Summary
holding plaintiffs proved antitrust standing after they plausibly alleged conspiracy and traded bonds directly with several defendants during class period
Summary of this case from In re SSA Bonds Antitrust Litig.Opinion
19-CV-1704 (JSR)
2019-08-29
IN RE GSE BONDS ANTITRUST LITIGATION
OPINION AND ORDER
JED S. RAKOFF, U.S.D.J. This putative class action alleges a conspiracy among several large banks to fix the secondary market prices of bonds issued by government-sponsored entities ("GSEs"). Now before the Court is defendants' joint motion to dismiss the consolidated complaint for failure to state a claim upon which relief can be granted. See Joint Motion to Dismiss, ECF No. 220; Mem. Supp. Joint Mot. Dismiss ("JMD Mem."), ECF No. 221. For the reasons that follow, the motion to dismiss is granted in part and denied in part.
I. Factual Background
The parties' familiarity with the procedural history of this case is presumed. The facts alleged by the Second Amended Complaint ("2AC"), ECF No. 244, are as follows. GSEs, including the Federal National Mortgage Association ("Fannie Mae"), the Federal Home Loan Mortgage Corporation ("Freddie Mac"), the Federal Farm Credit Banks ("FFCB"), and the Federal Home Loan Banks ("FHLB"), are privately run enterprises sponsored by the federal government. 1AC ¶¶ 1, 108. GSEs issue bonds to finance their operations. Id. ¶ 111. GSE bonds are not backed by the federal government, so they are riskier than treasury bonds, but they are perceived to be relatively stable due to GSEs' close ties to the government. Id. ¶ 114. GSE bonds are unregulated, unregistered issuances and are exempt from the registration and disclosure provisions of the federal securities laws. Id. ¶ 115.
Originally, this motion was addressed to the First Amended Complaint, ECF No. 171. The Second Amended Complaint was filed solely for the purpose of adding Deutsche Bank Securities Inc. as a defendant. It is otherwise substantively identical to the First Amended Complaint, and Deutsche Bank joined the pending motion to dismiss, see ECF No. 247. The Court therefore addresses the Second Amended Complaint.
GSEs issue bonds through pre-approved bond dealers, including defendants. Approved bond dealers both underwrite GSE bonds and also trade the bonds with investors in the secondary market. Id. ¶ 123. To win the right to trade bonds, approved dealers participate in regularly-scheduled auctions in which they submit proposals for bringing a given bond issue to market. Id. ¶ 127. Typically, dealers will work together as a "syndicate" at this stage. Id. Each dealer in the syndicate with the winning bid receives an allocation of newly issued GSE bonds. Id.
The defendants named in the Second Amended Complaint are Barclays Capital Inc.; BNP Paribas Securities Corp.; Citigroup Global Markets Inc.; Credit Suisse Securities (USA) LLC; Deutsche Bank Securities Inc.; First Tennessee Bank, N.A. and FTN Financial Securities Corp.; Goldman Sachs & Co. LLC; J.P. Morgan Securities LLC; Merrill Lynch, Pierce, Fenner & Smith Inc.; UBS Securities LLC; Morgan Stanley & Co. LLC; HSBC Securities (USA) Inc.; Nomura Securities International, Inc.; TD Securities (USA) LLC; Cantor Fitzgerald & Co.; and SG Americas Securities, LLC. 2AC ¶¶ 27-106.
The bonds then enter the "syndication" phase. Id. ¶ 128. The goal in this phase is to place the bonds with a bulk buyer in the "primary market" - typically, a regional bank that will hold the bonds as an investment, rather than reselling to other investors. Id. Dealers are permitted to communicate with one another during this phase and utilize multi-bank chat rooms to do so. Id. At this stage, the syndicate members are acting together.
The syndication phase ends when the dealers declare the newly issued bonds "free to trade," or "FTT." Id. ¶ 129. At this point, secondary market trading begins, and the individual dealers are competing against one another. Id. The secondary market for GSE bonds is relatively opaque. Prices are not listed publicly; investors have to reach out to a dealer, who will send back a list of prices. Id. ¶ 132.
Defendants are approved bond dealers. They collectively traded 77.16% of all GSE bonds issued during the proposed class period, January 1, 2009 through January 1, 2016. Id. ¶ 134. Plaintiffs are investment and retirement funds that transacted with defendants during the class period. They bring this suit on behalf of a putative class of all persons and entities that transacted in unsecured GSE Bonds with defendants during the class period. Id. ¶ 263.
Specifically, the plaintiffs are Joseph Torsella, in his official capacity as Treasurer of Pennsylvania; the City of Birmingham Retirement and Relief System; Electrical Workers Pension Fund Local 103, I.B.E.W.; and Local 103, I.B.E.W. Health Benefit Plan.
A. Proof of Conspiracy
1. Direct Evidence
The thrust of plaintiffs' theory is that defendants agreed to keep prices high for newly issued bonds when they were released to the secondary market. Plaintiffs claim to have received information from a "cooperating co-conspirator," who "implicated" the defendants in this scheme. Id. ¶ 146. As direct evidence of the conspiracy, plaintiffs offer what are alleged to be transcripts of chatroom conversations between dealers acting on behalf of several defendants. The Court will briefly recap each conversation as transcribed in the complaint.
On August 31, 2011, BNP, Deutsche Bank, and Cantor Fitzgerald jointly won a bid for FFCB bonds. Id. at ¶ 152. The next day, traders from Deutsche Bank and BNP agreed to "go FTT," so they could sell the bonds at the agreed-upon price of 99.925 rather than the syndication price of 99.90. Id.
On February 13, 2012, Morgan Stanley, Deutsche Bank, and BNP won a joint bid for newly-issued FHLB bonds. Id. ¶ 147. In a chat on February 17, a Morgan Stanley trader expressed concern over "creat[ing] a race to the bottom between the 3 of us." Id. ¶ 148. A BNP trader indicated that they "d[id]n't want to see spreads gap out." Id. The Deutsche Bank trader suggested that they "go out FTT 99.985?" Id. The Morgan Stanley trader responded "Sure FTT at 99.985." Id.
On July 17, 2012, traders from BNP, Deutsche Bank, Goldman Sachs, and Merrill Lynch discussed setting the price of newly-acquired FFCB bonds. Id. ¶ 150. The Deutsche Bank representative suggested "99.985 on the FFCB's?" and the BNP representative answered "Sounds good here." Id. The Merrill Lynch trader indicated that they "might be a touch higher than u guys but def not going lower." Id. The Merrill Lynch trader added "We were 99.875 yesterday, right," which the Deutsche Bank trader confirmed. Id. The Deutsche Bank trader then solicited input from the other two banks, asking "You guys all cool w/ that?" Id. The BNP representative responded "Our live mark right now is 99.66." Id. The Deutsche Bank trader remarked that although "anyone can hit any bid," the trader "figure[d] maybe we at least try and stay on the same page ... less volatile." Id. BNP then asked if their price was "too cheap." Id. Finally, the Goldman Sachs trader jumped in to remind everyone that "if we are free to trade, we cannot talk about prices." Id. The Deutsche Bank trader responded "Fair enough. Just don't want anyone getting annoyed if someone hits a bi[d]." Id.
On September 18, 2013, Goldman Sachs and Deutsche Bank submitted a winning joint bid for FHLB bonds. Id. ¶ 149. The same day, the Deutsche Bank trader suggested going "FTT 99.90," to which the Goldman Sachs trader responded "Let[']s just go 99. 925." Id.
2. Indirect and Statistical Evidence
In addition to this direct evidence, plaintiffs allege that the nature of the GSE bonds market facilitated unlawful coordination. Specifically, because the same traders are involved in all phases of acquiring and selling GSE bonds, the same people will at one point be working together (to bid on bonds and then to place them with bulk buyers) and then working as competitors (to sell the bonds on the secondary market). Id. ¶ 160-62. Moreover, the opaque nature of the market from the investors' point of view hides the heightened pricing from buyers. Id. ¶ 164.
Plaintiffs also offer economic analysis that they contend corroborates the claim of conspiracy. For example, from March 1, 2010 through December 31, 2015, during the class period, approved GSE bond dealers sold newly issued Benchmark and Reference Notes (a particularly common type of GSE bond) for, on average, 10.6 basis points more than the dealers had paid. Id. ¶¶ 184-87. In other words, dealers sold these Notes for an average of 0.106% more than they paid during this time. From January 1, 2016 through December 31, 2017, after the class period ended, dealers sold for only an average of 1.2 basis points more than they paid. Thus, the buy-sell differential was nearly nine times higher, on average, during the class period than after. Id. ¶ 188. When the analysis is expanded to all GSE bonds traded during this time (instead of just Notes), the result is similar, although less extreme. Defendants charged an average premium of 2.58 basis points to investors during the class period, but an average of only 1.15 basis points afterwards. Id. ¶ 193.
A "basis point" is 1/100 of one percent. 2AC ¶ 193 n.8.
Plaintiffs also compare the prices charged by defendants for GSE bonds to the prices charged for Treasury securities with comparable maturity periods. Since Treasury securities carry a similar amount of risk as GSE bonds, plaintiffs allege that, in a normal market, they should be priced similarly. Id. ¶ 194. According to plaintiffs, the defendants imposed a substantial markup on GSE bonds over Treasury securities, compared to other, non-defendant approved bond dealers. Defendants charged a premium of 100 basis points for GSE bonds over Treasury securities during the class period, and roughly 40 basis points after the class period. Other, non-defendant dealers charged a premium of 40 basis points during the class period and only about 10 basis points afterwards. Id. ¶ 197. Similarly, the bid-ask spread - that is, the difference between what defendants offered to pay for a bond versus what they offered to sell it for, id. ¶¶ 140-41 - for defendants' price quotes was much higher during the class period than after. Id. ¶ 220.
Plaintiffs mostly focus on evidence that defendants manipulated the price of newly-issued bonds, which are referred to as "on-the-run." Plaintiffs also claim, however, that defendants manipulated the price of bonds that were about to go "off-the-run" - i.e. they would hike up the price of old bonds right before a new issuance, because the most recent off-the-run bonds are used as a benchmark to price new issuances. Id. ¶ 200. Plaintiffs claim there is a "statistically significant prince increase" in bonds about to go off-the-run "in the two days immediately leading up to a new issuance." Id. ¶ 204. Plaintiffs further claim that this cannot be explained by normal economic factors, because typically bonds about to go off-the-run become less desirable as a new issuance approaches. Id. Plaintiffs supplement this claim with two further allegations. First, in sales of bonds about to go off-the-run, dealers charged customers significantly higher prices than they charged other dealers. Id. ¶ 208. Second, in looking at the three-day period surrounding a new auction (and so new release of bonds), the prices dealers charged to other dealers for the now off-the-run bonds dropped noticeably (as you would expect), but the prices charged to customers dropped far less. Id. ¶ 209.
B. Injury and Claim for Relief
Each of the named plaintiffs claims to have been overcharged in their transactions with various defendants. Id. ¶¶ 238-262. They seek damages under the Sherman Act. Id. ¶ 279-80. They also seek certification of a class of all persons who traded in GSE bonds with the defendants during the class period. Id. ¶ 263.
Plaintiffs claim the statute of limitations was tolled because the defendants fraudulently concealed their price-fixing until, at the earliest, June 2018, when it was publicly reported that the Department of Justice was investigating price-fixing in the GSE bonds market. Id. ¶ 270.
II. Defendants' Joint Motion to Dismiss
Defendants jointly moved to dismiss the consolidated complaint. They argue that (1) the chat logs are not "direct evidence" of a conspiracy; (2) plaintiffs have not pleaded parallel conduct or "plus factors" sufficient to sustain the complaint without direct evidence; (3) the complaint fails to make specific allegations against certain defendants; (4) the statistical analysis is flawed and unreliable; (5) the conduct alleged is not per se unlawful; (6) plaintiffs failed to plead antitrust injury; and (7) most of the claims are time-barred.
As explained more fully below, the Court finds that the Second Amended Complaint adequately alleges the existence of a conspiracy to fix the price of GSE Bonds, at least among those defendants who appear in the chat room transcripts - i.e. Deutsche Bank Securities Inc., BNP Paribas Securities Corp., Morgan Stanley & Co., Goldman Sachs & Co., and Merrill Lynch, Pierce, Fenner & Smith Inc. (collectively, "the Chatroom Defendants"). The Court further concludes that the Second Amended Complaint plausibly pleads that the conspiracy extended beyond just these defendants. The Court agrees with defendants, however, that the remaining allegations do not adequately link any of the particular named defendants beyond the Chatroom Defendants to the conspiracy. Accordingly, the motion to dismiss is denied as to the Chatroom Defendants. It is granted as to the remaining defendants - i.e. Barclays Capital Inc.; Citigroup Global Markets Inc.; Credit Suisse Securities (USA) LL; First Tennessee Bank, N.A. and FTN Financial Securities Corp.; UBS Securities LLC; Jefferies Group LLC; HSBC Securities (USA) Inc.; Nomura Securities International, Inc.; TD Securities (USA) LLC; Cantor Fitzgerald & Co.; and SG Americas Securities, LLC - but with leave to replead.
A. Sufficiency of the Pleading
1. Legal Standard
"A plaintiff's job at the pleading stage, in order to overcome a motion to dismiss, is to allege enough facts to support the inference that a conspiracy actually existed." Mayor and City Council of Baltimore v. Citigroup, Inc., 709 F.3d 129, 136 (2d Cir. 2013). In the context of a complaint alleging antitrust conspiracy, "there are two ways to do this. First, a plaintiff may, of course, assert direct evidence that the defendants entered into an agreement in violation of the antitrust laws. Such evidence would consist, for example, of a recorded phone call in which two competitors agreed to fix prices at a certain level." Id. (citation omitted). Second, because "this type of ‘smoking gun’ can be hard to come by, especially at the pleading stage," a complaint may instead "present circumstantial facts supporting the inference that a conspiracy existed." Id. To do so, the plaintiff must allege (1) parallel conduct and (2) "plus factors." Id.
In reviewing the sufficiency of a complaint, the Court "accept[s] all factual allegations as true and draw[s] every reasonable inference from those facts in the plaintiff's favor." Id. at 135. "To survive dismissal, the plaintiff need not show that its allegations suggesting an agreement are more likely than not true or that they rule out the possibility of independent action, as would be required at later litigation stages such as a defense motion for summary judgment, or a trial." Gelboim v. Bank of America Corp., 823 F.3d 759, 781 (2d Cir. 2016) (internal quotation marks omitted) (quoting Anderson News, LLC v. Am. Media, Inc., 680 F.3d 162, 184 (2d Cir. 2012) ). So long as the allegations are plausible, plaintiffs burden has been met; the Court may not "cho[ose] between two plausible inferences that may be drawn from factual allegations" at this stage. Id. (quoting Anderson News, 680 F.3d at 184 ).
2. The Chatroom Defendants
Here, we have the rare smoking gun, at least as to the Chatroom Defendants. The chats unmistakably show traders, acting on behalf of those defendants, agreeing to fix prices at a specific level before bringing the bonds to the secondary market. See In re London Silver Fixing, Ltd., Antitrust Litig. ("Silver Fix II"), 332 F. Supp. 3d 885, 901 (S.D.N.Y. 2018) ("The chat messages included in the TAC are direct evidence of an anticompetitive agreement to manipulate the silver market.").
Defendants offer a slew of arguments against relying on the chat logs, but none of them is persuasive. First, defendants emphasize that there are only four chats, representing conversations about four bonds during a period when tens of thousands were traded. JDM Mem. 13. But plaintiffs are not expected to marshal evidence (especially at this early stage) of every single time defendants unlawfully conspired. Moreover, the tone of the conversations suggests that these were not isolated instances. At no point do any of the involved traders say, for example, that they should not be fixing prices. The only time anyone voices concern is the Goldman Sachs trader during the July 17, 2012 chat, but that objection is solely about the timing of the conversation, not about its happening at all.
Contrary to defendants' argument, there is no rule that isolated occurrences of conspiratorial conduct do not qualify as direct evidence. Direct evidence can even be "a recorded phone call in which two competitors agreed to fix prices at a certain level." Mayor and City Council of Baltimore, 709 F.3d at 136. Here, plaintiffs have produced, not one chat log, but four. That is more than adequate to allege the existence of a conspiracy, at least against those defendants implicated by the chats. Such communications have repeatedly been held sufficient for pleading purposes. E.g., Silver Fix II, 332 F. Supp. 3d at 901-03 ; In re Foreign Exchange Benchmark Rates Antitrust Litig. ("FX I"), 74 F. Supp. 3d 581, 591-92 (S.D.N.Y. 2015).
Second, defendants argue that these communications were lawful for co-underwriters acting in a syndicate. JMD Mem. 13-14. Plaintiffs appear to agree that traders are permitted to communicate during the syndication phase for certain purposes. PI. Mem. Opp. Joint Mot. Dismiss ("JMD Opp.") 18, ECF No. 35. But, plaintiffs argue, while defendants are allowed to communicate about primary market sales during this time - sales that the defendants make jointly - they are not permitted to communicate during this period (or otherwise) about secondary market sales when they are competitors. The Court agrees.
Specifically, the Second Circuit has held that the "corruption" of a "cooperative endeavor" can be actionable under the antitrust laws. In Gelboim, the district court dismissed a complaint alleging a LIBOR-setting conspiracy, in part because it found that the defendants were engaged in a "cooperative endeavor." 823 F.3d at 775. The Second Circuit reversed, because "the Banks circumvented the LIBOR-setting rules, and that joint process thus turned into collusion." Id. This case is closely analogous.
Defendants have no serious answer for this objection. They simply insist that any kind of communication during syndication is permitted, as though that excuses any conversations about price-fixing at any stage. The Court cannot agree. If it is illegal to fix the secondary prices for bonds once they are on the market - and defendants do not dispute that it would be - it cannot be legal to fix such prices through conversations that occur right before the bonds go on the market.
Moreover, at least one of the conversations happened during the FTT phase. And while defendants are eager to emphasize that the Goldman trader shut down that conversation, it went on for some time, with active participation by representatives of defendants Deutsche Bank, Merrill Lynch, and BNP. That suggests that defendants were willing to talk prices, even during the FTT phase. Defendants also argue that "the dealers did not agree on a price." JMD Mem. 14. But this is wrong both as to the facts and as to the law. Legally, an attempt to conspire, while not itself independently actionable, still supports the allegation of a broader conspiracy on other occasions. And factually, defendants mischaracterize the conversation; in actuality, one defendant, BNP, immediately agrees to the price suggested by the Deutsche Bank trader, while Merrill Lynch says its prices might go higher but that it would not undercut the others by going lower. This, on its face, is blatant price-fixing.
Next, defendants argue that the bonds discussed in many of the cited chats actually regularly sold for less than the agreed-upon price. JMD Mem. 15. Although the Court may take judicial notice of the TRACE data on this motion to dismiss, see Ganino v. Citizens Utilities Co., 228 F.3d 154, 166 n.8 (2d Cir. 2000), the pricing trends and their relation to the dates of these chats is a factual question ill-suited for resolution at this early stage. Moreover, even assuming defendants are correct, an antitrust claim may proceed "based on the influence that a conspiracy exerts on the starting point for prices." Gelboim, 823 F.3d at 776 ; see also In re High Fructose Corn Syrup Antitrust Litig., 295 F.3d 651, 656 (7th Cir. 2002) ("An agreement to fix prices is ... a per se violation of the Sherman Act even if most or for that matter all transactions occur at lower prices.").
Defendants then contend that a "broad, market-wide conspiracy is implausible" because there are some 65 authorized bond dealers. JMD Mem. 17. But defendants concede they collectively traded some 77% of the market. That they were a numerical minority of dealers - a fact they hammer repeatedly in their papers - is totally irrelevant; they had control over most of the market. Nor is the nature of the conspiracy implausible; contrary to defendants' claims, this particular conspiracy does not seem to have required much coordination. All that was necessary was for traders to use already-established channels of communication to discuss prices before selling. While defendants question why investors would not simply trade with other dealers offering lower prices, JMD Mem. 17, that is plausibly explained by the relatively opaque nature of the market. And, as plaintiffs point out, JMD Opp. 24, these are all factual objections that are not really proper at this stage. Mere "[s]kepticism of a conspiracy's existence is insufficient to warrant dismissal." Gelboim, 823 F.3d at 781.
Additionally, the plausibility of the alleged conspiracy is bolstered, at least to some extent, by the ongoing Department of Justice investigation into the same alleged misconduct. See Starr v. Sony BMG Music Enter., 592 F.3d 314, 325 (2d Cir. 2010) (finding inference of conspiracy plausible in part because DOJ launched investigations into possible price fixing by defendants); In re Propranolol Antitrust Litig., 249 F. Supp. 3d 712, 723 (S.D.N.Y. 2017) ("The presence of an ongoing investigation into the same subject matter as alleged in the pleadings here raises an inference of conspiracy.").
Defendants also point out that all of the chats involved callable bonds. From this, they argue that there is no evidence of a conspiracy involving other types of bonds. JMD Mem. 18. This is not persuasive. It is more than plausible that these kinds of conversations also happened about other types of bonds.
"Callable" bonds can be redeemed prior to maturity. 2AC ¶ 122.
At bottom, most of defendants' arguments simply "urge the Court to pick-and-choose between plausible inferences." Silver Fix II, 332 F. Supp. 3d at 903. That is not permissible at this stage of the litigation. Plaintiffs are not required to prove that their theory is the most plausible, but only that it is plausible. See Gelboim, 823 F.3d at 781. The Court finds that plaintiffs have met that burden, at least as to the Chatroom Defendants. The chat logs are exactly what they appear to be: direct evidence of a conspiracy to fix prices.
3. The Remaining Defendants
The Court reaches a different result, however, as to the remaining defendants. The chatroom transcripts do not so much as mention any of those defendants. Cf. Sonterra Capital Master Fund Ltd. v. Credit Suisse Group AG, 277 F. Supp. 3d 521, 556-57 (S.D.N.Y. 2017) (finding complaint adequately alleged conspiracy against defendant Royal Bank of Scotland, which appeared in chat messages, but not against other defendants who did not appear in those messages). And while the Court finds it entirely plausible that the conspiracy evidenced by the chatroom logs may have extended beyond the specific defendants participating in those conversations, plaintiffs must still adduce some reason to believe that the particular defendants named in this suit were involved. "[E]ach defendant is entitled to know how he is alleged to have conspired, with whom and for what purpose." In re Zinc Antitrust Litig., 155 F. Supp. 3d 337, 384 (S.D.N.Y. 2016). "Mere generalizations as to any particular defendant - or even defendants as a group - are insufficient." Id.
In the absence of direct evidence against the remaining defendants, plaintiffs must "present circumstantial facts supporting the inference that a conspiracy existed." Mayor and City Council of Baltimore, 709 F.3d at 136. To do so, the plaintiff must allege (1) parallel conduct and (2) "plus factors." Id. But defendants correctly point out that plaintiffs have failed to plead either parallel conduct or plus factors. JMD Mem. 19-20. The complaint does not, for example, include any evidence that defendants all priced their bonds similarly.
The Second Amended Complaint alleges that a "cooperating co-conspirator" has provided examples of price-fixing conversations, and that all of the various defendants were "directly implicated in conspiratorial multi-bank chats." 2AC ¶ 4. But this unadorned allegation, without any specifics, cannot salvage the pleading against these defendants. To be clear, the Court does not hold that, to state a claim for relief, a plaintiff must adduce direct evidence (such as a chatroom transcript) for each and every defendant named. There are other ways to plausibly allege participation in a conspiracy. But there must be something in the complaint that ties each defendant to the conspiracy.
Without any allegations relating to specific actions taken by the remaining defendants, plaintiffs are left to rely on their statistical allegations. Defendants contend that plaintiffs' statistics are "flawed and unreliable." JMD Mem. 29. The Court disagrees; although there are some problems with the statistics, they generally support the allegation of a price-fixing conspiracy. At this stage, a statistical analysis, like any other allegation, need only be plausible. Merely pointing out that there are problems with the analysis, or that a better method is available, will not suffice. See Wacker v. JP Morgan Chase & Co., 678 F. App'x 27, 30 (2d Cir. 2017) (summary order) ("[O]ur precedents caution against assessing the choice of a benchmark at the pleading stage because it involves an inherently fact-intensive inquiry into the relationship between the benchmark and the market it allegedly tracks.").
First, defendants complain that plaintiffs' statistical analysis in the predecessor complaints differed, allegedly based on the same data. JMD Mem. 30. Inconsistency, however, is not a reason to discount the current statistics, which are the only ones that matter at this stage. Shields v. Citytrust Bancorp., Inc., 25 F.3d 1124, 1128 (2d Cir. 1994) ("[A]n amended complaint ordinarily supersedes the original, and renders it of no legal effect.") (quoting Int'l Controls Corp. v. Vesco, 556 F.2d 665, 668 (2d Cir. 1977) ). That is especially true when, as here, there is a reasonable explanation for a change in the complaint - to wit, that the cases were consolidated, and a cooperating co-conspirator allegedly helped plaintiffs to refine their case. In any event, because the statistics need only be plausible, different interpretations are not necessarily fatal at this stage. There may be more than one plausible interpretation of a single set of data. Gelboim, 823 F.3d at 781.
Next, defendants complain that plaintiffs use averages for multi-year periods. JMD Mem. 31. The Court agrees that this is an issue. An average can flatten or hide trends that might tell a different story, and one can manipulate an average by picking the cutoff point between two periods. However, averages still have some value as a measure of analysis. The price differences shown in the complaint appear to be relatively large; thus, it is plausible that, even if a more granular method were used, there would still be measurable differences.
Defendants also argue that the Court should not credit statistical analysis based on undisclosed data. JMD Mem. 35. The Court is aware of no law requiring plaintiffs to produce the data underlying their statistical analysis at the time they file a complaint. Moreover, much of the data is publicly available. Next, defendants argue that plaintiffs' before-and-after price comparison is flawed because it does not account for underwriter fees. JMD Mem. 35. But, as plaintiffs point out, underwriter fee are charged in the transactions both before and after the class period, so that fee cannot possibly account for the difference observed between the two periods. JMD Opp. 33-34.
Finally, defendants argue that plaintiffs failed to control for other factors, such as macroeconomic conditions. JMD Mem. 38. But this type of fact-bound argument, which necessarily relies on complicated decisions regarding statistical modeling and analysis, is ill-suited for resolution on the pleadings. See Gelboim, 823 F.3d at 781.
The Court therefore concludes that, while defendants' objections to the reliability of plaintiffs' statistics have some weight, the statistics are not so unreliable as to be useless at this very early stage of the litigation in supporting the allegation of a price-fixing conspiracy.
The deeper problem is that the statistics do not plausibly suggest that the particular defendants named in this suit were part of that conspiracy. Even assuming that the price-fixing conspiracy extended beyond the banks appearing in the chatroom logs, there is no particular reason to believe that the other defendants named in this suit were involved apart from plaintiffs' say-so. The conspiracy could well have involved some of them, or none of them, or a mix of the named defendants and other GSE bond dealers. The Court has no meaningful way of distinguishing, and plaintiffs' statistics do not help.
Several of plaintiffs' figures do not distinguish at all between defendant and non-defendant dealers. JMD Mem. 33. And even for those that do, because the statistics are quite granular and require comparing a great many dealers against one another, it is impossible to have any confidence that the statistics actually capture something different about each and every defendant. For example, plaintiffs allege that difference in price between GSE bonds and Treasury instruments was noticeably higher for defendant dealers than non-defendant dealers, and that the gap significantly closed after the class period ended. 2AC ¶ 197. That is sufficient to plausibly imply that defendants, as a group, were charging higher prices than non-defendants. But the Court cannot infer, from this, that each individual defendant was charging higher prices. What would the comparison look like if, say, Cantor Fitzgerald were moved into the non-defendant column? Or if one of the non-defendants were made a defendant? There are a dozen traders named as defendants in this case who are not implicated by the chat logs, and some of them have a comparatively small market share. See 2AC ¶ 134. The effects of any given defendant's trading activity and pricing choices might well be swallowed in plaintiffs' aggregated statistics.
Accordingly, the Court finds, based on the direct evidence of price-fixing activity, as supplemented by the statistical evidence, that plaintiffs have adequately pleaded an antitrust conspiracy against Deutsche Bank Securities Inc., BNP Paribas Securities Corp., Morgan Stanley & Co., Goldman Sachs & Co., and Merrill Lynch, Pierce, Fenner & Smith Inc. Plaintiffs have not pleaded a claim upon which relief can be granted against the remaining defendants.
At oral argument, however, plaintiffs represented that, at the time of filing the Second Amended Complaint, they were in possession of only the four chats quoted in the complaint, but that they have since received more, implicating additional defendants. See Tr. July 23, 2019 at 15:24-16:11, ECF No. 251. Defendants have never argued that leave to amend would be futile. Accordingly, the Court grants plaintiffs leave to file an amended complaint, provided that the new complaint is filed no later than September 10, 2019. Any of the non-Chatroom Defendants who then wishes to renew its motion to dismiss may so move - in a single joint filing by all defendants concerned, not to exceed 25 pages - by September 17. Plaintiffs may respond by September 23. No reply papers will be permitted, but the Court will hear oral argument on September 27 at 2:00pm.
B. Defendants' Remaining Arguments
Defendants raise several additional arguments. The Court finds none of them persuasive.
1. Failure to Plead a Rule-of-Reason Claim
First, defendants argue that price-fixing in this context is not per se unlawful and that plaintiffs have not pleaded a rule of reason claim. JMD Mem. 41. Defendants argue that the rule of reason "applies to alleged agreements among co-underwriters working together in a syndicate about the price of the bonds they underwrote." Def. Mem. 42. But that only applies to the pricing while the dealers are working as a syndicate - in this case, while selling in the primary market. See generally Credit Suisse Securities (USA) LLC v. Billing, 551 U.S. 264, 127 S.Ct. 2383, 168 L.Ed.2d 145 (2007). It has no application to sales in the secondary market. See JMD Opp. 42 (explaining that, in Credit Suisse, "the joint underwriting banks were always in ‘syndicate,’ and thus never competitors with respect to the financial product").
Similarly, defendants claim that plaintiffs "challenge syndicate activity aimed at bringing a new product - a GSE bond - to the market." Def. Mem. 44. Not so. Plaintiffs do not challenge the syndicate activity; they challenge the exploitation of the existence of the syndicate to permit anti-competitive non-syndicate activity, i.e. the individual sales of bonds at a coordinated price. In any event, the product is not "new" by the time it is sold on the secondary market.
Defendants argue that, because they were free to sell the bonds at par (i.e. for their face value), it could not be illegal to agree to sell the bonds at a discount to par. JMD Mem. 45. The Court is skeptical that it would, in fact, be legal for the defendants to agree to sell the bonds at par. Mutually agreeing to keep the price of the bonds as high as possible sounds like a classic case of price-fixing. But even assuming, arguendo, that it would be legal for defendants to mutually agree to sell bonds at an unattractively-high price, it does not necessarily follow that it is also legal for them to set a more attractive (but still higher than competitive) price.
Plaintiffs have adequately alleged, for pleading purposes, that defendants conspired to fix prices in a marketplace in which they were competitors. "Horizontal price-fixing conspiracies among competitors are unlawful per se, that is, without further inquiry." Gelboim, 823 F.3d at 771. Accordingly, plaintiffs' failure to plead a rule-of-reason claim is irrelevant.
2. Anti-Trust Standing
Defendants next claim that plaintiffs have failed to plead injury-in-fact because they failed to establish that any of the transactions they entered into were affected by the alleged conspiracy. JMD Mem. 48. But this argument depends almost entirely on the preceding argument that plaintiffs failed to adequately plead a conspiracy at all. Since that argument is meritless, this one is too. See In re Aluminum Warehousing Antitrust Litig., 833 F.3d 151, 158 (2d Cir. 2016) ("[C]onsumers in the market where trade is allegedly restrained are presumptively the proper plaintiffs to allege antitrust injury.") (quoting Serpa Corp. v. McWane, Inc., 199 F.3d 6, 10 (1st Cir. 1999) ). There is no dispute that the complaint alleges that plaintiffs participated in GSE bond transactions during the class period with at least several of the defendants, including all of the Chatroom Defendants. See 2AC ¶¶ 17-26.
3. Statute of Limitations
Finally, defendants argue that the majority of the class period is time-barred and that equitable tolling does not apply. JMD Mem. 52. The statute of limitations is four years. 15 U.S.C. § 15b. "An antitrust action accrues and the statute of limitations begins to run when the defendant commits an act that injures the plaintiff." Merced Irrigation Dist. v. Barclays Bank PLC, 165 F. Supp. 3d 122, 134 (S.D.N.Y. 2016) (quoting In re Nine West Shoes Antitrust Litig., 80 F. Supp. 2d 181, 191 (S.D.N.Y. 2000) ). "In an alleged price-fixing conspiracy, each overt act that is part of the violation and that injures the plaintiff starts the statutory period running again, regardless of the plaintiff's knowledge of the alleged illegality at much earlier times." Id. at 134-35 (internal quotation marks and alteration omitted) (quoting In re Nine West Shoes, 80 F. Supp. 2d at 192 ).
Here, the class period runs from January 1, 2009 through January 1, 2016. Plaintiffs allege that the conspiracy continued through that period. E.g., 2AC ¶ 4. The first individual complaint in this case was filed on February 22, 2019, and the First Amended Complaint was filed on May 23, 2019. Thus, if even a single overt act in furtherance of the conspiracy injured the plaintiffs in the second half of 2015, this action is timely. Given the nature of the alleged conspiracy, the Court finds this entirely plausible.
Even if the limitations period were not revived by continuing violations, it would be extended by the doctrine of fraudulent concealment. "[A]n antitrust plaintiff may prove fraudulent concealment sufficient to toll the running of the statute of limitations if he establishes (1) that the defendant concealed from him the existence of his cause of action, (2) that he remained in ignorance of that cause of action until some point within four years of the commencement of his action, and (3) that his continuing ignorance was not attributable to lack of diligence on his part." New York v. Hendrickson Bros., Inc., 840 F.2d 1065, 1083 (2d Cir. 1988). "[T]he plaintiff may prove the concealment element by showing either that the defendant took affirmative steps to prevent the plaintiff's discovery of his claim or injury or that the wrong itself was of such a nature as to be self-concealing." Id.
Here, the wrong itself was self-concealing. Defendants are accused of using private chat rooms to coordinate pricing. "Allegations of price-fixing conspiracies in violation of antitrust law constitute the type of unlawful activity that is inherently self-concealing." Merced Irrigation Dist., 165 F. Supp. 3d at 135 ; see also Sonterra Capital Master Fund, Ltd. v. Barclays Bank PLC, 366 F. Supp. 3d 516, 526 (S.D.N.Y. 2018) ; cf. Hendrickson Bros., 840 F.2d at 1084 ("[A] bid-rigging conspiracy is the kind of enterprise that ... must remain concealed from the victim of the collusive bids."). Defendants' only counterargument is that plaintiffs themselves rely on public data for their statistical analysis. JMD Mem. 54. But requiring potential plaintiffs to conduct exhaustive statistical analysis of millions of transactions, just on the off chance that it would reveal some suspicious behavior, would be absurd. And this ignores the fact that many patterns in plaintiffs' data would not have become clear until after 2016, when the conspiracy supposedly ended and the pricing suddenly changed dramatically.
Defendants argue that plaintiffs "cannot have it both ways. They cannot rely on widely available data as evidence of a plausible conspiracy, while simultaneously contending that the same data did not at least put them in inquiry notice years ago." JMD Mem. 54. But exactly the same argument could be leveled at defendants: they cannot simultaneously claim the pricing data is innocent and reveals nothing, yet should have galvanized plaintiffs to action immediately. See Sonterra Capital Master Fund Ltd. v. Credit Suisse Group AG, 277 F. Supp. 3d 521, 568 (S.D.N.Y. 2017).
Plaintiffs allege that they became aware of the potential conspiracy in June 2018, when it was first reported that DOJ was investigating possible price-fixing in the bonds industry. That is well within the four-year statute of limitations. At a minimum, defendants' arguments do not conclusively establish that the statute of limitations applies, and so dismissal on that ground would be premature. Merced, 165 F. Supp. 3d at 136-37.
III. Conclusion
For the foregoing reasons, the Court finds, based on the direct evidence of price-fixing activity, as supplemented by the statistical evidence, that plaintiffs have adequately pleaded an antitrust conspiracy against Deutsche Bank Securities Inc., BNP Paribas Securities Corp., Morgan Stanley & Co., Goldman Sachs & Co., and Merrill Lynch, Pierce, Fenner & Smith Inc. The motion to dismiss is therefore denied as to those defendants.
The Court further finds that plaintiffs have not adequately pleaded a claim against the remaining defendants, i.e. Barclays Capital Inc.; Citigroup Global Markets Inc.; Credit Suisse Securities (USA) LL; First Tennessee Bank, N.A. and FTN Financial Securities Corp.; UBS Securities LLC; Jefferies Group LLC; HSBC Securities (USA) Inc.; Nomura Securities International, Inc.; TD Securities (USA) LLC; Cantor Fitzgerald & Co.; and SG Americas Securities, LLC. The motion to dismiss is therefore granted as to those defendants. However, because the Court does not find that amendment would be futile, leave to amend is granted as to the dismissed defendants. Any such amended complaint must be filed by September 10, 2019. Any non-Chatroom Defendant named in the amended complaint may then renew its motion to dismiss - in a single joint filing by all defendants concerned, not to exceed 25 pages - by September 17. Plaintiffs may respond by September 23. No reply papers will be permitted, but the Court will hear oral argument on September 27 at 2:00pm. The Court will consider all arguments previously raised to be incorporated by reference in the new motions; the renewed motions should focus exclusively on the new allegations.
SO ORDERED.