Loreley Fin. (Jersey) No. 3 Ltd. v. Wells Fargo Sec. LLC, 797 F.3d 160 (2d Cir. 2015):

This case, like so many others of late, concerns liability for investment losses. Specifically, it asks who, if anyone, ought to shoulder legal blame for losses suffered as part of the recent financial crisis. Plaintiffs-Appellants--whose names are all numbered variants of "Loreley Financing" (collectively, "Plaintiffs")--are special-purpose investment entities operated by the German bank IKB Deutsche Industriebank AG and domiciled in the Bailiwick of Jersey, Channel Islands. In late 2006 and early 2007, Plaintiffs invested millions of dollars in the notes of three financial products known as collateralized debt obligations ("CDOs"). Two of the CDOs were named for constellations: Octans CDO II ("Octans") and Sagittarius CDO I ("Sagittarius") (together, the "constellation CDOs"). The third [**3]  was Longshore CDO Funding 2007-3 ("Longshore"). Each CDO was created and sold by three Wachovia subsidiaries (collectively, "Wachovia"). Between late 2007 and mid-2008, all three CDOs went into default, failing to make payments owed to Plaintiffs.

In April 2012, in the wake of these losses and the larger financial crisis, Plaintiffs filed suit in New York state court against several parties responsible for structuring, offering, and managing the CDOs (collectively, "Defendants"). Plaintiffs allege, among other things, fraud in connection with disclosures about the construction of the three CDOs. According to the complaint, Defendants represented to "long" investors like Plaintiffs that the constellation CDOs would be handled by judicious collateral managers, even though Defendants knew that, in reality, these CDOs had been built at the direction of a powerful "short" investor who stood to profit massively if the CDOs failed. As to Longshore, the non-constellation CDO, Plaintiffs allege that despite similar representations  [*164]  it was used to dump toxic assets from Wachovia's own balance sheets at above-market prices.

After this case was removed to federal court, the United States District [**4]  Court for the Southern District of New York (Sullivan, J.) dismissed the complaint under Rule 12(b)(6), denying Plaintiffs' request to replead. See Loreley Fin. (Jersey) No. 3 Ltd. v. Wells Fargo Secs., LLC, 12 Civ. 3723 (RJS), 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668 (S.D.N.Y. Mar. 28, 2013). Because we conclude that the district court erred in aspects of its dismissal of Plaintiffs' fraud claim and also exceeded the bounds of its discretion in denying Plaintiffs leave to amend the complaint as to the remaining claims, we reverse in part, vacate in part, and remand the case for further proceedings consistent with this opinion.

BACKGROUND

Plaintiffs' fraud allegations are only intelligible if one has some understanding of the basic structure and function of CDOs. We offer a brief description before turning to the particulars of this case.

A. The Structure of a CDO¹

1   Our description draws from and supplements the helpful accounts already given by other courts, including panels of our Court as well as district courts in this Circuit. See, e.g., Fin. Guar. Ins. Co. v. Putnam Advisory Co., LLC, 783 F.3d 395, 398 (2d Cir. 2015); Gearren v. McGraw-Hill Cos., Inc., 690 F. Supp. 2d 254, 258 n.2 (S.D.N.Y. 2010); In re Am. Int'l Grp., Inc. 2008 Sec. Litig., 741 F. Supp. 2d 511, 520 (S.D.N.Y. 2010). For a more comprehensive introduction to CDOs, see generally Douglas J. Lucas, Laurie S. Goodman & Frank J. Fabozzi, Collateralized Debt Obligations: Structures and Analysis (2d ed. 2006).

The [**5]  construction of a CDO begins, at least conceptually, with asset-backed loans, such as mortgages or car loans. These loans are, of course, contracts in which the lender trades capital up front for the borrower's promise, secured by the borrower's asset, to make monthly payments. Banks frequently sell their secured rights to the monthly payments to the makers of financial products known as "asset-backed securities," the most prominent of which are mortgage-backed securities ("MBSs").

An MBS is created when a financial institution bundles a large number of mortgage loans into a special-purpose entity. The resulting entity owns the rights to a large pool of borrowers' monthly payments. The institution simultaneously sells notes backed by the MBS, i.e., by the bundle of loans, and may also sell equity interests in the MBS. When the maker of an MBS does this--when it sells the rights to the cash flow generated by the mortgages in its bundle--it may do so by creating different classes, or "tranches," of notes. "Tranching" allows the bank to create notes with different risk-and-return profiles and thereby to attract a variety of buyers, from the most risk-averse to the least. Such tranches are [**6]  often classified by letter, with first priority in receiving payment given to the holders of tranche "A" notes, second priority to "B," and so on. The riskier, lower-priority notes will receive higher interest rates. (Although lettering conventions differ across MBSs, the mechanics are roughly the same, regardless of how the various tranches are denominated.) At the bottom of the hierarchy is a small class of investors who have purchased equity interests in the MBS.

The payment scheme for the different tranches is typically known as a "waterfall." As mortgage payments come into the MBS entity, they cascade, "watering" tranche A noteholders first, then B, and so on down to the equity. No part of the  [*165]  borrowers' payments will go to the holders of equity in the MBS unless those payments exceed what the MBS must pay to its noteholders. The brunt of any borrower's default on one of the underlying mortgages is therefore borne first by the equity interest, then by the most junior notes, intermediate notes, and so on. It takes a large number of defaults to impair the cash flow to holders of tranche A notes--which is what makes those notes the least risky. If, however, the MBS performs well, receiving [**7]  full payment, the holders of the riskier tranches (and especially the equity) will receive higher returns.

By bundling large numbers of mortgages together into tranched MBS notes, a bank can achieve a number of goals. For one, it can create securities that enable non-lending institutions to invest in the housing market. In addition, it can create relatively safe investment opportunities through the senior tranches, because it takes widespread mortgage defaults to impair the cash flow to those tranches. Needless to say, the word "relatively" bears emphasizing in light of the real estate market collapse that lies behind this case and the many other cases like it.

In the same way that an MBS comprises a bundle of mortgage notes, a CDO comprises a bundle of MBS (or other asset-backed security) notes. Thus, where an MBS is a financial product backed by mortgages, a CDO is, in a sense, simply a second-order MBS, backed by those first-order financial products. A CDO is likewise built by creating a special-purpose entity that takes possession of a large group of notes--say, tranche B notes of various MBSs. The CDO will then sell to investors tranches of notes with diminishing priority, paying [**8]  out the funds collected on the securities held by the CDO to noteholders in the order of the tranches' relative priority.

A related type of derivative security available to investors in the mortgage market is a "credit default swap" ("CDS"). A CDS is known as a "derivative" because it transfers the risk associated with owning a particular security without necessarily transferring ownership of that security. In general, derivatives are purely financial contracts that call for payment by one contracting party to the other based on a specific event, such as fluctuation in the value of a selected security, interest rate, market index, or the like. Investment in a mortgage-based CDS is the opposite of investment in mortgage notes, in that it benefits the investor only if the borrowers do not make their mortgage payments. More precisely, the purchaser of the CDS promises to make regular monthly payments to the issuer in exchange for the issuer's promise to pay the purchaser in the event--and roughly to the extent--that borrowers default in making payments on the selected category of mortgage notes. Unless such defaults occur, the CDS buyer gets nothing in return for her regular payments.

Investment [**9]  in mortgage-based CDSs can serve two purposes. First, it may function as a speculative bet against the mortgage market. In other words, an investor who believed the housing market to be unrealistically inflated could purchase a CDS in anticipation of borrowers' defaults. Such an investor is essentially shorting the mortgage market, while the issuer of the CDS is taking a "long" position in that market.² Indeed, an investor eager to  [*166]  capitalize on an expected downturn in the market could increase its short position by purchasing multiple CDSs, thereby placing what amounts to a very large bet on impending defaults by borrowers. A second use for a CDS is as a hedge, or insurance against such defaults. Thus, investors in MBSs or CDOs, whose cash flow and value depend on borrowers' making their payments, can lessen the consequences of defaults by purchasing CDSs keyed to a similar class of mortgages.

2   The buyer of a CDS is sometimes referred to as holding the long position. See, e.g., Anouk G. P. Claes & Mark J. K. De Ceuster, Single Name Credit Default Swap Valuation: A Review, in Credit Risk: Models, Derivatives and Management (Niklas Wagner ed., 2008). While we understand the motivation [**10]  for that usage, we think that, at least in the securities litigation context, it makes more sense to use "long" to refer to the position taken by the party who stands to benefit from the success of the reference security--here, the CDS seller. See Lucas et al., supra, at 221 ("The protection seller under a CDS is said to be the seller of the CDS, but he is also long the CDS and long the underlying credit. The logic behind selling protection and being long the CDS is that the protection seller is in the same credit position as someone who owns, or is long, a bond.").

As pertinent here, some CDOs contain--in addition to asset-backed securities like MBSs--derivative securities like CDSs. A CDO might contain, for instance, not only specific tranches of MBS notes but also the long side of CDS contracts related to those tranches. In that case, the cash flow into the CDO would come from the regular payments by the CDS buyers--the short investors--as well as payments on the underlying mortgages. The CDO would also bear the corresponding risk both of defaults by borrowers and of the payouts to CDS buyers triggered by such defaults.

Given that CDOs consist of a portfolio of assets, a crucial matter for the makers of a CDO is [**11]  deciding who will pick the assets, e.g., the MBSs and CDSs that will be bundled together to form the CDO's collateral. Generally, this job is performed by a "collateral manager," an entity or person who has discretion to select assets that further the goals, and fulfill the requirements, of the CDO. Such requirements may concern various characteristics of the collateral securities, including their ratings by credit ratings agencies, their contractual structure, and their performance to date.

With that basic structure in mind, we turn to the particulars of this case.

B. Plaintiffs' Fraud Allegations

Plaintiffs invested millions in three CDOs created and offered by Wachovia: Octans, Sagittarius, and Longshore. Wachovia marketed these CDOs to Plaintiffs and also sold CDSs on each CDO. Structured Asset Investors, LLC ("SAI"), a Wachovia subsidiary, and Harding Advisory LLC ("Harding"), an independent company, served as collateral managers--Harding for Octans, SAI for Sagittarius and Longshore. All three CDOs held MBS notes as well as the long side of CDS contracts.

With respect to Octans and Sagittarius, Plaintiffs allege that, contrary to representations made to investors, Harding and SAI [**12]  selected shoddy, high-risk assets at the urging of Magnetar Capital LLC ("Magnetar"), a hedge fund that stood to profit massively if the CDOs failed. As to Longshore, Plaintiffs allege that, without telling investors, Wachovia used it to dump, at above-market prices, riskier MBS notes that had been on Wachovia's own balance sheets. Each alleged scheme is recounted more fully below.

 

1. The Constellation CDOs: Octans and Sagittarius

According to the complaint, Magnetar colluded with several banks and collateral managers to "orchestrate" at least 27 CDOs named after constellations, including Octans and Sagittarius. J.A. at 112. Magnetar would purchase the equity  [*167]  tranches in exchange for clandestine influence over the selection of assets. Although, as an equity holder, Magnetar thus stood to benefit from the CDOs' success, it used its position and influence to advance a contrarian investment strategy. In reality, Magnetar was betting heavily against the CDOs and the underlying MBSs by buying the short side of CDS contracts, which it had the ability to purchase at below-market prices from banks like Wachovia as a condition of its equity investment in the CDOs. Magnetar was simultaneously [**13]  able to fund these short positions with payments from its equity stake for a longer period of time, until the CDOs actually went into default, because the CDOs had been structured--at Magnetar's insistence--so that they would not divert cash from the equity holders to senior tranches in the face of early warning signs that the value of the portfolio collateral was deteriorating.

Plaintiffs, however, did not bring this suit against Magnetar. The instant litigation concerns Magnetar's helpmeets, Defendants, who are alleged to have conspired in structuring the deals and attracting long investors like Plaintiffs by masking Magnetar's central and adverse role.

As to Octans, the offering documents touted Harding's experience and skill as a collateral manager, stating that Harding would "[i]nvest in high quality assets with stable returns" and "minimize losses through rigorous upfront credit and structural analysis, as well as ongoing monitoring of asset quality and performance." Id. at 121. The documents also specified numerous procedures to be used by Harding in asset selection, including "detailed loan-level analysis." Id. at 121-22.

Harding, however, allegedly acted entirely contrary to these representations. It acceded [**14]  to Magnetar's requests, knowing that Magnetar's interests were directly at odds with the CDOs' success. In support of this claim, Plaintiffs assert several facts regarding particular email exchanges between Magnetar and Harding in August and September 2006, as assets were being selected. In one such exchange--between James Prusko, Magnetar's Senior Vice President, and Wing Chau, Harding's founder and president--Prusko requests to be copied on the trade approval process and updated daily if any trading activity occurred, adding, "We should also discuss CDO exposure as I will source the CDO CDS." Id. at 124. Chau responds, "Sounds good." Id. In another such exchange, a Harding employee asks Prusko "to let [them] know if [Magnetar] plan[s] on shorting any names into any of the [Octans] transactions." Id. at 125. An email to Prusko three days later from the same employee lists shorting opportunities that Harding was "able to source for [Magnetar]." Id. Based on these and other such exchanges, Harding is alleged not only to have known of Magnetar's shorting activities but also to have facilitated and concealed them.

Ignorant of Magnetar's role, Plaintiffs invested $94 million in October 2006 in the notes of various [**15]  Octans tranches. All of these notes became virtually worthless when the respective tranches went into default in May 2008.

Plaintiffs' factual allegations regarding Sagittarius are similar. The Sagittarius term sheets outlined the same type of goals as the Octans term sheets, detailing similarly rigorous procedures for managing the CDO. The "conservative approach" of the chosen collateral manager, SAI, was a key selling point, as Defendants knew. Id. at 128. With respect to both CDOs, Defendants stressed the collateral manager's expertise and its approach to asset selection because, as acknowledged in the  [*168]  offering documents, each CDO's performance ultimately turned on these two variables.

Plaintiffs allege that, despite these representations, Magnetar exerted control over SAI's asset selection, as it did over Harding's, and that Magnetar's strategy of betting against Sagittarius was known to SAI. For example, Prusko assertedly emailed Wachovia's managing director early on, stating that while he "didn't mean to kill [SAI] off," he did want it to be "more user friendly." Id. at 131. A few months later, a Wachovia trader allegedly emailed Prusko with a list of especially weak MBSs that were proposed for inclusion [**16]  in the CDO, inviting him to express "any thoughts or concerns." Id. at 132. Prusko responded, "Let[']s test the waters!" Id. Finally, in March 2007 Prusko sent Wachovia an email to which he attached a document that graphed Magnetar's returns for different projected loss scenarios. The graph showed that the worse the CDO performed, the larger Magnetar's profit. In the body of the email Prusko himself described Sagittarius as "not a pretty bond." Id. at 134.

In March 2007 Plaintiffs invested $5 million in Sagittarius Class A and B notes each. Both tranches defaulted in October of that year.

2. Longshore

The third CDO at issue, Longshore, was not among the constellation CDOs created as part of Magnetar's so-called "long-short" strategy. As with the other two CDOs, however, the Longshore offering documents highlighted the high quality of the asset selection and due diligence procedures that would be used by the collateral manager--here SAI. Contrary to these representations, Wachovia allegedly used Longshore as a dumping ground for MBS assets that it knew faced an imminent and steep decline in value, including assets on Wachovia's own books that were being transferred into Longshore from the warehouse of another, [**17]  canceled CDO deal.

As detailed in the complaint, these allegations were the subject of a fraud investigation by the SEC. In an order issued as part of the settlement of that investigation, the SEC found that while Wachovia represented in its offering documents that Longshore assets would be acquired in deals resembling arm's-length transactions, the assets from the collapsed CDO were transferred at their original cost basis despite, according to Wachovia's own internal valuations, a significant decline in their fair market value.

In April 2007 Plaintiffs bought notes of various Longshore tranches with a total face value of $59.1 million. In February 2008 these notes went into default.

C. Procedural History

In April 2012, in the wake of these losses and the larger financial crisis, Plaintiffs filed suit in New York state court against Defendants--namely, the CDO entities, SAI and Harding, and certain subsidiaries of Wells Fargo, which acquired Wachovia in 2008. The case was removed to federal court pursuant to the Edge Act. See 12 U.S.C. § 632; 28 U.S.C. § 1441(a).³

3   The Edge Act gives original jurisdiction to federal district courts over any civil suit involving a corporate party "organized under the laws of the United States" [**18]  and "arising out of transactions involving international or foreign banking [or] other international or foreign financial operations." 12 U.S.C. § 632. The present suit falls under the Edge Act because Wells Fargo Bank, N.A. (a party to this action as the successor in interest of Wachovia Bank, N.A.) is a federally chartered corporation, and this suit "arise[s] out of [its] offshore banking or financial transaction[s]." Am. Int'l Grp., Inc. v. Bank of Am. Corp., 712 F.3d 775, 784 (2d Cir. 2013).

 [*169]  In July 2012, following the voluntary dismissal of certain defendants, but before the remaining defendants moved to dismiss, the district court held a "pre-motion conference." In three-page letters and at oral argument, the parties previewed their arguments in support of and opposition to the remaining defendants' anticipated Rule 12(b)(6) motion. At that conference, the district court also inquired of Plaintiffs whether they wished to amend the complaint in light of this preview, stating that it was not necessarily the court's practice to "give them another opportunity later." J.A. at 415. The court indicated that it considered Defendants' pre-motion letter and the points raised at conference to provide "fair warning" of Defendants' arguments and the potential need for amendment. Id. Plaintiffs declined the court's [**19]  invitation to amend, arguing that the complaint was legally sufficient to state a claim for fraud.

Shortly thereafter, Defendants moved to dismiss the complaint for failure to state a claim under Rule 12(b)(6).⁴ While vigorously opposing the motion, Plaintiffs also requested leave, in the alternative, to amend the complaint. See J.A. at 584 (citing Fed. R. Civ. P. 15(a)(2)). In February 2013, the district court held oral argument on the motion, at the end of which Defendants raised the issue of amendment, reading back the relevant portions of the pre-motion conference transcript. Plaintiffs' counsel twice sought to be heard on the issue and was twice denied a chance to respond. See J.A. at 732 ("THE COURT: All [defense counsel] did was quote to me what I said at the premotion conference. I was there. And the time to respond was then.").

4   Defendants simultaneously moved, under Rule 12(b)(2), to dismiss the complaint as to Wells Fargo Securities International Limited for lack of personal jurisdiction, which motion the district court granted. See Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *5-6. As Plaintiffs do not contest this part of the district court's decision, that specific Wells Fargo subsidiary is not a party to this appeal.

In March 2013 the district court issued a memorandum [**20]  opinion dismissing the complaint in its entirety, with prejudice. See Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *16 & n.3. The instant appeal followed.

DISCUSSION

Plaintiffs challenge on appeal the district court's determination that they inadequately pleaded their fraud claim as well as the court's concomitant denial of their request to replead. We review de novo the district court's dismissal under Rule 12(b)(6), accepting all factual allegations in the complaint as true and drawing all reasonable inferences in Plaintiffs' favor. Adelson v. Harris, 774 F.3d 803, 807 (2d Cir. 2014). We review the district court's denial of leave to amend the complaint for abuse of discretion. See Lotes Co. v. Hon Hai Precision Indus. Co., 753 F.3d 395, 403 (2d Cir. 2014).

I

Before turning to particular aspects of Plaintiffs' complaint, we briefly address the law that applies to pleading fraud in general, including a threshold choice-of-law question in this somewhat unusual case.

When a federal district court sits in diversity, it applies the Federal Rules of Civil Procedure, as it does in all but a few civil actions, see Fed. R. Civ. P. 1, and it generally applies the decisional law of the state in which it sits, including the forum state's choice-of-law rules. See 28 U.S.C. § 1652; Klaxon Co. v. Stentor Elec. Mfg. [*170]  Co., 313 U.S. 487, 496, 61 S. Ct. 1020, 85 L. Ed. 1477 (1941); In re Coudert Bros. LLP, 673 F.3d 180, 186 (2d Cir. 2012). The district court's jurisdiction here, however, was not predicated on diversity. The case was removed from New York state court pursuant [**21]  to the Edge Act, which similarly confers federal jurisdiction but neither provides any substantive federal law for us to apply nor guides our choice-of-law inquiry. See 12 U.S.C. § 632; A.I. Trade Fin., Inc v. Petra Int'l Banking Corp., 62 F.3d 1454, 1459-65, 314 U.S. App. D.C. 122 (D.C. Cir. 1995) (offering in-depth discussion of choice of law in Edge Act litigation); cf. Barkanic v. Gen. Admin. of Civil Aviation of the People's Republic of China, 923 F.2d 957, 959 (2d Cir. 1991) (analyzing Foreign Sovereign Immunities Act's comparable lack of choice-of-law provisions). The difficulties inherent in choice-of-law questions under the Edge Act need not, however, detain us. In the instant case, the district court applied the substantive law of New York. Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *8. And, as neither party objects, we will do the same. See Vacold LLC v. Cerami, 545 F.3d 114, 122-23 (2d Cir. 2008).⁵

5   In decades past, panels of our Court analyzed choice of law in Edge Act cases under federal common law. See, e.g., Aaron Ferer & Sons Ltd. v. Chase Manhattan Bank, N.A., 731 F.2d 112, 120-21 (2d Cir. 1984); Corporacion Venezolana de Fomento v. Vintero Sales Corp., 629 F.2d 786, 791-92 (2d Cir. 1980). It is not clear, however, that employing federal common law was determinative of the choice-of-law issue in either Aaron Ferer or Corporacion Venezolana. And the Supreme Court's curtailment of federal common lawmaking in the years since those cases were decided casts doubt on the durability of their approach. See, e.g., Atherton v. F.D.I.C., 519 U.S. 213, 218, 117 S. Ct. 666, 136 L. Ed. 2d 656 (1997) (requiring "significant conflict" between federal policy and use of state law as "precondition" for federal common lawmaking); O'Melveny & Myers v. F.D.I.C., 512 U.S. 79, 87, 114 S. Ct. 2048, 129 L. Ed. 2d 67 (1994) (same); see also In re Gaston & Snow, 243 F.3d 599, 605 (2d Cir. 2001) ("[F]ederal choice of law rules are a [**22]  species of federal common law . . . . [And they] are no different from any other judicially created rule of decision. . . ." (citations omitted)). In light of the Supreme Court's more recent pronouncements, at least one circuit has endorsed the opposite approach in Edge Act cases--namely, applying the forum state's choice of law. See Petra Int'l, 62 F.3d at 1463-64 ("[W]here the 'federal question' giving rise to federal jurisdiction need not appear upon the face of a well-pleaded complaint, there is no reason for the federal court to conduct any different choice-of-law inquiry than would a court of the forum state in deciding the same issue.").

We need not decide the question here, however. The complaint asserts, and Defendants do not dispute, that "all Defendants transacted business within New York giving rise to Plaintiffs' causes of action," and that the conduct complained of was "orchestrated . . . in and from New York." J.A. at 105. And the result is therefore likely the same, regardless of whether we analyze choice of law under federal or New York law. Compare Corporacion Venezolana, 629 F.2d at 793 (2d Cir. 1980) (inquiring into which state has most significant contacts with litigation), and Aaron Ferer, 731 F.2d at 121 (same), with Schultz v. Boy Scouts of Am., Inc., 65 N.Y.2d 189, 197, 480 N.E.2d 679, 491 N.Y.S.2d 90 (1985) (inquiring into which state has "greatest interest" in litigation). [**23]

Under New York law, fraud requires proof of (1) a material misrepresentation or omission of a fact, (2) knowledge of that fact's falsity, (3) an intent to induce reliance, (4) justifiable reliance by the plaintiff, and (5) damages. Eurycleia Partners, LP v. Seward & Kissel, LLP, 12 N.Y.3d 553, 559, 910 N.E.2d 976, 883 N.Y.S.2d 147 (2009); see Lama Holding Co. v. Smith Barney, 88 N.Y.2d 413, 421, 668 N.E.2d 1370, 646 N.Y.S.2d 76 (1996). At the pleading stage, to withstand a Rule 12(b)(6) challenge in federal court, Plaintiffs must "assert facts that plausibly support the inference of fraud." Cohen v. S.A.C. Trading Corp., 711 F.3d 353, 360 (2d Cir. 2013).

 [*171]  Additionally, in conjunction with the facial plausibility standard of Rule 12(b)(6), Plaintiffs must satisfy the heightened pleading standard set forth in Rule 9(b), which reads:

   In alleging fraud or mistake, a party must state with particularity the circumstances constituting fraud or mistake. Malice, intent, knowledge, and other conditions of a person's mind may be alleged generally.

Fed. R. Civ. P. 9(b).

In essence, Rule 9(b) places two further burdens on fraud plaintiffs--the first goes to the pleading of the circumstances of the fraud, the second to the pleading of the defendant's mental state. As to the first, we have held that the complaint must "(1) detail the statements (or omissions) that the plaintiff contends are fraudulent, (2) identify the speaker, (3) state where and when the statements (or omissions) were made, and (4) explain why the statements [**24]  (or omissions) are fraudulent." Eternity Global Master Fund Ltd. v. Morgan Guar. Trust Co. of N.Y., 375 F.3d 168, 187 (2d Cir. 2004) (internal quotation marks omitted). As to the second, though mental states may be pleaded "generally," Plaintiffs must nonetheless allege facts "that give rise to a strong inference of fraudulent intent." Lerner v. Fleet Bank, N.A., 459 F.3d 273, 290-91 (2d Cir. 2006).

In determining the adequacy of Plaintiffs' fraud pleadings under these various requirements, we view the alleged facts in their totality, not in isolation. See Tellabs, Inc. v. Makor Issues & Rights, Ltd., 551 U.S. 308, 322-23, 127 S. Ct. 2499, 168 L. Ed. 2d 179 (2007). As always at the Rule 12(b)(6) stage, we credit all non-conclusory factual allegations in the complaint and draw all reasonable inferences in Plaintiffs' favor. Nielsen v. Rabin, 746 F.3d 58, 62 (2d Cir. 2014). The question that governs our de novo review is whether such allegations and inferences plausibly indicate Plaintiffs' entitlement to relief. See Ashcroft v. Iqbal, 556 U.S. 662, 678-80, 129 S. Ct. 1937, 173 L. Ed. 2d 868 (2009).

The district court gave several grounds for its dismissal of Plaintiffs' fraud claim: (1) the complaint as a whole did not differentiate among the Wachovia entities, Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *9, *13; (2) the pleadings as to the constellation CDOs raised neither a plausible inference of a material misrepresentation nor a strong inference of scienter, 2013 U.S. Dist. LEXIS 49665, [WL] at *10-13; and (3) the pleadings as to Longshore impermissibly relied on the SEC order and were otherwise insufficiently particular, 2013 U.S. Dist. LEXIS 49665, [WL] at *13-15. Upon our own review, we find Plaintiffs' fraud pleadings [**25]  sufficient to state a claim against defendants Wachovia and Harding, but not against defendant SAI. Moreover, for the reasons given in Part III, infra, we conclude that even as to SAI, whose dismissal was proper, dismissal with prejudice was improper. Accordingly, we reverse the dismissal in part, as to Wachovia and Harding, and we vacate it in part, as to SAI, remanding the case to the district court to determine on the basis of an amended complaint whether repleading will cure the defects identified by us below. See infra Sections I.B.2, I.C, I.D.2.

A. Identification of the Speaker

Under Rule 9(b), Plaintiffs must "identify the speaker" of the allegedly fraudulent statements. Eternity Global, 375 F.3d at 187. According to Defendants, the complaint fails because it attributes these statements to a cluster of subsidiaries collectively referred to as "Wachovia" rather than to any specific Wachovia entity. In response, Plaintiffs argue that their pleadings satisfy Rule 9(b) on this score because the speaker is a group of affiliates, and an appropriate factual basis for such  [*172]  grouping exists. In rejecting Plaintiffs' argument, the district court concluded that the allegations of interrelatedness were insufficient to treat the Wachovia entities [**26]  as a group for purposes of the identification requirement. See Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *9, *12. We disagree.

The complaint identifies three Wachovia entities who acted together to structure and offer the securities in question: (1) Wachovia Capital Markets, LLC, as the initial purchaser of the notes issued by all three CDOs; (2) Wachovia Securities International Limited, as this initial purchaser's agent for the sale of the Sagittarius and Longshore notes; and (3) Wachovia Bank, N.A. as the initial CDS counterparty for the three CDOs, their warehouse financing provider, and the liquidity provider for Octans and Sagittarius. See J.A. at 104.⁶ When read together with the complaint as a whole, these allegations suffice, in our view, to "inform each defendant of the nature of [its] alleged participation in the fraud." DiVittorio v. Equidyne Extractive Indus., Inc., 822 F.2d 1242, 1247 (2d Cir. 1987). The complaint states at the outset that it will refer to these entities collectively as "Wachovia," J.A. at 99, and we are hard-pressed to see how Plaintiffs could have done otherwise in the context of the present litigation, or why they ought to have done otherwise based on our cases.

6   As noted in the Procedural History, supra, the corresponding defendants in this case are three Wells Fargo subsidiaries, one of whom--Wells Fargo [**27]  Securities International Limited, the successor in interest of Wachovia Securities International Limited--was dismissed for lack of personal jurisdiction. See Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *5-6.

Our Circuit first addressed the issue of group-produced misrepresentations in Luce v. Edelstein, 802 F.2d 49 (2d Cir. 1986). There, disgruntled investors in an "ill-fated real estate partnership" alleged securities fraud, suing the partnership, its general partners (themselves partnerships), affiliated entities, and the individuals who tightly controlled all of them. Id. at 51. The complaint attributed to "defendants" as an undifferentiated group several statements about the securities in question--in particular, (1) statements made in an offering document and (2) other oral and written statements made outside the offering documents themselves.

The treatment in Luce of the two types of statements is instructive. We held that while, as to the second category, the complaint lacked the specificity required by Rule 9(b), such group pleading was sufficient as to the first category, i.e., the statements made in the offering document. Id. at 55 ("[N]o specific connection between fraudulent representations in the Offering Memorandum and particular defendants is necessary where . . . defendants are insiders [**28]  or affiliates participating in the offer of the securities in question."); see Ouaknine v. MacFarlane, 897 F.2d 75, 80 (2d Cir. 1990) ("[R]eference to an offering memorandum satisfies 9(b)'s requirement of identifying time, place, speaker, and content of representation where . . . defendants are insiders or affiliates participating in the offer of securities."); DiVittorio, 822 F.2d at 1247.⁷

7   Defendants suggest in their brief that group pleading may no longer be viable in the wake of the PSLRA. While other circuits have held that the PSLRA eliminated group pleading for private actions under the federal securities laws, see, e.g., Winer Family Trust v. Queen, 503 F.3d 319, 336 (3d Cir. 2007), we note that numerous district courts in our Circuit have reached the opposite conclusion. See City of Pontiac Gen. Employees' Ret. Sys. v. Lockheed Martin Corp., 875 F. Supp. 2d 359, 373 (S.D.N.Y. 2012) (collecting cases). The vitality of the group pleading doctrine as to federal securities fraud is an open question in our Circuit, and one that is not before us in this case. Plaintiffs' fraud claim here is governed by state common law rather than the PSLRA, and at least in the present context it seems clear to us that Luce and its progeny remain precedential.

 [*173]  The statements here belong to the first category treated in Luce: official materials produced in connection with the sale of securities. Plaintiffs allege that Wachovia itself--acting through three [**29]  different affiliates--structured the CDOs, was the initial purchaser of the notes, provided financing, and acted as the initial CDS counterparty. The entities named in the complaint and treated collectively were thus "insiders or affiliates participating in the offer of securities," Ouaknine v. MacFarlane, 897 F.2d 75, 80 (2d Cir. 1990), and they were allegedly "link[ed] . . . in . . . specific way[s] to . . . fraudulent representation[s] or omission[s]" made in those offering documents. DiVittorio, 822 F.2d at 1249.

Even under the heightened pleading standard of Rule 9(b), Plaintiffs are not obliged to disaggregate these affiliates to pursue their fraud claim. Where a plural author is implied by the nature of the representations--for instance, where, as here, (1) the alleged fraud is based on statements made in the offering materials and (2) the complaint gives grounds for attributing the statements to the group--group pleading may satisfy the source identification required by Rule 9(b). Following Luce, we hold that there is no fixed requirement in such circumstances to identify a single entity within the group on pain of dismissal.

Moreover, in the instant case, each Wachovia entity was a member of a corporate subgroup that operated together and communicated with Plaintiffs under a shared trade name: "Wachovia Securities." [**30]  J.A. at 263, 309, 466.⁸ Each employee involved in the CDO transactions was listed on a Wachovia Securities phone list without reference to a specific entity in the subgroup. And the logo emblazoned on the marketing materials was that of Wachovia Securities. As a result, Plaintiffs' designation in the complaint of these three defendants by a group shorthand rather than their individual entity names amounts, at most, to excusable mislabeling. Cf. Datskow v. Teledyne, Inc., Cont'l Prods. Div., 899 F.2d 1298, 1300-01 (2d Cir. 1990). And the costs of such mislabeling are better borne in this situation by those who authored the offering documents, which were characterized by that same slippage between the collective trade name and the entities acting under it. It would be strange indeed to demand greater precision of Plaintiffs in pleading the author's identity than they received as readers of these documents.

8   Wells Fargo continues to use a trade name ("Wells Fargo Securities") to cover the successor entities.

In sum, given that the alleged fraud focuses (1) on specific misrepresentations in the CDO offering documents and (2) on the coordinated activity by specific Wachovia affiliates in constructing and offering these CDOs, Plaintiffs' identification of the group suffices [**31]  to meet the particularity of attribution required by Rule 9(b). Wachovia's own lack of transparency in identifying which entity was communicating to prospective investors only bolsters our conclusion in this regard.

B. Material Misrepresentations and Omissions

Having concluded that the complaint sufficiently identifies Wachovia as the source of many of the statements at issue, we next consider whether "the complaint assert[s] facts that plausibly support the inference of fraud." Cohen, 711 F.3d at 360. Because the substance of the fraud alleged with respect to Octans and Sagittarius  [*174]  differs in kind from that alleged with respect to Longshore, we treat the two types of fraud allegations separately. We discuss the constellation CDOs in this section and the next, Section I.C (Scienter), before turning to Longshore in Section I.D.

As to Octans and Sagittarius, the district court determined that Plaintiffs had inadequately pleaded a material misrepresentation or omission. See Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *10-12. In so doing the district court erred, in our view, by imposing its own reading of Plaintiffs' substantive fraud allegations without considering them in the light most favorable to Plaintiffs. While Rule 9(b) requires that "the circumstances constituting fraud" be [**32]  "state[d] with particularity," Fed. R. Civ. P. 9(b), it does not require factual pleadings that demonstrate the probability of wrongdoing. See Iqbal, 556 U.S. at 678 ("The plausibility standard is not akin to a 'probability requirement' . . . ." (quoting Bell Atl. Corp. v. Twombly, 550 U.S. 544, 556, 127 S. Ct. 1955, 167 L. Ed. 2d 929 (2007)). At the pleadings stage, the alleged fraud need only be plausible based on the complaint; it need not be more likely than other possibilities. Twombly, 550 U.S. at 556 ("[A] well-pleaded complaint may proceed even if it strikes a savvy judge that actual proof of those facts is improbable, and that a recovery is very remote and unlikely." (citation and internal quotation marks omitted)); cf. Cohen, 711 F.3d at 360 ("Iqbal . . . requires assertions of facts supporting a plausible inference of fraud--not of facts which can have no conceivable other explanation.").

Upon our de novo review, we conclude that the facts asserted in the complaint are sufficiently particular and plausibly support the existence of a material misrepresentation or omission with respect to Wachovia and Harding, but not with respect to SAI.

1. Wachovia and Harding

The gravamen of Plaintiffs' complaint is that, as to the two constellation CDOs, the offering documents (1) misrepresented that SAI and Harding would serve as judicious collateral managers, who would [**33]  employ certain selection and monitoring procedures, and (2) omitted both Magnetar's role in selecting the collateral and Magnetar's adverse position relative to the CDOs.

The complaint plausibly suggests that these alleged misrepresentations were made by both Wachovia and Harding. First and foremost, Wachovia (as a group) authored the Octans and Sagittarius offering documents in which the allegedly misleading statements appeared. In addition, Harding itself stated that it "w[ould] be responsible for selecting and monitoring [Octans'] collateral," with no mention of Magnetar. J.A. at 122. And this statement came from a page of the Octans Offering Circular prepared by Harding, at the top of which Harding claimed "responsibility for the information contained in this section" and expressly represented that it had "not omit[ted] anything likely to affect the import of such information." Id. at 253.

Plaintiffs allege that had they known that the collateral managers would not exercise independent judgment and would instead accede to the desires of a powerful short investor, they would not have invested in either CDO.

It is not for us to say at this stage whether Plaintiffs' account of Magnetar's role and of Defendants' [**34]  sleights of hand regarding that role is true, nor is it for us to say whether, at a later stage, a judge or jury might find that such misrepresentations were immaterial to sophisticated investors  [*175]  like Plaintiffs. The question properly before us is whether the facts pleaded are sufficiently detailed to satisfy Rule 9(b)--in particular, whether these facts "plausibly support the inference of fraud." Cohen, 711 F.3d at 360.

Our conclusion that they do rests, first, on the numerous alleged exchanges between Magnetar and Defendants. As to Octans, for example, various exchanges in August and September 2006 between Prusko and Chau--senior officers of Magnetar and Harding, respectively--may be read to suggest (1) Magnetar's considerable influence over the CDO and (2) high-level discussions regarding Magnetar's long-short strategy. An email from a Wachovia trader to Prusko also refers to "a few trades [Wachovia] did on behalf of Magnetar," while, in another email, a Harding employee asks Prusko to "let [them] know if [he] plan[s] on shorting any names into any of the [Octans] transactions." J.A. at 125. Together, such emails plausibly indicate not only a cozy relationship with Magnetar but also specific actions taken by Wachovia and [**35]  Harding at Magnetar's direction based on Magnetar's short positions, such as selecting certain CDSs for inclusion in Octans.

Exchanges between Magnetar and Wachovia pertaining to Sagittarius support a similar reading of Magnetar's undisclosed role with respect to this CDO. For example, Prusko emailed Wachovia's managing director a request to make SAI more "user friendly." Id. at 131. In another email to a team of Wachovia employees a few months later, Prusko attached a graph showing higher profits to Magnetar from the CDO's failure and described Sagittarius as "not a pretty bond." Id. at 134.

In addition, both Octans and Sagittarius had built-in features conducive to Magnetar's alleged strategy, and the presence of these features lends further support to Plaintiffs' account of Magnetar's role. For example, Plaintiffs allege that Magnetar received sourcing fees for every CDS that it chose for the asset bundle, and that Magnetar also negotiated for favorable waterfall rules, which would permit equity holders like itself to continue receiving payments longer in the face of early signs of default. While at least some of these features were disclosed to investors and may not form a basis for fraud in themselves, [**36]  they may be read to suggest favoritism towards Magnetar and thereby to put Plaintiffs' other allegations in context.

Construed in the light most favorable to Plaintiffs, the emails regarding Octans and Sagittarius--together with the structural elements advantageous to Magnetar--plausibly support an inference that the offering documents materially misled investors by falsely holding out the skill and rigorous asset selection methods of the respective collateral managers while failing to disclose Magnetar's antagonistic influence.

The district court reached the opposite conclusion by discounting Plaintiffs' account as "read[ing] too much" into the emails and by proffering benign alternative explanations. Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *10. In the district court's view, the emails show Magnetar merely to have been "an active and involved equity investor." 2013 U.S. Dist. LEXIS 49665, [WL] at *11. The excerpts of the emails quoted in Plaintiffs' complaint are scarcely unequivocal and may well be susceptible of plausible alternative readings. But it is not our task at this stage to construe the abundant industry jargon here in any definitive fashion. Rule 9(b) requires only that Plaintiffs plead, with particularity, facts from which it is plausible to infer fraud; it does not require [**37]  Plaintiffs to plead facts that make fraud more probable than other explanations. Iqbal, 556 U.S. at 678;  [*176]  Twombly, 550 U.S. at 556. Drawing all reasonable inferences favorable to Plaintiffs, we conclude that such emails are sufficiently particular to render Plaintiffs' account at least plausible. In concluding otherwise, the district court erred by requiring Plaintiffs to show that their reading was superior to the court's own benign reading, thereby imposing a de facto probability requirement at the pleadings stage.⁹

9   It is possible that the district court took itself to be required to consider "plausible opposing inferences" in order to apply the standard not of mere plausibility but of "at least as compelling as any opposing inference." Tellabs, Inc. v. Makor Issues & Rights, Ltd., 551 U.S. 308, 323-24, 127 S. Ct. 2499, 168 L. Ed. 2d 179 (2007). That standard, however, reflects the requirement that there be a "strong inference" of scienter, Id. (emphasis added), and this comparative inquiry applies only to the scienter element, cf. Lerner, 459 F.3d at 290-91, which we discuss in the next section, and not to the material misrepresentation element discussed here.

2. SAI

Plaintiffs' pleadings of material misrepresentation falter, however, with respect to SAI. As explained above, Wachovia (as a group) is alleged to have misled investors in the offering documents by [**38]  misrepresenting that Harding and SAI were judicious collateral managers without disclosing in those documents Magnetar's part in the CDOs' design and asset selection. Harding, too, is alleged to have misrepresented its role--on a page of the Octans Offering Circular for which it took express responsibility--by stating that it was responsible for selecting and monitoring Octans' assets and omitting any mention of Magnetar.

The complaint does not attribute any similar statement to SAI. Plaintiffs allege only that

   SAI, as a subsidiary of Wachovia, shared premises with Wachovia, acted under Wachovia's direction, and worked closely with Wachovia to develop a close relationship with Plaintiffs' investment advisor, for the purpose of selling CDO investments to Plaintiffs.

J.A. at 129.

Whether SAI, like Harding, ever held itself out in official sales materials to be in charge of selecting assets for the CDO is not clear from the complaint, nor does the complaint allege facts from which it would be reasonable, at this point, to infer that SAI--separate and apart from the entities operating as "Wachovia Securities"--misled investors as to its authority over asset selection by failing to disclose Magnetar's [**39]  influence. To sustain a cause of action for fraud against SAI, Plaintiffs will need to plead, with the requisite particularity, a material misrepresentation or omission by SAI.

C. Scienter

Under New York law, Plaintiffs must ultimately prove that Defendants possessed "knowledge of [their misstatements'] falsity" and "an intent to induce reliance." Eurycleia Partners, 12 N.Y.3d at 559. While Rule 9(b) allows mental states to be "alleged generally," Fed. R. Civ. P. 9(b), this relaxation of the heightened pleading requirement is not to be mistaken "for a license to base claims of fraud on speculation and conclusory allegations." Lerner, 459 F.3d at 290. To survive a Rule 12(b)(6) challenge, Plaintiffs must state facts sufficient to "give rise to a strong inference of fraudulent intent." Id.; cf. 15 U.S.C. § 78u-4(b)(2)(A) (using similar language as part of PSLRA's pleading standard for federal securities fraud). An inference [*177]  is "strong" if it is "cogent and at least as compelling as any opposing inference one could draw from the facts alleged." Tellabs, 551 U.S. at 324. In determining whether this strength-of-inference requirement is met, "[w]e consider the complaint in its entirety and 'take into account plausible opposing inferences.'" Stratte-McClure v. Morgan Stanley, 776 F.3d 94, 106 (2d Cir. 2015) (quoting Tellabs, 551 U.S. at 323).

The district court concluded that the facts asserted in Plaintiffs' complaint failed to raise [**40]  a strong inference of scienter because those facts raised no plausible inference of misrepresentation in the first place. See Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *12-13. We agree with the district court that Plaintiffs have yet to allege a misrepresentation by SAI and, hence, that, as to SAI, scienter has been inadequately pleaded. As indicated above, however, we disagree with the district court's conclusion that the facts in the complaint are insufficient to infer that Wachovia and Harding materially misled Plaintiffs in the Octans and Sagittarius offering documents. We must, therefore, proceed to analyze the sufficiency of the pleadings of scienter as to Wachovia and Harding. In examining these defendants' alleged knowledge and intent, we draw all reasonable inferences favorable to Plaintiffs and take into account any plausible competing inferences.

At the pleading stage, under Rule 9(b), a fraud plaintiff may establish a "strong inference" of scienter, among other ways, "by alleging facts that constitute strong circumstantial evidence of conscious misbehavior or recklessness." Lerner, 459 F.3d at 290-91 (quoting Shields v. Citytrust Bancorp, Inc., 25 F.3d 1124, 1128 (2d Cir.1994)). In the securities fraud context, we have typically found it sufficient to state a claim based on recklessness if the complaint "specifically allege[s] defendants' [**41]  knowledge of facts or access to information contradicting their public statements." Novak v. Kasaks, 216 F.3d 300, 308 (2d Cir. 2000).¹⁰ Where the defendant at issue is a corporation, it is possible to plead corporate scienter by pleading facts sufficient to create a strong inference either (1) that "someone whose intent could be imputed to the corporation acted with the requisite scienter" or (2) that the statements "would have been approved by corporate officials sufficiently knowledgeable about the company to know" that those statements were misleading. Teamsters Local 445 Freight Div. Pension Fund v. Dynex Capital Inc., 531 F.3d 190, 195-96 (2d Cir. 2008) (citation and internal quotation marks omitted).

10   Because the parties have confined their arguments regarding scienter to whether the complaint satisfies Rule 9(b), and have not addressed whether the substantive scienter requirement under New York law has been met, we likewise consider only whether the complaint adequately pleads scienter under Rule 9(b).

With respect to Wachovia, the email exchanges and circumstances of the constellation CDOs' design suffice to create a strong inference that Michael Thompson, a managing director, knew of Magnetar's alleged position and role vis-à-vis the CDOs and, hence, also knew (or should have known) that omitting that fact from the Octans and Sagittarius offering [**42]  documents made the representations of the collateral managers' skill and careful selection methods in those documents misleading.

For example, early on, as the Sagittarius CDO was being created and marketed, Prusko asked Thompson directly to improve the user-friendliness of SAI, and Thompson replied that they were "working on that angle." J.A. at 131. A few weeks later, a Wachovia trader emailed Prusko,  [*178]  stating that Thompson wanted him to forward a "list of names" of potential assets, adding, "Please let us know if you have any thoughts or concerns." Id. at 132. (Prusko replied, "Let[']s test the waters!" Id.) Prusko likewise emailed unidentified Wachovia employees a graph showing increased profits to Magnetar as Sagittarius failed. In the body of that email, Prusko described the CDO as "not a pretty bond." Id. at 134. By containing specific facts showing (1) that Wachovia knew Magnetar's profits would increase as Sagittarius failed and (2) that Magnetar was being accorded a role in asset selection, the complaint adequately pleads Wachovia's awareness of undisclosed facts that rendered the company's representations to investors materially misleading.

With respect to Harding, similar factual allegations [**43]  give rise to a strong inference of corporate scienter. Specifically, Plaintiffs plausibly allege that Wing Chau, Harding's principal and owner, knew of Magnetar's part in constructing Octans and selecting its assets and, hence, also knew that Harding's own statement in the Octans offering documents about its "responsib[ility] for selecting and monitoring the collateral portfolio" was misleading. Id. at 122. Not only does the email correspondence between Chau and Prusko in August and September 2006 suggest that Harding explicitly catered to Magnetar; some of these emails may also be read to imply discussions with Magnetar about CDSs that would be selected for inclusion in Octans on the basis of Magnetar's broader investment goals. See id. at 123-124. An email from one of Chau's subordinates asking Prusko for specific names that Magnetar would be "shorting . . . into" Octans supports this reading. Id. at 125.

From these emails, together with the structural features of the constellation CDOs that favored Magnetar's supposed long-short strategy, it is reasonable to infer that Thompson and Chau--high-level employees of Wachovia and Harding, respectively--knew or should have known that the disclosures in the offering documents were [**44]  misleading because of their omission of Magnetar's influence over the CDOs' asset selection. In discussing the elements of material misrepresentation and omission, the district court drew from Chau's emails a contrary non-culpable inference--namely, that Harding was, at most, "happy to accommodate" certain requests by "an active and involved equity investor." Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *11. Whether Plaintiffs can ultimately prove their account of Magnetar's short investment strategy and of its control over the CDOs is, of course, another matter. But the facts pleaded in the complaint plausibly support Wachovia's and Harding's knowing omission of that control. And, in our view, the inference of corporate scienter here is "cogent and at least as compelling as" the innocent picture painted by the district court. Tellabs, 551 U.S. at 324.

In sum, Plaintiffs' allegations of fraud by Wachovia and Harding regarding the constellation CDOs satisfy Rules 9(b) and 12(b)(6). As to the role played by SAI, however, Plaintiffs have pleaded neither a material misrepresentation nor, a fortiori, scienter. They must replead both elements if they wish to proceed against SAI for fraud in connection with Sagittarius, the constellation CDO of which it was the collateral manager. [**45]

D. Longshore

The alleged fraud with respect to Longshore differs in kind and, therefore, requires separate treatment. The thrust of Plaintiffs' claim is that Wachovia used this non-constellation CDO to dispose of devalued  [*179]  assets on Wachovia's books, including assets from a canceled CDO deal, and that Wachovia did so by transferring the assets at their original cost basis, without accounting for their true (sinking) value. Although the complaint does not allege precisely when the toxic assets were transferred or at precisely what diminished value, it does assert (1) that in transferring these assets, Wachovia was aware of "impending changes to the ratings methodologies" that would result in major downgrades to the assets' value, J.A. at 137, and (2) that the Longshore Offering Memorandum misrepresented that SAI, as the collateral manager, would "cause any acquisition or sale" of assets "to be conducted on an arm's length basis" or--if carried out with SAI or an affiliate--to be done at least "on terms as favorable to [Longshore] as would be the case" in a transaction between unrelated parties, Id. at 136.

The issue is whether these allegations suffice to raise a plausible inference of a material [**46]  misrepresentation by Wachovia and SAI as well as a strong inference of scienter. While the question is a close one, we find the complaint sufficient to support the requisite inferences as to Wachovia. As to SAI, however, the complaint fails to plead a fraudulent misrepresentation with the particularity required by Rule 9(b) and, hence, also fails to allege scienter. Cf. supra Sections I.B.2 & I.C.

1. Wachovia

The SEC itself investigated the underlying charge of fraud by Wachovia in connection with the transfer into Longshore of assets from the canceled CDO. In its final order, the SEC set forth findings that Plaintiffs, in turn, recite in their complaint. See J.A. 138-39. Hence, a threshold question in our de novo review is what weight, if any, to give the quoted findings, which--as the district court correctly observed--were not admitted by Wells Fargo (Wachovia's successor in interest) when it settled the matter with the SEC. Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *14.

Citing our decision in Lipsky v. Commonwealth United Corp., 551 F.2d 887 (2d Cir. 1976), Defendants maintain that because the SEC order is inadmissible to prove fraud,¹¹ it should likewise be disregarded in deciding the sufficiency of the fraud allegations under Rule 12(b)(6). In response Plaintiffs argue that Lipsky does not categorically bar courts [**47]  from considering the content of such orders at the pleading stage, regardless of their admissibility. And they argue this while not disputing that the assertions of fact in question, having been neither admitted nor denied, represent only allegations of wrongdoing, with precisely no collateral estoppel effect. See Lipsky, 551 F.2d at 893-94 ("[C]onsent decrees . . . are not true adjudications of the underlying issues."); U.S. S.E.C. v. Citigroup Global Markets Inc., 827 F. Supp. 2d 328, 333 (S.D.N.Y. 2011) ("As a matter of law, an allegation that is neither admitted nor denied is simply that, an allegation."), vacated and remanded on other grounds, 752 F.3d 285 (2d Cir. 2014).

11   Although we have consistently held SEC orders like the one at issue here inadmissible to prove the facts of liability, our cases have not been uniform in explaining this result in terms of the Federal Rules of Evidence. Compare Lipsky, 551 F.2d at 893 (stating that such orders, like criminal nolo contendere pleas, are excluded under Rule 410), with United States v. Gilbert, 668 F.2d 94, 97 (2d Cir. 1981) (reasoning that because consent decrees are akin to the settlement of civil claims, their exclusion as proof of liability is actually governed by Rule 408, which explicitly permits use for other purposes), and Brady v. Wal-Mart Stores, Inc., 531 F.3d 127, 136 (2d Cir. 2008) ("A consent decree may properly be admitted to demonstrate that a defendant was aware of its legal obligations.").

 [*180]  Whatever cognizance of secondhand [**48]  allegations courts may take at the pleading stage, it seems to us clear that the portions of the SEC order quoted in the complaint are in the nature of allegations "upon information and belief," which cannot ordinarily form the basis of a fraud claim "except as to matters peculiarly within the opposing party's knowledge." Luce, 802 F.2d at 54 n.1 (citation and internal quotation marks omitted). Even as to the latter, a fraud plaintiff must generally state the facts upon which her belief is founded. Id.

While a complaint that merely recites others' allegations may therefore be insufficient, we are satisfied that in this case Plaintiffs do also allege non-conclusory facts and that these additional factual pleadings are sufficient to render unproblematic any implied reliance on the SEC findings. In particular, the complaint asserts that:

   (1) Wachovia had been preparing two CDO deals in February 2007, one of which became Longshore while the other was canceled;

(2) because Wachovia was the warehouse provider for both deals, assets from the canceled CDO remained on its books;

(3) based on superior insider knowledge, "Wachovia was aware of significant problems in the []MBS sector," J.A. at 137;

(4) despite Defendants' [**49]  representations that all acquisitions for Longshore would be carried out in arm's length-type transactions, the assets from the canceled CDO deal "were sold to Longshore for $4.6 million over their then-current market value," Id. at 137-38; and

(5) this decline in value would otherwise have been borne by Wachovia.

These allegations--albeit clearly overlapping with the SEC order--are made directly by Plaintiffs, and were signed by Plaintiffs' counsel subject to the requirements of Rule 11. See Fed. R. Civ. P. 11(b)(3). Taken together, they are, in our view, adequate to survive a motion to dismiss.

Defendants object that the complaint requires the court to infer their knowledge of a decline in the value of the specific assets transferred into Longshore from broader conditions in the market. That inference, however, seems to us reasonable under the circumstances and, thus, is fair to draw in Plaintiffs' favor at the Rule 12(b)(6) stage. See Nielsen, 746 F.3d at 62. This is not a case in which a securities fraud plaintiff alleges a mere downturn in the market, asking the court to infer defendant's knowledge of the poor performance of a given company. Plaintiffs pleaded scienter by asserting that Wachovia was aware of a high probability that at the time the assets were [**50]  transferred into Longshore, the original acquisition cost no longer represented those assets' fair market value. The basis for that alleged knowledge is not vague; it is Wachovia's status as a major participant in the MBS market, a participant who was, consequently, aware of a broad drop in value in the very class of securities to which the transferred assets belonged.

In sum, while the complaint could be more detailed as to the timeline and valuation of the securities in question, there is enough particularity to withstand Defendants' Rule 12(b)(6) motion. Plaintiffs' direct allegations (1) that Wachovia knew the MBS market well, and knew that it was in decline over the "months" in question, J.A. at 137, and (2) that Wachovia nevertheless transferred assets on its books at their original cost basis while telling investors that SAI would ensure  [*181]  that the transactions were conducted as if between unrelated parties, suffice to raise a plausible inference of material misrepresentation by Wachovia as well as a strong inference of scienter.

2. SAI

In contrast to the allegations regarding Wachovia's role, Plaintiffs' claims regarding SAI's role are less detailed and fail to satisfy the particularity required [**51]  by Rule 9(b). Indeed, although the Longshore scheme differs in kind from the constellation CDOs scheme, as to SAI, the complaint suffers from the same infirmity that we found earlier when dealing with the constellation CDOs--namely, that Plaintiffs have alleged neither a fraudulent misrepresentation by SAI nor, a fortiori, scienter on SAI's part.

In the case of Longshore, the plausibly fraudulent statement in the offering documents is that SAI, as the collateral manager, would "cause any acquisition or sale" of assets "to be conducted on an arm's length basis," or at least "on terms as favorable to [Longshore] as would be the case" in a transaction between unrelated parties." Id. at 136. That statement is clearly attributable to Wachovia, as the author of the offering documents. But as we noted before in the case of the constellation CDOs, the problem here is that Plaintiffs have not adequately pleaded any such misrepresentation by SAI. Cf. supra Section I.B.2. And Plaintiffs' allegations regarding the close relationship between SAI and Wachovia lack particulars from which it would be reasonable to infer that SAI itself misled investors about its role in the acquisition of assets by Longshore.¹²

12   It is [**52]  true that SAI bore responsibility for some information contained in certain sections of the Longshore offering documents. See J.A. at 309, 317-18, 325-28. But Plaintiffs identify--and our own examination reveals--no statement in those sections to the effect that SAI would ensure that all transactions were conducted as if between unrelated parties.

Absent facts plausibly indicating a misrepresentation by SAI, the complaint necessarily lacks allegations giving rise to a strong inference of scienter on SAI's part. Cf. supra Section I.C. Hence, as to SAI, the complaint fails to state a claim in connection with either the constellation CDOs or Longshore. To proceed against SAI, Plaintiffs must replead their claim with sufficient particularity to give rise to the requisite inferences of misrepresentation and scienter.

II

With respect to all three CDOs, Plaintiffs lost millions of dollars. But that loss coincided with the recent financial crisis, which affected large swaths of securities in and beyond the MBS market. For that reason, the question arises whether Defendants' misstatements met the requirements of loss causation as well as of transaction causation.¹³ Loss causation was not  [*182]  pleaded in [**53]  great detail below, but neither was its absence relied on by the district court in dismissing the complaint. Nevertheless, Defendants argue on appeal, as they argued before the district court, that Plaintiffs' failure to plead it is fatal to the complaint.

13   Transaction causation is also known as "reliance" in fraud cases. See Halliburton Co. v. Erica P. John Fund, Inc., 134 S. Ct. 2398, 2418, 189 L. Ed. 2d 339 (2014) ("To prove reliance, the plaintiff must show transaction causation, i.e., that the specific misstatement induced the investor's decision to engage in the transaction." (internal quotation marks omitted)); Dura Pharms., Inc. v. Broudo, 544 U.S. 336, 341, 125 S. Ct. 1627, 161 L. Ed. 2d 577 (2005) (referring to "reliance" as "transaction causation"); ATSI Commc'ns, Inc. v. Shaar Fund, Ltd., 493 F.3d 87, 106 (2d Cir. 2007) (using two terms interchangeably); Eurycleia Partners, 12 N.Y.3d at 559 (referring to requirement as "justifiable reliance"). In the securities fraud context, transaction causation, or reliance, "refers to the causal link between the defendant's misconduct and the plaintiff's decision to buy or sell securities. It is established simply by showing that, but for the claimed misrepresentations or omissions, the plaintiff would not have entered into the detrimental securities transaction." Emergent Capital Inv. Mgmt., LLC v. Stonepath Grp., Inc., 343 F.3d 189, 197 (2d Cir. 2003) (citations omitted).

At the outset of our review of Plaintiffs' pleadings as to loss causation are two unsettled questions: (1) whether this element must be [**54]  pleaded (rather than simply supported by evidence at a later stage) and, if so, (2) with what level of particularity. Although loss causation must certainly be pleaded to state a claim for federal securities fraud, see Lentell v. Merrill Lynch & Co., 396 F.3d 161, 172 (2d Cir. 2005), our Circuit has not had occasion to decide whether a plaintiff making an analogous claim under state common law must likewise plead this element.¹⁴ The Private Securities Litigation Reform Act ("PSLRA"), which codified loss causation as a separate element of federal securities fraud actions, see 15 U.S.C. § 78u-4(b)(4), draws on--but also departs from--the common law. See Dura Pharms., Inc. v. Broudo, 544 U.S. 336, 341, 125 S. Ct. 1627, 161 L. Ed. 2d 577 (2005) ("[A] private damages action [for securities fraud under federal law] resembles, but is not identical to, common-law tort actions for deceit and misrepresentation."). It is thus scarcely clear that pleading is identical in the two contexts.

14   While the substantive elements of common-law fraud that must be proven are a matter of state law, what must be pleaded and with what level of particularity are governed by Rules 9(b) and 12(b)(6). See Rules Enabling Act, 28 U.S.C. § 2072; Hanna v. Plumer, 380 U.S. 460, 472, 85 S. Ct. 1136, 14 L. Ed. 2d 8 (1965) ("[T]he constitutional provision for a federal court system (augmented by the Necessary and Proper Clause) carries with it congressional power to make rules governing the practice and pleading in those courts."). We [**55]  note that the First Department of New York's Appellate Division has sometimes required the pleading of loss causation and held its absence to be grounds for dismissal. See Mosaic Caribe, Ltd. v. AllSettled Grp., Inc., 117 A.D.3d 421, 985 N.Y.S.2d 33, 34 (1st Dep't 2014); Greentech Research LLC v. Wissman, 104 A.D.3d 540, 961 N.Y.S.2d 406, 407 (1st Dep't 2013). We also note, however, that in Plaintiffs' several parallel cases against other defendants in New York state courts, the First Department has recently found allegations similar to those before us sufficient to state a claim for fraud, and that it has done so with nary a mention of loss causation. See Loreley Fin. (Jersey) No. 4 Ltd. v. UBS Ltd., 123 A.D.3d 413, 998 N.Y.S.2d 172 (1st Dep't 2014); Loreley Fin. (Jersey) No. 3 Ltd. v. Citigroup Global Markets Inc., 119 A.D.3d 136, 987 N.Y.S.2d 299 (1st Dep't 2014); Loreley Fin. (Jersey) No. 28, Ltd. v. Merrill Lynch, Pierce, Fenner & Smith Inc., 117 A.D.3d 463, 985 N.Y.S.2d 499 (1st Dep't 2014). Finally, we note that the district court viewed loss causation as an issue typically "reserved for summary judgment," J.A. at 411, and that Defendants agreed with this view of the general course of securities fraud litigation, see id. ("It's usually not a driver on the motion to dismiss."). See also Allstate Ins. Co. v. Stanley, 117 A.D.3d 546, 985 N.Y.S.2d 562, 563 (1st Dep't 2014) ("Defendants' argument that the general collapse of the residential mortgage-backed securities market bars plaintiffs from proving loss causation is not ripe for determination at the pleading stage.").

We recently assumed the existence of a requirement that loss causation be pleaded in a similar common-law case in which loss causation had, in fact, been pleaded in considerable detail. See Fin. Guar. Ins. Co. v. Putnam Advisory Co., LLC, 783 F.3d 395, 402-05 (2d Cir. 2015). We there left open whether [**56]  a plaintiff must plead it with the specificity required by Rule 9(b). Id. at 403. And this second question--as to the level of particularity--is an open  [*183]  one in our Circuit even in the PSLRA context, where the obligation to plead loss causation is well settled. See Acticon AG v. China N.E. Petrol. Holdings Ltd., 692 F.3d 34, 37-38 (2d Cir. 2012); Wilamowsky v. Take-Two Interactive Software, Inc., 818 F. Supp. 2d 744, 753 & n.7 (S.D.N.Y. 2011) ("The question of whether Rule 9(b) applies to loss causation has not yet been definitively addressed by the Second Circuit, but the vast majority of courts in this district have required that loss causation only meet the notice requirements of Rule 8."); see also Dura, 544 U.S. at 346 (assuming that ordinary notice pleading applies).

Since, however, we find the pleadings of loss causation here--while much less detailed than those in Financial Guaranty--to be sufficient at this preliminary stage regardless of the applicable pleading standard, we need not decide these questions today. Yet the meaning of loss causation remains a source of much misunderstanding. And that perplexity warrants a review of this element before we turn to the specifics of this case and explain our view that Plaintiffs' admittedly slim pleadings on the subject nevertheless are sufficient.

A. The Nature of Loss Causation

Loss causation has long been a requirement in securities and other fraud cases. See [**57]  Schlick v. Penn-Dixie Cement Corp., 507 F.2d 374, 380 (2d Cir. 1974); Bastian v. Petren Res. Corp., 892 F.2d 680, 683 (7th Cir. 1990) (Posner, J.) ("Indeed what securities lawyers call 'loss causation' is the standard common law fraud rule, merely borrowed for use in federal securities fraud cases." (parenthetical citation omitted)); Moore v. PaineWebber, Inc., 189 F.3d 165, 174 (2d Cir. 1999) (Calabresi, J., concurring) (RICO fraud). What loss causation is, however, has not always been expressed with great precision and clarity. Perhaps the best way to describe it is negatively, by adducing a few examples of its absence.

Take the classic torts case of Berry v. Sugar Notch Burrough, 191 Pa. 345, 43 A. 240 (1899), in which a negligently speeding trolley car is damaged by a falling tree. The wrongfulness--the speeding--is a but-for cause of the accident and injury: had the trolley car not been speeding, it would have been elsewhere when the tree fell. As a general matter, though, and apart from the chance occurrence in this case, speeding does not make it likelier that trees will fall on trolley cars. Indeed, speeding arguably reduces the likelihood of such accidents by reducing the amount of time that one is under any given tree. But-for cause is present; causal link or tendency is not. (It would, of course, be different if one could demonstrate that speeding trolley cars create vibrations that lead damaged trees to fall with greater [**58]  frequency. In that case, a causal relation could be said to exist between the speeding and the injury. See Guido Calabresi, Concerning Cause and the Law of Torts, 43 U. Chi. L. Rev. 69, 72 (1975).)

The requirements of transaction and loss causation are exactly analogous to but-for cause and causal tendency in this classic torts case. Suppose a real estate company misrepresents that a certain house belonged to Abraham Lincoln, and a buyer purchases the house because of this. Suppose the buyer can show that, absent the lie, she would not have bought the house and would instead have bought a house in another part of town. Subsequently, a flood destroys her house and others in the neighborhood, while leaving the "other part of town" unscathed. The loss to her would not have occurred but for the fraud. And yet, so long as the neighborhood was not more prone to flooding,¹⁵ the lie in no way increased the chances of the actual  [*184]  damage to her house. Transaction causation is present; loss causation is not. (To be sure, if the buyer paid a premium for Lincoln's house, that premium--i.e., the amount above the fair market value of the house in light of its true, non-presidential provenance--represents a separate harm whose causal [**59]  linkage to the lie is evident.¹⁶)

15   If the neighborhood turns out to be more prone to flooding than surrounding areas, then inducing people to buy in that neighborhood by lying (whether about Lincoln, the school system, or the drainage basin) actually increases the flood risk to these homebuyers, even if the liar is oblivious to this increased risk. As far as loss causation is concerned, the question is whether the fraud creates a risk that materialized in the ultimate harm to the plaintiff and, if repeated, would create a similar risk to others in the future. See Zuchowicz v. United States, 140 F.3d 381, 388 n.7 (2d Cir. 1998) (describing test as whether wrongful activity "enhanced (at the time the defendant acted) the chances of the harm occurring or . . . increase[d] the chances of a similar accident in the future if the defendant should repeat the same wrong"). Where such a risk exists--regardless of whether it is only discovered after the fact--there is a cost that is properly deemed a cost of the misstatement. Of course, how we wish to allocate that cost is a separate question, the answer to which depends on a host of factors having to do with what is normally analyzed under the rubric of "proximate cause." See infra note 18. For example, even if the neighborhood [**60]  is more prone to flooding, we may wish to allocate the cost of flood damage not to all lying brokers who attract buyers to this neighborhood but only to those who, say, lie about the drainage basin.

16   Relevant parallels can be drawn from the so-called "darting out" cases, in which speeding drivers hit children who run into the street from behind an object that blocks the driver's view. First, while the speeding is a but-for cause--but for the velocity the car would have been elsewhere--causal tendency is lacking unless a driver in the same position going only the posted speed limit could have avoided the collision. See Gulf Oil Corp. v. Reed, 334 F.2d 960, 963, 118 U.S. App. D.C. 212 (D.C. Cir. 1964); Underwood v. Fultz, 1958 OK 210, 331 P.2d 375, 378-79 (Okla. 1958); 4 Fowler V. Harper, Fleming James, Jr., & Oscar S. Gray, The Law of Torts § 20.5, at 165-66 & n.48 (2d ed. 1986); see id. at 80 (cumulative supplement 2004). Second, although speeding may have no bearing on the risk of the accident, to the extent the child was injured more than would have been the case had the driver been going more slowly, the unlawful velocity clearly bears on the additional injury. The risk of greater injury thus represents a cost of speeding that we may, if we so wish, allocate to the driver in tort.

Some authors have treated loss causation as if it were part of [**61]  proximate cause, whereas transaction causation is treated as part of cause-in-fact. See Jill E. Fisch, Cause for Concern: Causation and Federal Securities Fraud, 94 Iowa L. Rev. 811, 816-17 (2009) ("Subsequent courts have analogized loss causation to proximate or legal cause, while they analogize transaction causation to 'but-for' or factual cause."); Id. at 817 n.23 (citing examples).¹⁷ In truth, proximate cause is separate, and transaction causation and loss causation alike, to the extent that they are required in a given case, are each subject to the further requirement of proximity. Lack of proximity between the wrongful activity and the transaction will preclude liability, just as liability will not attach if the causal tendency--the thing that increases the chances of the actual occurrence of the harm--is too remote. In sum, proximity applies to both transaction and loss causation, and for both elements the additional element of proximity must be present before liability may be imposed. It is only because some of the considerations  [*185]  relevant to showing proximate cause (nature of the risk and existence of intervening causes) typically pertain to loss that courts frequently, albeit unnecessarily, analyze loss causation under the rubric [**62]  of proximate cause.¹⁸

17   Loss causation has also been described as the conclusion that all the requirements of common-law causation--but-for cause, causal link, and proximate cause--are present with respect to the loss. See Moore v. PaineWebber, Inc., 189 F.3d 165, 175 (2d Cir. 1999) (Calabresi, J., concurring). We believe it clearer to analogize loss causation to causal link or tendency, as we have done here, and to analyze it separately from both but-for and proximate cause.

18   The conflation of loss causation with proximate cause appears to result, in many cases, from the conflation of each, in turn, with foreseeability. But, whereas foreseeability is to some extent relevant to proximate cause, it is "aftseeability" that matters for loss causation. That is, to return to the Berry case, if speeding trolley cars can be shown to create vibrations that increase the chance of rotten branches falling, then causal tendency (loss causation) is present, even though the link may not have been anticipated and may only have been shown to exist after the fact. Foreseeability, on the other hand, goes partly to how strong or weak, near or remote, we take the causal connection to be and ultimately, then, to the question of whether we wish to hold those speeding liable [**63]  for an increased likelihood of falling branches damaging passing trolley cars. The unforeseeability of a particular risk, which may indicate the relative weakness of a causal connection, is among the factors courts consider under "proximate cause," but a decision not to impose liability due to lack of proximity does not negate the presence, in fact, of causal tendency. It may be that this is what Justice Loevinger meant when--conflating causal tendency with proximate cause, as was commonly done at the time--he wrote, "It is enough to say that negligence is tested by foresight but proximate cause is determined by hindsight." Dellwo v. Pearson, 259 Minn. 452, 456, 107 N.W.2d 859 (1961).

Some writers have also suggested that while the requirement of transaction causation (but-for cause) may occasionally be waived, as in cases of multiple or statistical causation, loss causation (causal tendency) is virtually always required. This is not so. See Calabresi, supra, at 100-01 & n.49. Both requirements are based, as much of law is, on policy grounds. There are jurisdictions that choose not to require loss causation in suits based on fraud. See Jane Stapleton, Benefits of Comparative Tort Reasoning: Lost in Translation, 1 J. Tort L., no. 3, 2007, at 2 ("[I]n the U.S. it seems to be [**64]  a principle that a defendant in the tort of deceit cannot be liable for coincidental consequences; but that principle is rejected in England.").

Additionally, in certain areas of the law, the requirement of loss causation is eliminated either by contract or by statute. To take a common example from the law of insurance, a decedent's estate may be denied payment on her life insurance policy if she lied about her health on her application, even if the risk of the specific illness or accident that killed her was totally unaffected by the undisclosed ailment. See Dormer v. Nw. Mut. Life Ins. Co., 408 F. App'x 452, 454 (2d Cir. 2011). The occasional "deviation" case that does not require causal tendency can be explained in similar fashion: a carrier is held liable for the destruction of goods in transit, having shipped them by a route other than that "specified in the contract or reasonably within the contemplation of the parties," even though the destructive event was not made likelier by the choice of route. Green-Wheeler Shoe Co. v. Chicago, R.I. & P.R. Co., 130 Iowa 123, 106 N.W. 498 (1906) (applying this approach to a shipping delay).

The relaxed requirements in suits brought under the Federal Employers' Liability Act bear a family resemblance to these cases, and indeed have been criticized on this very ground, as a departure from the [**65]  common-law requirement of causal link. See Gallick v. Baltimore & Ohio R.R. Co., 372 U.S. 108, 126-27, 83 S. Ct. 659, 9 L. Ed. 2d 618 (1963) (Stewart & Goldberg, JJ., dissenting); CSX Transp., Inc. v. McBride, 131 S. Ct. 2630, 2645, 180 L. Ed. 2d 637 (2011) (Roberts, C.J., dissenting) ("The test the Court would substitute--whether negligence  [*186]  played any part, even the slightest, in producing the injury--is no limit at all. It is simply 'but for' causation.").

Where does all of this leave us in the present case? In the securities fraud context in general, an investor may buy shares of a certain stock because her broker falsified--or neglected to mention--some detail but then suffer a loss due to a nationwide recession. Loss causation is lacking unless the fraudulent statement that induced her to invest can also be shown to have made her investment, in fact, more disposed to suffer the alleged harm--a catastrophic market collapse--than honestly described alternative investments. See Powers v. British Vita, P.L.C., 57 F.3d 176, 189 (2d Cir. 1995).

B. Plaintiffs' Loss Pleadings

Here, Plaintiffs clearly allege transaction causation, i.e., that they would not have invested in the three CDOs but for Defendants' misrepresentations.¹⁹ They also allege that the CDOs were all in some way designed to fail and did fail. Specifically, Plaintiffs assert that contrary to the picture painted in the offering documents of purportedly independent [**66]  collateral managers selecting high-quality assets for the benefit of long investors, toxic assets were purposefully chosen for each CDO--in the case of the constellation CDOs, in order to advance Magnetar's long-short strategy; in the case of Longshore, in order to offload these assets from Wachovia's own books. Defendants contend, however, that the subsequent market crash was of such dramatic proportions that Plaintiffs' losses would have occurred at the same time and to the same extent regardless of the alleged fraud.

19   As with loss causation, transaction causation was not discussed by the district court, but Defendants argue that failure to plead it is similarly fatal here. We do not address this element separately because (1) Plaintiffs' pleadings as to transaction causation seem to us plainly adequate on their face, and (2) the possible defense that any such reliance was unreasonable as a matter of law is premature based solely on the complaint and incorporated documents.

In general, the reasonableness of a plaintiff's reliance is a "nettlesome" and "fact-intensive" question, Schlaifer Nance & Co. v. Estate of Warhol, 119 F.3d 91, 98 (2d Cir. 1997), which we, like our Circuit's many district courts, will not lightly dispose of at the motion-to-dismiss stage. [**67]  See, e.g., Bayerische Landesbank, New York Branch v. Barclays Capital, Inc., 902 F. Supp. 2d 471, 474 (S.D.N.Y. 2012) ("Whether or not reliance on alleged misrepresentations is reasonable in the context of a particular case is intensely fact-specific and generally considered inappropriate for determination on a motion to dismiss." (internal quotation marks omitted)); Robinson v. Deutsche Bank Tr. Co. Americas, 572 F. Supp. 2d 319, 322-23 (S.D.N.Y. 2008) (citing several cases to the same effect).

Defendants argue, however, that Plaintiffs' reliance was unreasonable as a matter of law because of disclaimers in the offering circulars and the fact that Plaintiffs were sophisticated investors. Yet the offering circulars' disclaimers are so general that we cannot be confident they bear on the misrepresentations alleged in the complaint, much less that they render Plaintiffs' fraud claim implausible under Rule 12(b)(6). See Caiola v. Citibank, N.A., New York, 295 F.3d 312, 330 (2d Cir. 2002) ("[A] valid disclaimer provision must contain explicit disclaimers of the particular representations that form the basis of the fraud claim." (internal quotation marks omitted) (emphasis added)); cf. Citigroup Global Markets, 987 N.Y.S.2d at 304-06. Nor is it obvious that ordinary due diligence by sophisticated investors would uncover the sort of scheme at issue, which is alleged to have involved deliberate concealment of facts known only to Magnetar and Defendants. Cf. Steinhardt Grp. Inc. v. Citicorp, 272 A.D.2d 255, 708 N.Y.S.2d 91, 93 (1st Dep't).

Lack of reasonable reliance may ultimately prove [**68]  to be a successful defense. It does not, however, warrant dismissal at the pleading stage in this case.

If Defendants are right, and the alleged fraud in no way increased the chance of  [*187]  Plaintiffs' ultimate losses, then loss causation is lacking. Fraud claims under New York law being an area in which loss causation is required, see Greentech Research LLC v. Wissman, 104 A.D.3d 540, 961 N.Y.S.2d 406, 407 (1st Dep't 2013), Plaintiffs must ultimately prove it; that is, they bear the burden of convincing the finder of facts that this element was present. But the question at this preliminary stage is whether the pleadings suffice to withstand Defendants' Rule 12(b)(6) challenge. We conclude that they do.

Assuming both that loss causation must be pleaded in fraud actions brought under state common law, as in federal securities fraud actions, and that the two pleading requirements are similar, we note that Plaintiffs' burden is not a heavy one. See Dura, 544 U.S. at 347. The complaint must simply give Defendants "some indication" of the actual loss suffered and of a plausible causal link between that loss and the alleged misrepresentations. Id.; see Fin. Guar., 783 F.3d at 404.

What sort of pleading is sufficient will depend on the factual circumstances of the case. See Lentell, 396 F.3d at 174 ("Loss causation is a fact-based inquiry and the degree of difficulty [**69]  in pleading will be affected by circumstances. . . ."). In the federal securities fraud context, we have held that "when the plaintiff's loss coincides with a marketwide phenomenon causing comparable losses to other investors," Id. (internal quotation marks omitted), the plaintiff may be required to plead facts from which it would be reasonable to infer that the risks which materialized in her loss were risks concealed by the fraud rather than risks evident on the face of the investment disclosures. see id. at 172-78. At the same time, we have observed that where the question, at bottom, is one of intervening events--a consideration properly analyzed under proximate cause--"the chain of causation is a matter of proof at trial and not to be decided on a Rule 12(b)(6) motion to dismiss." Id. at 174 (citation, ellipsis, and internal quotation marks omitted).

We think it possible to distinguish here three broad types of fraud complaints. Our list is not meant to exhaust the possibilities but only to illustrate considerations that a court may deem relevant at different stages of the litigation. To adapt the prior real estate example, assume three houses are destroyed in an earthquake. The first is the one already encountered--the [**70]  house that Lincoln was falsely said to have owned. This situation is most clearly akin to the tree and speeding trolley car in Berry. The causal connection, if any exists, is not readily apparent. Hence, in assessing the facial plausibility of the claim under Rule 12(b)(6), a court may be unwilling to permit the complainant to proceed without additional facts from which it would be reasonable to infer that the defendant's wrongful activity in some way increased the likelihood of the eventual harm.²⁰

20   However one feels about loss causation, a plaintiff must ultimately prove this element to prevail on a fraud claim under New York law. See Wissman, 961 N.Y.S.2d at 407. One may, of course, wish to hold liars liable regardless of whether the type of lie at issue imposed on the plaintiff an increased risk of the harm suffered or would impose on similarly situated others such a risk in the future. To do so, however, would require us to ignore the substantive law of New York and to dispense with one of its common-law requirements. Cf. supra discussion in Section II.A (giving examples of contractual and statutory contexts in which loss causation is not required).

The second and third types of fraud involve, in contrast, misrepresentations [**71]  as to the solidity of the house and thus differ  [*188]  in kind from the first in a way that bears directly on loss causation in the present case. In the second type of fraud complaint, the misrepresentation is that the house was well built. In the third--a further variation of the second--the buyer is told that the house was, in fact, "earthquake proof." It may turn out that shoddy and well-built houses alike were destroyed in the earthquake. It may even be that the earthquake was so vast as to destroy houses that were "earthquake proof," as "earthquake proof" is normally defined. Cf. Bastian, 892 F.2d at 685-86 (describing hypothetical loss on investment represented by broker to be "risk-free"). In both such situations, however, it is enough at the pleading stage for the plaintiff to allege the particular misrepresentation and the destruction of the house in an earthquake. Cf. Fisch, supra, at 852 n.223 ("Similarly, it is plausible that companies involved in fraud are especially susceptible to ruinous harm upon the occurrence of adverse economic events. Indeed, many of the companies that collapsed due to the bursting of the dot-com bubble--as opposed to weathering it--were those engaged in financial accounting manipulations and similar practices."). [**72]  It then falls to the defendant to proffer facts indicating that a well-built house, or even an earthquake-proof one, would have been destroyed in this earthquake.

What we are here describing is the burden of pleading, which has clearly been met by a plaintiff's allegation of a misrepresentation that goes to how well the house was built, as well as the burden of introducing some counterevidence (onus procedendi), which has shifted to the defendant. Cf. Liriano v. Hobart Corp., 170 F.3d 264, 272 (2d Cir. 1999) ("This shifting of the onus procedendi has long been established in New York."). The essential point for present purposes is that where a potential causal link is evident, it is not necessary for the plaintiff to plead loss causation in detail to render her fraud claim plausible at the motion-to-dismiss stage.

The instant suit lies somewhere between these second and third cases in which the sufficiency of the loss causation pleadings follows from the nature of the fraud alleged and the harm suffered. In that way, this case fundamentally differs from the speeding trolley car and the Lincoln house cases, and the issue of causal link is consequently distinctly easier. Moreover, as the difference between the second and third is a matter of [**73]  degree, or of the relative strength of causal tendency inferable from the type of fraud and surrounding circumstances, that difference will principally be relevant at later phases of the litigation, e.g., on a motion for summary judgment, in the court's instruction on causation to the jury, or in a post-trial challenge to the sufficiency of the evidence.

Here, while Defendants by no means represented the CDOs to be free of risk, they represented the CDOs to be more than simply "well built," and far from what Plaintiffs claim the CDOs actually were: designed to fail. Indeed, according to the complaint, Defendants hid from investors (1) that Magnetar was actively undermining the constellation CDOs by selecting marginal collateral to capitalize on eventual defaults and (2) that Wachovia was dumping into Longshore worsening assets at their original cost to recoup the losses and pass the risk to investors. Whether Plaintiffs can prove these allegations--and whether defendants in turn can proffer evidence that the CDOs would have collapsed regardless, due to the larger crash in the MBS market--are evidentiary matters for later phases of this lawsuit. It is sufficient under Rule 12(b)(6) that the allegations [**74]  themselves give Defendants "some indication" of the risk concealed by the  [*189]  misrepresentations that plausibly materialized in Plaintiffs' ultimately worthless multimillion-dollar investment in these CDO notes. Fin. Guar., 783 F.3d at 404 ("The purpose of the loss causation element is to require a plaintiff 'to provide a defendant with some indication of the loss and the causal connection that the plaintiff has in mind,' not to make a conclusive proof of that causal link." (quoting Dura, 544 U.S. at 347)).

While Plaintiffs did not plead that the alleged fraud caused their losses independently of the larger financial events of 2007 and 2008, they were not required to do so under our precedents. The requirement, if any, to plead a causal link does not place on Plaintiffs a further pleading obligation to rule out other contributing factors or alternative causal explanations. see id. ("Nor is [the plaintiff] required to allege that its losses were caused solely by [the defendant's] misrepresentations . . . ."); King Cnty., Wash. v. IKB Deutsche Industriebank AG, 708 F. Supp. 2d 334, 342 (S.D.N.Y. 2010) ("Neither Dura nor Lentell . . . imposes on plaintiffs the heavy burden of pleading facts sufficient to exclude other non-fraud explanations." (internal quotation marks omitted)). Plaintiffs must only allege enough facts regarding [**75]  their loss to support the inference that they "would have been spared all or an ascertainable portion of that loss absent the fraud." Lentell, 396 F.3d at 175.²¹ We are satisfied that Plaintiffs have done so in this case.

21   Defendants' argument to the contrary misreads our decision in Lentell. See 396 F.3d at 172-78. The plaintiffs there invested in internet companies at the height of the dot-com bubble on the strength of bullish ratings by brokers, and then sued the brokers when the bubble burst and stock prices plummeted. The complaint, however, lacked any allegations attributing the drop in price either to a subsequent disclosure correcting the ratings or to a "risk of price volatility" not already apparent on the face of the market analysis published by the brokers concurrently with their ratings. Id. at 176. To put Lentell in real estate terms, the brokers were alleged to have misrepresented the house to be "worth buying," even as a detailed prospectus given to buyers disclosed a risk of earthquakes. The brokers' professed, but allegedly insincere belief in their ratings did not conceal the otherwise manifest risk. In that way, the opinion that the house is "worth buying" also clearly differs from the misrepresentation that a given [**76]  house is "well built" because "worth buying," at least without additional context, implies no knowledge of facts in manifest conflict with a risk of earthquake damage. Cf. Omnicare, Inc. v. Laborers Dist. Council Const. Indus. Pension Fund, 135 S. Ct. 1318, 1328-32, 191 L. Ed. 2d 253 (2015).

While, unlike Lentell, the present fraud action was brought under state common law, our assumption here and in Financial Guaranty that a plaintiff must give fair notice of the causal link between the alleged fraud and her injury is in full accord with our holding in Lentell that federal securities complaints lacking such notice may be dismissed for failure to state a claim. See Lentell, 396 F.3d at 176 ("[P]laintiffs allege no loss resulting from the market's realization that the opinions were false, or that [the defendant] concealed any risk that could plausibly . . . have caused plaintiffs' loss."). As we noted in Lentell, questions of causation are often complex, and where the issue is the chain of causation, that issue is not appropriate for resolution on a motion to dismiss. Id. at 174. The difficulties of causation do not, however, relieve fraud plaintiffs of the obligation to satisfy Rule 12(b)(6)'s more general plausibility standard.

III

Although, for the reasons discussed in Part I, supra, we disagree with several of the district court's conclusions as to the sufficiency [**77]  of Plaintiffs' fraud allegations, we hold--separate and apart from these errors--that the district court exceeded the bounds of its discretion in denying Plaintiffs leave to amend their complaint.  [*190]  An amended complaint, moreover, may cure the remaining defects as to SAI that we noted above. See supra Sections I.B.2, I.C, I.D.2.

In so holding, we hew to the liberal standard set forth in Rule 15, which states that "[t]he court should freely give leave [to amend] when justice so requires." Fed. R. Civ. P. 15(a)(2). As we have explained, the "permissive standard" of Rule 15 "is consistent with our strong preference for resolving disputes on the merits." Williams v. Citigroup Inc., 659 F.3d 208, 212-13 (2d Cir. 2011) (per curiam) (internal quotation marks omitted); see also Foman v. Davis, 371 U.S. 178, 182, 83 S. Ct. 227, 9 L. Ed. 2d 222 (1962) ("If the underlying facts or circumstances relied upon by a plaintiff may be a proper subject of relief, he ought to be afforded an opportunity to test his claim on the merits.").

Here, the procedure by which the district court denied leave to amend was improper. The court required the parties to attend a pre-motion conference and to exchange, in preparation, letters of no more than three pages regarding Defendants' anticipated motion to dismiss for failure to state a claim. The Federal Rules of Civil Procedure do not speak [**78]  to the use of pre-motion conferences. Such conferences are not in themselves problematic, however, and indeed may in many instances efficiently narrow and/or resolve open issues, obviating the need for litigants to incur the cost of more extensive filings. The impropriety occurred not when the district court held the pre-motion conference but when, in the course of the conference, it presented Plaintiffs with a Hobson's choice: agree to cure deficiencies not yet fully briefed and decided or forfeit the opportunity to replead. Without the benefit of a ruling, many a plaintiff will not see the necessity of amendment or be in a position to weigh the practicality and possible means of curing specific deficiencies.

Our opinion today, of course, leaves unaltered the grounds on which denial of leave to amend has long been held proper, such as undue delay, bad faith, dilatory motive, and futility--none of which were a basis for the denial here. See Foman, 371 U.S. at 182; Williams, 659 F.3d at 213-14. No improper purpose is alleged. And while leave may be denied where amendment would be futile, Williams, 659 F.3d at 214, the approach taken by the district court was not rooted in futility. Rather, the court treated Plaintiffs' decision to stand by the complaint after [**79]  a preview of Defendants' arguments--in the critical absence of a definitive ruling--as a forfeiture of the protections afforded by Rule 15. This was, in our view, premature and inconsistent with the course of litigation prescribed by the Federal Rules.

Defendants argue on appeal that the denial of leave was proper because of the informality of the request, which was raised "in the alternative" at the end of Plaintiffs' brief opposing the motion to dismiss. J.A. at 584. Generally, we will not deem a request for leave to amend insufficient on the basis of form alone. See Porat v. Lincoln Towers Cmty. Ass'n, 464 F.3d 274, 276 (2d Cir. 2006) ("[A] lack of a formal motion is not a sufficient ground for a district court to dismiss without leave to amend." (citing Oliver Sch., Inc. v. Foley, 930 F.2d 248, 252-53 (2d Cir. 1991)). Denial of leave might be proper where a plaintiff's request was inconspicuous and never brought to the court's attention, see In re Tamoxifen Citrate Antitrust Litig., 466 F.3d 187, 220 (2d Cir. 2006), abrogated on other grounds by F.T.C. v. Actavis, Inc., 133 S. Ct. 2223, 186 L. Ed. 2d 343 (2013), or where the request gives no clue as to "how the complaint's defects would be cured," Porat, 464 F.3d at 276. Even in those situations, however,  [*191]  our analysis is directed at other underlying issues like notice and futility. See Tamoxifen, 466 F.3d at 220. Here, both the district court and Defendants were aware of Plaintiffs' request, and at the time it was made, as at the time of the pre-motion [**80]  conference, there was no decision yet as to the precise defects, such that a proposed cure would have been in order as part of the request to amend. Given the posture of the case, Plaintiffs were not obliged to formulate their request differently. Moreover, since the district court based its denial on Plaintiffs' timing, see Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *16 n.13, we doubt that a more formal motion for leave to amend would have fared any differently.

The present case combines a complex commercial reality with a long, multi-prong complaint. In such situations, pleading defects may not only be latent, and easily missed or misperceived without full briefing and judicial resolution; they may also be borderline, and hence subject to reasonable dispute. As discussed in Part I, supra, dismissal was partly based on the district court's determination that Plaintiffs' fraud allegations raised neither plausible inferences of material misrepresentations nor strong inferences of scienter. 2013 U.S. Dist. LEXIS 49665, [WL] at *10-15. These determinations entail judgment calls on which reasonable minds can differ in a not insignificant number of cases. Cf. Iqbal, 556 U.S. at 679 ("Determining whether a complaint states a plausible claim for relief will . . . be a context-specific task that requires the reviewing [**81]  court to draw on its judicial experience and common sense."). The district court's rejection of Plaintiffs' position that the strength-of-inference requirements had been met by the facts set forth in the original complaint was, without more, insufficient reason to bar Plaintiffs from repleading. Indeed, that our opinion today partially vindicates Plaintiffs' position is, we think, some measure of the potential for reasonable disagreement here.

Yet even were we not persuaded that--at least as to some elements--the district court erred in its dismissal under Rule 12(b)(6), its denial of leave to amend would still, under the present circumstances, necessitate our remanding the case with instructions to grant the requested leave. As to SAI, there may remain hurdles in pleading fraud that Plaintiffs cannot clear. By not giving Plaintiffs a chance to replead, however, the district court not only "violated the liberal spirit of Rule 15," Williams, 659 F.3d at 214; it also skirted the key issue of futility. At this stage, the district court has not identified any issue as to which (i) the complaint is deficient and (ii) amendment would be futile. Neither, on the record before us, do we identify any such issue.

Accordingly, we instruct the [**82]  district court to grant Plaintiffs leave to amend their complaint in light of the remaining pleading defects as to SAI identified by our opinion today. It is too early for us to say whether amendment will cure these defects. That question should be addressed to the district court in the first instance. That court denied Plaintiffs the opportunity to demonstrate that their claims deserve to be decided on the merits, and Plaintiffs should be given that opportunity now.²²

22   Plaintiffs' complaint asserts--in addition to fraud--several related claims: rescission of contract, conspiracy to defraud, aiding and abetting fraud, fraudulent conveyance, and unjust enrichment. The district court dismissed the rescission, conspiracy, and aiding and abetting claims for "fail[ure] to state an underlying claim for fraud." Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *15. Our opinion today holds, on the contrary, that Plaintiffs have adequately pleaded fraud except as to SAI. Because the adequacy of the pleadings as to the claims directly predicated on fraud (rescission of contract, conspiracy to defraud, and aiding and abetting fraud) was not well briefed on appeal, and because we are already remanding to the district court with instructions to grant Plaintiffs [**83]  permission to replead the fraud claim as to SAI, we decline to address whether dismissal was proper on some alternative ground and leave it to the district court to reassess, in light of our fraud holding, the viability of these claims on the basis of fuller briefing and an amended complaint.

We note that in parallel suits filed by Plaintiffs in New York state court against other defendants based on allegations similar to those in the complaint before us, the First Department of the Appellate Division has uniformly permitted the cases to go forward on the fraud claims, but not on many of the related causes of action. See UBS Ltd., 998 N.Y.S.2d at 173 ("Because plaintiffs are only limited-recourse creditors, their fraudulent conveyance claim was properly dismissed."); Citigroup Global Markets, 987 N.Y.S.2d at 307 ("We find . . . that the unjust enrichment cause of action should have been dismissed because the CDO transactions were governed by written agreements."); Merrill Lynch, 985 N.Y.S.2d at 504 (same). For much the same reasons given by the First Department, as well as the district court below, see Wells Fargo, 2013 U.S. Dist. LEXIS 49665, 2013 WL 1294668, at *15-16, we find Plaintiffs' fraudulent conveyance and unjust enrichment claims to be inadequately pleaded at present. We see no harm, however, in permitting Plaintiffs to replead these claims as [**84]  well, and we see much wisdom in having the district court determine in the first instance--with the benefit of an amended complaint--(1) whether the repleaded claims fail on the same or other grounds and, if so, (2) whether further amendment would be futile.

CONCLUSION

For the foregoing reasons, we REVERSE so much of the district court's  [*192]  judgment as concerns the fraud claim against Wachovia and Harding; we VACATE the judgment as to the remainder of Plaintiffs' claims; and we REMAND the case for further proceedings consistent with this opinion.