Opinion
No. 21-853 August Term 2021
07-20-2022
GAMMA TRADERS - I LLC, individually and on behalf of all others similarly situated, Vega Traders LLC, individually and on behalf of all others similarly situated, Michael Patterson, Plaintiffs-Appellants, Robert Charles Class A, L.P., Robert L. Teel, Yuri Alishaev, Abraham Jeremias, individually and on behalf of all others similarly situated, Morris Jeremias, individually and on behalf of all others similarly situated, Consolidated Plaintiffs-Appellants, v. MERRILL LYNCH COMMODITIES, INC., Bank of America Corporation, Morgan Stanley & Co. LLC, Edward Bases, John Pacilio, Defendants-Appellees.
Eric F. Citron, Goldstein & Russell, P.C., Bethesda, MD (Deborah Clark-Weintraub, Max R. Schwartz, Thomas L. Laughlin, IV, Jeffrey P. Jacobson, Scott+Scott Attorneys at Law LLP, New York, NY; Daniel Woofter, Goldstein & Russell, P.C., Bethesda, MD; Vincent Briganti, Margaret MacLean, Lowey Dannenberg, P.C., White Plains, NY, on the briefs), for Plaintiffs-Appellants Gamma Traders - I LLC, et al. Richard F. Schwed (Adam S. Hakki, on the brief), Shearman & Sterling LLP, New York, NY, for Defendants-Appellees Merrill Lynch Commodities, Inc., and Bank of America Corporation. Scott D. Musoff, Skadden, Arps, Slate, Meagher & Flom LLP, New York, NY, for Defendant-Appellee Morgan Stanley & Co. LLC. David H. McGill, Kobre & Kim LLP, Washington, DC, for Defendant-Appellee John Pacilio. Evan Cohen, Finn Dixon & Herling LLP, Stamford, CT, for Defendant-Appellee Edward Bases.
Eric F. Citron, Goldstein & Russell, P.C., Bethesda, MD (Deborah Clark-Weintraub, Max R. Schwartz, Thomas L. Laughlin, IV, Jeffrey P. Jacobson, Scott+Scott Attorneys at Law LLP, New York, NY; Daniel Woofter, Goldstein & Russell, P.C., Bethesda, MD; Vincent Briganti, Margaret MacLean, Lowey Dannenberg, P.C., White Plains, NY, on the briefs), for Plaintiffs-Appellants Gamma Traders - I LLC, et al.
Richard F. Schwed (Adam S. Hakki, on the brief), Shearman & Sterling LLP, New York, NY, for Defendants-Appellees Merrill Lynch Commodities, Inc., and Bank of America Corporation.
Scott D. Musoff, Skadden, Arps, Slate, Meagher & Flom LLP, New York, NY, for Defendant-Appellee Morgan Stanley & Co. LLC.
David H. McGill, Kobre & Kim LLP, Washington, DC, for Defendant-Appellee John Pacilio.
Evan Cohen, Finn Dixon & Herling LLP, Stamford, CT, for Defendant-Appellee Edward Bases.
Before: Lynch, Carney, And Sullivan, Circuit Judges.
Richard J. Sullivan, Circuit Judge:
Plaintiffs-Appellants Gamma Traders - I LLC; Vega Traders LLC; Robert Charles Class A, L.P.; and various individuals (collectively, "Gamma") brought this suit pursuant to the Commodity Exchange Act ("CEA"), 7 U.S.C. § 1 et seq. , alleging that they were harmed by the defendants’ manipulation of futures markets for four precious metals – gold, silver, platinum, and palladium – as well as the options markets on those futures contracts. The defendants, who previously admitted to having manipulated prices in those markets, are three corporations – Merrill Lynch Commodities, Inc.; Bank of America Corporation; Morgan Stanley & Co. LLC – and two individuals who were employed by the corporate defendants (collectively, "Defendants"). The district court (Liman, J. ) dismissed Gamma's complaint as time-barred and, in the alternative, for failing to adequately plead that Defendants’ conduct had actually injured Gamma. Because we agree with the district court that Gamma has not adequately pleaded that it was "actual[ly] damage[d]" by Defendants, as required to state a claim under the CEA, 7 U.S.C. § 25(a)(1), we affirm the district court's dismissal of Gamma's complaint. I. BACKGROUND
A. Facts
Defendants manipulated the prices for precious metals futures and option contracts by "spoofing," a fraudulent practice in which the spoofing traders send false supply and demand signals to the market by placing orders to buy or sell that they never intend to execute. For instance, a trader who wishes to sell gold at a favorable price might place an offer to buy a large quantity of gold at just below the market price. The spoofing trader does not intend to actually execute this offer but instead hopes that other traders, who can see the number of orders available in the market, will respond to the perceived uptick in demand for gold by temporarily pushing the actual market price higher. At that point, the spoofing trader can sell gold at an artificially high price. After selling the gold at inflated prices, the spoofing trader will then cancel her original offer to buy gold before it can be executed. Conversely, a spoofing trader who wishes to purchase the commodity at a lower price will place a large sell order, which she does not intend to execute, and when the market price falls in reaction to this perceived movement toward a sell-off, she can buy at an artificially depressed price. The process may also be used, as is alleged here, to manipulate the price of commodity futures and options contracts.
In addition to publicly placing orders she does not intend to execute, a spoofing trader can also conceal her actual intention – to trade in the direction opposite her spoofing – by placing "iceberg" orders. J. App'x at 64 ¶ 37. An iceberg order entails placing a large order but allowing it to become visible to the market only in incremental portions. Once a portion of the order is filled, another predetermined portion becomes visible to the market, and so on until the whole order has been filled. The purpose of this technique is to avoid price movements that would otherwise result from placing a single, large order. A spoofing trader can therefore use iceberg orders to avoid sending accurate supply and demand signals to the market about orders that she does intend to fill. Thus, a trader wishing to sell a commodity might attempt to spoof the market by placing many buy orders she does not intend to execute – falsely signaling high demand for the commodity and thus driving up the price – while simultaneously placing sell orders above market price using the iceberg technique to prevent the downward pricing pressure that would otherwise be generated by revealing a large sell order to other traders. Moreover, there is no mechanism that discloses the identity of traders who have placed orders in the market, so other market participants will not be able to detect or react to the suspicious trading activity of one trader placing both buy- and sell-side orders in the same market.
One might wonder why a trader would ever not use an iceberg order to mask the true volume a trader seeks to buy or sell. On many exchanges, including at least one on which the parties traded here, bids are filled using a "first in, first out" system. For instance, if there are ten orders to buy gold at $1,000, the first one placed will be the one filled by the first seller to offer a price of $1,000, while the other would-be buyers must wait. With an iceberg order, only the visible part of the order receives this timing priority; after it is filled, the remainder of the order must wait at the back of the line, as if it had just been placed for the first time. That forfeiture of priority in the timing hierarchy discourages investors from exclusively using iceberg orders.
Defendants in this action used these spoofing techniques repeatedly in their trading activity, for which they have faced criminal and regulatory enforcement actions that resulted in the imposition of criminal penalties and fines. Specifically, on June 25, 2019, Defendant-Appellee Merrill Lynch Commodities, Inc. entered into a non-prosecution agreement with the U.S. Department of Justice and a settlement with the Commodity Futures Trading Commission (the "CFTC"), in which it agreed to pay $25 million in fines, restitution, and forfeiture because of its traders’ spoofing activities, including those of Defendants-Appellees Edward Bases and John Pacilio, who were indicted individually in the Northern District of Illinois in July 2018. Defendants’ spoofing activities were frequent and widespread. The cases brought by the Department of Justice and CFTC have uncovered thousands of instances of spoofing over a period lasting more than four years and affecting markets for the four precious metals at issue in this case. Additionally, as these government investigations and prosecutions were proceeding, private plaintiffs brought suits alleging price-fixing and market manipulation, including by spoofing, in various metals markets. See, e.g. , Third Amended Complaint at 6, 121, In re London Silver Fixing, Ltd., Antitrust Litig. , No. 14-md-2573 (VEC) (S.D.N.Y. June 6, 2017), ECF No. 258.
B. Procedural History
On June 27, 2019, Gamma commenced this action, alleging that Defendants had spoofed in the futures and options markets for gold, silver, platinum, and palladium. Gamma sought certification of a class on behalf of all persons who bought or sold futures contracts, or options on those contracts, in any of the four metals markets at issue between January 1, 2007, and December 31, 2014. Gamma alleged four separate causes of action, including: (1) market manipulation in violation of the CEA; (2) employing a manipulative and deceptive device in violation of the CEA; (3) principal-agent liability under the CEA, to the extent that any of the Defendants’ agents, representatives, or others acting on their behalf assisted in the spoofing schemes; and (4) unjust enrichment for profits derived by Defendants as a result of their spoofing activities. See generally 7 U.S.C. § 1 et seq. After Defendants filed a motion to dismiss the initial complaint, Gamma responded by filing an amended complaint, which is now the operative complaint under review.
The amended complaint maintained the same class period and four claims for relief. It alleged that "Defendants spoofed the market for precious metals futures contracts thousands of times throughout the Class Period." J. App'x at 56 ¶ 11. In particular, Gamma relied on information revealed by various government investigations and prosecutions to allege that between February 4, 2011, and April 17, 2014, there were nine separate dates on which at least one plaintiff traded, and at least one defendant spoofed in the opposite direction, in the same market. In other words, on each of those nine dates in question, at least one defendant was spoofing the price down and a plaintiff was attempting to sell, or a defendant was spoofing the price up and a plaintiff was attempting to buy.
Gamma states in its briefing that the complaint identified fourteen dates on which at least one Defendant spoofed and at least one plaintiff traded in the opposite direction in the same market, but the relevant portions of the complaint reflect that although Gamma alleges fourteen specific dates when Defendants spoofed, it makes the additional allegation that it traded only on nine, not fourteen, of the dates identified in the complaint.
After Defendants moved to dismiss the amended complaint, the district court granted their motion in full, finding that Gamma's claims were time-barred and, alternatively, that Gamma failed to state a claim because it did not plausibly allege that it was harmed by Defendants’ spoofing.
The district court held, in relevant part, that Gamma failed to state a claim because it did not adequately plead damages under the CEA. The court noted that a party seeking to recover under the CEA must plead – and, ultimately, prove – that the CEA violations harmed it by causing it to trade at a manipulated price that redounded to its detriment. See Harry v. Total Gas & Power N. Am., Inc. (Total Gas ), 889 F.3d 104, 112 (2d Cir. 2018). While "[t]he most direct way to plead such [a] harm is to point to a specific manipulated transaction or set of transactions between a plaintiff and a defendant with the plaintiff on the (net) losing end and the defendant on the (net) winning end," id. , the district court recognized that pleading this type of privity is not strictly necessary. It nonetheless concluded that Gamma's pleadings did not meet the Total Gas standard.
Because we affirm on the basis that Gamma did not adequately plead damages under the CEA, we have no need to address the district court's determination that Gamma's claims were untimely.
In reaching that conclusion, the district court rejected Gamma's argument that a CEA injury could be inferred from the facts that: (1) there were at least nine days on which Defendants spoofed and Gamma traded on the same day; and (2) Defendants’ spoofing extended to thousands of manipulated trades. The district court rejected the first theory of injury because Gamma never pleaded that it traded after Defendants spoofed on a particular day – and if it traded before the spoofing activity, there was no basis in the complaint to infer that the spoofing affected the price at which Gamma traded. Moreover, in the district court's view, the complaint provided no basis to infer that Gamma was harmed by Defendants’ spoofing as opposed to having benefited from it.
The district court further concluded that Gamma's second theory – that Gamma traded sufficiently actively, and Defendants spoofed sufficiently frequently, to render plausible the inference that the spoofing affected the price at which Gamma traded on at least one occasion – was too speculative. According to the district court, the relative magnitude of the thousands of trades alleged by Gamma dwindled when placed in the context of an eight-year class period involving four different futures markets and four additional corresponding options markets. The district court concluded that "[a]bsent either a far greater volume of trading or some additional facts pleaded," it was merely speculative – as opposed to plausible – that Defendants’ spoofing activities had a negative effect on Gamma. J. App'x at 163.
The district court also dismissed Gamma's state-law unjust enrichment claim, but because Gamma does not challenge that dismissal on appeal, we do not pass on it here.
Gamma timely appealed.
II. DISCUSSION
We review de novo the dismissal of a complaint for failure to state a claim. Deutsche Bank Nat'l Tr. Co. v. Quicken Loans Inc. , 810 F.3d 861, 865 (2d Cir. 2015). In assessing a motion to dismiss, we "accept[ ] all factual allegations in the complaint as true[ ] and draw[ ] all reasonable inferences in the plaintiff's favor." Shomo v. City of New York , 579 F.3d 176, 183 (2d Cir. 2009) (citation omitted). Because we hold that Gamma did not adequately plead that Defendants’ spoofing injured it, we do not reach the question of whether its claims are timely.
A. Failure to Plead Damages
Because the CEA ultimately makes offending parties "liable for actual damages," 7 U.S.C. § 25(a)(1), CEA plaintiffs must establish that they were personally harmed by the defendant's fraudulent trading activity, see Total Gas , 889 F.3d at 109–10. Thus, to state a claim for relief under the CEA, a plaintiff must plausibly plead that the defendant "t[ook] an action that had an impact on the [plaintiff's] position," and that "that impact [was] negative." Id. at 112. While the most direct way to plead such an injury is to allege privity (i.e., that the defendant directly traded with the plaintiff while manipulating the market price), other factual allegations that give rise to an inference of harm can also be sufficient. See id. For instance, "should a plaintiff plead that she traded and lost money ... during a bout of defendant's alleged market manipulation in the same contract type in the same exchange for delivery at [the] same time and place, her pleading of injury is likely to be nearly as good as if she had pled privity." Id. (footnote omitted).
Gamma argues that it would be appropriate to infer that Defendants’ conduct harmed Gamma because: (1) the parties engaged in such a high volume of trading that it is at least plausible (if not probable) that Defendants’ spoofing influenced the price at which Gamma bought and sold futures and options contracts on at least one occasion; and (2) on nine specific dates, Defendants spoofed and Gamma took positions opposite the Defendants’ in the same markets at some point on each of those days. Neither of Gamma's theories, either alone or in combination, is sufficient to support a reasonable inference that Defendants’ spoofing conduct injured it.
1. The Parties’ High Trading Volume
Gamma contends that since Defendants spoofed thousands of times, and Gamma made thousands of trades, it is implausible as a matter of sheer probabilities that Defendants’ spoofing activities never once affected the price at which Gamma traded. This type of pleading based on rote probabilities is hardly novel, but it is generally disfavored. For example, the Supreme Court has disparaged this approach to pleading Article III injury in fact, see Summers v. Earth Island Inst. , 555 U.S. 488, 498–99, 129 S.Ct. 1142, 173 L.Ed.2d 1 (2009), and we have explained that, though the analyses are similar, pleading CEA injury is more demanding than pleading an Article III injury, see Total Gas , 889 F.3d at 111–12.
A plaintiff cannot simply plead that out of one million trades, there might be at least one in which he came out on the losing end of a spoof. Yet Gamma follows precisely that approach, alleging conclusorily that there must have been at least one trade – though it has no idea which one or when it may have occurred – in which it came out on the net losing end of Defendants’ market manipulation. In other words, Gamma's pleadings are far indeed from those that "point to a specific manipulated transaction or set of transactions ... with the plaintiff on the (net) losing end and the defendant on the (net) winning end." Total Gas , 889 F.3d at 112 (emphasis added).
The standard of review requires us to draw all reasonable inferences in Gamma's favor, see Shomo , 579 F.3d at 183, but the inference that Gamma urges us to draw is unreasonable. If allegations such as Gamma's were enough to state a claim, then any person who traded with even modest frequency could state a CEA claim against a market-manipulating defendant, without ever plausibly alleging that she suffered a loss as a result of the defendant's conduct. This is exactly the type of "citizens’ arrest[ ] for commodities fraud" that we have held is impermissible. Total Gas , 889 F.3d at 110. Put differently, even if there were a 99.9% probability that any given trade was free of fraudulent influence, after enough trades, it could be argued as a mathematical matter that the frequent trader will have bought or sold at a spoof-influenced price simply because the volume of trades is sufficiently large. Gamma's view implies that even these probabilistic allegations are enough to plead a plausible claim to relief. The contention that such meager allegations – which do not even exclude the possibility that the net effect of the defendant's spoofing was beneficial for the plaintiff – are sufficient to plead a CEA claim cannot be squared with the CEA's requirement of proving "actual damages," 7 U.S.C. § 25(a)(1), or our admonition in Total Gas that CEA plaintiffs may recover only upon "establish[ing] that they themselves have been harmed by Defendants’ activities," 889 F.3d at 110.
Gamma cites no authority to support its statistical-probability style of pleading, instead relying solely on Total Gas ’s recitation that a CEA plaintiff must plead only that "at least one" trade was negatively affected by a defendant. Gamma's Br. at 24 (quoting 889 F.3d at 112 ) (emphasis omitted). But that analysis from Total Gas indicates only that pleading damages from one fraudulent trade, as opposed to some larger number of trades, is sufficient to state a claim. It says nothing about the method of pleading damages from that trade or whether damages can be inferred based on statistical probabilities across a large set of trades.
Gamma's reliance on antitrust cases is also misplaced, since price-fixing behavior – unlike spoofing – results in harms to all market participants, without the possibility of some market participants inadvertently benefitting from the illegal conduct. Cf. New York v. Hendrickson Bros., Inc. , 840 F.2d 1065, 1075–77 (2d Cir. 1988) (explaining that the evidence amply demonstrated that the State was injured by a collusive bid-rigging scheme). Here, by contrast, it is not at all clear whether – even assuming that Gamma was affected by Defendants’ spoofing scheme – Gamma incurred a net loss or a net benefit from it. Our antitrust law is therefore inapposite in this context, since a price-fixing conspiracy aims to consistently push the market price in a single direction, whereas Gamma alleges that spoofing can artificially move the market in either direction, and yesterday's market sellers can become tomorrow's market buyers. See Total Gas , 889 F.3d at 113 (A CEA plaintiff who does not allege privity "will have to plead additional facts to make it plausible that the impact on her was harmful rather than neutral or beneficial.").
Finally, even if we were to indulge Gamma's statistical-probability approach to pleading, it would fail on its own terms. Gamma alleges that it traded thousands of times and Defendants spoofed thousands of times during the class period. But as Gamma acknowledges, "[c]ommodities and futures exchanges are busy places" and "[m]any transactions occur every minute and even every second." Gamma's Br. at 4. Given the enormous trading volume during the eight-year class period in these highly liquid markets, Gamma's trades and Defendants’ spoofs represented a tiny share of the overall activity in these markets over the class period. Thus, Gamma's allegations do not make it plausible – rather than merely speculative – that Gamma's own trades interacted with Defendants’ transactions to Gamma's detriment. See Ashcroft v. Iqbal , 556 U.S. 662, 678–79, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009). 2. Nine Specific Instances of Spoofing
Gamma alternatively argues that damages may be inferred from the existence of nine occasions when it traded on the same day that Defendants spoofed and it took a market position opposite the Defendants’ spoofing. But Gamma does not plead, even in general terms, how long it takes for the market price to return to a non-artificial level after a spoof. As the district court pointed out, if Gamma traded before Defendants spoofed, the spoofing could not have affected the price at which Gamma traded. And even if we were to assume that Gamma traded after the spoofs on one or more of these days – which is nowhere alleged – the complaint provides no factual basis that would justify an inference that the market price was still artificial by the time Gamma traded. In the absence of such factual allegations, we cannot reasonably infer that spoofing's effects last throughout the day. Even pleading same-day, post-spoof trades does not justify an inference of injury without any factual allegations to support the inference that the effects of the spoof linger for the remainder of the trading day. See Total Gas , 889 F.3d at 112 n.3 ("A plaintiff and defendant need not have been trading simultaneously so long as a plaintiff pleads facts indicating that the defendant's actions caused price artificiality during the time in which plaintiff was trading.").
Gamma argues in its reply brief that the effects of a spoof and the duration of the resulting market distortions are pure questions of fact that cannot be resolved on a motion to dismiss. The first assertion is true enough – the effects of spoofing pose questions of fact. But federal pleading standards require Gamma – as plaintiff – to allege some facts that support an inference of actual injury. See Iqbal , 556 U.S. at 678–79, 129 S.Ct. 1937. All Gamma pleads is that "Defendants’ manipulation of the markets for precious metals futures contracts caused prices to be artificial throughout the Class Period." J. App'x at 69 ¶ 46. But this is precisely the sort of "mere[ly] conclusory statement[ ]" that we need not credit, even on a motion to dismiss. Iqbal , 556 U.S. at 678, 129 S.Ct. 1937. And while Gamma's appellate briefs refer to various economic reports it claims indicate that spoofing may have continuing effects, the law is clear "that a party may not amend pleadings through a brief." Kleinman v. Elan Corp., plc , 706 F.3d 145, 153 (2d Cir. 2013). Gamma is limited to factual contentions it alleged in its operative complaint.
As the district court pointed out, Gamma's allegations, spanning an eight-year class period covering four different markets, reflect an average of one spoof every few days. Absent some factual allegations supporting Gamma's contention that spoofs representing such a small share of overall market activity during this extended period nevertheless could have had a continuous effect on the price in every market at issue for eight years , Gamma's threadbare allegations do not state a claim for damages attributable to Defendants’ spoofing activity.
Even without a factual pleading about how long the effects of spoofing last, Gamma might still be able to state a claim if it pleaded that its trades occurred so close in time to Defendants’ spoofing as to permit us to infer as a matter of common sense that the market prices were artificial when Gamma traded. See Irrera v. Humpherys , 859 F.3d 196, 198 (2d Cir. 2017) (explaining that "[j]udges ... rely on their ‘experience and common sense[ ]’ " in "draw[ing] the line between speculative allegations and those of sufficient plausibility to survive a motion to dismiss") (quoting Iqbal , 556 U.S. at 679, 129 S.Ct. 1937 ); see also Total Gas , 889 F.3d at 112 n.3. But Gamma comes nowhere close to making such allegations here, since it never actually alleges when its own trades took place. To the contrary, while Gamma pleads with specificity when some of Defendants’ spoofing occurred – based on information made available from government investigations and criminal complaints – Gamma concedes that it no longer has ready access to records indicating the time of its own trades.
Although brokers must maintain records of their clients’ trades that are time-stamped to the nearest minute, brokers are required to keep this information for only five years. See 17 C.F.R. § 1.31(b). Gamma did not seek its own trading records during that time period because, it claims, it was unaware of its potential injury until after its brokers’ duties to maintain the records had ended. But Gamma does not cite any authority suggesting that the fact that the passage of time has rendered it more difficult for Gamma to access these records entitles it to lower pleading standards. Nor has Gamma pleaded any facts to suggest that its brokers were involved in the spoofing scheme or were otherwise complicit in Defendants’ fraud, such that we might consider whether the brokers’ possible failure to maintain records would warrant an inference at this stage somewhat akin to an adverse-inference jury instruction. Cf. Residential Funding Corp. v. DeGeorge Fin. Corp. , 306 F.3d 99, 107 (2d Cir. 2002) (discussing standards for an adverse-inference jury instruction).
Gamma urges that discovery is the remedy for this problem, and that the district court wrongly inferred that the trades on these nine particular days occurred before the spoofing episodes. But before proceeding to discovery, it is the plaintiff's burden "to provide facts sufficient to allege a plausible connection between their trading and [ Defendants’ spoofing]," Total Gas , 889 F.3d at 114, and that they were harmed as a result of Defendants’ spoofing. Without pleading how long the effects of spoofing last, or that the trades happened so close in time to the spoofing episodes that we may reasonably infer price artificiality affecting Gamma's trading, Gamma has failed to state a claim for relief.
* * *
Finally, Gamma repeatedly asserts that the district court erred by not considering its two theories of damages "together." Gamma argues that its allegations are stronger when considered together because the government identified only thirty specific dates on which Defendants spoofed, and at least one plaintiff traded at least once in the same market on nine of those thirty dates. In Gamma's view, this substantial "hit rate" suggests that the parties engaged in same-day trading on far more occasions in the full universe of Defendants’ spoofing activity, increasing the likelihood that Gamma was injured on at least one occasion. Gamma's Reply Br. at 29.
It is true that we determine whether a complaint states a claim based on " ‘all of the facts alleged, taken collectively,’ not whether an inference [of liability] is permissible based on ‘any individual allegation, scrutinized in isolation.’ " Kaplan v. Lebanese Canadian Bank, SAL , 999 F.3d 842, 854 (2d Cir. 2021) (quoting Tellabs, Inc. v. Makor Issues & Rights, Ltd. , 551 U.S. 308, 323, 127 S.Ct. 2499, 168 L.Ed.2d 179 (2007) ). Gamma's argument is ultimately unpersuasive, however, because it does not overcome the deficiencies in its approach to pleading already discussed. Fundamentally, Gamma's "hit rate" theory relies on the same type of probabilistic allegations as its argument based on the sheer number of trades. But Gamma's hit rate does not support a reasonable inference that it must have traded in close proximity to Defendants’ spoofing – particularly given Gamma's failure to allege any facts to support its theory about the length of time that spoofing affects the market or the timing of any of its trades in relation to the spoofs. Considering Gamma's damages theories "together" therefore is not enough to salvage its claim.
Because Gamma's allegations as a whole fail to plausibly plead that Defendants’ conduct damaged Gamma, and because pleading damages is an element of each of the three claims Gamma presses on appeal, we affirm the district court's judgment dismissing the complaint in its entirety.
B. Leave to Amend
At oral argument, Gamma requested for the first time leave to amend its complaint in the event that we determined that it had failed to plead damages. We would typically review a district court's denial of leave to amend for abuse of discretion. See, e.g. , Feinman v. Dean Witter Reynolds, Inc. , 84 F.3d 539, 542 (2d Cir. 1996) ). But a district court cannot be said to err by "not permitting an amendment that was never requested." Horoshko v. Citibank, N.A. , 373 F.3d 248, 250 (2d Cir. 2004) ; see also City of Harper Woods Empls.’ Ret. Sys. v. Olver , 589 F.3d 1292, 1304 (D.C. Cir. 2009) ("When a plaintiff fails to seek leave from the [d]istrict [c]ourt to amend its complaint, either before or after its complaint is dismissed, it forfeits the right to seek leave to amend on appeal.").
Counsel for Gamma suggested at oral argument that its failure to seek leave to amend before the district court should be excused in light of the district court's principal holding that the complaint was untimely – a holding that would have made any request to amend with respect to damages futile. This rationale is unpersuasive. Gamma would have been equally permitted to amend its complaint to strengthen its case for timeliness – by, for instance, pleading more specifically that it had acted with reasonable diligence in pursuing its claims, such that equitable tolling is available, see Koch v. Christie's Int'l PLC , 699 F.3d 141, 157 (2d Cir. 2012) – so the district court's timeliness holding did not necessarily render a request for amendment futile.
Furthermore, the parties were aware that the district court's rulings as to timeliness and damages were independent of each other. See, e.g. , Gamma's Br. at 36. As a result, it was Gamma's burden to anticipate that we might reach only one of them. If Gamma had additional factual allegations to bolster its theory as to damages, it should have requested the opportunity to amend its complaint to strengthen its argument in the event that a future appellate panel reversed or did not reach the district court's timeliness holding. Instead, Gamma's request for leave to amend came in rebuttal at oral argument on appeal – far too late for us to consider it. See Horoshko , 373 F.3d at 250 ; City of Harper Woods Empls.’ Retirement Sys. , 589 F.3d at 1304.
III. CONCLUSION
The CEA does not deputize traders to rove the commodities markets hunting for bad behavior. Rather, it makes fraudsters "liable for actual damages." 7 U.S.C. § 25(a)(1). Here, Gamma has not plausibly alleged that it was damaged. Instead, it theorizes that its regular participation in the relevant commodities markets supports an inference that it was injured by Defendants’ spoofing at least once. But this argument is so broad that endorsing it would permit any regular market participant to proceed to discovery any time a significant market player has repeatedly committed fraud – contravening both the statute, see id. , and our caselaw, see Total Gas , 889 F.3d 104. And although Gamma also contends that Defendants occasionally spoofed in particular markets on the same day that Gamma took opposite positions in those very markets, these allegations do not support an inference of damages since Gamma does not allege facts regarding precisely when its trades took place or the duration of a spoof's effects on the market price. Accordingly, we AFFIRM the district court's judgment dismissing this action.