antitrust Litig., No. 13 MD 2481 (KBF), 2014 WL 4277510 (S.D.N.Y. Aug. 29, 2014) (“Aluminum I”); In re Aluminum Warehousing Antitrust Litig., 95 F. Supp. 3d 419 (S.D.N.Y. 2015) (“Aluminum II”); In re Aluminum Warehousing Antitrust Litig., 833 F.3d 151 (2d Cir. 2016) (“Aluminum III”); Aluminum VI, 936 F.3d 86; In re Aluminum Warehousing Antitrust Litig., 336 F.R.D. 5 (S.D.N.Y. 2020) (“Aluminum VII”). The Court’s account of the facts here draws mainly from the parties’ submissions in connection with the pending motion for summary judgment. The Court has considered the following submissions by defendants in support of their motion: defendants’ Local Rule 56.1 statement, Dkt. 1291 (“Def. 56.1”); the declaration of John S. Playforth, Esq., Dkt. 1290 (“Playforth Decl.”), and exhibits thereto; and defendants’ reply to plaintiffs’ counter-statement to defendants’ Local Rule 56.1 statement, Dkt. 1305 (“Def. Reply 56.1”). The Court has also considered the following submissions by plaintiffs in opposition to the motion: plaintiffs’ counter-statement to defendants’ Local Rule 56.1 statement, Dkt. 1300 (“Pl. 56.1”); the declaration of Patrick J. Coughlin, Esq., Dkt. 1298 (“Coughlin Decl.”), and exhibits thereto, in support of the FLPs’ opposition; and the declaration of Derek Y. Brandt, Esq., Dkt. 1301 (“Brandt Decl.”), and exhibits thereto, in support of the IPs’ opposition. To the extent cited by the parties in their summary judgment papers, the Court also considers the evidence submitted in connection with the FLPs’ motion for class certification. In addition, solely to provide context and without The Individual Purchaser plaintiffs are Agfa, Mag, Kodak, and Fujifilm. See JAC ¶¶ 31–49; Fujifilm Compl. ¶¶ 34–39. Plaintiffs seek damages arising from their purchases of aluminum at allegedly inflated prices. They have sued six sets of defendants, three of which traded in primary aluminum and primary-aluminum derivatives on the London Metals Exchange (“LME”) during the relevant period (“Financial Defendants”), and three of which owned and operated LME-certified warehouses for the storage of metal during that period (“Warehousing Defendants”). See TAC ¶¶ 85–127; see also Aluminum VI, 936 F.3d at 89. The Financial Defendants are each affiliated with Goldman Sachs & Co., J.P. Morgan Securities plc, or Glencore Ltd. Each Warehousing Defendant became associated with one of the Financial Defendants during the relevant period. Defendant Goldman Sachs & Co. is an international financial company headquartered in New York, New York. See TAC ¶¶ 86–87. Defendant J. Aron & Company, a commodities trading firm, is a New York corporation with the same headquarters as Goldman Sachs & Co. Id. ¶ 88. Defendant Goldman Sachs International is an international financial services provider headquartered in London, United Kingdom. Id. ¶ 89. Each of these entities (collectively, the accepting those allegations as true, the Court draws on allegations in the Third Amended Complaint, Dkt. 738 (“TAC”), the consolidated complaint of the Individual Purchasers, Dkt. 745 (“JAC”), and Fujifilm’s separate amended complaint, 15 Civ. 8307, Dkt. 35 (“Fujifilm Compl.”). Citations to a party’s 56.1 statement incorporate the evidentiary materials cited therein. When facts stated in a party’s 56.1 statement are supported by testimonial, video, or documentary evidence and not denied by the other party, or denied by a party without citation to conflicting admissible evidence, the Court finds such facts to be true. See S.D.N.Y. Local Civil Rule 56.1(c) (“Each numbered paragraph in the statement of material facts set forth in the statement required to be served by the moving party will be deemed to be admitted for purposes of the motion unless specifically controverted by a correspondingly numbered paragraph in statement required to be served by the opposing party.”); id. Rule 56.1(d) (“Each statement by the movant or opponent . . . controverting any statement of material fact[] must be followed by citation to evidence which would be admissible, set forth as required by Fed. R. Civ. P. 56(c).”). “Goldman Sachs defendants”) is a subsidiary of The Goldman Sachs Group, Inc. (together with the Goldman Sachs defendants, “Goldman Sachs”). Id. ¶¶ 85–89.4 Defendant J.P. Morgan Securities plc (“J.P. Morgan”), headquartered in London, provides securities brokerage services. Id. ¶ 97. J.P. Morgan is a wholly-owned subsidiary of J.P. Morgan Chase Bank, N.A. (“JPMCB”). Id.5 Defendant Glencore Ltd. (“Glencore”), headquartered in Stamford, Connecticut, trades aluminum and other metals, as well as financial derivatives tied to the underlying asset prices of such metals. Id. ¶¶ 113, 121.6 The Warehousing Defendants are Metro International Trade Services LLC (“Metro”), Mitsi Holdings LLC (“Mitsi”), Henry Bath LLC (“Henry Bath”), and Pacorini Metals USA LLC (“Pacorini”). Id. ¶¶ 92–93, 107, 122. Each Warehousing Defendant “owns and operates aluminum warehouses certified by the LME, and each was owned during the relevant time by one of the Financial Defendants—Henry Bath by J.P. Morgan, Metro [and Mitsi] by Goldman Sachs, and Pacorini by Glencore.” Aluminum VI, 936 F.3d at 89; see Pl. 56.1 ¶ 62. 4 The Honorable Katherine B. Forrest, to whom this case was previously assigned, dismissed all claims against The Goldman Sachs Group, Inc. In re Aluminum Warehousing Antitrust Litig., No. 13 MD 2481 (KBF), 2015 WL 1344429, at *3 (S.D.N.Y. Mar. 23, 2015). 5 Judge Forrest denied the FLPs’ motion to add JPMCB as a defendant, In re Aluminum Warehousing Antitrust Litig., No. 13 MD 2481 (KBF), 2016 WL 1629350, at *7–9 (S.D.N.Y. Apr. 25, 2016), and dismissed all claims against J.P. Morgan Chase & Co., In re Aluminum Warehousing Antitrust Litig., 2015 WL 1344429, at *3. 6 The TAC named several additional entities as “Glencore Defendants.” TAC ¶ 110. Judge Forrest dismissed Glencore plc for lack of personal jurisdiction, In re Aluminum Warehousing Litig., 90 F. Supp. 3d 219, 232 (S.D.N.Y. 2015), and denied leave to amend to add Glencore International AG and Glencore UK Ltd., In re Aluminum Warehousing Litig., 2016 WL 1629350, at *6–9; In re Aluminum Warehousing Antitrust Litig., 2015 WL 1344429, at *3. B. Factual Background The Court assumes familiarity with the factual background of this case. See Aluminum VII, 336 F.R.D. at 10–22. It summarizes here only the background that is necessary context for the pending motion. 1. The Market for Primary Physical Aluminum Primary aluminum is created by large, integrated producers (smelters) such as Alcoa, Inc. (“Alcoa”), UC Rusal (“Rusal”), and Rio Tinto Alcan, Inc. (“Rio Tinto Alcan”). Pl. 56.1 ¶¶ 3 & n.4, 4. Such producers generally sell primary aluminum to (1) manufacturers and processors (“industrial users” or “fabricators”), such as the FLPs and IPs, which use aluminum in either industrial processes or to fabricate finished products; and (2) traders and financial institutions, such as defendants, which hold and resell aluminum through various methods. Industrial users buy aluminum both under long-term contracts, which may include annual supply arrangements, and in one-off purchases on the spot market. See, e.g., Def. 56.1 ¶¶ 12–13, 18–19, 23–24, 29, 45–46, 51, 58. In most cases, contracts for the physical delivery of aluminum contain a purchase price comprised of at least two main components: (1) the LME settlement price for aluminum; and (2) a premium reflecting, inter alia, the costs of storing and delivering the aluminum. Def. 56.1 ¶ 54; Pl. 56.1 ¶ 54. Such contracts also may include another, smaller price component—a “conversion cost”—based the specifics of the purchased product, e.g., metal grade or shape. Def. 56.1 ¶ 54; Pl. 56.1 ¶ 54. According to defendants, the LME settlement price is the largest component of the total price of aluminum, the regional premium is the second largest, and the conversion price is the smallest. See Dkt. 1311 (“Arg. Tr.”). at 18–19; see also Def. 56.1 ¶ 56 (during the relevant period, the MWP ranged between five and 23 percent of the MWTP). The LME settlement price represents the current value of the metal. It fluctuates in accord with, inter alia, the daily global intersection of supply and demand in LME trading. Pl. 56.1 ¶ 57. It does not reflect the costs of delivery from a seller to a buyer. Id. The inclusion of a regional premium in the price of aluminum incorporates into the overall price of aluminum the costs of delivery, storage, financing, and other costs associated with the logistics of physical aluminum sales. Def. 56.1 ¶ 57; Pl. 56.1 ¶ 57. In the United States, the major benchmark regional premium is the Midwest Premium (“MWP”) reported by the Platts publishing company. (Despite its name, the MWP is the predominant benchmark throughout the country.) To calculate the MWP, Platts first surveys the all-in prices paid on the spot market for aluminum to be delivered in the midwestern United States on a given day. Def. 56.1 ¶ 54. That surveyed price, known as the Midwest Transaction Price (“MWTP”), reflects the “[d]aily all- inclusive or ‘all-in’ price for spot physical 99.7% high-grade P1020A aluminum.” Id. The MWP, in turn, represents the mathematical difference between the MWTP and the LME settlement price for aluminum on any given day. Id. ¶¶ 54–55; Pl. 56.1 ¶¶ 54–55. During the relevant period, the MWP ranged between about five and 23 percent of the overall MWTP. Def. 56.1 ¶ 56. As discussed below, however, not all contracts for primary aluminum explicitly refer to the LME settlement price, MWP, or MWTP. 2. The LME System The LME is the world’s leading exchange for non-ferrous industrial metals, including aluminum. Dkt. 1014 (“Liu Decl.”), Ex. 1 (“First Hausman Decl.”) ¶ 8. Its trading and clearing platform allows buyers and sellers to trade standardized futures contracts that provide for the physical delivery of metal on specified maturity dates. Id. The LME also certifies warehouses around the world, including in Detroit, Baltimore, and New Orleans in the United States, and Vlissingen in the Netherlands. Id. ¶ 9. LME futures contracts can be settled either through an offsetting trade or through the delivery of “warrants” issued by LME-certified warehouses. Id. ¶¶ 9–10. An LME warrant entitles its bearer to a specified lot of metal stored at one of the LME warehouses. Id. The vast majority of LME aluminum trading occurs among traders and financial institutions who speculate on price movements and have no expectation of taking delivery of the actual physical metal. And the overwhelming majority of physical aluminum never passes through an LME warehouse. Generally, industrial users buy physical aluminum directly from smelters, not through the LME. Id. ¶ 19. As a result, the aluminum stored in LME warehouses usually is not delivered out. Still, a buyer of an LME futures contract can choose to settle that contract through delivery. In that instance, the buyer receives a warrant for physical metal, which they can then “cancel,” in industry parlance, to order their aluminum for delivery out of the warehouse. Id. ¶ 10. Holders of LME warrants are also responsible for paying rent to the LME warehouse where the warranted metal is located. See Aluminum III, 833 F.3d at 155 (“LME-warehouses make money by charging rent while the aluminum is stored, and exit penalties when the aluminum leaves[.]”). A party that cancels a warrant must continue to pay rent to the relevant LME warehouse until the warehouse finishes loading out the cancelled metal. 3. The Alleged Conspiracy The Second Circuit has summarized plaintiffs’ conspiracy allegations as follows: Because of the convention to use the Midwest Premium as a price element in sales of primary aluminum, the Financial Defendants allegedly devised a conspiratorial plan to artificially inflate the Midwest Premium, thereby artificially inflating the prices they could realize in later selling off their large aluminum holdings. Working together with the Warehousing Defendants they had recently acquired, the Financial Defendants undertook to inflate the Midwest Premium by artificially lengthening the delays involved in taking delivery from LME warehouses. Their strategies to accomplish this included offering incentives to producers to move even more aluminum into their warehouses; agreeing to treat the minimum load-out rate set by the LME as the maximum rate of load-outs (effectively capping the number of lots that would be loaded out of their warehouses each day); and, most importantly, needlessly canceling their warrants for aluminum they had purchased, causing aluminum to be shuttled back and forth among the Defendant Warehouses. The consequence was to cause long delays between cancellation of a warrant and actual load-out of a covered lot. The plaintiffs allege that in this fashion, the defendants increased delivery delays from about six weeks pre-2011 to nearly two years (approximately 625 days) by 2014. This caused the Midwest Premium to triple, from about 6.45 cents per pound in 2011 to 20 cents per pound by 2014. Accordingly, because the Midwest Premium was a component of the price the Financial Defendants would receive on selling their aluminum holdings, that price increased to the same extent. Aluminum VI, 936 F.3d at 92. As reflected in the summary judgment record, plaintiffs allege that defendants began their conspiracy in 2009—when the Financial Defendants began to gain control of the LME warehousing system. Pl. 56.1 ¶ 62. By 2010, Goldman Sachs had acquired Metro, J.P. Morgan had acquired Henry Bath, and Glencore had acquired Pacorini. Id. After doing so, the Financial Defendants also began to acquire millions of metric tons in aluminum holdings across the globe, valued in the billions of dollars. Id. ¶¶ 64–68. They stored much of that aluminum at Metro’s Detroit warehouses. Id. ¶ 69; Def. Reply 56.1 ¶ 69. Next, defendants began to cancel large numbers of LME warrants, allegedly intending to create massive loadout queues from their warehouses and thereby “restrain physical supply and increase the price of aluminum.” Pl. 56.1 ¶ 83; see id. ¶¶ 69–82; see also Aluminum VII, 336 F.R.D. at 19–22 (summarizing allegations of “concerted multi-pronged efforts of the defendants” to lengthen queues). According to plaintiffs, those elongated queues served to “drive up the MWP.” Dkt. 1296 (“Pl. Opp’n”) at 7; see Pl. 56.1 ¶ 81 (defendants “knew that queues drive up the regional premia”); id. ¶ 82 (defendants “intended to and did drive up the MWP”).7 Last, defendants sold physical aluminum to customers—largely not including plaintiffs— at prices allegedly inflated by the elevated MWP component. See Def. Reply 56.1 ¶ 84 (denying accuracy of amounts sold and price terms, but admitting that each defendant “sold physical aluminum to consumers during the relevant period”); Pl. 56.1 ¶ 84 (J.P. Morgan sold aluminum to consumers “frequently”). In short, by hoarding aluminum and lengthening loadout delays at LME warehouses, plaintiffs allege, defendants caused the MWP to rise, “thereby artificially inflating the prices they could realize in later selling off their large aluminum holdings.” Aluminum VI, 936 F.3d at 92. 4. Plaintiffs’ Aluminum Purchases a. Sources of Purchased Aluminum For the most part, defendants did not sell aluminum to plaintiffs. During the alleged conspiracy, plaintiffs instead bought the vast majority of their aluminum from aluminum smelters such as Rio Tinto Alcan, Alcoa, or Rusal. See Def. 56.1 ¶¶ 2–4, 7, 31; Pl. 56.1 ¶¶ 2–4, 7, 31. 7 Defendants dispute that their warrant cancellations were coordinated, see Def. Reply 56.1 ¶ 69, that lengthened queues caused the overall price of aluminum to rise, id. ¶¶ 73–74, and that they intended to and did inflate the MWP, id. ¶¶ 81–83. They admit that their warrant cancellations could have caused queues to lengthen, all else equal, but maintain that “canceling warrants,” and the particular decisions that caused the largest queue increases, “made unilateral economic sense” under the circumstances. Id. ¶ 75; see id. ¶¶ 76–80. They also admit that lengthened queues “in some circumstances could increase the portion of all-in aluminum prices attributable to a regional premium by depressing the LME cash price relative to the all-in price,” but dispute that queues inflate the all-in prices for aluminum. Id. ¶ 81. Because it is “useful to distinguish the question of whether an antitrust violation occurred from whether plaintiffs have standing to pursue it,” the Court, for purposes of the present motion, resolves these disputes in plaintiffs’ favor, “assum[ing] the alleged violation and assess[ing] only plaintiffs’ standing to pursue their claim.” Daniel v. Am. Bd. of Emergency Med., 428 F.3d 408, 437 (2d Cir. 2005); see Def. 56.1 ¶ 83 (confirming that “merits issues have no relevance to Plaintiffs’ efficient enforcer standing”). Plaintiffs do not allege, nor have they adduced evidence, that any of those entities participated in the above-described scheme. Pl. 56.1 ¶¶ 6, 33. Ampal is the only plaintiff at issue to have bought any aluminum from a defendant. Between February 1, 2010 and March 25, 2016, Ampal bought 50,048 metric tons of aluminum for which it seeks damages. Def. 56.1 ¶ 9; Pl. 56.1 ¶ 9; Ex. 1 (“Second Hausman Decl.”) tbl.1.8 More than 95% of those purchases were from non-defendants Rio Tinto Alcan or Alcoa. Fewer than 5% were from defendants or an alleged co-conspirator. Second Hausman Decl., tbl.1 (47,876 metric tons from Alcoa and Rio Tinto Alcan; 2,172 metric tons from Glencore and Century). No other plaintiff bought any such aluminum from a defendant or alleged co-conspirator during the relevant period. See Def. 56.1 ¶¶ 2–3, 7–8, 31, 34; Pl. 56.1 ¶¶ 2–3, 7–8, 31, 34. The parties agree that Claridge, Custom, Extruded, Agfa, Fujifilm, Kodak, and Mag each bought 100% of the relevant aluminum from non-party entities not alleged to have participated in any scheme to manipulate aluminum prices. See Def. 56.1 ¶¶ 17, 22, 28, 36, 40, 44, 50; Pl. 56.1 ¶¶ 17, 22, 28, 36, 40, 44, 50. Thus, of the roughly 435,800 metric tons of first-level aluminum purchases for which plaintiffs seek damages, they bought only about 2,200, or 0.5%, from any entity involved in the alleged conspiracy. See Def. Reply 56.1 ¶ 3; Def. 56.1 ¶¶ 8, 34; Pl. 56.1 ¶¶ 8, 34. b. Terms of Aluminum Purchases The above facts about the sources of plaintiffs’ aluminum purchases are undisputed. But the parties vigorously disagree as to the significance of various surrounding facts. Defendants 8 As the parties have sequentially numbered their exhibits, the Court refers to each only by number. For the sake of clarity, exhibits 1–74 refer to the exhibits attached to the Playforth Declaration; exhibits 75–145 refer to the exhibits attached to the Coughlin Declaration; and exhibits IP1–IP28 refer to the exhibits attached to the Brandt Declaration. contend that at least three features of plaintiffs’ aluminum transactions with non-conspiring parties attenuate their injuries from defendants’ actions. First, defendants note that many purchase contracts at issue do not refer to the MWP or MWTP at all. Consistent with the existence of contracts of this nature, defendants contend that non-conspirator aluminum producers made pricing decisions independent of changes in the MWP and often did not charge, or had the option not to charge, the allegedly manipulated MWP in full. Second, defendants contend that plaintiffs were able to, and often did, hedge against fluctuations in the MWP by entering into “fixed-forward” contracts—agreements specifying a set price for up to a year’s worth of aluminum purchases—that avoided the effects of changes in either the LME or MWP in the meantime. Third, defendants maintain that plaintiffs were able to, and successfully did, negotiate all- in prices for their aluminum purchases, which could compensate for increases in the MWP, even if plaintiffs could not directly negotiate the MWP component itself, assuming it was charged. Plaintiffs disagree with each proposition. The Court briefly summarizes the relevant evidence on each point. i. MWP Terms in Contracts Defendants observe that most of the relevant plaintiffs’ aluminum purchases included at least some purchases made under agreements that lacked an express MWP or MWTP price component (or such component based on another regional premium). See Def. 56.1 ¶¶ 12–13 (Ampal), 18–19 (Claridge), 23–24 (Custom), 45–46 (Kodak), 51 (Mag), 58. More broadly, they note that, as plaintiffs have admitted, roughly 31% of Alcoa and Rio Tinto Alcan contracts during the relevant period did not include any express reference to the MWP or MWTP. See id. ¶ 58. As to the FLPs specifically, defendants cite several specific agreements in which a plaintiff contracted to pay a specified price per pound of aluminum, without reference to the MWP or MWTP.9 As to their spot purchases, defendants note that such were often made pursuant to informal or verbal contracts, without evidence that the MWP or MWTP were incorporated into their terms.10 As to the IPs, defendants similarly note that Kodak and Mag entered into fixed-price contracts without a floating MWP or MWTP component.11 Thus, as the Court observed in recently denying the FLPs’ motion for class certification, “many purchase contracts do not reference the MWTP or MWP at all, and those that do reference the MWTP or MWP do so in a variety of ways.” Aluminum VII, 336 F.R.D. at 14; see id at 64. Although defendants do not contend that all plaintiffs entered into contracts that lacked a reference to either the MWP or MWTP, they contend that many did, and that all at least could 9 See, e.g., Ex. 8 at 2–3 (Alcoa quoted to Ampal “a fixed (Hedge) price for” deliveries “instead of the MidWest Transaction Price for December average”); Ex. 15 (Claridge contract with smelter charging dollar amount per pound without reference to MWTP or MWP); Ex. 19 at 2 (contract between Custom and Noble for three months’ of deliveries at fixed price each month, without reference to MWP or MWTP). Defendants state that Extruded also entered into such contracts, but the agreements cited include a “firm MWTP” in each, representing “ingot plus [] billet conversion” (though not referring to the MWP itself), covering several months’ worth of deliveries. See, e.g., Ex. 24–26. 10 See Ex. 2 (“Def. Ramsey Tr.”) at 6–7 (Ampal Rule 30(b)(6) witness describing “verbal,” or “phantom contract”); Ex. 4 (“Def. Dillett Tr.”) at 159 (Custom Rule 30(b)(6) witness explaining “[a] spot purchase would not have a contract; therefore, the pricing is not fixed to us”); Def. 56.1 ¶ 19 (describing Claridge spot purchases). 11 See, e.g., Ex. IP16 at 5 (four-year supply agreement between Kodak and Alcoa which included two pricing methodologies, between which Kodak could choose, and defining one, “Fixed Forward Price,” as the LME “plus contango,” and the other, “Month Prior Pricing,” as the Platt’s reported average, while including price adjustment if MWP falls outside certain range during contract term); Ex. 48 (“Def. Zecchini Tr.”) at 96 (Mag 30(b)(6) witness agreeing that fixed- forward contract “locks in the all-in price”). But see Ex. 45 at 1 (Kodak contract with Alcoa fixing “firm metal price” but also noting that such price “will be the Midwest US Transaction price and fixed forward for each purchase order”). have done so, given the prevalence of producer contracts that lacked a MWP or MWTP term. See Dkt. 1287 (“Def. Mem.”) at 17–18.12 Plaintiffs acknowledge, as the face of the contracts makes undisputable, that plaintiffs entered into fixed-price contracts that lacked express reference to the MWP or MWTP. But they contend, relying on their own deposition testimony, that each transaction nevertheless implicitly incorporated the MWP, and that all aluminum contracts, in one way or another, did so. For example, Ampal’s Rule 30(b)(6) witness testified that “long-term contracts” are “based on [the] metals weight transaction price,” that “[a]ll purchases are based on the Midwest transaction price plus a premium,” and that “fixed forward contracts” are always “based on a Midwest transaction price, which includes a premium.” Ex. 88 (“Pl. Ramsey Tr.”) at 36–37, 68–69, 109.13 Similarly, plaintiffs cite testimony that, even absent an express contract term, sellers of aluminum in spot purchases also always based their prices on the MWP.14 12 Defendants also note that the IPs’ contracts often included European regional premiums, not the MWP. See Def. 56.1 ¶¶ 37, 41, 45–46, 51. Plaintiffs argue that this fact is immaterial given that the defendants have not otherwise discussed the relevance of such premiums or based their motion on their import. See Pl. 56.1 ¶¶ 37, 41. On this point, the Court agrees with plaintiffs, and does not view the existence of contracts including European regional premiums as material to the pending motion. 13 See also, e.g., Ex. 94 (“Pl. Hall Tr.”) at 67 (Extruded Rule 30(b)(6) witness confirming that all its purchases included the MWP); Ex. IP28 (“Pl. Zecchini Tr.”) at 123 (Mag Rule 30(b)(6) witness testifying that, for “forward buys,” “the aluminum cost is based on the LME, Midwest conversion cost”). 14 See, e.g., Pl. Ramsey Tr. at 37 (Ampal stated that spot-purchase price “would always be the same, metals weight transaction price, plus the premium, depending on the delivery method”); id. at 47 (spot purchases “would always be based on an LME plus the Midwest”); Ex. 93 (“Pl. Dillett Tr.”) at 157 (“components of the spot price” were “the current LME for the day, the current Midwest premium for the day, and the conversion charge to make that into billet”). Although not material to the pending motion, plaintiffs’ testimony suggests an odd circularity in aluminum pricing. The MWTP, as reported by Platts, purports to reflect actual prices paid in spot-market transactions; but each such transaction, according to plaintiffs, invariably reflects the MWTP. In sum, defendants argue that the contracts associated with many of plaintiffs’ aluminum purchases did not expressly refer to the MWP or MWTP. Plaintiffs do not dispute that fact, but contend that the price for every aluminum purchase at issue in this case was based on the MWTP and included the MWP, regardless of the presence of any such contract term. ii. Hedging Next, defendants contend that the plaintiffs were able to, and did, hedge against changes in the MWP. Def. 56.1 ¶¶ 14, 18, 25, 29, 47, 52. In support, they cite several “fixed-forward” contracts plaintiffs entered into, which “lock[ed] in” specific prices they paid for many months at a time.15 Plaintiffs do not appear to dispute that they could, and did, hedge at least part of the all-in price they paid for aluminum.16 But they cite testimony suggesting that they were unable to specifically hedge the MWP.17 Thus, although plaintiffs dispute whether they could fix the MWP term in their contracts, it appears undisputed that some could “lock[] in the all-in price” of aluminum, Def. Zecchini Tr. at 96, of which the MWP was a component. 15 Ex. 8 at 3; See, e.g., Def. Ramsey Tr. at 71 (fixed contracts were “a way that [Ampal] would hedge”); Ex. 10 (Ampal “hedged” agreement with Alcoa providing a “fixed, firm price”); Ex. 3 (“Def. Hardcastle Tr.”) at 16 (Claridge Rule 30(b)(6) witness described “forward bias” contracts as “contracts or transactions whose intent was to hedge the price of aluminum for Claridge”); Ex. 30 (“Starfinger Tr.”) at 212–13 (Kodak Rule 30(b)(6) witness agreeing that, under fixed-price contract, “during that period, if . . . the Midwest Premium rose, that would not have affected Kodak’s purchase price”); id. at 215 (“Q. So there are some cases in which Kodak is able to enter into aluminum purchase agreements that fix the premium? A. Yes.”); Pl. 56.1 ¶ 52 (agreeing that “Mag’s forward contracts generally provide stability of the all-in price for aluminum of which the Midwest Premium is only one part”). 16 See, e.g., Pl. 56.1 ¶¶ 14, 18, 25 (“[O]nly the LME price could be hedged.”), 47 (Kodak could “fix the LME price portion of the aluminum expense.”), 52 (“Mag would agree to a fixed all-in price for future prices of aluminum with its suppliers” when they allowed it.). 17 See Pl. 56.1 ¶¶ 18, 25, 47; Ex. IP14 (“Martin Tr.”) at 151 (“Q. Was there ever a contract in which Kodak fixed-forward the premium, a regional premium as well? A. No.”); Ex. IP8 (“Walde Tr.”) at 164 (Q. “Did the proposal also contemplate [Fujifilm] fixing a regional premium? A. No, the only thing that was fixed was the LME portion.”). iii. Negotiations Last, defendants argue that each plaintiff had the ability to—and many successfully did— negotiate the all-in price they would pay for aluminum to the producers from whom they bought the vast majority of their aluminum. Def. 56.1 ¶¶ 15, 20, 26, 30, 38, 42, 48, 53. In response, plaintiffs do not appear to dispute that they could negotiate portions of the price they ultimately paid for aluminum.18 But they do contest whether such negotiations could have affected the MWP component of the all-in price. In support, they cite testimony that they could never negotiate the MWP term in any given contract.19 For example, when asked if Ampal ever tried to negotiate the MWP term in its contracts with a supplier, Ampal’s Local Rule 30(b)(6) witness responded: “Fait accompli. Here it is, you want ingot, this is what you pay. If you don’t want ingot, we don’t have to sell it to you.” Pl. 56.1 ¶ 15. But Kodak, at least, reported negotiating price components, such as conversion fees, directly “in response to the rising premiums.” Martin Tr. at 102; see id. at 138 (“There is no negotiating regional premiums. So we did everything we could to lower our total net cost. . . . [We] didn’t really care where that came from, [we] just cared about the net . . . price.”). Thus, plaintiffs admit that each had the ability to negotiate certain aspects of the all-in price of aluminum, and therefore the total price 18 See, e.g., Pl. 56.1 ¶¶ 15 (describing negotiations “as to shape, purity, location, cash payment, and delivery method”), 20 (same), 26 (noting successful negotiations over “conversion rate”). 19 See also, e.g., Ex. 92 (“Pl. Hardcastle Tr.”) at 85–86 (“[Claridge] tried, but . . . there was very little negotiation” as to “[f]abrication and length premiums,” and no negotiations on the MWP); Pl. Hall Tr. at 287 (Extruded could negotiate over the “direct sale price, the fabrication price, whatever you want to call it for the billet conversion and/or toll. Those are the numbers that we negotiate. The prior month Midwest average as, again, stated earlier, we have no control over.”); Ex. IP1 (“Pellegroms Tr.”) at 136 (Agfa Rule 30(b)(6) witness testified that “[t]here are two elements that we can negotiate. That is the conversion costs and transport costs.”); id. at 41–42 (“[O]ur aluminum suppliers could not do anything about the premium pricing” or “for the metal price part. . . . They are not allowed to sell us aluminum without the premium.”); Walde Tr. at 112–13 (“The regional premiums were always fixed . . . once it was settled, the original discussion . . . . We negotiated conversion fees, and that is really it.”). they paid for aluminum. But, plaintiffs emphasize, they could not do so as to the specific MWP term, which was pre-determined by market forces outside plaintiffs’ or producers’ control. C. Pertinent History of This Litigation 1. Relevant MDL History In December 2013, this MDL was created and transferred to this District for consolidated and coordinated pretrial proceedings and assigned to then-Judge Forrest. Dkt. 1. The ensuing history of the case is complex. The case was reassigned to this Court in September 2019, after Judge Forrest’s retirement and after the Second Circuit reversed her decision dismissing plaintiffs’ claims. See Dkt. 1142. The Court highlights here only the decisions germane to the instant motion for summary judgment—in particular, aspects of decisions by Judge Forrest or the Second Circuit that plaintiffs contend make it law of the case that they have antitrust standing. Of most significance since this Court assumed supervision, the Court, on July 23, 2020, in Aluminum VII, denied plaintiffs’ motion for class certification. a. Aluminum II On March 26, 2015, Judge Forrest held that first-level purchasers were efficient enforcers, and therefore had antitrust standing, because they made purchases “directly from a producer (and, in one instance, from one of the defendants),” with “no buyer of aluminum other than defendants . . . higher up or more direct in their respective distribution chains.” Aluminum II, 95 F. Supp. 3d at 444. Judge Forrest recognized that only one plaintiff had bought aluminum directly from a defendant or alleged co-conspirator. Id. at 442 n.24, 444 n.27. Nonetheless, she held that, because “the contracts between the producer-seller and plaintiff-buyer allegedly contain provisions tying the contract prices to the Midwest Premium,” and because the IPs alleged that “in the United States it is not possible to negotiate spot or long-term aluminum contracts with major producer Alcoa that do not incorporate the Midwest Premium,” plaintiffs’ harms were direct enough to support antitrust standing. Id. at 444. In so holding, Judge Forrest rejected defendants’ argument that “users of defendants’ warehouse services” were better- positioned to prosecute the alleged violations. Id. And she held that plaintiffs’ damages, as pled, were non-speculative because they could be “defined by the amount by which the Midwest Premium was inflated.” Id. Her decision, however, left “for another day” more specific findings about whether such damages had been incurred and, if so, whether “any amount is determinable.” Id. b. Aluminum IV On October 5, 2016, after an intervening decision by the Second Circuit regarding certain plaintiffs who had been dismissed in Aluminum II, Judge Forrest dismissed the TAC. See In re Aluminum Warehousing Antitrust Litig., No. 13 MD 2481 (KBF), 2016 WL 5818585, at *1 (S.D.N.Y. Oct. 5, 2016) (“Aluminum IV”). She held that, because defendants’ allegedly anticompetitive conduct had occurred in one market (for aluminum warehousing services) but plaintiffs’ alleged harms occurred in another (for physical aluminum), plaintiffs had not sufficiently pled antitrust injury. Id. at *8–9. Any alleged inflation in the MWP, Judge Forrest reasoned, was accomplished “first and foremost” by restraining trade in the warehousing market, precluding plaintiffs from seeking damages sustained in the separate market for physical aluminum. Id. at *1. c. Aluminum VI On August 27, 2019, the Second Circuit reversed Aluminum IV. See Aluminum VI, 936 F.3d 86. It held that Judge Forrest had erred in attributing defendants’ conduct “first and foremost” to the warehousing market. Id. at 95–96. Instead, the Circuit held, “the burden inflicted by the defendants on the warehousing market was merely the means to accomplish the defendants’ anticompetitive objective,” i.e., “inflat[ing] the prices of the metal.” Id. at 96. As a result, the Circuit held, plaintiffs had adequately pled antitrust injury. No party raised, and Aluminum VI did not address, whether any or all plaintiffs were efficient enforcers. Id. at 94 (“[T]he only issue on appeal is whether the plaintiffs have suffered an antitrust injury.”). 2. Class Certification and Daubert Decision After the Second Circuit’s reversal and the case’s remand and reassignment, this Court considered the FLPs’ motion for class certification, and their associated motion under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), to exclude the declaration of the defense’s expert, Dr. David P. Kaplan, on market impact. These had been pending since 2016. See Dkts. 916–17, 1037. In late 2019, the parties filed supplemental briefing on those motions. See Dkts. 1157, 1194, 1198. And in summer 2020, the Court ordered, and the parties submitted, further briefing on discrete topics related to class certification. Dkts. 1248, 1253, 1256, 1258. On June 25, 2020, the Court held argument. On July 23, 2020, the Court denied both motions. See Aluminum VII, 336 F.R.D. at 66. The Court held that Dr. Kaplan was qualified to opine as he had, that his critiques of plaintiffs’ theory of unitary market impact were relevant and reliable, and that the FLPs’ arguments to the contrary went to the weight of Kaplan’s opinions, not their admissibility. Id. at 35. As to class certification, the Court held that the FLPs had failed to establish that common questions of law or fact predominated over individualized inquiries, and thus did not satisfy Federal Rule of Civil Procedure 23(b)(3). Id. at 44.20 For numerous reasons, the Court found that plaintiffs’ expert, Dr. Gilbert, had failed to reliably prove the class-wide economic impact of defendants’ alleged 20 The Court found that the FLPs had satisfied the first three elements of Rule 23(a) (numerosity, commonality, and typicality), see Aluminum VII, 336 F.R.D. at 37–39, and declined to address the disputed fourth element (adequacy of representation), in light of its holding that the FLPs failed to show predominance under Rule 23(b)(3), id. at 43. conduct. See id. at 51–53, 55–63. And because the plaintiffs had not adduced documentary or lay evidence establishing the “complex[] . . . proposition that plaintiffs must establish here,” id. at 51, the Court found that they had not met their burden of proving predominance, id. at 66. 3. Motion for Summary Judgment After the Court denied the class certification motion, it scheduled a conference to discuss next steps, and directed the parties to submit a joint letter with their views, Dkt. 1276, which, on August 7, 2020, the parties did, Dkt. 1278. The FLPs reported that they had petitioned the Second Circuit for leave to appeal the denial of class certification, but did not seek a stay pending that appeal. Id. at 1.21 Relevant here, defendants proposed briefing a “discrete summary judgment motion”—the motion at issue here—directed at “umbrella claims,” i.e., “claims based on aluminum purchases made from sellers that were not participants in the alleged conspiracy.” Id. At a conference held on August 19, 2020, the Court discussed the anticipated motion, and adopted the parties’ proposed briefing schedule. Dkt. 1283.22 On September 2, 2020, defendants filed their motion for summary judgment, Dkt. 1286; a memorandum of law in support, Def. Mem.; a Local Rule 56.1 statement, Def. 56.1; and the Playforth Declaration with exhibits. On October 20, 2020, the FLPs and IPs filed a consolidated 21 On December 16, 2020, the Second Circuit denied the FLPs’ petition. See Dkt. 1317. 22 Defendants previewed that the anticipated motion would be a “short and narrow,” “surgical motion,” implicating few factual disputes. See Ex. 86 (“Conf. Tr.”) at 12 (identifying factual issues as “who purchased from whom” and whether any plaintiff could “negotiate the price it [] paid its smelters”). Accordingly, the Court did not require the parties to submit a joint statement of stipulated facts, but rather “encourage[d] but [did] not require some degree of consultation” among them over potentially disputed facts. Id. at 20. In their Local Rule 56.1 counter- statement, plaintiffs fault defendants for, ostensibly, departing from their representations about the extent of factual disputes, and for relying on disputed facts without having first consulted plaintiffs. Pl. 56.1 at n.1. By and large, the Court finds defendants’ representations to have been accurate and the material factual disputes presented to be discrete. opposition brief, Pl. Opp’n; a counter-statement to defendants’ Local Rule 56.1 statement, Pl. 56.1; the Coughlin Declaration with exhibits; and the Brandt Declaration with exhibits. On November 20, 2020, the defendants filed a reply, Dkt. 1303 (“Def. Reply”), and a reply to plaintiffs’ Local Rule 56.1 counter-statement, Def. Reply 56.1. On December 11, 2020, the Court heard argument. II. Legal Standards Governing Motions for Summary Judgment To prevail on a motion for summary judgment, the movant must “show[] that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a); see Celotex Corp. v. Catrett, 477 U.S. 317, 322–23 (1986). The movant bears the burden of showing the absence of a question of material fact. In making this determination, the Court must view all facts “in the light most favorable” to the non-moving party. Holcomb v. Iona Coll., 521 F.3d 130, 132 (2d Cir. 2008). If the movant meets its burden, “the nonmoving party must come forward with admissible evidence sufficient to raise a genuine issue of fact for trial in order to avoid summary judgment.” Jaramillo v. Weyerhaeuser Co., 536 F.3d 140, 145 (2d Cir. 2008). “[A] party may not rely on mere speculation or conjecture as to the true nature of the facts to overcome a motion for summary judgment.” Hicks v. Baines, 593 F.3d 159, 166 (2d Cir. 2010) (citation omitted). Rather, to survive a summary judgment motion, the opposing party must establish a genuine issue of fact by “citing to particular parts of materials in the record.” Fed. R. Civ. P. 56(c)(1)(A); see also Wright v. Goord, 554 F.3d 255, 266 (2d Cir. 2009). “Only disputes over facts that might affect the outcome of the suit under the governing law” will preclude a grant of summary judgment. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248 (1986). In determining whether there are genuine issues of material fact that preclude summary judgment, a court is “required to resolve all ambiguities and draw all permissible factual inferences in favor of the party against whom summary judgment is sought.” Johnson v. Killian, 680 F.3d 234, 236 (2d Cir. 2012) (quoting Terry v. Ashcroft, 336 F.3d 128, 137 (2d Cir. 2003)). III. Discussion In pursuing summary judgment here, defendants rely on a line of cases in this District that, drawing upon the 2016 decision in Gelboim v. Bank of Am. Corp., 823 F.3d 759, 777–78 (2d Cir. 2016), have dismissed antitrust claims in benchmark-manipulation cases on the grounds that plaintiffs who did not transact directly with defendants were not efficient enforcers of the antitrust laws. The plaintiffs in these cases have alleged that the defendants conspired to fix or manipulate benchmark rates incorporated into products that plaintiffs bought or sold, thereby corrupting prices of those products in the market as a whole. At least some of those transactions, however, were not with the defendants whom plaintiffs sued, but rather with third-party non- wrongdoers at prices allegedly affected by the corrupted benchmark. Such plaintiffs pursued what have become known as “umbrella” claims for damages against the alleged antitrust violators, on the theory that the market-wide harms flowing from the rigged benchmark extended to transactions to which the cartel members were not parties and from which they did not profit. In such circumstances, courts in this District—drawing on dicta in Gelboim—have tended to hold that plaintiffs are not efficient enforcers to the extent that they sued entities with whom they had not directly transacted. These decisions have largely held that plaintiffs’ claimed injuries are most proximately attributable to the intervening decisions by plaintiffs and their counterparties to incorporate the allegedly rigged benchmark into their transactions. Defendants argue here that the IPs and three of the four FLPs23 are similarly not efficient enforcers. They emphasize that these plaintiffs made almost all their aluminum purchases from non-party industrial smelters who incorporated the allegedly manipulated MWP into their contracts. Thus, defendants argue, as in these other “umbrella” cases, defendants’ conduct in allegedly rigging benchmark prices did not directly cause plaintiffs’ alleged injuries, given the intervening transactions between plaintiffs and third-party sellers, who remained free at all times not to charge the full MWP when selling aluminum to plaintiffs.24 Plaintiffs disagree. First, they contend that it is law of the case that they are efficient enforcers as to their claims against defendants. They argue that defendants are seeking an improper second bite at the apple on this point. Second, on the merits, they argue that they their claims are distinguishable from those dismissed in recent benchmark-manipulation cases in this District because defendants’ illicit conduct aimed at driving up the MWP premium demonstrably and directly caused them injury. 23 Ampal is the sole exception because, unlike the other plaintiffs at issue, it bought roughly 2,200 metric tons of primary aluminum from Glencore and Century (an alleged Glencore affiliate) during the conspiracy period. 24 The parties disagree at length about whether plaintiffs’ claims are aptly characterized as “umbrella” claims. See Pl. Opp’n at 27–31; Def. Reply at 6–10. But the umbrella metaphor that some cases and commentators have used in this area does not carry legal significance. As have many other courts in this Circuit, see infra note 25, the Court finds the term to be descriptively useful, in that it captures the situation, present here, in which “a cartel controls only part of a market, but a consumer who dealt with a non-cartel member alleges that he sustained injury by virtue of the cartel’s raising of prices in the market as a whole.” Gelboim, 823 F.3d at 778. But a determination of standing in an individual antitrust case is highly fact-specific. Associated Gen. Contractors of Cal., Inc. v. Cal. State Council of Carpenters (“AGC”), 459 U.S. 519, 536–37 (1983). The use of the umbrella label does not predetermine how the efficient-enforcer factors apply in any particular case, or tacitly impose an inflexible “privity” requirement for there to be antitrust-standing inquiry. Cf. Pl. Opp’n at 28–29. The Court first sets forth the legal principles governing the instant motion, and canvasses the post-Gelboim cases in this District on which defendants rely. Next, the Court addresses plaintiffs’ argument that it is law of the case that they are efficient enforcers and have antitrust standing. Finally, concluding that that question remains open, the Court addresses it. A. Applicable Legal Principles Section 4 of the Clayton Act establishes a private right of action to enforce section 1 of the Sherman Act. 15 U.S.C. § 15. Although that provision “appears to confer a broad private right of action for antitrust damages, ‘Congress did not intend the antitrust laws to provide a remedy in damages for all injuries that might conceivably be traced to an antitrust violation.’” Daniel, 428 F.3d at 436–37 (quoting AGC, 459 U.S. at 534). Rather, as the Supreme Court has explained in its foundational decision in this area, “the judicial remedy cannot encompass every conceivable harm that can be traced to alleged wrongdoing.” AGC, 459 U.S. at 536. Accordingly, “[j]ust as in common-law tort and contract litigation, concepts such as ‘foreseeability and proximate cause, directness of injury, certainty of damages, and privity of contract’ circumscribe a party’s right to recovery,” so too in antitrust actions “‘the plaintiff’s harm, the alleged wrongdoing by the defendants, and the relationship between them,’ can limit the right to sue.” Daniel, 428 F.3d at 437 (quoting AGC, 459 U.S. at 532–33). In such cases, “the court must draw a line beyond which a defendant will not be held responsible for harm experienced by a plaintiff.” In re LIBOR-Based Fin. Instruments Antitrust Litig. (“LIBOR VI”), No. 11 MD 2262 (NRB), 2016 WL 7378980, at *15 (S.D.N.Y. Dec. 20, 2016). Following AGC, courts have developed limiting contours to the Clayton Act’s private cause of action. These are “embodied in the concept of ‘antitrust standing.’” Gatt Commc’ns, Inc. v. PMC Assocs., LLC, 711 F.3d 68, 76 (2d Cir. 2013). That doctrine “prevents private plaintiffs from ‘recover[ing] damages under § 4 . . . merely by showing injury causally linked to an illegal presence in the market.’” Id. (quoting Atl. Richfield Co. v. USA Petrol. Co., 495 U.S. 328, 334 (1990)). Instead, a private antitrust plaintiff must show (1) “a special kind of ‘antitrust injury’”; and (2) “that it is a suitable plaintiff to pursue the alleged antitrust violations and thus is an ‘efficient enforcer’ of the antitrust laws.” Id.; see Gelboim, 823 F.3d at 777–78 (“‘A showing of antitrust injury is necessary, but not always sufficient,’ to establish standing.” (quoting Cargill, Inc. v. Mofort of Colo., Inc., 479 U.S. 104, 110 n.5 (1986))). Plaintiffs bear the burden of showing antitrust standing. See, e.g., In re Am. Express Anti-Steering Rules Antitrust Litig., 433 F. Supp. 3d 395, 407–08 (E.D.N.Y. 2020) (citing Gelboim, 823 F.3d at 770). Only the second of these requirements—that plaintiffs be “efficient enforcers” of the antitrust laws—is at issue here. See Aluminum VI, 936 F.3d at 95–97 (holding that plaintiffs’ plausibly alleged antitrust injury). The efficient-enforcer inquiry turns on four factors: (1) whether the violation was a direct or remote cause of the injury; (2) whether there is an identifiable class of other persons whose self-interest would normally lead them to sue for the violation; (3) whether the injury was speculative; and (4) whether there is a risk that other plaintiffs would be entitled to recover duplicative damages or that damages would be difficult to apportion among possible victims of the antitrust injury. Gelboim, 823 F.3d at 772. These factors are meant to guide a court in deciding “whether the putative plaintiff is a proper party to ‘perform the office of a private attorney general’ and thereby ‘vindicate the public interest in antitrust enforcement.’” Id. at 780 (quoting Gatt, 711 F.3d at 80). “[T]he Supreme Court has noted that the first factor, requiring proximate causation, ‘must be met in every case.’” In re Commodity Exch., Inc. (“Gold”), 213 F. Supp. 3d 631, 653 (S.D.N.Y. 2016) (quoting Lexmark Int’l, Inc. v. Static Control Components, Inc., 134 S. Ct. 1377, 1392 (2014)). But the other efficient-enforcer factors do not provide “an independent basis for denying standing where it is adequately alleged that a defendant’s conduct has proximately injured an interest of the plaintiff’s that the statute protects.” DNAML Pty, Ltd. v. Apple Inc., 25 F. Supp. 3d 422, 430 (S.D.N.Y. 2014) (quoting Lexmark, 134 S. Ct. at 1392 (emphasis in Lexmark)). 1. Directness of the Alleged Injury In considering the first efficient-enforcer factor, courts examine “whether the violation was a direct or remote cause of the injury.” Gelboim, 823 F.3d at 772. This “requires evaluation of the ‘chain of causation’ linking [plaintiffs’] asserted injury and the [defendants’] alleged price-fixing.” Id. at 778. “Where the chain of causation between the asserted injury and the alleged restraint in the market ‘contains several somewhat vaguely defined links,’ the claim may be insufficient to provide antitrust standing.” Nypl v. JPMorgan Chase & Co., No. 15 Civ. 9300 (LGS), 2017 WL 3309759, at *5 (S.D.N.Y. Aug. 3, 2017) (quoting AGC, 459 U.S. at 540). “The antitrust laws do not require a plaintiff to have purchased directly from a defendant in order to have antitrust standing.” In re Foreign Exch. Benchmark Rates Antitrust Litig. (“FOREX”), No. 13 Civ. 7789 (LGS), 2016 WL 5108131, at *9 (S.D.N.Y. Sept. 20, 2016). But “one consideration in determining causation is whether plaintiffs transacted with defendants directly.” LIBOR VI, 2016 WL 7378980, at *15 (quoting 2A Areeda & Hovenkamp, Antitrust Law ¶ 335c(3) (2014) (“Beyond the actual customers, most other plaintiffs would be classified as ‘remote’ and denied standing even though they have suffered injury-in-fact.”)). So, too, is “the scope of the relevant market” and defendants’ role in it. Sonterra Cap. Master Fund, Ltd. v. Barclays Bank PLC (“Sonterra v. Barclays”), 366 F. Supp. 3d 516, 533 (S.D.N.Y. 2018) (citing Gelboim, 823 F.3d at 778); see Sullivan v. Barclays PLC, No. 13 Civ. 2811 (PKC), 2017 WL 685570, at *16 (S.D.N.Y. Feb. 21, 2017); see also Dennis v. JPMorgan Chase & Co., 343 F. Supp. 3d 122, 165 (S.D.N.Y. 2018) (plaintiffs were efficient enforcers despite not purchasing directly from defendants because “the conspiracy alleged here involves essentially all of the entities that contribute to setting [the relevant benchmark]”). In the post-Gelboim decisions assessing this factor, courts in this District giving this issue focused attention have found it to weigh heavily against antitrust standing where plaintiffs had “purchased products from non-defendants but allege that defendants’ actions raised their prices.” LIBOR VI, 2016 WL 7378980, at *15; see Silver II, 332 F. Supp. 3d 885, 905 (S.D.N.Y. 2018) (collecting cases holding that “plaintiffs who are not direct purchasers are not efficient enforcers in a benchmark manipulation case” and noting one “exception to this rule”). Generally, these courts have rejected that such “umbrella purchasers”—those who purchased from sellers outside the alleged cartel, who were free to set their own prices—are efficient enforcers. See, e.g., Silver II, 332 F. Supp. 3d at 905; see also Gelboim, 823 F.3d at 778 (“Umbrella standing concerns are most often evident when a cartel controls only part of a market, but a consumer who dealt with a non-cartel member alleges that he sustained injury by virtue of the cartel’s raising of prices in the market as a whole.”).25 25 See also, e.g., Sonterra v. Barclays, 366 F. Supp. 3d at 533 (noting Gelboim’s concern with “umbrella standing” involving “a plaintiff injured by dealing with a non-defendant by virtue of a defendant’s raising of prices in the market as a whole”); Silver II, 332 F. Supp. 3d at 905 (“Plaintiffs who do not deal directly with the defendants are referred to as ‘umbrella purchasers’ or ‘umbrella plaintiffs.’”); 7 W. 57th St. Realty Co. v. Citigroup, Inc., 314 F. Supp. 3d 497, 512 (S.D.N.Y. 2018) (“umbrella purchasers” are “plaintiffs who did not deal directly with” defendants (quoting Gelboim, 823 F.3d at 778)), aff’d, 771 F. App’x 498 (2d Cir. 2019) (summary order); Sonterra Cap. Master Fund Ltd. v. Credit Suisse Grp. AG (“Sonterra v. Credit Suisse”), 277 F. Supp. 3d 521, 559 (S.D.N.Y. 2017) (“Plaintiffs who purchased products from non-defendants but allege that defendants’ actions raised their prices are called ‘umbrella purchasers.’” (quoting LIBOR VI, 2016 WL 7378980, at *15)); In re Platinum & Palladium Antitrust Litig. (“Platinum I”), No. 14 Civ. 9391 (GHW), 2017 WL 1169626, at *21 (S.D.N.Y. Mar. 28, 2017) (“Plaintiffs who do not have direct dealings with the defendants, but purchased products allegedly affected by defendants’ price fixing, are referred to as ‘umbrella purchasers.’”); Sullivan, 2017 WL 685570, at *15 (rejecting “‘umbrella’-type damages claims” where “plaintiffs had no direct dealings with the defendants”); LIBOR VI, 2016 WL 7378980, at *15 (“Plaintiffs who purchased products from non-defendants but allege that defendants’ actions raised their prices are called ‘umbrella purchasers.’”). In particular, these cases have reasoned, an umbrella plaintiff’s purchase from an outsider to the cartel raises the concern “that ‘significant intervening causative factors, most notably, the independent pricing decisions of non-conspiring retailers, attenuate the causal connection between the violation and the injury.’” Sonterra v. Barclays, 366 F. Supp. 3d at 533 (quoting LIBOR VI, 2016 WL 7378980, at *15). “In such circumstances, ‘the defendants secured no illegal benefit at [the plaintiff’s’] expense,’ and permitting recovery in such a transaction ‘could subject antitrust violators to potentially ruinous liability, well in excess of their illegally-earned profits.’” LIBOR VI, 2016 WL 7378980, at * 15 (quoting Mid-W. Paper Prods. Co. v. Cont’l Grp., Inc., 596 F.2d 573, 583, 586 (3d Cir. 1979)). For this reason, courts following Gelboim have “[drawn] a line between plaintiffs who transacted directly with defendants and those who did not,” reasoning that the “independent decision of contracting parties to incorporate [a benchmark price] breaks the chain of causation between defendants’ actions and plaintiff’s injury.” Sonterra v. Barclays, 366 F. Supp. 3d at 533 (quoting Sullivan, 2017 WL 685570, at *17); see also Gelboim, 823 F.3d at 779 (voicing skepticism of efficient-enforcer status where plaintiffs’ claims seek “damages disproportionate to wrongdoing”).26 26 Plaintiffs discount Gelboim’s concern with “damages disproportionate to wrongdoing” as “dicta” and a mere “prudential concern, not a statutory concern, that should not bear on standing under Lexmark.” Pl. Opp’n at 23–24. But the Second Circuit’s guidance to this effect, which post-dated Lexmark by two years, was considered. And this factor rationally bears on whether the plaintiffs here, as opposed to a party more proximate to wrongdoers, are efficient prosecutors of a section 1 claim. Following Gelboim’s lead, courts assessing whether a plaintiff is an efficient enforcer have accordingly considered whether the plaintiff’s suit could exact outsized damages from the alleged-malefactor defendant. See, e.g., LIBOR VI, 2016 WL 7378980, at *15; Platinum II, 449 F. Supp. 3d at 308–11. This Court is guided by this persuasive authority. In all events, although the possibility of outsized damages reinforces the Court’s holding that the indirect- purchaser plaintiffs here are not efficient enforcers, it is not dispositive. The Court would reach 2. Existence of More Direct Victims The second factor inquires “whether there is an identifiable class of other persons whose self-interest would normally lead them to sue for the violation,” Gelboim, 823 F.3d at 772, and who “suffered an antitrust injury more directly than the present” plaintiffs, LIBOR VI, 2016 WL 7378980, at *17. “Implicit in the inquiry is recognition that not every victim of an antitrust violation needs to be compensated under the antitrust laws in order for the antitrust laws to be efficiently enforced.” Gelboim, 823 F.3d at 779. In benchmark-manipulation cases, courts have tended to give this factor “diminished weight.” Id. That is because, at least in a paradigmatic such case, there should be “little difference between direct and umbrella purchasers,” as true “benchmark-manipulation affects all market participants equally,” Silver II, 332 F. Supp. 3d at 909 (citing Platinum I, 2017 WL 1169626, at *23). In such cases, assuming a rigged benchmark equally drives up the aggregate price paid by direct and indirect purchasers, even purchasers who did not transact directly with defendants would have been harmed in the same manner as those who did. See, e.g., Gelboim, 823 F.3d at 779 (benchmark cases are “peculiar” because plaintiffs who purchased from non-defendants “would be injured to the same extent and in the same way as direct customers” of a defendant). 3. Speculative Damages The third efficient-enforcer factor inquires whether the plaintiff’s damages are “highly speculative,” as such is “a sign that a given plaintiff is an inefficient engine of enforcement.” Gelboim, 823 F.3d at 779. To be sure, “[t]he most elementary conceptions of justice and public policy require that the wrongdoer shall bear the risk of the uncertainty which his own wrong has the same result absent that consideration, based on the considerable lacuna between defendants’ allegedly collusive actions and the pricing harm allegedly worked on plaintiffs. created.” In re DDAVP Direct Purchaser Antitrust Litig., 585 F.3d 677, 689 (2d Cir. 2009); see Gelboim, 823 F.3d at 780 (“Some degree of uncertainty stems from the nature of antitrust law.”). And “potential difficulty in ascertaining and apportioning damages is not . . . an independent basis for denying standing where” plaintiffs have been proximately injured by defendants’ conduct. Lexmark, 134 S. Ct. at 1392 (emphasis in Lexmark). But courts may still consider this factor in determining whether plaintiffs are efficient enforcers, and the presence of speculative damages can favor dismissal on efficient-enforcer grounds. See, e.g., Silver II, 332 F. Supp. 3d at 911 (“[T]his factor clearly weighs against Plaintiffs.”); Sonterra v. Credit Suisse, 277 F. Supp. 3d at 564 (finding “that the speculative damages factor weighs strongly against the standing of the umbrella plaintiffs but less strongly as to the Direct Transaction Plaintiffs”). Courts in this Circuit typically assess four considerations in deciding whether damages are unduly speculative in price-fixing cases: (1) the extent to which the damages claim is conclusory in nature; (2) whether the injury is “so far down the chain of causation from defendants’ actions that it would be impossible to untangle the impact of the fixed price from the impact of intervening market decisions,” which relates to the causation factor; (3) whether external market factors affected the “relationship between the fixed price and the price that the plaintiffs ultimately paid”; and (4) whether “the non-fixed components of a transaction were heavily negotiated between the parties in relation to the fixed component.” Sonterra v. Barclays, 366 F. Supp. 3d at 546 (quoting LIBOR VI, 2016 WL 7378980, at *17–18). 4. Duplicative Recovery or Complex Apportionment Last, the Court must assess whether plaintiffs’ claims present a risk of duplicative recovery or complex apportionment. “Under this factor courts are traditionally concerned with the prospect of different groups of plaintiffs attempting to recover for the same exact injury[.]” LIBOR VI, 2016 WL 7378980, at *23 (noting that such prospect was not present there). 5. Review of Applicable Case Law In large part, defendants base their arguments on what they describe as a consensus that has crystallized in this District during the past five years in cases resolving efficient-enforcer challenges in the context of benchmark manipulation. See Silver II, 332 F. Supp. 3d at 905 (noting “critical mass” of recent decisions on this point). They argue that plaintiffs’ claims here are flawed in the same way as—and indeed, in some respects have greater infirmities than—the claims dismissed in these cases. Plaintiffs counter that their stature as efficient enforcers is law of the case, and, regardless, that they have a better claim to standing than in the cases on which defendants rely. The fountainhead of this recent line of cases is the Second Circuit’s decision in Gelboim, although, even before then, courts in this District and elsewhere had voiced skepticism of “umbrella” claims under section 1 of the Sherman Act. See, e.g., Mid-W. Paper Prod. Co., 596 F.2d at 583–87; Gross v. New Balance Athletic Shoe, Inc., 955 F. Supp. 242, 245–46 (S.D.N.Y. 1997). In Gelboim, the Circuit reviewed the dismissal of plaintiffs’ claims arising from the alleged rigging of the London Interbank Offered Rate (“LIBOR”), which at the time approximated the average rate at which a group of designated banks could borrow money. See 823 F.3d at 764. Plaintiffs there had sued those banks, accusing them of having conspired to artificially depress LIBOR, so as to reduce their costs of borrowing and enable them to project greater financial health. Id. at 766. But LIBOR—which had been called “the world’s most important number”—was used in countless other financial transactions, including in many to which no defendant bank was party. Id. at 765. And plaintiffs in Gelboim sought to recover from the defendant banks not only damages from deflated LIBOR in transactions with those banks. They also sought to recover for other transactions of theirs that incorporated LIBOR, for example, where they had been paid artificially low interest payments from non-conspiring counterparties as a result of the defendant cabal’s alleged depression of LIBOR. Id. The Second Circuit in Gelboim reversed the district court’s holding that the plaintiffs had failed to demonstrate an antitrust injury. However, in doing so, the Circuit raised questions about whether plaintiffs qualified as efficient enforcers, which the district court had not addressed. It stated that “close attention” was required to the efficient enforcer factors in that case. Id. at 778. In particular, the Circuit questioned “the antitrust standing of those plaintiffs who did not deal directly with the Banks.” Id. It recognized that, in some respects, victims of benchmark manipulation are situated similarly to entities that transact directly with defendants: “At first glance, here there appears to be no difference in the injury alleged by those who dealt in LIBOR-denominated instruments, whether their transactions were conducted directly or indirectly with the Banks.” Id. at 779; see id. (noting that a “peculiar feature” of such cases is that even remote victims “would be injured to the same extent and in the same way as direct customers of [] the Banks”). But, the Circuit stated, given the potential for disproportionate liability arising from transactions removed from those in which the defendants took part, the likelihood of highly speculative damages, and risks of potential duplicative recoveries, close attention to efficient-enforcer status on remand was necessary. See id. at 779–80. After Gelboim, but before the district court (Buchwald, J.) addressed these questions on remand, three decisions, mindful of Gelboim’s guidance, held that, based on the pleadings, plaintiffs in similar benchmark-manipulation cases satisfied the efficient-enforcer factors. In the first, plaintiff sellers of gold and financial products based on the price of gold alleged that defendant banks had conspired to artificially depress the global benchmark price for gold, causing them damages in the form of lower profits from sales. Gold, 213 F. Supp. 3d at 644. The district court (Caproni, J.) acknowledged “substantial challenges to Plaintiffs’ causation theory,” but “deferred” resolution of those issues for later in the litigation. Id. at 656. The court held that, because plaintiffs had plausibly alleged that defendants “rigged the ‘entire . . . market’ for gold” and that “all market participants ‘moved in line’ according to Defendants’ price manipulation, leaving little room for any interfering price impact due to the actions of nonculpable entities or exogenous market forces,” plaintiffs’ alleged injuries were sufficiently direct, and the other three factors did not independently defeat their antitrust standing. Id. at 655. Soon after, another district court decision (Schofield, J.) denied a motion to dismiss aimed partially at umbrella-purchaser plaintiffs, who had alleged a conspiracy to manipulate foreign-currency exchange rates, which plaintiffs then incorporated into financial instruments. See FOREX, 2016 WL 5108131, at *1. As to such plaintiffs, the court held that their injuries were sufficiently direct because the complaint plausibly alleged that the rates defendants manipulated tracked the rates plaintiffs paid “at near parity.” Id. at *9. And even though plaintiffs traded on an exchange rather than directly with defendants, the Court held that concerns of disproportionate damages had diminished weight because defendants “dominated the [relevant] market with a combined market share of over 90%.” Id. Further, the court held, damages would not be speculative because, unlike in Gelboim, the rates were not “only a component of price,” or subject to “negotiated rates in reference to” the manipulated benchmark. Id. at *8 (noting “conduct in this case is more straightforward”). In the third case—also involving the alleged fixing of a metal price (this time silver)—the district court (Caproni, J.) held that plaintiffs alleging a scheme similar to the one in Gold had antitrust standing, even where they transacted with non-defendants. See In re London Silver Fixing, Ltd., Antitrust Litig. (“Silver I”), 213 F. Supp. 3d 530, 542 (S.D.N.Y. 2016). The Court, tracking the analysis in Gold, held that, on the pleadings, plaintiffs had shown they were efficient enforcers. See id. at 552–57.27 In December 2016, the district court (Buchwald, J.), on remand from Gelboim, resolved LIBOR VI. Primarily relying on the directness factor, the court held that “where a plaintiff’s counterparty is reasonably ascertainable and is not a defendant bank, a plaintiff is not an efficient enforcer.” LIBOR VI, 2016 WL 7378980, at *16 (footnote omitted). The court found dispositive that “[a] plaintiff and a third party could, and did, easily incorporate LIBOR into a financial transaction without any action by defendants whatsoever.” Id. That “independent decision,” the court held, was sufficient to “break[] the chain of causation between defendants’ actions and a plaintiff’s injury.” Id. The court distinguished FOREX, where the court had used market control “as a proxy for the question of direct causation” given the anonymity of their counterparties on an exchange, on the basis that the defendant banks in LIBOR VI had nowhere near the 90% market share at issue in FOREX. Id. As a result, given plaintiffs’ many LIBOR-denominated transactions “over which defendants had no control, in which defendants had no input, and from which defendants did not profit,” there was a much greater risk of disproportion between defendants’ illicit profits and plaintiffs’ potential recovery than in FOREX. Id. The court held that the second and fourth efficient-enforcer factors carried little weight in the analysis, id. at *17, 23, and that the third factor applied to different degrees as to different classes of plaintiffs, id. at *17–23. But, it held, the indirectness of plaintiffs’ injuries alone “persuaded 27 In a 2018 decision in the same case, see infra, Judge Caproni dismissed an amended complaint on several grounds, including that plaintiffs who transacted only with non-conspiring defendants were not efficient enforcers. Silver II, 332 F. Supp. 3d at 905. In support, she cited the overwhelming majority of decisions that had held similarly in benchmark-manipulation cases, identifying FOREX as the only “exception to this rule.” Id. at 905 n.13. [the court] to draw a line between plaintiffs who transacted directly with defendants and those who did not.” Id. Only the former, not the latter qualified as efficient enforcers. After LIBOR VI, all but one decision in this District to confront umbrella claims in cases of benchmark manipulation has, at the pleading stage, dismissed them on efficient-enforcer grounds. In several, the court considered LIBOR-like manipulation of benchmark interest rates tied to various currencies. In Sullivan, the district court (Castel, J.), considering whether an alleged conspiracy to manipulate an average interest rate on Euro deposits (“Euribor”), “comfortably conclude[d] that four plaintiffs” were not efficient enforcers because they “had no direct dealings with the defendants on any Euribor-based transactions.” 2017 WL 785570, at *15. In Sonterra v. Credit Suisse, the district court (Stein, J.) addressed an alleged conspiracy in which banks influenced the Swiss Franc London Interbank Offered Rate (“CHF LIBOR”) “in whatever direction benefited their own trading positions.” 277 F. Supp. 3d at 539. Holding that plaintiffs who transacted with defendants were better situated to enforce the antitrust laws, and that the directness inquiry “strongly disfavors those plaintiffs who transacted with an identifiable third party,” the court dismissed umbrella plaintiffs notwithstanding the absence of risk of duplicative recovery. Id. at 561, 565. The same result followed in a number of structurally similar cases. See Sonterra v. Barclays, 336 F. Supp. 3d at 548 (dismissing claims “based solely on financial transactions with third-parties, and not with Defendants,” where plaintiffs sued over alleged scheme to manipulate British Pound Sterling LIBOR); see 7 W. 57th St., 314 F. Supp. 3d at 515 (similar); In re European Gov’t Bonds Antitrust Litig., No. 19 Civ. 2601 (VM), 2020 WL 4273811, at *14 (S.D.N.Y. July 23, 2020) (similar for European bond instruments); see also FrontPoint Asian Event Driven Fund, L.P. v. Citibank, N.A., No. 16 Civ. 5263 (AKH), 2017 WL 3600425, at *12–13 (S.D.N.Y. Aug. 18, 2017) (dismissing umbrella plaintiff as inefficient enforcer on grounds unrelated to whether it purchased directly from defendant, and declining to decide whether the latter fact independently weakened that plaintiff’s claim to antitrust standing). In one such case, however, the court held that plaintiffs who had incorporated a benchmark interest rate—the Bank Bill Swap Rate (“BBSW”), which the court described as “somewhat similar in concept to [LIBOR]”—into contracts with non-defendants had antitrust standing to sue the defendant banks, at least at the pleading stage. Dennis, 343 F. Supp. 3d at 149. There, plaintiffs alleged that, as in LIBOR VI, bank defendants had conspired to manipulate BBSW by, inter alia, engaging in manipulative transactions and making false rate submissions. Id. at 149. Because BBSW rates had been incorporated into plaintiffs’ derivative transactions, plaintiffs argued that they had been underpaid and overcharged on those contracts, depending on the direction of defendants’ manipulations at any given time. Id. In holding that plaintiffs who had not transacted with defendants were efficient enforcers, the district court (Kaplan, J.) distinguished the case from those in which “a group of conspirators, without controlling the entire market for a commodity, fixed prices for that commodity.” Id. at 165. In such cases, the court noted, “the question of causation is whether the conspirators that did not transact directly with a plaintiff nonetheless affected the price of the commodity in which the plaintiff transacted.” Id. In contrast, the court noted that the conspiracy as alleged “involves essentially all of the entities that contribute to setting BBSW,” and thus had directly caused plaintiffs’ injuries, such that “at this most preliminary stage,” plaintiffs’ claims to be efficient enforcers survived. Id. Other cases have expanded these holdings outside the context of interest-rate manipulation, including into commodities markets. In Silver II, the district court (Caproni, J.) addressed price-fixing allegations in markets for silver and derivative financial products, after plaintiffs in that multi-district litigation filed a third amended complaint. Plaintiffs had alleged that the banks responsible for daily setting the “fix price” of silver, as well as other “non-fixing banks,” had engaged in a “comprehensive scheme of market manipulation” to manipulate the price of silver, which in turn caused harm to plaintiffs who had entered into transactions based, in part, on that benchmark price. 332 F. Supp. 3d at 890–92. The non-fixing banks, who were alleged to have taken actions in silver markets that allegedly indirectly affected the price of silver, moved to dismiss, in part, on efficient-enforcer grounds. Recognizing that decisions in a “critical mass” of similar post-Gelboim cases had held that plaintiffs who had transacted only with non-defendants in similar cases lacked antitrust standing, the court found that all four efficient-enforcer factors disfavored plaintiffs’ standing and dismissed their claims. Id. at 905. Finally, in Platinum I and II, the district court (Woods, J.) addressed a price-rigging conspiracy similar to those analyzed in Gold and Silver I. See 2017 WL 1169626, at *22; In re Platinum & Palladium Antitrust Litig. (“Platinum II”), 449 F. Supp. 3d 290 (S.D.N.Y. 2020). Plaintiffs alleged that a number of banks and other entities had conspired to fix, at low levels, “the daily market prices of platinum and palladium,” and profited by taking advantage of resulting dips in prices for the metals and related financial products. Platinum I, 2017 WL 1169626, at *3, 8. Plaintiffs alleged that, as a result, they had obtained lower prices when selling those metals and derivative financial contracts because that global benchmark price was “hardwired” into their contracts. Id. at *2–5; Platinum II, 449 F. Supp. 3d at 309. The court held in Platinum I that, although all plaintiffs had plausibly alleged an antitrust injury, only plaintiffs who had transacted directly with defendants were efficient enforcers. 2017 WL 1169626, at *20, 22. Because umbrella plaintiffs’ harms were indirect and presented substantial risks of speculative damages, the Court found it appropriate “to draw a line between persons who transacted directly with defendants and those who did not.” Id. at *22, 25. The district court reprised that analysis in Platinum II, on reconsideration. See 449 F. Supp. 3d at 303–11. As to the one plaintiff who had “sold physical platinum and palladium” but not to any alleged defendant or co-conspirator, the court held it to be an inefficient enforcer. See id. at 304 (“Because the TAC does not allege that the OTC Plaintiff transacted directly with any defendant, it is not an efficient enforcer.”). Recognizing some division of authority—i.e., between decisions such as Gold and Dennis, on one hand, and LIBOR IV, Silver II, and other interest-rate-manipulation cases on the other—see id. at 304–05, the court found the case more comparable to the latter, with the gap between defendants’ actions and plaintiffs’ harms breaking the causal chain between the two, id. at 304, 308. And, it held, the potential for disproportionate damages—even if not as “ruinous” as in interest-rate cases such as LIBOR IV—strongly disfavored allowing indirect-seller plaintiffs of physical platinum or palladium to pursue their claims. Id. at 310–11 & n.14. In sum, with limited exceptions, courts in this District in recent years overwhelmingly have held that plaintiffs alleging benchmark manipulation by defendants who did not control the relevant market, and with whom they did not transact, are not efficient enforcers. In one case, the court held otherwise where plaintiffs’ counterparties were not readily ascertainable and defendants controlled the vast majority of the relevant market. FOREX, 2016 WL 5108131, at *9. In others, crediting allegations that made the plaintiff’s claims direct and/or non- speculative, courts have permitted claims to survive challenges at the pleading stage. See, e.g., Gold, 213 F. Supp. 3d at 656. But in the main, courts have drawn a clear line between plaintiffs who transacted directly with defendants and those who did not, holding that only the former are efficient enforcers, and often noting the risk of disproportionate liability that would exist were the latter permitted to pursue recovery from diffuse, market-wide injuries. B. Law of the Case The Court first considers plaintiffs’ argument that that their status as efficient enforcers is law of this case. 1. Legal Standard The law-of-the-case doctrine has two branches. In re Interest Rate Swaps Antitrust Litig., 351 F. Supp. 3d 698, 702 (S.D.N.Y. 2018). The first requires a trial court to follow an appellate court’s previous ruling on an issue in the same case. United States v. Uccio, 940 F.2d 753, 757 (2d Cir. 1991). This is the so-called “mandate rule.” United States v. Tenzer, 213 F.3d 34, 40 (2d Cir. 2000). The mandate rule “ordinarily forecloses relitigation of all issues previously waived by the defendant or decided by the appellate court.” United States v. Quintieri, 306 F.3d 1217, 1225 (2d Cir. 2002). The second is more flexible. Maraschiello v. City of Buffalo Police Dep’t, 709 F.3d 87, 97 (2d Cir. 2013) (“The doctrine of law of the case is discretionary and does not limit a court’s power to reconsider its own decisions prior to final judgment.” (citation omitted)). It applies when any court reconsiders its own ruling on an issue without any intervening ruling on the same issue by a higher court. “[W]hen a court has ruled on an issue, that decision should generally be adhered to by that court in subsequent stages in the same case, unless cogent and compelling reasons militate otherwise.” Quintieri, 306 F.3d at 1225, 1230 (citations omitted). “Compelling reasons to revisit a decision include an intervening change in law, availability of new evidence, or the need to correct a clear error or prevent manifest injustice.” Doe v. E. Lyme Bd. of Educ., 962 F.3d 649, 662–63 (2d Cir. 2020) (citation omitted). And the doctrine does not control where “relevant issues are governed by different standards of review” at different stages of the case. United States v. Johnson, 616 F.3d 85, 93 (2d Cir. 2010), abrogated on other grounds, 576 U.S. 591 (2015). Fidelity to the law of the case is as important, if not more so, when one judge “is asked to consider the ruling of a different judge.” Lillbask ex rel. Mauclaire v. State of Conn. Dep’t of Educ., 397 F.3d 77, 94 (2d Cir. 2005). “[U]pon reassignment, ‘the new judge is well advised to pay particular heed to the doctrine of “law of the case,” and not to attempt a de novo review of . . . decisions made over a lengthy period by diligent and experienced judicial officers who have handled the case previously.’” Waverly Props., LLC v. KMG Waverly, No. 09 Civ. 3940 (PAE), 2011 WL 13322667, at *1 (S.D.N.Y. Dec. 19, 2011) (quoting Peyser v. Searle Blatt & Co., No. 99 Civ. 10785 (GEL), 2004 WL 307300, at *1 (S.D.N.Y. Feb. 17, 2004)). 2. Analysis Plaintiffs contend that each branch of the law-of-the-case doctrine favors, or requires, the denial of defendants’ motion for summary judgment. a. Aluminum II Plaintiffs first argue that Judge Forrest’s 2015 decision in Aluminum II, which denied defendants’ motion to dismiss on efficient-enforcer grounds, controls the outcome here because there has not been any intervening change in controlling law or material new evidence. The Court is unpersuaded, for two reasons. First, the law-of-the-case doctrine does “not preclude a district court from granting summary judgment based on evidence after denying a motion to dismiss based only on the plaintiff’s allegations.” Maraschiello, 709 F.3d at 97. Here, defendants argue that several categories of evidence—which were either unavailable or not cognizable when Aluminum II was decided—undermine plaintiffs’ claim to be efficient enforcers. In particular, in assessing the directness of plaintiffs’ injuries in 2015 in Aluminum II, Judge Forrest relied principally on three of plaintiffs’ factual allegations: that (1) plaintiffs had bought “aluminum directly from a producer (and, in one instance, from one of the defendants)”; (2) “the contracts between the producer-seller and plaintiff-buyer allegedly contain provisions tying the contract prices to the Midwest Premium”; and (3) “it is not possible to negotiate spot or long-term aluminum contracts with major producer Alcoa that do not incorporate the Midwest Premium.” Aluminum II, 95 F. Supp. 3d at 444. Plaintiffs’ first factual allegation—that they bought aluminum largely from producers, not defendants—is borne out by the summary judgment record. Defendants, in fact, embrace that fact in pursuing dismissal. But plaintiffs’ second and third factual allegations—that plaintiffs’ aluminum-purchase contracts always contained the MWP as a price component and that paying the MWP was unavoidable for aluminum purchasers—sit uneasily against the summary judgment record. Judge Forrest accepted, for purposes of defendants’ motion to dismiss, the TAC’s factual allegations on these points. See id. at 444 (alleging that it is “not possible” to buy Aluminum from Alcoa without a contract incorporating the MWP); TAC ¶ 176 (“Nearly all industrial contracts for the physical delivery of primary aluminum express the price for aluminum” using the LME settlement price and a regional premium (emphasis added)). But the evidence adduced now makes clear that many aluminum contracts lack such a term—so much so, in fact, that plaintiffs have jettisoned an aspect of their claims that involve such contracts. See, e.g., Def. 56.1 ¶¶ 58–59; Pl. 56.1 ¶ 58 (admitting that 31% of Alcoa contracts “appeared in the data as not having paid the full MWP” but now disclaiming such contracts as a basis for relief, stating that “Plaintiffs do not seek damages on those types of contracts”).28 The Court in 28 Plaintiffs maintain that all primary-aluminum sales required the payment of the MWP, even where contract documents failed expressly to reference it. See, e.g., Pl. 56.1 ¶ 58; id. ¶ 59 (dismissing as “irrelevant” whether contracts include “MWP or MWTP on the face of the contract for sale . . . because on all first-level purchases of aluminum, the MWP is paid”). But that claim, too, diverges from the allegations that the Court accepted as true at the pleading stage. See Aluminum II, 95 F. Supp. 3d at 44 (crediting allegations that contracts “contain provisions tying the contract prices to the Midwest Premium” and that “it is not possible to negotiate” contracts that do not do so); see also TAC ¶¶ 166, 176; JAC ¶¶ 32, 44; Fujifilm Compl. ¶ 28 (aluminum contracts “expressly incorporate the benchmark regional premiums”). Aluminum II also did not have occasion to consider the extent to which plaintiffs could negotiate all-in prices with smelters, potentially in order to offset increases in the MWP. See, e.g., Def. 56.1 ¶¶ 15, 20, 26, 30, 38, 42, 48, 53; Def. Reply 56.1 ¶ 48. Thus, important facts here differ from those pled. And the standard of review on summary judgment differs from that on the motion to dismiss. Accordingly, the 2015 denial of defendants’ motion to dismiss does not require the denial of their 2020 motion for summary judgment. Second, there has been, as canvassed above, significant intervening legal authority bearing on antitrust standing. This provides a “cogent and compelling reason” for the Court, in its discretion, to revisit Aluminum II’s efficient-enforcer holding. Quintieri, 306 F.3d at 1225. The Second Circuit issued its decision in Gelboim more than a year after Judge Forrest decided Aluminum II. Its pointed instructions as to the inquiries to be conducted on remand, see 823 F.3d at 778–80, catalyzed not only the later dismissal of the claims in that case, but a substantial body of decisions in this District assessing the efficient-enforcer factors in the context of benchmark- manipulation claims. To be sure, the Circuit’s decision in Gelboim did not alter the governing legal standards, and it noted factual “features of this case that make it like no other.” Id. at 778. Gelboim thus does not itself represent an “intervening change in law.” Doe, 962 F.3d at 662–63. Nonetheless, the “critical mass” of judges in this District who “have concluded that plaintiffs who are not direct purchasers are not efficient enforcers in a benchmark manipulation case,” Silver II, 332 F. Supp. 3d at 905 (collecting cases), does reflect a change in the consensus application of how these doctrinal principles apply in the context at hand. These varied, nuanced, and thoughtful decisions have almost uniformly held that plaintiffs akin to those implicated by this motion—i.e., entities that transacted exclusively with non-defendants in transactions allegedly influenced by defendants’ conduct—are not efficient enforcers. See supra pp. 31–38. Although these decisions do not change the governing legal framework from that applied in Aluminum II, compare, e.g., Silver II, 332 F. Supp. 3d at 905, 906–12 (applying four efficient-enforcer factors from AGC), with Aluminum II, 95 F. Supp. 3d at 441, 444 (same), the “overwhelming weight of later decisions considering [an] issue” can justify reconsideration of a prior holding even without an intervening change of law. See In re MTBE Prod. Liab. Litig., 148 F. Supp. 3d 309, 316 (S.D.N.Y. 2015) (citing three contrary decisions); see also Uccio, 940 F.2d at 758 (confirming district court’s “discretion” whether to adhere to law of the case). And the body of district-court precedents that has accreted since Gelboim is consequential. See supra pp. 31–38. The Aluminum II Court could not have foreseen the near-consensus that has developed on this point among its peer courts. These developments provide good reasons to examine anew the efficient-enforcer question, informed by the reasoning in those decisions. The law-of-the-case doctrine thus does not oblige the Court to adhere to the assessment in Aluminum II of plaintiffs’ efficient-enforcer status. b. Aluminum VI Plaintiffs next argue that the Second Circuit’s decision in Aluminum VI bars defendants’ bid for summary judgment. They note that the Circuit was aware at the time of that decision that plaintiffs had largely bought aluminum from producers, not defendants. They note that, while such was not a basis for the Court’s dismissal of plaintiffs’ claims in Aluminum IV, the Circuit could have affirmed that decision on the alternative ground that plaintiffs were not efficient enforcers, rather than reversing based on the error in Judge Forrest’s antitrust-injury analysis. Pl. Opp’n at 15–16. That the Circuit did not do so, plaintiffs contend, “strongly suggests” that it continued to agree with Judge Forrest’s holding in Aluminum II that plaintiffs were efficient enforcers. Id. at 6. Relatedly, plaintiffs argue that, because defendants failed to cross-appeal the efficient-enforcer holding in Aluminum II when plaintiffs appealed from Aluminum IV, defendants waived the right now to reopen that issue now. Id. at 15. Neither argument is persuasive. No party raised the efficient-enforcer issue on appeal in Aluminum VI.29 As a result, the Second Circuit did not consider, let alone decide, that question. See Aluminum VI, 936 F.3d at 94 (“Here, the only issue on appeal is whether the plaintiffs have suffered an antitrust injury.”). To be sure, the Circuit may affirm a dismissal on grounds not considered below. See, e.g., Scott v. Fischer, 616 F.3d 100, 105 (2d Cir. 2010). But plaintiffs do not cite authority requiring a district court to treat a Circuit’s failure to sua sponte take up an alternative ground for affirmance as implicitly repudiating it.30 And even had Aluminum VI implicitly affirmed Aluminum II’s efficient-enforcer holding, the Circuit “focus[ed] on” the viability of plaintiffs’ “legal theory”—not the sufficiency of their proof. Aluminum VI, 936 F.3d at 93 n.3. At the very least, Aluminum VI thus leaves defendants free, with fact discovery now complete, to argue that the evidence adduced is incompatible with efficient-enforcer status. See Maraschiello, 709 F.3d at 97. 29 On the contrary, defendants made clear that, although the efficient-enforcer question was not at issue on appeal, they intended to pursue that issue in the district court if the Second Circuit reversed on antitrust injury. See Joint Br. of Defendants-Appellees at 37 n.11, Aluminum VI, 936 F.3d 86 (No. 16-4230) (“Although these appeals do not present the question of whether Plaintiffs are ‘efficient enforcers’ of the antitrust laws, Defendants would expect to show that Plaintiffs do not satisfy that requirement in any further proceedings in the District Court.”). 30 The one case plaintiffs cite, Glendora v. Cablevision Sys. Corp., 893 F. Supp. 264, 266 (S.D.N.Y. 1995), is inapposite. There, in reversing the summary dismissal of a pro se plaintiff’s claims, the Circuit directed the district court to consider an issue—whether a statute provided an implied cause of action—that it had previously disregarded. The district court took the Circuit’s sua sponte action, and specific instruction, to connote that it agreed there was such a cause of action. Id. Even so, the district court did not treat the Circuit as having already decided the question, but instead undertook its own independent analysis, id. at 267–71, as this Court proceeds to do here. Nor does defendants’ decision, in seeking affirmance in Aluminum VI, not to pursue the alternative ground that most plaintiffs were not efficient enforcers waive that argument. The Second Circuit has rejected such a broad conception of waiver. “The role of the appellee is to defend the decision of the lower court. This Court has not held that an appellee is required, upon pain of subsequent waiver, to raise every possible alternate ground upon which the lower court could have decided an issue.” Brown v. City of New York, 862 F.3d 182, 188 (2d Cir. 2017). Accordingly, Aluminum VI is also not law of the case on the efficient-enforcer question. C. Efficient-Enforcer Factors The Court now considers, afresh, whether plaintiffs are efficient enforcers. 1. Directness of the Alleged Injury The first efficient-enforcer factor decisively favors defendants’ motion. The vast majority of plaintiffs’ claims—those against defendants with whom they did not transact—are indirect. And the evidence adduced in discovery, even considered in the light most favorable to plaintiffs, shows that the decision to charge, or not to charge, the MWP was ultimately made by non-conspiring smelters, not defendants, breaking the chain of causation between defendants’ allegedly collusive steps to manipulate the MWP and plaintiffs’ injuries. At the outset, the Court notes that plaintiffs’ section 1 claims, even as to the limited number of transactions in which plaintiffs dealt directly with defendants, involve an unusually indirect chain of causation. Unlike benchmark cases in which defendants are alleged to have colluded to directly rig an industry-wide interest rate or benchmark price through false submissions about transactions or similar subterfuges, plaintiffs here posit a more attenuated link between defendants’ actions and the prices plaintiffs ultimately paid. Plaintiffs’ theory of harm posits that, as a result of steps defendants took to lengthen warehouse queues, spot-market participants adjusted their purchase prices upward, which ultimately refracted the MWP price at the relevant time, affecting transactions in which defendants did not participate but which embedded the MWP as a price component.31 In this respect, plaintiffs’ claim resembles those in which courts have held plaintiffs not to be efficient enforcers where their claims depended on “a complicated series of market interactions” involving the decisions of “innumerable individual decision- makers.” Reading Indus., Inc. v. Kennecott Copper Corp., 631 F.2d 10, 13 (2d Cir. 1980); see Silver II, 332 F. Supp. 3d at 907 (“[W]here the chain of causation between the asserted injury and the alleged restraint in the market contains several vaguely defined links, the claim is insufficient to provide antitrust standing.” (quoting Laydon v. Mizuho Bank, Ltd., No. 12 Civ. 3419 (GBD), 2014 WL 1280464, at *9 (S.D.N.Y. Mar. 28, 2014))). The Court nevertheless will assume for the purposes of this motion that the first several steps in this causal chain occurred as plaintiffs imagine: i.e., that defendants’ queue machinations caused the MWP to increase. That is because, although plaintiffs’ theory is complex and laden with disputed assumptions, whether longer warehouse queues in fact increased the MWP at any given time presents a merits or damages question, not a question of whether plaintiffs are the proper parties to recover damages for harms inflicted by that scheme. And for purposes of assessing antitrust standing, courts are generally to “assume the alleged violation and assess only plaintiffs’ standing to pursue their claim.” Daniel, 428 F.3d at 437. 31 As portrayed by defendants, this chain entailed, in sequence: (1) defendants’ acts lengthening warehouse queues; (2) the longer queues causing third-party spot-market participants to enter into transactions for the delivery of physical primary aluminum at higher prices, even though such transactions often did not involve LME-warehoused aluminum; (3) those participants reporting such transactions to Platts, which incorporated them into its tabulation of the MWP; (4) longer queues not reducing the LME settlement price in an offsetting manner; and (5) the inflated MWP causing plaintiffs to pay a higher all-in aluminum price because their counter- parties, i.e., smelters and industrial producers, charged the MWP without any adjustment or offsets to account for its inflation. See Def. Mem. at 6, 15. However, even assuming in plaintiffs’ favor all prior steps in the chain of causation that their theory of an aluminum price-fixing conspiracy posits, the fifth step alone—the independent decision by non-defendant sellers to charge plaintiffs a price containing the allegedly inflated MWP—“breaks the chain of causation between defendants’ actions and plaintiffs’ injury.” Sonterra v. Barclays, 366 F. Supp. 3d at 533 (alterations omitted) (quoting LIBOR VI, 2016 WL 7378980, at *17). In this critical respect, this case aligns with the post-Gelboim decisions in this District dismissing benchmark claims on the ground that “plaintiffs who are not direct purchasers are not efficient enforcers in a benchmark manipulation case.” Silver II, 332 F. Supp. 3d at 905. In enforcing that line, these courts have noted “that ‘plaintiffs who did not purchase directly from defendants continue to face the same hurdle: they made their own decisions to incorporate [the benchmark] into their transactions, over which defendants had no control, in which defendants had no input, and from which defendants did not profit.” Platinum I, 2017 WL 1169626, at *22 (quoting LIBOR VI, 2016 WL 7378980, at *16). Plaintiffs’ claims relating to purchases of aluminum from non-conspiring smelters face that hurdle, too: “the independent pricing decisions of non-conspiring retailers . . . attenuate the causal connection between the violation and the injury.” Sonterra v. Barclays, 366 F. Supp. 3d at 533. This case, in fact, supplies an unusually good illustration of the pitfalls that may arise when responsibility for prosecuting Sherman Act claims in a benchmark-price-fixing case is taken on by plaintiffs who transacted at a remove from the allegedly conspiring defendants, especially on such a complex theory of market manipulation. That is because fact discovery is now complete, enabling—to a degree greater than in other post-Gelboim cases, decided largely on the pleadings—a thorough and evidence-based assessment. And the evidence adduced makes concrete the theoretical concerns articulated in those cases. Most significantly, the evidence adduced strongly indicates that charging customers the MWP was an independent pricing decision by smelters, not the inevitable result of defendants’ alleged conspiracy. Many aluminum contracts did not expressly incorporate the MWP or MWTP. Plaintiffs have acknowledged that at least 31% of Alcoa and Rio Tinto Alcan contracts lacked such express price terms, and “appeared in the data as not having paid the full MWP.” Pl. 56.1 ¶ 58; see also, e.g., Def. 56.1 ¶¶ 12–13, 18–19, 23–24, 29, 45–46, 51, 58.32 Plaintiffs have recently sought to distance themselves from those transactions, excluding them in moving for class certification and clarifying that they do not “seek redress for any purchases where the MWP was not paid.” Pl. 56.1 ¶ 58. But whether or not plaintiffs now forgo damages from those transactions, the fact that so many such transactions took place reveals that it was readily possible for sellers of primary aluminum to decide not to charge the full MWP. As a result, plaintiffs’ alleged harms from paying an inflated MWP, even assuming that that price term was inflated due to defendants’ machinations, are most proximately attributable to the pricing decisions of third parties to the alleged conspiracy, not to defendants’ conduct. Important, too, plaintiffs have not adduced evidence of any constraint—legal, economic, or otherwise—that would have required smelters to charge plaintiffs the full MWP. See, e.g., Arg. Tr. at 77–79. Plaintiffs do contend that the MWP was effectively incorporated into all aluminum prices “[b]y industry convention.” Pl. Opp’n at 27; see also, e.g., Pl. 56.1 ¶¶ 12–13, 18–19, 23–24, 29, 45–46, 51, 58. But that fact, even if assumed true, does not make plaintiffs’ claims direct. First, even if third-party smelters, in every instance, chose to impose the inflated 32 Plaintiffs now assert, without citations to evidence and for the first time in this litigation, that the absence of the MWP in those contracts “is a function of internal accounting at Alcoa, not reflective of transactions where the MWP was not paid.” Pl. 56.1 ¶ 58. Absent supporting admissible evidence, the Court does not credit that statement. See Fed. R. Civ. P. 56(c). MWP on plaintiffs, either by incorporating the MWP expressly or tacitly allowing it to influence other pricing terms, that decision remained up to the sellers. Plaintiffs do not argue that an enterprising smelter or producer could not have chosen to charge a lower (or no) premium, perhaps to undercut competitors and grow their own market share. The critical fact is not the pricing decisions that sellers and their customers (i.e., plaintiffs) ultimately made, but the fact that they were independent actors at liberty to negotiate these terms. See, e.g., Sonterra v. Barclays, 366 F. Supp. 3d at 533 (emphasizing “the independent pricing decisions of non- conspiring retailers,” which “attenuate the causal connection between the violation and the injury”). Here, it was the non-party smelters who, to the extent the MWP appeared in contracts, chose to charge plaintiffs this premium on the purchases at issue, even as the MWP was on the rise. See, e.g., Martin Tr. at 66, 102 (noting awareness of rising premiums). That plaintiffs might have paid the MWP to third parties pursuant to industry convention, or pursuant to contracts with the MWP “hardwired” in, Pl. Opp’n at 2, does not distinguish this case from other benchmark-manipulation actions where courts have found plaintiffs’ transactions with non-defendants too indirect to support antitrust standing. In Platinum II, for example, plaintiffs alleged that defendants had manipulated “the worldwide benchmark price of platinum and palladium,” which directly depressed the price any consumers would pay for plaintiffs’ metals, and that such manipulated benchmark price was “hardwired into legal relationships” to which plaintiffs were parties. 449 F. Supp. 3d at 297, 309. And in Silver II, plaintiffs alleged manipulation of “the Fix price,” i.e., “the price of silver bullion” that all market participants paid. 332 F. Supp. 3d at 890–91. Nonetheless, in each, the court held that plaintiffs who did not transact with defendants were not efficient enforcers. See Platinum II, 449 F. Supp. 3d at 309–11; Silver II, 332 F. Supp. 3d at 909; see also LIBOR VI, 2016 WL 7378980, at *16 (not inquiring whether industry convention was to base financial instruments on LIBOR benchmark). Plaintiffs do not offer good reason to depart from these precedents, or to hold that, where plaintiffs claim that the inclusion of benchmarks was an industry-wide pricing convention, pricing decisions by a third party may be treated as if made by the remote defendants. In fact, the summary judgment record shows plaintiffs’ claims to be even less direct than those above, where plaintiffs alleged manipulation of a single benchmark price, absent other price components. Here, even where the MWP was paid, the non-conspiring smelters retained latitude to negotiate prices, including by offering offsetting discounts on other portions of the all-in aluminum price. See, e.g., Ex. 22 (Custom representative: “We have negotiated with each supplier and have had some success.”); Ex. 23 (“Def. Hall Tr.”) at 134–35 (portion of aluminum price was “a negotiated rate . . . like [Extruded] did with all our other vendors when we have to negotiate supply agreements with them for the upcoming year”). One plaintiff’s representative testified that such negotiations specifically responded to changes in the MWP. See Martin Tr. at 102, 138 (confirming that Kodak was able “to negotiate lower conversion prices in response to the rising premiums” and that Kodak “just cared about the . . . net price” (emphasis added)). Plaintiffs emphasize that the MWP price itself was non-negotiable and cite substantial evidence to that effect. But the availability of negotiations as to other price components, and their capacity to offset the inflated MWP, reinforces that the prices plaintiffs paid to non-defendants were decided by the smelters (and the plaintiffs), not by defendants. Gelboim, 823 F.3d at 780 (same for “negotiated . . . increment above LIBOR,” which was itself not negotiable). Allowing plaintiffs to pursue defendants with whom they did not do business would also create the risk of disproportionate liability. See, e.g., Platinum II, 449 F. Supp. 3d at 308–11; Silver II, 332 F. Supp. 3d at 908–99. The summary judgment record reflects that defendants sold only 2,200 metric tons of primary aluminum to the IPs and FLPs, the plaintiffs implicated by the pending motion. Pl. 56.1 ¶¶ 8, 34. But enabling these plaintiffs to pursue such defendants would expand defendants’ liability in this case alone more than 200-fold—and expose them to trebled damages on all 435,800 metric tons of aluminum the eight FLPs and IPs bought from smelters. See Def. Reply 56.1 ¶¶ 3, 8, 34; LIBOR VI, 2016 WL 7378980, at *16. Finding plaintiffs to be efficient enforcers with respect to remote transactions would further risk expanding defendants section 1 liability for ostensible price inflation to cover the more than 10 million metric tons’ worth of aluminum that smelters sold to other potential plaintiffs during the relevant period. See Def. 56.1 ¶ 61 (sales by Alcoa, Rio Tinto Alcan, and Rusal to U.S. customers totaled “more than ten million metric tons of primary aluminum”); Pl. 56.1 ¶ 61 (noting that estimate is “incomplete” in light of sales by other major smelters).33 Such sales, plaintiffs maintain, make up the “vast majority” of the primary-aluminum market, “dwarf[ing]” defendants’ sales of such aluminum. TAC ¶ 166; Arg. Tr. at 50–51. Allowing plaintiffs to pursue these claims would thus invite a significant risk of disproportionate liability: based on the record here, the transactions in which defendants did not participate and on which they did not profit overwhelm in numbers those to which defendants were parties. See also infra pp. 54–55. 33 In practice, given this longstanding litigation, the likelihood that new plaintiffs—who bought directly from smelters, but not defendants—will come forward with claims appears low. That said, with the Court having recently denied the FLPs’ motion for class certification, such claims may yet be timely for entities that purchased directly from smelters and which were included within the FLPs’ proposed class. See Am. Pipe & Constr. Co. v. Utah, 414 U.S. 538, 552 (1974). In any event, the Second Circuit has held that such “external or practical considerations” are “not substantially material” in assessing the efficient-enforcer factors. IQ Dental Supply, Inc. v. Henry Schein, Inc., 924 F.3d 57, 67 (2d Cir. 2019). That inquiry is concerned with the “overall scheme of antitrust enforcement”; therefore, “[t]he existence of other plaintiffs who could lay claim to precisely the same damages, whether in theory or in actuality, indicates that the would- be antitrust plaintiff might not be well positioned to vindicate the antitrust laws for the benefit of the public.” Id. (emphasis added). Plaintiffs’ counterarguments why they are, in fact, efficient enforcers are unpersuasive. First, plaintiffs argue cite the Second Circuit’s Aluminum VI holding to establish the directness of their injuries. For the reasons above, that decision is not controlling. Second, plaintiffs argue that theirs are not true “umbrella” claims, and differ from those dismissed in post-Gelboim cases, because, here, defendants rigged “the entire aluminum market.” E.g., Pl. Opp’n at 28. The analysis of this factor, however, does not turn on the “umbrella” label, see supra note 24, but on the extent of attenuation between defendants’ alleged price fixing and plaintiffs’ injuries. And plaintiffs’ denunciation that defendants rigged “the entire aluminum market” elides plaintiffs’ specific theory: that defendants took actions involving warehouse queues that tended to drive up the MWP, which in turn resulted in an increase in the prices plaintiffs paid for aluminum in agreements incorporating the MWP. See, e.g., Aluminum II, 95 F. Supp. 3d at 431 (“Plaintiffs do not allege that defendants have fixed a particular spot price for aluminum—but rather that they have taken a variety of actions that have caused the Midwest Premium to increase.”); id. at 434. Plaintiffs’ claim of market-wide “rigging” is not distinguishable from those in the recent line of cases that similarly alleged—often by more direct means of price- fixing—collusive setting of an industry-wide benchmark, which refracted through the industry and influenced prices for all market participants. See, e.g., See Platinum II, 449 F. Supp. 3d at 309–11; Silver II, 332 F. Supp. 3d at 909; Platinum I, 2017 WL 1169626, at *22.34 34 Plaintiffs note that a conspiracy formed with the “purpose and with the effect of raising, depressing, fixing, pegging, or stabilizing the price of a commodity . . . is illegal per se.” Pl. Opp’n at 28 (quoting Aluminum VI, 936 F.3d at 95–96). Judge Forrest stated early in this litigation, however, that it is “highly likely” that plaintiffs’ claims are subject to a “rule of reason,” not per se, analysis. Aluminum II, 95 F. Supp. 3d at 448–49. Regardless, even if defendants’ manipulation of warehouse queues with the goal of driving up the MWP were treated as an instance of per se illegal price fixing, that would not answer the efficient-enforcer question at issue here. See, e.g., Daniel, 428 F.3d at 437 (allegation of per se violation “does not absolve [plaintiffs] of the obligation to demonstrate [antitrust] standing”). And plaintiffs’ claims are well afield of those in the limited cases in which efficient- enforcer status was found, notwithstanding that plaintiffs had not transacted with defendants. In FOREX, for example, plaintiffs could not ascertain their counterparties because they had transacted on an exchange, but the defendants “dominated” the relevant market “with a combined market share of over 90%.” 2016 WL 5108131, at *9; see Silver II, 332 F. Supp. 3d at 909 (explaining that there was little risk of disproportionate liability in FOREX because “the defendants [were], de facto, ‘the market’”). Here, however, plaintiffs’ counterparties—i.e., industrial smelters such as Alcoa, Rio Tinto Alcan, and Rusal—are readily ascertainable, and plaintiffs have not adduced any evidence that defendants held such a dominant position in the market for primary physical aluminum. To the contrary, they have claimed that smelters sell the “vast majority” of such aluminum in the United States “directly to industrial users.” TAC ¶ 166; JAC ¶ 102; Fujifilm Compl. ¶ 89; see also Arg.. See, e.g., Arg. Tr. at 50–51 (defendants’ sales “dwarfed by the purchases from smelters”). Gold is similarly inapposite. Decided on a motion to dismiss and before the decision on remand from Gelboim in LIBOR VI, the Gold court declined to dismiss umbrella claims, relying on allegations that “all market participants . . . ‘moved in line’ according to Defendants’ price manipulation, leaving little room for any interfering price impact due to the actions of non- culpable entities or exogenous market forces.” Gold, 213 F. Supp. 3d at 656. Here, the fulsome summary judgment record shows the opposite—that nonculpable entities and exogenous market forces played a substantial role in determining the prices plaintiffs ultimately paid for aluminum. See supra pp. 45–46, 48–50; infra pp. 58–63; see also Gold, 213 F. Supp. 3d at 656 (noting “substantial challenges to Plaintiffs’ causation theory,” but deferring resolution of such questions to later stages of the litigation absent a “sufficiently developed” record). Finally, in Dennis, also decided at the pleading stage, plaintiffs’ alleged conspiracy “involve[d] essentially all of the entities that contribute to setting” BBSW. 343 F. Supp. 3d at 165. Here, on the other hand, defendants did not “set” the MWP, and plaintiffs themselves emphasize that a plethora of factors influenced the MWP and the LME settlement price, only some of which (delivery and storage costs) defendants’ conduct even affected. See, e.g., Ex. 75 (“Vazquez Decl.”) at 3–6. In sum, this case far more closely resembles the majority of post- Gelboim decisions that have held indirect purchasers not to be efficient enforcers than it does the few that have held otherwise.35 Third, plaintiffs argue that—considered in light of the scope of defendants’ alleged wrongdoing and their wrongful gains from sales to non-plaintiffs—their claims would not impose disproportionate liability. See Pl. Opp’n at 23 (“[D]amages are not disproportionate to Defendants’ activities in fixing the entire aluminum market”). They also note that the damages they pursue from defendants based on remote or “umbrella” claims would not reach the trillions of dollars, in contrast to some benchmark-manipulation lawsuits. Id. at 23–27. But plaintiffs have not adduced any evidence that defendants’ sales of primary aluminum approach the 10 million metric tons that smelters sold during the relevant period, and have confirmed that, generally, such purchases from defendants were “dwarfed by the purchases from smelters.” 35 Loeb Industries, Inc. v. Sumitomo Corp., 306 F.3d 469, 483 n.4 (7th Cir. 2002), which plaintiffs cite, does not counsel a different result. Loeb did not decide the umbrella-standing issue, which it left “open for further exploration at the district court level.” Id. And Loeb’s dicta suggesting that plaintiffs should have antitrust standing to sue defendants with whom they did not transact but who “rigg[ed] product standards” is inapposite. Id. The MWP is not a product standard, and plaintiffs accuse defendants of indirectly influencing, not “rigging,” the premium. And Loeb is factually afield. In Loeb, the contracts at issue were “directly and explicitly based on the” manipulated benchmark, whereas here, in many instances, that was not so. Id. at 488; see id. at 489 (“It is this contractual linkage, absent in Reading, that prevents other market variables from” complicating the causal analysis). Arg. Tr. at 50–51; see Pl. 56.1 ¶ 61. Were defendants liable for all sales by those non-conspiring parties, as holding for plaintiffs would authorize, defendants’ potential damages would far outstrip their ill-gained profits.36 Even if that exposure would fall short of bankrupting a defendant, Sullivan, 2017 WL 685570, at *16, a multiple of U.S. GDP, id., or “astronomical,” Sonterra v. Credit Suisse, 277 F. Supp. 3d at 559–61, that does not resolve whether it would be disproportionate. “The question is not . . . whether defendants are subject to ‘ruinous liability,’” but “whether defendants—no matter their size—are subject to liability that is disproportionate to their allegedly ill-gotten gains.” Platinum II, 449 F. Supp. 3d at 310 n.14. The evidence here strongly suggests that disproportionate liability would result from holding defendants liable for every sale of aluminum in the United States that, explicitly or implicitly, involved the MWP. Accordingly, because plaintiffs’ claims would subject defendants with whom they had done business to the risk of damages that are disproportionate to their ill-gotten gains, this first factor strongly weights against antitrust standing. 2. Existence of More Direct Victims Although defendants focus mainly on the first efficient-enforcer factor—the directness of plaintiffs’ injury—they argue that the remaining three factors also favor dismissal of plaintiffs’ umbrella claims. As to the second factor, more direct purchasers are indeed readily ascertainable. Indeed, one is present among the plaintiffs implicated by the instant motion: Ampal, which directly bought 2,200 metric tons of primary aluminum from defendants. See Def. 56.1 ¶¶ 9–11. And, as defendants note, in a related action consolidated before this Court, two plaintiffs against whom 36 Plaintiffs identify certain large-scale sales of aluminum by defendants, Pl. 56.1 ¶ 84, but do not specify whether those included the MWP, Def. Reply 56.1 ¶ 84 (many sales identified by plaintiffs did not involve MWP). Defendants also counter that plaintiffs misleadingly present defendants’ aluminum holdings as sales figures, and skew data so as to suggest greater sales by defendants (e.g., citing global versus domestic sales and profit figures). Id.; Def. Reply at 17–18. defendants have not moved for summary judgment—Reynolds Consumer Products LLC (“Reynolds”) and Southwire Company LLC (“Southwire”)—made a substantial portion of their purchases from defendants or their affiliates. See Def. Mem. at 20. Assuming defendants’ misconduct as alleged, those purchasers were more directly injured than were the IPs or most of the FLPs. That is because, in these transactions, defendants—not a third party—made the decision to charge the price component which defendants allegedly colluded to inflate. As a result, such claims, unlike those of most plaintiffs here, implicate defendants’ illicit profit at plaintiffs’ alleged expense.37 See, e.g., Platinum II, 449 F. Supp. 3d at 306 (“[T]hose who transacted directly with defendants may be conceived of as more direct victims . . . than those who did not.”); Sonterra v. Credit Suisse, 277 F. Supp. 3d at 560 (such purchasers are “obviously the most efficient enforcers”); Sullivan, 2017 WL 685570, at *18 (second factor favored defendants because a “counterparty to a defendant employing the price-fixers [was] in the immediate impact zone of the defendant’s unlawful conduct”). 37 Plaintiffs, curiously, argue that claims by direct purchasers—apparently including Ampal, Reynolds, and Southwire—against defendants could be barred by Illinois Brick Co. v. Illinois, 431 U.S. 720, 730 (1977), and thus that plaintiffs are the only viable enforcers here. That is because, they argue, those purchases were “secondary” in the sense that defendants had first bought aluminum from a smelter, and paid the MWP, before selling it to those customers at a price including the then-inflated MWP. See Pl. Opp’n at 20. Plaintiffs thus suggest that the holding in Illinois Brick—that those who were harmed through “pass-on” price increases generally cannot recover antitrust damages, see Ill. Brick, 431 U.S. at 734–35—might limit direct- purchaser plaintiffs’ ability to recover on purchases from the antitrust violators themselves. Plaintiffs do not cite any authority for that proposition. And defendants appear to concede that Illinois Brick does not bar the claims asserted by Ampal, Reynolds, or Southwire. See Def. Reply at 15 (plaintiffs’ argument under Illinois Brick “get[s] it backwards”); Arg. Tr. at 80 (“We never said that . . . the direct purchasers were barred by Illinois Brick, and any such statement in my view would be completely erroneous.”). In any event, the Supreme Court has recently underscored that such a claim would be available: “[W]e have consistently stated that ‘the immediate buyers from the alleged antitrust violator’ may maintain a suit against the antitrust violators.” Apple Inc. v. Pepper, 139 S. Ct. 1514, 1520 (2019) (emphasis added) (quoting Kansas v. UtiliCorp. United Inc., 497 U.S. 199, 207 (1990)). To be sure, in benchmark-manipulation cases, courts have tended to give this second factor “diminished weight.” Gelboim, 823 F.3d at 779. That is because, in the typical such case, where a fixed benchmark applies equally to direct and indirect purchasers, the effects of a rigged benchmark apply equally to all purchasers. But here, as noted, the premise that the benchmark equally affected direct and indirect purchasers is far from apparent, as reflected in the many contracts lacking an explicit MWP term, and because the parties to aluminum contracts, including innocent smelters, could negotiate other price components to mitigate the impact of the inflated MWP. Direct purchasers, as a category, are thus in a superior position to pursue antitrust claims. As to those purchasers, the pricing decisions can fairly be attributed most proximately to defendants’ conduct. This factor thus also favors dismissal of plaintiffs’ claims based on transactions with non- defendants. There are more efficient enforcers who could and, indeed have, sued: those who bought primary physical aluminum directly from defendants. See Def. Reply 56.1 ¶ 84 (confirming that “Goldman Sachs, JPMorgan and Glencore sold physical aluminum to consumers during the relevant period”); Def. Reply at 2 (“Reynolds and Southwire . . . are far more efficient enforcers of the antitrust laws.”); see also Sonterra v. Credit Suisse, 277 F. Supp. 3d at 563 (second factor did not help umbrella plaintiffs, “as either this factor should be assigned reduced weight in this context or the umbrella plaintiffs should be categorized as less direct victims”). At the same time, because the inflated price here was a widely (though not universally) used benchmark in the industry, a hypothetical indirect purchaser who demonstrably paid the MWP, without offset, suffered a harm similar to those who transacted with defendants. Such a purchaser, however, would have a higher degree of difficulty in establishing injury traceable to collusion than one who transacted directly with a colluding defendant. 3. Speculative Damages Defendants also argue that plaintiffs are not efficient enforcers because their claims implicate unduly speculative damages. Plaintiffs counter that their claims as less speculative than those dismissed in other benchmark-manipulation decisions. That is unpersuasive. As the summary judgment record reveals, the process of determining a plaintiff’s damages on an aluminum purchase from a non-defendant is rife with complicating factors. These include intervening pricing and contracting decisions by the nonculpable sellers with whom plaintiffs negotiated and from whom they purchased, and the challenging inquiry into whether, but for the inflated MWP, other price terms would have been different. Critically, too, unlike in other cases in which courts have dismissed umbrella plaintiffs’ claims, plaintiffs’ theory here is not even that defendants directly manipulated the benchmark price of the good in question. See, e.g., Platinum II, 449 F. Supp. 3d at 299, 306–07 (defendants allegedly conspired to rig the “benchmark price for platinum and palladium,” not a component thereof). Rather, plaintiffs’ theory is that defendants conspired to take certain actions tending to elongate the queue at certain aluminum warehouses, which in turn tended to cause a benchmark component of aluminum’s price—the MWP—to rise. See Aluminum II, 95 F. Supp. 3d at 431 (“Plaintiffs do not allege that defendants have fixed a particular spot price for aluminum—but rather that they have taken a variety of actions that have caused the Midwest Premium to increase.”); id. at 434 (“This conspiracy was not a traditional ‘price fix’—no price was fixed.”). And, as the Court noted in denying class certification—in part because defendants’ alleged manipulation of queues was “occasional and lumpy”—the analysis of the market impact, if any, from defendants’ machinations unavoidably must be evaluated based on conditions and conduct at the particular moment in time of each transaction. See Aluminum VII, 336 F.R.D. at 56–57. As a result, even a plaintiff like Ampal, suing on the basis of its limited direct purchases of primary physical aluminum from a defendant, would confront a considerable challenge in attempting reliably to quantify its damages. It is worth reviewing concretely the challenge that a direct purchaser would face before examining the heightened challenge faced by an indirect- purchaser plaintiff. A direct-purchaser plaintiff like Ampal would, first, have to isolate the impact of the alleged conspiracy on the MWP. Doing so would require, at least, identifying (1) how much of the lengthening in warehouse queues was attributable to defendants’ warrant cancellations or other alleged illicit activity, as opposed to extrinsic market factors;38 (2) the extent to which longer warehouse queues actually caused spot-market prices to rise, thereby causing Platts to report an inflated MWP; and (3) disentangling the effects of defendants’ actions from the many other drivers of the MWP. See, e.g., Vazquez Decl. at 5 (identifying, as only “[t]he most important underlying market determinants of the MWP”: “inland, barge and ocean freight rates,” “storage fees,” “loading/unloading charges,” “port fees,” “cost of financing,” and “duties”). A direct-purchaser plaintiff, next, would have to show how much of any defendant- caused increases in the MWP were not offset by resulting decreases in the LME settlement price attributable to defendants’ conduct. See Def. 56.1 ¶ 57 (collecting sources discussing potential downward pressure on LME price of longer queues); Pl. 56.1 ¶ 57 (disputing whether queues might drive down LME price); Aluminum VII, 336 F.R.D. at 57–58 (explaining how lengthened 38 Plaintiffs’ expert at the class-certification stage has already characterized such cancellations as “difficult to model” given the “occasional and lumpy” nature of warrant cancellations. Aluminum VII, 336 F.R.D. at 52, 56 (quoting report of Dr. Christopher L. Gilbert). That “lumpiness” severely undermines plaintiffs’ unsupported claim that “Defendants’ manipulation of the MWP . . . was sustained and universal.” Pl. Opp’n at 21–22. queues might depress value and price of LME warrants); see also Sonterra v. Credit Suisse, 277 F. Supp. 3d at 563 (collecting cases holding that “any damages would need to be netted out as to each plaintiff to offset any benefit from defendants’ manipulation”).39 Uncertainty as to the interplay between the MWP and the LME settlement price itself sets this case apart from others, in which the allegedly fixed price comprised the “only ‘benchmark’” that determined the price plaintiffs ultimately paid or accepted. See, e.g., Platinum I, 2017 WL 1169626, at *24; LIBOR VI, 2016 WL 7378980, at *20 (distinguishing Loeb on the basis that, in Loeb, “the evidence show[ed] that as the [manipulated] Comex price increased, the premium also increased” and that there was thus “no possibility that the two components ‘offset’ or that the premium somehow compensated for the defendants’ manipulated price inflation.” (emphasis added) (quoting Loeb, 306 F.3d at 487–88)). 39 Plaintiffs contend that defendants’ acknowledgment of disagreement on this point defeats their argument because, at summary judgment, such disputes “must be resolved in favor of Plaintiffs.” PL. 56.1 ¶ 57. That broad principle, of course, is true. But the material issue under the third efficient-enforcer factor is the presence of speculation and, hence, uncertainty. Thus, even at the more forgiving pleading stage, courts have found this factor to favor dismissal in benchmark- manipulation cases. See, e.g., Silver II, 332 F. Supp. 3d at 911; Platinum I, 2017 WL 1169626, at *24–25; Sullivan, 2017 WL 685570, at *18–19. And here, there is no genuine dispute that the relationship between longer queues and the all-in prices paid by customers is complex and debatable. See Pl. Opp’n at 18 (characterizing dispute as “a hotly contested factual matter”). Compare, e.g., Ex. 55 at 2 (“[T]he impact of premium inflation is not to increase the [all-in] price but rather to cause the LME price to trade at a discount to the exogenously-set [all-in] price.”), Ex. 63 at 6 (Alcoa stating that “it is not the case that long queues have resulted in higher overall metal prices,” and that “the total price for delivered aluminum continues to be well below the pre-financial crisis levels”), and Ex. 64 (Rusal stating that it “agrees with the LME’s analysis that any rise in aluminum premiums would lead to a corresponding fall in the exchange price”), with Vazquez Decl. at 2 (contending that “the LME price does not inherently move inversely to the MW Premium,” owing to independent inputs into each), id. at 3–9. But see Pl. Opp’n at 6 (showing MWP rising and LME price falling between 2010 and 2014); Ex. 76 (“Supp. Zona Rpt.”) ¶ 9 (concluding only that changes in the MWP were “not fully offset by any changes in the LME Price,” not that the LME did not decline in response to longer queues (emphasis added)). A direct-purchaser plaintiff would then need to identify the role, if any, that the inflated MWP played in the plaintiff’s actual purchases of aluminum. Plaintiffs minimize this challenge, depicting aluminum contracts as “relatively straight-forward” and urging that “damages can readily be calculated by determining the amount by which the premiums were inflated.” Pl. Opp’n at 21. But although that proposition sufficed when the Court was limited to considering plaintiffs’ pleadings, see Aluminum II, 95 F. Supp. 3d at 444, there is today an evidentiary record, which belies this easy formula. That record reflects that, even for aluminum purchases that included the payment of a price term based on the MWP, plaintiffs could lock in long-term aluminum contracts that provided a stable all-in price—including the MWP—for up to a year out. See, e.g., Ex. 46 at 17–22 (providing Kodak fixed price for all of 2013). Other contracts locked in fixed prices over different time frames. See, e.g., Ex. 10 at 2 (Ampal contract providing “fixed, firm price” for several months out); Ex. 15 (similar Claridge agreement). Plaintiffs have confirmed that such forward buys “lock[ed] in the all-in price” for aluminum— including the MWP. Def. Zecchini Tr. at 96 (emphasis added); see Pl. 56.1 ¶ 52 (agreeing that “Mag’s forward contracts generally provide stability of the all-in price for aluminum of which the Midwest Premium is only one part”). Any damages estimate would have to account for the nuances of the particular contract, and assess in light of these the impact of the incremental inflation, if any, of the MWP at the relevant point over the course of that contract term. Undoubtedly, the uncertainty attendant to such an exercise would be compounded by contracts that do not expressly indicate the role played by the MWP in the all-in price paid. The exercise of reliably determining damages, difficult even as to a direct-purchaser plaintiff like Ampal, would be more daunting still as to indirect-purchaser plaintiffs. The intervening decisions by a non-cartel seller complicate the injury, particularly because the all-in prices plaintiffs paid to such sellers were subject to negotiations. See LIBOR VI, 2016 WL 7378980, at *19 (noting that the “Second Circuit expressed skepticism about the measure of damages in such highly negotiated transactions.”). That defendants did not have any role in plaintiffs’ aluminum transactions with third- party smelters like Alcoa “makes it even harder to determine what role, if any, [the MWP] played in the ultimate price” plaintiffs paid to such third parties, especially where those parties did not expressly charge the MWP in their contracts. Sonterra v. Credit Suisse, 277 F. Supp. 3d at 564. As discussed, many of the major smelters’ aluminum contracts do not expressly incorporate the MWP as a price term. And as plaintiffs have shown, reconstructing those third parties’ pricing decisions is a fraught exercise. Even after close examination of several such contracts, plaintiffs’ summary judgment expert was effectively left to conjecture as to the extent to which the MWP had influenced the all-in prices the plaintiff purchaser had paid. See, e.g., Vazquez Decl. at 16–19 (inferring that, because contract prices did not equal the LME price plus the MWP as of the relevant dates, the final sales prices includes “an Alcan margin on the transaction” or “a likely upcharge for immediate delivery”). Although it may be that the defendants’ contracts for the sale of aluminum also fail expressly to incorporate the MWP, they, as parties to this litigation, are much better situated to explain their pricing decisions than are smelters and producers whose motivations are unknown. See, e.g., Sullivan, 2017 WL 685570, at *19 (for transactions “in which a . . . defendant was a counterparty, . . . that defendant, based upon its own records, ought to be able to intelligently respond”). The special challenges an indirect-purchaser plaintiff would face in reliably estimating damages from third-party transactions are heightened by the presence of negotiations as to the all-in price they paid. See Sonterra v. Barclays, 336 F. Supp. 3d at 547 (“[H]ighly negotiated contracts . . . make the effect of [benchmark] manipulation highly speculative.”). Although the record evidence supports that plaintiffs were unable to negotiate the calculation of the MWP to the extent it was included in their contracts, plaintiffs admit that they were able to negotiate other aspects of the all-in prices they paid for aluminum. See Pl. 56.1 ¶¶ 15, 20, 26, 30, 38, 42, 48, 53; e.g., Ex. 22 (“We have negotiated with each supplier to the best of our ability and have had some success.”). Plaintiffs also could, and Kodak apparently did, negotiate “in response to the rising premiums.” Martin Tr. at 102. A reliable damages calculation would need to account for such decisions, which, as Kodak’s witness recognized, may specifically have taken place in order to protect purchasers for otherwise inflexible MWP terms. See Sonterra v. Barclays, 366 F. Supp. 3d at 546 (damages speculative where it might be “impossible to untangle the impact of the fixed price from the impact of intervening market decisions”); see also Gelboim, 823 F.3d at 780 (negotiations made damages speculative “notwithstanding that the negotiated component was the increment above” the manipulated benchmark); LIBOR VI, 2016 WL 7378980, at *20 (damages speculative because “the negotiated component [of transactions] compensated for” the manipulated benchmark). As to direct-purchaser plaintiffs, on the other hand, all pricing decisions, including negotiations, can fairly be attributed to a conspiring defendant. In sum, for the plaintiffs implicated by this motion, the damages inquiry would unavoidably be speculative. It would “require a complex multi-step analysis to quantify the indirect effect of Defendants’ alleged manipulation” of the MWP on the prices plaintiffs paid to third-party sellers of aluminum. 7 W. 57th St., 314 F. Supp. 3d at 514. Although much of that complexity applies to direct-purchaser and umbrella plaintiffs alike, the inquiry “will be simplified considerably by confining it to those transactions in which defendants actually took part.” Sonterra v. Credit Suisse, 277 F. Supp. 3d at 564; see Sullivan, 2017 WL 685570, at *15 (damages inquiry, as to direct purchasers, would be “cabined and discernable”). As to such transactions, defendants, for example, can testify as to bases of their own pricing decisions, without the need to inject a third-party participant’s recollections, or the lack thereof, into the mix. See, e.g., Sullivan, 2017 WL 685570, at *19 (“[T]he two plaintiffs ought to be able to identify specific transactions in which a remaining named defendant was a counterparty and that defendant, based upon its own records, ought to be able to intelligently respond.”). Accordingly, the speculative nature of plaintiffs’ damages also supports a finding that the umbrella plaintiffs in this litigation are not efficient enforcers of the section 1 claims at issue. 4. Duplicative Recovery or Complex Apportionment Defendants concede that plaintiffs’ claims present “no obvious risk of duplicative recovery,” given the absence of parallel litigation against them by other market participants. Def. Mem. at 23. And given the long history of this litigation, there is no reason at this late date to anticipate that a competing claimant will come forward. Defendants state, conclusorily, that they are concerned about “the question of complex apportionment.” Id. at 23–24. But they do not explain the basis of that concern, with no other party pursuing claims alongside the purchasing plaintiff on each transaction in which they engaged. Accordingly, this factor does not weigh against plaintiffs’ status as efficient enforcers. Defendants’ inability to establish this factor is not, however, dispositive, insofar as the other efficient-enforcer factors disfavor plaintiffs as to the claims at issue. See, e.g., 7 W. 57th St., 771 F. App’x at 503 (affirming denial of leave to amend as futile, on efficient-enforcer grounds, even where “[d]efendants do not offer any serious argument why” the case presented a risk of “complex apportionment of damages or would run the risk of duplicative recovery”); Sonterra v. Barclays, 366 F. Supp. 3d at 548 (dismissing on efficient-enforcer grounds despite no apparent risk of duplicative or complex apportionment); LIBOR VI, 2016 WL 7379890, at *23. * * * Because plaintiffs other than Ampal (as to its one, 2,200-metric-ton purchase) have failed to satisfy three of the four efficient-enforcer factors, the Court dismisses, for want of antitrust standing, the MWP-manipulation claims asserted by those plaintiffs who did not transact directly with defendants. CONCLUSION For the foregoing reasons, the Court grants defendants’ motion for summary judgment as to the IPs’ claims and most of the FLPs’ claims. All claims brought by plaintiffs Custom, Claridge, Extruded, Agfa, Mag, Kodak, and Fujifilm are dismissed. Ampal’s claims are dismissed to the extent they do not arise from transactions directly with Glencore, Century, or any other defendant or co-conspirator. The Clerk of Court is respectfully directed to close the following cases: 14 Civ. 211; 14 Civ. 217; 14 Civ. 6849; and 15 Civ. 8307. The Clerk of Court is further directed to terminate the following plaintiffs as parties to cases 13 MD 2481 and 14 Civ. 3116: (1) Custom Aluminum Products, Inc.; (2) Claridge Products and Equipment, Inc.; and (3) Extruded Aluminum Corporation. The Clerk of Court is also directed to terminate the following plaintiffs as parties to case 13 MD 2481: (1) Agfa Corporation; (2) AGFA Graphics, NV; (3) Mag Instrument Inc; (4) Eastman Kodak Company; (5) Fujifilm Manufacturing U.S.A., Inc. This case will now proceed solely on the claims asserted by plaintiffs who transacted directly with defendants during the relevant period, i.e., Ampal in 14 Civ. 3116, and Reynolds and Southwire in 16 Civ. 5955. An order will issue shortly as to next steps in this case. SO ORDERED. ____________________________ Paul A. Engelmayer United States District Judge Dated: February 17, 2021 New York, New York