From Casetext: Smarter Legal Research

Fin. Guar. Ins. Co. v. Putnam Advisory Co.

Sep 14, 2020
12-cv-7372 (LJL) (S.D.N.Y. Sep. 14, 2020)


12-cv-7372 (LJL)





Plaintiff asserts that Defendant committed fraud, negligent misrepresentation, and negligence in its capacity as the collateral manager for a collateralized debt obligation. For the reasons that follow, the Court holds that Defendant prevails on all counts.


1. Background

1. On October 1, 2012, Financial Guaranty Insurance Company ("FGIC" or "Plaintiff") brought this action against The Putnam Advisory Company, LLC ("Putnam" or "Defendant"). Dkt. No. 1 ("Complaint"). The Complaint alleged that Putnam, as the collateral manager of a complex financial product named Pyxis ABS CDO 2006-1 ("Pyxis"), intentionally and/or negligently made false and misleading statements that defrauded FGIC into participating in Pyxis. Id. In particular, the Complaint alleged that Putnam had falsely informed FGIC that "it—and it alone—would select the collateral for [Pyxis] and that it would do so acting independently and in good faith in the interests of long investors (i.e., investors who profit when the investment performs as designed and succeeds)," when, "[i]n fact," Putnam "allowed the collateral selection process to be controlled by Magnetar Capital LLC ('Magnetar'), a hedge fund manager with a significant net short investment in Pyxis—i.e., an investment that would pay off when Pyxis defaulted." Id. ¶ 1. The Complaint also alleged that Putnam communicated a falsely inflated credit quality of targeted assets for the portfolio and, on August 9, 2006, "provided" FGIC "an updated target portfolio" showing "that at least $145 million of the Pyxis Portfolio would be prime RMBS assets," when "ultimately there were no prime RMBS assets in the Pyxis Portfolio." Id. ¶¶ 72, 76, 77.

2. The Honorable Robert W. Sweet was assigned to the case. Dkt. No. 2.

3. On November 16, 2012, FGIC filed an amended complaint. Dkt. No. 4 ("Amended Complaint"). The Amended Complaint asserted the same claims and made the same allegation regarding the August 9, 2006 "updated target portfolio." Id. ¶ 77.

4. Putnam moved to dismiss the Amended Complaint for failure to state a claim upon which relief could be granted. Dkt. No. 9. Judge Sweet granted that motion with leave to replead. Fin. Guar. Ins. Co. v. Putnam Advisory Co. ("Putnam I"), 2013 WL 5230818 (S.D.N.Y. Sept. 10, 2013). Judge Sweet held that FGIC had failed to adequately plead loss causation to support its fraud claim because FGIC had not buttressed its assertions "with facts sufficient to demonstrate that there was any pool of collateral that could have avoided default while still conforming to Pyxis' detailed eligibility criteria." Id. at *3. Judge Sweet held that FGIC failed to state a claim for negligent misrepresentation because it had "failed to allege facts showing the existence of a prior relationship with Putnam that was so special and close in nature so as to be 'approaching that of privity,'" as required under New York law. Putnam I, 2013 WL 5230818, at *5 (S.D.N.Y. Sept. 10, 2013) (quoting Anschutz Corp. v. Merrill Lynch & Co., 690 F.3d 98, 108 (2d Cir. 2012)).

5. FGIC filed its second amended complaint on September 30, 2013, alleging the same claims. Dkt. No. 22 ("Second Amended Complaint"). The Second Amended Complaint made "new allegations" about the purported "special relationship" between FGIC and Putnam, "including references to a meeting at Putnam's offices and phone interviews between Putnam and FGIC." Fin. Guar. Ins. Co. v. Putnam Advisory Co., LLC ("Putnam II"), 2014 WL 1678912 (S.D.N.Y. Apr. 28, 2014). Like its predecessors, the Second Amended Complaint included the same allegation about the August 9, 2006 "updated target portfolio." Dkt. No. 22 ¶ 79. Putnam moved to dismiss the Second Amended Complaint, Dkt. No. 23, and Judge Sweet again granted that motion. Putnam II, 2014 WL 1678912. Judge Sweet concluded that FGIC again had failed to plead loss causation to support the fraud claim—in other words, FGIC had not stated facts sufficient to support that "the collateral selection process for Pyxis caused FGIC's losses, as opposed to the global financial crisis." Id. at *10. And, "[a]s with the [original complaint]," Judge Sweet concluded that FGIC had failed to allege "facts sufficient to infer that there was any pool of collateral that could have avoided default while still conforming to Pyxis' detailed eligibility criteria." Id. at *12. Judge Sweet further held that FGIC had failed to adequately plead the "special relationship" required to sustain its claim for negligent misrepresentation. Id. at *13. FGIC appealed that decision. Dkt. No. 35.

6. The Second Circuit vacated Judge Sweet's dismissal order and remanded all of FGIC's claims for further proceedings. Fin. Guar. Ins. Co. v. Putnam Advisory Co., LLC ("Putnam III"), 783 F.3d 395 (2d Cir. 2015). The Second Circuit held that FGIC had "alleged particular facts that, when considered as a whole, plausibly allege[d] that Putnam's alleged misrepresentations and omissions caused at least some of the economic harm it suffered." Id. at 403. Rejecting the district court's standard as overly stringent at the motion to dismiss stage, the Second Circuit explained that FGIC was "not required to establish that the collateral it ha[d] identified as selected by Magnetar was the exclusive cause of its losses; rather, it need[ed] only [to] allege sufficient facts to raise a reasonable inference that Magnetar's overall involvement caused an ascertainable portion of its loss." Id. at 404. On negligent misrepresentation, while acknowledging the "undisputed" fact that "there was no actual contractual privity between FGIC and Putnam," the Second Circuit held that FGIC had "plausibly allege[d] facts evincing Putnam's understanding that FGIC would 'rely on [Putnam's] care and competence in managing' the Pyxis portfolio," thereby leaving open the possibility that the requisite "special relationship" could be proven. Id. at 405-407 (quoting Bayerische Landesbank v. Aladdin Capital Mgmt. LLC, 692 F.3d 42, 59-61 (2d Cir. 2012)); see id. at 407 (noting that "determination of whether a special relationship exists is a factual inquiry") (internal quotation marks and citation omitted).

7. The parties commenced discovery upon remand. See Dkt. No. 46. Discovery concluded in July 2018, see Dkt. No. 105, and in September 2018, Putnam and FGIC filed motions for full and partial summary judgment, respectively. Dkt. Nos. 110, 114. Those motions were argued in November 2018. Dkt. No. 146 at 1, 82.

8. On April 9, 2019, upon the death of Judge Sweet, the case was reassigned to the Honorable Analisa Torres. On September 10, 2019, Judge Torres issued an opinion that partially granted Putnam's motion for summary judgment and denied FGIC's motion entirely. Dkt. No. 160. The opinion began by addressing Putnam's motion for summary judgment on claims arising from the alleged misrepresentations in the updated target portfolio, labeled by FGIC the "Peach Colored Spreadsheet," or the "PCS." Id. at 14. The PCS, as described by Judge Torres, was a spreadsheet listing the "already-acquired and yet-to-be-acquired collateral assets chosen by Putnam" for Pyxis. Id. It was "sent to FGIC . . . on August 8, 2006" and its namesake derived from the fact that it "highlighted in a peach color" those assets that were "target[ed] for the Pyxis portfolio" (as opposed to already acquired). Id. By the time of the summary judgment motion, it was undisputed that the PCS was not provided directly by Putnam to FGIC, as suggested by the Second Amended Complaint. It was sent to FGIC in an email from the structuring bank for the Pyxis transaction, Calyon Corporate and Investment Bank ("Calyon"). Id. Judge Torres nonetheless ruled that FGIC had put forth sufficient evidence that the PCS was "attributable to Putnam," noting that it was "hotly-contested" whether Putnam, with knowledge of the PCS, referred FGIC to Calyon for it. Id. at 17. Judge Torres also found sufficient evidence to raise a genuine issue of fact as to FGIC's assertions that the PCS was material to FGIC's decision to participate in the Pyxis transaction, that FGIC had actually and reasonably relied on the PCS in deciding to participate in that transaction, and that Putnam had the requisite scienter. Id. at 18-28. Turning to loss causation, Judge Torres rejected Putnam's argument that FGIC had failed to connect the alleged misstatements to losses suffered by FGIC. Id. at 33. The opinion identified a few different theories of loss causation: that there was a "sufficiently direct [relationship] between Putnam's alleged misrepresentations" and the ultimate decision FGIC made to commute its liability; that "Putnam's alleged misrepresentations caused FGIC to enter into a different commutation agreement than it would have entered into if it knew the truth"; and that Putnam's misrepresentations caused FGIC to insure Pyxis at a higher attachment point than FGIC otherwise would have. Id. at 35-37. Judge Torres concluded that "a disputed issue of material fact exist[ed] over whether [FGIC would have insured Pyxis] on different, less risky terms if FGIC knew the truth about Pyxis's collateral," which would have "result[ed] in lower loss reserves and an attendant reduction in its . . . commutation payment." Id. at 38.

Shortly before trial commenced, the parties disputed (1) whether Judge Torres's order established as "law of the case" that FGIC could prove loss causation by showing that, but for Putnam's alleged misrepresentations, it would have entered into a transaction alternative to the Pyxis transaction, with the result that it would have paid less by way of commutation to Calyon; and (2) if so, whether the Court should reconsider that ruling as clearly erroneous. Dkt. No. 351. The Court acknowledged that there was "language in Judge Torres' explanation of her holding with respect to loss causation that could benefit each side," but explained that the Court would consider as "law of the case" only Judge Torres's answer to the question of whether there was "a genuine issue of fact requiring trial as to Plaintiff's claims regarding fraudulent misrepresentation, negligent misrepresentation, and negligence." Id. at 2. Having received no motion to strike on the loss causation matter, the Court expressly declined to "impose a limit on evidence or argument" regarding loss causation "via eleventh-hour interpretation of Judge Torres' opinion." Id. at 2-3. The Court directed "both sides" to "present the evidence and assert the arguments they believe[d] [were] necessary to prove their cases." Id. at 3.

9. Judge Torres granted Putnam's motion for summary judgment with respect to the remainder of FGIC's claims that did not arise from the PCS, including those arising from alleged misstatements in target portfolios sent on July 13, 2006 and August 7, 2006, in the Pyxis pitchbook, in the launch email for the Pyxis deal, in the target portfolios sent by Calyon to rating agencies, and with respect to the selection of assets and Putnam's performance as a collateral manager. Id. at 28-33. In closing, Judge Torres set Plaintiff's "negligence, negligent misrepresentation, and fraud claims arising from the PCS" for trial on April 27, 2020. Id. at 48.

Judge Torres made clear in a subsequent order that she had "granted Defendant's summary judgment motion with respect to Plaintiff's misrepresentation claims except for those arising from a document emailed to Plaintiff on August 8, 2006, and referred to as the 'Peach Colored Spreadsheet' or 'PCS.'" Dkt. No. 281 at 2. In a March 10, 2020 letter to Judge Torres, FGIC confirmed its understanding, in light of that "clear statement," that "the Court ha[d] indeed dismissed its claims based on Putnam's collateral selection misrepresentations" and that "FGIC [would] therefore proceed to trial only on its claims for fraud, negligent misrepresentation, and negligence arising out of the PCS." Dkt. No. 293 at 3.

10. In January 2020, FGIC filed a letter stating that it was "willing to withdraw its jury demand . . . subject to" Putnam's consent. Dkt. No. 231 at 4. Putnam consented to the withdrawal of the jury demand in March 2020, Dkt. No. 292, and Judge Torres ordered the case to be tried to the bench. Dkt. No. 294.

11. Shortly after, the trial was adjourned sine die in light of the COVID-19 pandemic. Dkt. No. 295. In April, during the pandemic, the trial was rescheduled to commence on July 6, 2020. Dkt. No. 297.

12. On May 28, 2020, defense counsel wrote to Judge Torres that both parties were "actively preparing for trial" and asked whether the trial would "proceed virtually to protect the safety of the witnesses, the Court, and the parties during the pandemic." Dkt. No. 298.

13. On June 2, 2020, the case was reassigned to the undersigned.

14. Three days later, this Court held a telephone conference with the parties. Dkt. No. 303. At that conference, FGIC consented to a trial conducted remotely via videoconference. Trans. of 6.5.2020 Conf. at 4. Three days later, Putnam filed notice of consent to the same. Dkt. No. 309. Over no objection, the Court ordered the trial to proceed remotely and the parties to submit direct testimony from their witnesses in the form of written declarations. Trans. of 6.5.2020 Conf. at 7.

The Court also obtained consent for the trial to proceed with the judge located outside the district. Trans. of 6.23.2020 Hearing.

15. On June 10, 2020, the Court ordered each side to file a letter-brief on whether the Court should set down only a subset of issues for the forthcoming trial, pursuant to Fed. R. Civ. P. 42(b). Dkt. No. 310. The parties filed letters, Dkt. Nos. 311, 312, and the Court permitted response letters, Dkt. No. 313, which the parties also filed. Dkt. Nos. 319, 320.

16. On June 23, 2020, the Court held a second telephone conference with the parties and heard argument on trial bifurcation. Trans. of 6.23.2020 Conf. The same day, the Court ordered that the July 6, 2020 trial would constitute "Phase 1" of the trial in this case, during which the parties should present their cases on the following issues with respect to the PCS: (1) whether Putnam is responsible for the PCS or had a duty to correct misinformation in it; (2) whether Putnam had a duty to correct the PCS under the "special circumstances" doctrine; (3) actual reliance or transaction causation; (4) reasonable reliance; and (5) loss causation. Dkt. No. 323. Phase 2, if there were a Phase 2, would address falsity, scienter, and damages. Id. Although the Court recognized that the issues to be tried in Phase 1 overlapped in certain respects with those in Phase 2, it made clear to the parties that they should present all of their evidence with respect to the Phase 1 issues in Phase 1 and that it would not exclude evidence on the grounds that it was also relevant to Phase 2 issues. Trans. of 6.23.2020 Conf. at 33-34; Dkt. No. 351. The Court also elicited Putnam's consent that, in deciding Phase 1 issues, it would assume that FGIC would be able to prove its case as to the Phase 2 issues of scienter. Trans. of 2.23.2020 Conf. at 42-44. Finally, with Putnam's consent, the Court granted FGIC's request that, in its case-in-chief, it be able to cross examine (or do a hostile examination) of two witnesses Putnam intended to call in its case—Carl Bell and John Van Tassel—even though FGIC had not subpoenaed those witnesses and even if Putnam would otherwise decide not to put on a case. Trans. of 6.23.2020 Conf. at 68-69; Dkt. No. 323.

2. Summary of Trial

17. As scheduled, trial began on July 6, 2020. The Court heard witness testimony from July 6, 2020 through July 22, 2020. The parties presented summations on July 30, 2020.

18. The trial proceeded remotely, Monday to Thursday, from 9:30 a.m. until the end of the day, with arrangements made by the parties for remote teleconferencing equipment and management of that equipment, and with the assistance of TrialGraphix. The Court was provided a laptop computer and the parties arranged for proceedings to be transmitted through Zoom. The lawyers for each side, the witness, and the judge were all in separate locations but were all able to view one another continuously during the proceedings. Pursuant to order of the Court, each witness was permitted only two sets of documents with him or her—a binder containing the witness's declaration and exhibits on direct examination and a separate binder in a sealed envelope (to be opened only after the cross-examination began) containing cross-examination documents. Dkt. No. 322. At the outset of the testimony of each witness, the courtroom deputy appeared to swear the witness remotely. The Court permitted the proponent of a witness to offer the witness's declaration as direct testimony. Cross-examination and redirect examination then proceeded. The parties consented to the oath being administered remotely. The Court also directed the parties to inform it if there were any technological or other issues with the remote proceeding. Only very few and brief connectivity issues arose; during those times, the proceedings paused. The Court made repeated findings that the remote proceedings had worked well and that it was able to view the witnesses, hear what they said, and make credibility determinations.

19. Given the remote nature of the trial, the Court gave the parties latitude. It did not limit the length of the trial generally or the time that the parties could spend examining witnesses. It permitted the parties to spend extensive time on issues of only marginal, or cumulative, relevance to the case. Considering the nature of the proceeding as a bench trial, the Court invited offers of proof on evidentiary issues. Whenever an issue was close, the Court admitted evidence subject to a motion to strike and, even then, erred on the side of inclusion. In the end, however, as laid out below, the evidence was not close. Putnam clearly prevailed on all of the elements of FGIC's claims.

The Court did impose time limits on opening and closing statements.

The Court ruled on most of the parties' objections to testimony and exhibits at trial. To the extent that the Court cites in this opinion to any evidence to which a party objected, the objection is overruled. The Court need not and does not resolve the remaining objections not addressed at trial or in this opinion. The Court has reviewed all evidence that was subjected to an objection and has determined that it would not change any of the Court's findings or conclusions regarding whether or not the objected-to evidence was admitted.

20. FGIC called five fact witnesses—Elizabeth Menhenett, Dana Skelton, Lynn Finkel, Thomas Adams, and Alex Rekeda—and one expert witness, Fiachra O'Driscoll. In addition, FGIC called two Putnam witnesses in its direct case: Carl Bell and John Van Tassel. Putnam submitted the declarations of Bell and Van Tassel and called expert John Dolan as a witness.

Putnam originally stated its intent to call Francis A. Longstaff as an expert witness, but withdrew him midway through trial. FGIC offered excerpts from Longstaff's deposition after the close of its direct case and after Putnam withdrew Longstaff as a witness. The Court excluded that testimony and it hereby adheres to that ruling. FGIC did not designate Longstaff as an expert prior to trial and thus his testimony was inadmissible under Fed. R. Civ. P. 26. In addition, with FGIC having made no efforts itself to procure Longstaff's attendance at trial, the deposition testimony is classic hearsay and does not constitute the admission of a party opponent. See Fed. Hous. Fin. Agency v. Nomura Holding Am., Inc., 2015 WL 539489, at *10 (S.D.N.Y. Feb. 10, 2015). Moreover, based on FGIC's proffer, the Court previously concluded and now confirms that any probative value of the testimony was far outweighed by its prejudicial effect, the potential to cause delay in trial, and the cumulative nature of the testimony. In excess of caution, the Court permitted FGIC to make an offer of proof and to designate the testimony it sought to offer and permitted Putnam to cross-designate testimony under the rule of completeness. The Court has now reviewed the designations and cross-designations and, even if they were admissible, they would not have made a difference in any of the Court's findings and conclusions.

21. The witnesses called by FGIC were generally not credible, lacked recollection of the important events of the case, and had little to add to what the documentary record showed.

22. As set forth below, Menhenett was an unreliable witness. Her testimony changed significantly from her deposition to her direct examination by declaration, and from her direct examination to her cross-examination. Her testimony was argumentative and evasive. She claimed precise recollection of 2006 events and conversations that FGIC claimed supported its case but, when confronted with evidence of equal importance from the same or later time periods that called her testimony into question, she professed a lack of recollection. Her selective memory of specific events and the evasive manner in which she answered questions called her general credibility into question and undercut the accuracy and veracity of her specific recollections.

23. Skelton, now self-employed as a freelance Russian linguist, was paid $300 per hour by FGIC to prepare her declaration and trial testimony. Tr. 718. At trial, she testified to the blanket statement, "I don't recall anything from 2006." Tr. 691. She also testified that she did not have "any active recollection of doing work on the Pyxis transaction," Tr. 674, or indeed, "any active recollection [of] the specifics of any transaction that [she] worked on at FGIC." Tr. 668. Her testimony did not add anything of substance to the record.

24. Lynn Finkel also faced challenges with recollection. She had few recollections or informative testimony other than reading from documents. Generally, her testimony was based on speculation and was not credible. When asked if she recognized documents from the relevant time period, she offered the memorable testimony, "I recognize my child's face. I don't necessarily recognize . . . a complex document from 14 years ago." Tr. 778. During another memorable portion of her testimony, Finkel was shown a clip from her 2017 deposition in which she spent approximately seven minutes reviewing what was one of the more important documents in the case (a loss analysis for Pyxis as to which she had some responsibility), only to admit ultimately that she did not recognize it. Id. In 2006, when Finkel worked on Pyxis, she had only worked with CDOs for two months. Tr. 781.

25. Adams, unlike Finkel and Skelton, did impress as to a recollection of events from the 2006 timeframe. However, he did not meet or speak with Putnam and he was more distant from the transaction than other FGIC witnesses. His testimony was probative only as to general practices. Even then, as detailed below, the qualified answers he gave on direct examination were severely undermined during cross-examination. The Court did not find him credible.

26. Rekeda, the only true third-party fact witness (as he was never employed by either party), had been out of the industry for ten years at the time of trial, Tr. 1028, and recalled very little about these events. His testimony about his current reading of documents, divorced from any recollection from the time they were authored, was not particularly probative. He proved most probative on general practices at Calyon.

27. The two fact witnesses from Putnam, Bell and Van Tassel, were generally credible. Bell's testimony and demeanor were identical on cross-examination and redirect examination. He admitted to facts that he did not recall and, without argument or reservation, answered questions that could have been perceived to be helpful to FGIC's case, including admitting that he could not rule out the existence of conversations that Menhenett said occurred but that he did not remember. He also gave logical, clear, and convincing explanations for practices and routines engaged in by Putnam. Although Bell's general value as a witness was limited by his lack of recollection of specific events, that same candor and his general demeanor and the direct way he answered questions made credible and convincing his testimony that he would not have engaged in conduct that FGIC accused him of. In the end, however, the Court need not rely on Bell's testimony. As laid out below, FGIC simply failed to make out its case.

28. Van Tassel was perhaps the most credible witness to testify. Although he too lacked specific recollection of conversations from 2006, he had a precise recollection of the markets in 2006 and of his trading strategy in approaching those markets. He answered questions on both cross-examination and redirect examination identically and without guile or evasion, and with no apparent thought as to which party his answer favored. His explanation of the markets and of his trading strategy and the independence with which he made decisions helped convince the Court that FGIC's trial theory was mistaken.

29. In the end, however, the Court does not rest its conclusions on the credibility or lack of credibility of one witness alone.

30. The experts from both sides offered helpful, and in many ways consistent, explanations of the key terms and economics in structured finance transactions. When answering questions that pointed to the specific events at issue—who was responsible for certain documents and whether FGIC relied on them—the experts had little more to say than to summarize the documents.

31. The following constitutes the Court's findings of fact. A word of caution to the reader. The Court delves into the details of events because the parties did so over the course of the two-and-a-half week trial. Those details are many and occasionally complex. They may also be unnecessary. They go largely to each item of circumstantial evidence the parties addressed at length. But in many instances, the circumstantial evidence was cumulative—and in a way helpful to Putnam and not to FGIC. Moreover, the most relevant facts and conclusions are few and simple. They are summarized here.

3. Brief Statement of Relevant Facts and Conclusions

32. FGIC's case depended in substantial part in tying Putnam to the PCS—proving that Putnam was either the source of the document or at least knew of its existence and the weight FGIC would put on it. But, the evidence established Putnam was not the source of information in the PCS and was not involved in its creation or transmission. Putnam did not know of FGIC's request for the information contained in the PCS, either before or after the fact, and did not know that Calyon delivered such information to FGIC. Calyon did so entirely on its own and for reasons of its own, without Putnam's knowledge or involvement.

33. FGIC did not prove that it relied on the allegedly false information in the PCS. FGIC's case depended on it convincing the finder of fact that it relied on the PCS as a representation of what would be in the final Pyxis portfolio when deciding to transact with Calyon. Only if the PCS represented, to FGIC, an approximation of what was in the final Pyxis portfolio would the alleged misrepresentations in the PCS would be important; that was also the testimony of FGIC's witnesses. But FGIC knew that the PCS constituted, at best, a point-in-time estimate of what could be placed in the Pyxis portfolio; the PCS was generated long before the assembly of the portfolio was complete. FGIC received different target lists, containing dramatically different apportionments of asset credit characteristics, both before and after receiving the PCS. It had no reaction, and asked Putnam to take no action, when—after the transaction closed—it learned that the assets in the final Pyxis portfolio were different from those reflected in the PCS. There is no convincing evidence that FGIC's decision to engage in the transaction with Calyon would have been different had the PCS represented the assets in the final Pyxis portfolio or had Calyon provided FGIC not the PCS but the list of assets that would end up in the final Pyxis portfolio.

34. FGIC also did not reasonably rely on the information in the PCS. FGIC asserts that in deciding to engage in the transaction with Calyon, it assumed that projected assets from August 8, 2006 would be in the portfolio come February 2007. But FGIC was a sophisticated market participant. It had engaged in several CDO transactions in the past (including with Calyon) and was represented by employees with substantial structured finance experience. It knew what was in the portfolio as of August 8, 2006, and that Putnam had no obligation to complete investing the Pyxis portfolio until six months later, in February 2007. The PCS came with a forward-looking disclaimer. It warned that neither Calyon nor or any of its affiliates were making "any representation or warranty, express or implied, as to the accuracy or completeness of the information contained herein" and that "nothing contained herein shall be relied upon as a promise or representation whether as to the past or future performance." PX253. It advised that "no representations are made as to the accuracy of any estimates or projections, or that all assumptions relating to such estimates or projections have been considered or stated, or that such estimates or projections will be realized." Id. More specifically, the cover email to the PCS stated that Rekeda had "not yet checked over [the] pool," such that FGIC should "be aware" of "changes" that might be made as the deal ramped up. Id. The day before FGIC received the PCS, it received an email from Calyon stating that the best estimate of the "intended portfolio" as of that time contained 4% prime residential mortgage-backed securities. Just a few weeks earlier, that figure had been 6%. FGIC thus knew that these numbers were fluctuating dramatically and would continue to do so, with the markets. FGIC has not claimed that those projections of a target portfolio were made in bad faith or were inaccurate; the evidence established that they were point-in-time estimates based on what Putnam had acquired and what it was seeing in the market when the projections were prepared. If the numbers and allocations changed dramatically from late-June to early August 2006, FGIC had no reason to believe that the numbers and allocations also would not change after August 7, 2006 and up to February 2007. Indeed, FGIC also knew that what would be in the final Pyxis portfolio would not and could not be accurately predicted or promised in August 2006. The final assets included were a function of what securities would be available in the market over the ensuing months, their price, how they fit into the constraints of the transaction, and Putnam's judgments about what was in the interest of Pyxis and derivatively in the interest of each of the levels of the capital structure—not just in the interest of FGIC. FGIC also did not reasonably rely on the PCS because, on September 8, 2006, the day of pricing, it received a term sheet that reflected Putnam's then-current view of what would be in the Pyxis portfolio, which differed dramatically from what was reflected in the PCS (but approximated what was in the final Pyxis portfolio). At that time, FGIC was not bound to the transaction. It could have walked away. Whether FGIC read the seven-page document and ignored it, or simply chose not to read the document and decided to rely on Calyon's earlier statements, its lack of care in connection with a nearly billion-dollar transaction undermines any claim of reasonable reliance. Finally, FGIC's reliance on the August 8, 2006 PCS as a prediction of what would be in the final Pyxis portfolio was unreasonable because in September 2006, at the time of pricing, and in October 2006, at the time of closing, FGIC could access information about what had already been acquired for the Pyxis portfolio. By the time of closing, approximately 85% of the Pyxis portfolio had been ramped (as compared to the 75% of the final portfolio that had been ramped at the time of the PCS). FGIC had access to information about what was in the actual Pyxis portfolio. It could have asked. Other investors did so. Its failure to do so, in reliance on the outdated August information, was not reasonable.

35. FGIC also has not proved loss causation. FGIC pressed two different theories in its opposition to summary judgment and in its pretrial briefing: (1) that the alleged lies in the PCS induced FGIC to forego an alternative, more attractive, transaction with a 50% attachment point; and (2) that FGIC incurred unnecessary loss because Putnam failed to invest the Pyxis portfolio as represented in the PCS. FGIC proved neither at trial. The evidence established that FGIC was never offered the opportunity to invest at the 50% attachment point. Calyon had a counterparty committed to insure it from the 28-40% attachment point. It needed an insurer to pick up the remainder of its exposure at the 40% attachment point. Thus, the single transaction offered to FGIC, after FGIC turned down a transaction attaching at the 28% attachment point, was attaching at the 40% attachment point. In addition, as to the second theory, FGIC offered no evidence that the portfolio and the CDO would have performed better or suffered fewer losses had Putnam invested in the securities reflected in the PCS rather than those in the Pyxis portfolio. Pyxis failed during the general credit collapse and financial crisis of 2007-2008. There is no evidence that FGIC's losses were caused as a result of Putnam's investment decisions.

36. FGIC's negligence claims fare no better. FGIC has not proved the existence of the special relationship necessary to establish a duty in negligence or for negligent misrepresentation. Putnam did not provide the PCS to FGIC, was not involved in it being provided to FGIC, and was not the source of the information in the PCS. Putnam did not know that FGIC had been given a detailed list of target assets for the Pyxis portfolio, or indeed any target portfolio for Pyxis. It therefore, and necessarily, did not know that FGIC was relying on a detailed list of assets to be targeted for the Pyxis portfolio. Even if Putnam had known that FGIC had received such a list, it would not have known that FGIC would be relying on such a list; to the contrary, every reasonable expectation would have been that the market and accordingly the targets for the Pyxis portfolio were dynamic. No prediction made in August 2006 could have been reliably steadfast with respect to the final portfolio as of February 2007. Finally, Putnam took no action evidencing an understanding that FGIC would be relying on a detailed targeted list supplied to FGIC by Calyon. To the extent that, outside of the marketing presentations, Putnam provided information to Calyon regarding assets that would be purchased into the portfolio, it was for Calyon's own use and not with the end and aim that it would be disseminated to FGIC.

37. Putnam had no freestanding duty of disclosure to FGIC. FGIC did not contract with Putnam. FGIC did not invest in the special purpose vehicle with which Putnam contracted to provide asset management services. Even if FGIC had, Putnam actually and necessarily disclaimed any fiduciary duties to the actual investors in Pyxis; its sole duties were to the Pyxis entity. It follows Putnam had no duties to FGIC, who was not even an investor but merely a contractual counterparty to an investor. As FGIC's own credit application proposing the transaction stated, the contract with Calyon would be "outside the [Pyxis] transaction." PX325.

38. Nor was Putnam in possession of any "special facts" that it knew or was reckless in not knowing would be important to FGIC. The special facts doctrine applies to communications between parties or potential parties to a transaction; Putnam does not fall into that category. Moreover, as explained above, Putnam knew neither of FGIC's request for a detailed list of target assets nor of Calyon's provision of such a list. Putnam also did not know, and had no responsibility to know, how FGIC made its insurance decisions. In other words, Putnam would not have known that FGIC used the PCS even if it had known that Calyon had provided the PCS to FGIC. Finally, Putnam did not have exclusive possession of the information that FGIC characterizes as crucial for it to have known before insuring the Pyxis portfolio. Putnam did not have some master plan or master list of the assets it intended to acquire for the Pyxis portfolio; its selection would depend on what was available in the market and what was necessary to satisfy the portfolio constraints at the time the assets were acquired. Putnam knew what had been acquired, but so did Calyon. So did any investor who chose to ask. FGIC chose not to ask. FGIC's failure does not require Putnam to become an interloper in the Calyon-FGIC relationship and volunteer information that it did not know FGIC wanted and that FGIC never even asked of Calyon. 4. General Terms, Definitions, and Concepts

39. Knowledge of certain financial terms and concepts is helpful to understanding the facts of this case and this Court's conclusions of law.

40. A collateralized debt obligation ("CDO") is a form of structured financial product. In broadest form, it issues notes and shares to investors in different tranches corresponding to differing rights and levels of risks. The CDO uses the proceeds that it generates from the sale of notes and shares to acquire a pool of assets that generate cash flow and serve as collateral for the notes. The assets may be in physical/cash form (e.g., notes) or in synthetic form (e.g., credit default swaps ("CDS") that generate cash flow in exchange for a promise by the CDO to pay cash to the CDS counterparty upon the occurrence of a specified credit event). The cash flow is distributed to the investors in a waterfall with the most senior generally paid first and the most junior (or those holding equity) paid last.

See Loreley Fin. (Jersey) No. 3 Ltd. v. Wells Fargo Sec., LLC, 797 F.3d 160, 164 (2d Cir. 2015) (describing structure and relevant features of CDOs).

41. The riskiness of an investment in a CDO is generally a function of: (1) the pool of collateral held by the CDO that generates cash flow (e.g., how likely the assets themselves are to default, and what level of diversification there is among the assets); (2) the structure of the transaction and the rights held by a particular class or tranche of security (e.g., whether it receives the first dollar generated by the pool of collateral, whether there are triggers that give it the right to priority over competing note holders upon the occurrence of specified credit events); and (3) in some instances, the skills and integrity of the firm purchasing, selling, and managing the collateral (e.g., whether it is selecting assets that generate large cash flow and bear little risk).

42. CDOs come in different forms. In a static pool CDO, the portfolio is fixed at the time of issue and the cash flow generated by the collateral is not used for reinvestment but to pay off noteholders. The investors know the securities that will be collateralizing their investments in advance, and take the risk that those securities will either perform or default.

43. In a CDO with a managed portfolio, such as would be the case with Pyxis, the collateral is not fixed and a collateral manager has discretion to buy and sell all or a certain percentage or portion of the assets that will comprise the collateral within contractually specified constraints. The portfolio is typically not fully invested when it is issued. The benefit of a CDO with a managed portfolio is that the manager can change the collateral securitizing the investment to pick those assets with the most attractive risk/reward ratio depending on the market conditions at a given time. The downside of a managed portfolio is that the investor does not know in advance which assets will securitize her investment. The investor relies to some extent (and within contractual guidelines) on the discretion and investment skill of the portfolio manager. In other words, the investor is investing, in part, in the portfolio manager.

44. As suggested by its complexity, numerous participants are necessary to the construction and success of a CDO.

45. The structuring bank or arranging bank or dealer typically structures the CDO, liaises with the rating agencies to obtain target ratings for the proposed tranches, markets the product's debt and equity, assumes the risk of underwriting the transaction, and funds the asset "warehouse" facility (explained below). The structuring bank is responsible for negotiating and determining the requirements for the collateral, the number and size of the tranches, the rights of each tranche, how the cash generated by the collateral will be distributed to different tranches of investors, what triggers are necessary, and what happens if the CDO is unable to pay a particular tranche of investors the cash that the tranche is due at any particular time.

46. As the funder of the warehouse facility, the structuring bank will also lend its balance sheet to the CDO. Before investors buy a note or purchase a share of a CDO, the investors want to know that the CDO has rights to a large amount of the collateral that will ultimately secure the notes and be used to pay them. Typically, the structuring bank will purchase that collateral and hold it in what the industry calls a "warehouse" base, on the promise that if the CDO is able to sell notes and shares, the CDO will purchase the collateral. It thereby takes the first risk of loss that the CDO will fail. If the CDO does fail and is unable to sell notes and shares, the structuring bank will be left holding the collateral that the CDO is not able to purchase, subject to any agreements it has with third parties to absorb all or part of that loss. Because the structuring bank bears the risk of loss on the warehoused assets, many structurers require a pre-commitment from a third party to purchase the CDO's equity tranche, which historically was the most difficult portion of a deal to sell.

47. The structuring bank will also be responsible for selling the notes and shares to investors, which generates the proceeds necessary to pay for the collateral. In that capacity, the structuring bank will take the actions necessary to make the securities saleable. It will liaise with the rating agencies, which rate the security of the notes; it will make sure the rating agencies are satisfied with the structure of the CDO, the waterfalls, and the collateral such that the rating agencies issue ratings on the notes that entice investors to purchase them; it will locate and market (through road show and investor presentations) the CDO to investors (typically, large hedge funds, pension funds, and other sophisticated investors); it will interface with those investors about any changes they require to the structure or collateral constraints for the transaction; and it will be responsible for negotiating many other contracts and arrangements, including the selection of a trustee and the preparation and negotiation of an indenture.

48. The structuring bank also acts as the CDO's underwriter, preparing the offering documents and obtaining the necessary approvals for CDO issuance. It is often the initial purchaser of notes issued by the CDO, and is responsible for marketing those notes to investors. The structuring bank generally serves as the conduit of information to potential investors and other market participants about topics such as spreads, ratings, and deal covenants.

49. The structuring bank also plays an important role in so-called synthetic and hybrid CDOs. CDOs can hold assets and generate cash flow in two different ways—synthetically or by holding a cash asset. A cash asset is a physical security; in the case of a bond, it is a fixed income instrument that represents a loan made by an investor to a borrower in return for the borrower's promise to pay to the lender a certain amount of interest (the coupon) and to return principal on specified terms. The interest and return of principal paid by the bond generates a portion of the cash flow necessary to satisfy the CDO's obligations to investors at different points in the capital structure.

50. A CDO can also generate cash flow by holding assets in a synthetic form known as a credit default swap. A CDS is a contract that gives its counterparty the right to a contractually-specified amount of periodic cash flow based upon the performance of a so-called "reference entity," in exchange for a promise that if the reference entity fails, the CDS holder will return a sum of cash to its counterparty. In essence, instead of owning an interest in the reference entity, the protection buyer obtains a contract right to the rewards of the reference entity, including the cash generated by it. In exchange, the protection seller receives from the protection buyer a periodic premium but also takes on the risk that the reference entity will fail. A CDS functions similarly to an insurance policy—the insurer (analogous to the protection seller) guarantees the insured (analogous to the protection buyer) against loss in exchange for a fixed premium but also foregoes the opportunity for gain. Holding a CDS as collateral is a way for the CDO to gain exposure to the performance of an asset without paying immediate cash for it or taking ownership of it (which is called holding collateral in "synthetic" form as opposed to "cash" form). The availability of assets synthetically is a function not of the number of bonds that have been issued and that are available for sale in the market but of the number and size of counterparties willing to write a contract with the bond as a reference entity. Synthetic assets thus are not subject to the same liquidity constraints as physical assets.

51. The "collateral manager" or "portfolio manager" contracts first with the structuring bank, and then after the transaction closes, with the CDO to select and manage the collateral. It selects the initial portfolio and manages the portfolio according to prescribed guidelines contained in the CDO's indenture. Before the transaction closes, the collateral manager will identify and recommend assets for the structuring bank to purchase within specified guidelines. Those assets will ultimately be sold to the CDO and used as the collateral to secure the notes and generate the cash flow to pay investors. The collateral manager also contracts with the CDO to manage the collateral after the CDO has issued the notes and shares, purchasing the remaining collateral that the structuring bank did not purchase before the notes were issued, and making decisions and recommendations regarding additional assets for the CDO to purchase or sell. In return, the collateral manager typically will receive a fee based, in part, on the assets' performance.

52. The equity investor takes the lowest tranche of a CDO. It absorbs the first losses and is paid generally only after all of the liabilities and obligations of the CDO are satisfied. It thus is entitled to the residual gains and losses from the transaction. As such, holding equity in a CDO is akin to holding a highly leveraged position in the underlying securities. During the relevant time for this case, equity investors demanded high returns commensurate with the risk associated with their equity position. Placing the equity tranche often proved to be one of the more challenging aspects of structuring an economically viable CDO. The equity investors ordinarily took their interest in the form of preferred shares.

53. In a sponsored CDO, the equity investors are the sponsors. They will become involved early in the transaction. They help structure the transaction, including the requirements for and the constraints on the collateral pool, and commit large sums of capital to the transaction.

54. The debt investors form the next layer in the capital structure of a CDO, above the equity. The debt investors are typically grouped in different tranches corresponding to their appetites and needs for the cash flow and the corresponding risk of the transaction. As noted above, the waterfall structure mandates that tranches receive payment generated by the CDO's assets in order of seniority, with investors in the most senior tranche paid first and those in the most junior (i.e., the equity tranche) paid last. If the cash flow is too limited to compensate every tranche, the tranches absorb losses in reverse order, with the most junior tranche exposed to the first loss. Yet, the greater the level of risk, the greater the reward. The tranched capital structure protects the CDO insofar as the senior debt holders suffer losses only after subordinate noteholders have absorbed the first losses (and indeed all subsequent losses up to the senior debt holders' position).

55. At the time of Pyxis's issuance, CDOs were sold to sophisticated investors in private, individually-negotiated transactions. Then and now, such investors possessed substantial resources to perform detailed analyses of a CDO prior to investing. As large and sophisticated investors, the debt investors often seek to influence the CDO structure by expressing their preferences to the structuring bank.

56. The capital structure may also include a super-senior tranche above the highest level of debt investors. The investment in the super-senior tranche may not be funded; it may come in the form of a contract promise. In synthetic and hybrid CDOs, for example, where the collateral manager can acquire assets either in cash or synthetically, the structuring bank may play an additional role as the "facing counterparty." In essence, the CDO's capital structure is comprised of investors (equity and debt), who provide cash funding at the time of inception that can be used to purchase collateral, and a CDS counterparty whose contractual promise to provide funding in the event of a credit loss while allowing the CDO to enjoy the rewards of a gain permits the CDO to gain exposure to assets synthetically. In that sense, the structuring bank takes the risk that the CDO portfolio will suffer losses and that the investment made at the lower levels of the capital structure will not be sufficient to absorb those losses in exchange for a periodic premium. Because the structuring bank may not be in the business of taking that risk (and for various other regulatory reasons), the structuring bank may seek to pawn off its risk on a third party—such as a monoline insurer—either by a separate credit default swap or by a financial guaranty.

57. Typically, an insurer will only agree to guaranty terms after: (1) it finds a suitable "attachment point"—the point at which it becomes liable for any of the CDO's losses; (2) it has scrutinized the credit quality of the assets in the CDO, the transaction structure, and key counterparties, to gauge the likelihood that the CDO will default and need an insurance payout to cover the loss; and (3) it has negotiated a sufficient fee in the form of premiums to make the risk/reward ratio of the guaranty attractive. The insurer profits if the structuring bank pays monthly premiums and never needs to submit a claim due to the CDO's default.

58. A debt tranche is said to have "attached" to the deal at the percentile where its credit enhancement ends. For example, if 50% of the capital structure is subordinate to a particular debt tranche, then that debt tranche is said to have "attached" at 50%. In broad summary terms, that means bonds occupying the bottom half of the capital structure would have to suffer 100% losses before the security at the 50% attachment point suffers a loss.

59. As a method of structural protection, a CDO may include "triggers." A trigger, broadly speaking, is simply an objective metric that, upon occurrence, causes a change in the rights and obligations of the parties to the CDO. For example, if the ratio of total collateral remaining in the CDO portfolio falls below a defined ratio in relation to a debt position (meaning that position has less "overcollateralization"), then the CDO's payment waterfall might be altered and the excess interest amount available at the triggered point could be diverted from equity or junior debt to senior debt.

60. The investors at different points in the capital structure share a common interest to some extent in the success of the CDO but also have divergent interests resulting from the differing points at which they will suffer losses.

61. The holders of the most senior tranches of a CDO will suffer a loss only if the CDO experiences multiple defaults across the portfolio. Its most significant risk concern is thus "tail risk"—the likelihood that all or many of the securities in the CDO portfolio will default all at once. All else being equal, then, the most senior tranches prefer that the CDO follows a conservative, diversified strategy. The more diversified the portfolio, the lower the likelihood that all of the securities fail at once, and thus, the greater the certainty that the most senior tranche will not suffer a loss.

62. By contrast, the holders of the equity may need an assurance that the CDO is designed to achieve a specified rate of return in order for them to invest. They thus benefit from riskier assets that can generate revenue if they are successful but, if they are unsuccessful, impose the first risk of loss on the debt investors.

63. One of the most important features of a CDO is the set of portfolio constraints that dictates the particular types of assets the collateral manager can purchase for the CDO as well as the general credit characteristics the portfolio as a whole must have. The CDO's portfolio constraints specify what collateral can be purchased by the CDO and how much discretion the collateral manager has within those specifications. Such constraints are usually negotiated heavily—understandably so, as they define the CDO and perform an important role in the ability to market the CDO and in the performance of the CDO once it is sold.

64. The portfolio constraints for the CDO will specify the types of securities that the collateral manager can purchase and how much. Types of debt or fixed income securities include corporate bonds, bank-preferred stock, asset-backed securities ("ABS"), and mortgage-backed securities ("MBS") that are either residential ("RMBS") or commercial ("CMBS").

65. A brief detour is necessary to explain some of these types of securities.

66. RMBS are created when a financial institution purchases pools of mortgage loans secured by residential properties with similar characteristics (e.g., prime loans, subprime loans) from mortgage originators, and sells them to a special purpose vehicle ("SPV"). The SPV in turn issues RMBS with multiple "classes" or "tranches," similar to the tranches of a CDO, each of which has a different priority when it comes to paying out principal and interest or allocating losses. The rating agencies play a critical role in determining the minimum amount of subordination for each class of RMBS, which the agencies base on their expectations of potential defaults and loss severity. In making these determinations, the rating agencies consider the types of loans in the collateral pool and require higher subordination for riskier assets.

67. RMBS come in different forms defined, in part, by the type of mortgage loans that collateralize the note, the vintage of seasoning of the RMBS (that is, the date the RMBS had when issued, which, in turn, provides some indication of when the underlying mortgages were issued), the issuers of the RMBS, the originators and servicers of the underlying mortgage loans or mortgage products, and the credit ratings of the RMBS or the amount of credit enhancement they enjoy.

68. Those factors determine both the creditworthiness and the selling or lending price of the RMBS. All else being equal, the more secure the RMBS, the less cash would flow to the CDO to secure the payments to the CDO noteholders and investors. Correspondingly, less secure RMBS would yield more cash flow. The security of an asset is often expressed in terms of "spread," which is a measure of the interest generated by a security and, correspondingly, its risk. Less secure RMBS have higher spreads—higher interest rates are paid to compensate for the comparatively higher likelihood of default. Correspondingly, more secure RMBS have lower spreads, because the default outcome is relatively less likely. For those reasons, spread is connected to the rating of a security; a more secure security will have a higher rating and lower spread. Spread is also often associated with yield, because the riskier an asset (hence, the higher its spread), the greater returns it yields (i.e., cash flow it generates).

69. The portfolio constraints for a CDO may also determine the amount of RMBS the CDO can purchase in each different "industry sector," which is defined based on the type of mortgages that collateralized the notes. For example, the portfolio constraints may require the CDO to have 90% subprime or midprime RMBS. Or it may require the CDO to have 75% prime RMBS. Or it may require the CDO to have assets not related to RMBS, such as CMBS or other types of debt securities.

70. "Subprime" RMBS (or "RMBS B/C") generally are backed by subprime mortgages, which are typically issued to borrowers who pose relatively greater risk of default.

71. "Midprime" RMBS generally refers to loans that incur medium risk, with borrowers who have slightly above average or average credit history.

72. "RMBS A" securities are frequently backed by "prime" mortgages, that is, mortgages issued to borrowers with good credit scores and other indicia of creditworthiness. However, there is no uniform definition of prime securities. Some RMBS A securities at the relevant time were backed by so-called "Alt-A" loans, which sometimes received "prime" designation but which many considered to be "liar loans" because they were issued to borrowers with incomplete indicia of creditworthiness.

73. The definitions applied to these RMBS may differ among ratings agencies. One rating agency may consider a particular type of bond to be prime while others might not. One rating agency at the relevant time did not have a definition of midprime and thus classified bonds that others might consider to be midprime as either prime or subprime.

74. The CDO portfolio constraints also will dictate the credit ratings an asset must have in order to be purchased into the CDO. Bonds are distinguished by their credit ratings, an assessment by the issuing credit rating agency of the financial soundness of the marketed debt instrument and the likelihood that it will default. Credit ratings incorporate considerations such as the amount of subordination the bond enjoys in the capital structure of the securitization in which it was issued, the average credit quality of the underlying borrowers, the performance capabilities of the mortgage issuers and underwriters, and other factors. Assets are rated alphabetically, valued in descending order, from A to D. Specifically, AAA is the highest rating, and proceeding down, the ratings are AA+, AA, AA-, then BBB+, BBB, and BBB-. Only A and B tranches are "Investment Grade."

75. As noted above, the rating of an RMBS or any other bond is a function not just of the strength of the assets that are in its collateral pool but also of the amount of credit enhancement the bond enjoys, i.e., its place in the credit structure and the other structural protections it enjoys. For example, a bond issued by an SPV that is collateralized by a relatively weak collateral pool but that is high in the capital structure might enjoy a stronger credit rating and might be more creditworthy than a bond issued by a SPV that is collateralized by a stronger collateral pool but that is lower in the capital structure. The collateral pool in the first example will be more likely to suffer losses but those losses will not impact the bonds in that example until after all of the bonds lower in the capital structure have been impaired. Thus, for example, in 2006, senior certificates were usually rated AAA (super-senior tranche) even if the RMBS was backed by non-prime loans. A prime RMBS may have a BB rating just as a subprime RMBS may have an AA rating.

76. RMBS can also be distinguished by their vintage. The least seasoned RMBS are those that had just been issued (known as "new issues"). More seasoned RMBS are backed by more seasoned mortgages. In 2006, more seasoned RMBS were considered to be more creditworthy and to generate lower yields than RMBS with newer vintages. That was in part because: (1) the weaker mortgages collateralizing seasoned RMBS had typically already suffered many of the losses they were expected to suffer by the time they were acquired for the CDO; (2) the mortgages included in seasoned RMBS had more time to benefit from house price appreciation since the date the mortgages originated, and so they had better loan-to-value ratios; (3) the remaining borrowers supporting seasoned RMBS had longer payment histories on their mortgages; and (4) more seasoned mortgages tended to have been underwritten by lenders using more restrictive underwriting guidelines than less seasoned mortgages.

77. The originators, issuers, and servicers of the RMBS can also make a difference in terms of the riskiness of the security. Broadly speaking, an originator is an institution that extends mortgage loans to homebuyers in exchange for the right to be repaid with interest. The originator sells those assets (the rights to be repaid with interest on the mortgage loans) to an issuer, which is an entity that creates a tradable security therefrom. At the relevant time, originators in different geographic areas originated loans pursuant to different underwriting guidelines and exercised differing due diligence in extending loans. Depending on the region where they originated, or the underwriting guidelines which addressed them, the loans—and the corresponding notes that were securitized by pools of the loans—were more or less risky.

78. Likewise, RMBS could be distinguished by issuer. Different issuers were known for, among other things, their differing due diligence standards and guidelines, and thereby the differing quality of loans underlying the securities they issued. Once again, depending on the issuer, the RMBS might be more or less risky.

79. Generally, mortgage servicers are the companies responsible for collecting on mortgage loans. Loans serviced by more diligent servicers yield higher-quality and less risky RMBS than loans serviced by unreliable or dilatory servicers.

80. Another factor that CDO investors consider is correlation—the relationship of the assets in the portfolio to one another and likelihood that all of the assets that form part of the portfolio will default at once. The greater the likelihood that all of the assets will default at the same time, the higher the correlation. The lower the likelihood that the assets would default at the same time, the lower the correlation. All else being equal, a portfolio of assets with a low correlation to one another is more secure than a portfolio of assets that were highly correlated with one another. Correlation, in sum, reflects the likelihood that a collection of assets will all fail at once.

81. In addition to determining and restricting the types of individual assets that may be purchased into the CDO, the CDO portfolio constraints dictate the credit characteristics that must be satisfied for the portfolio as a whole. Those include the spread, correlation, and overall risk of the CDO portfolio as a whole. Put broadly, portfolio constraints limit and guide the selection of assets for the collateral pool. Even if an asset does not have any characteristic that is prohibited by the portfolio constraints, it may nonetheless be ineligible for selection because of its impact on the collateral pool as a whole. For example, an asset that is too risky (its spread is too high), or conversely that yields too little cash flow (its spread is too low), or that is too highly or lowly correlated might not be eligible for purchase into a collateral pool even if it meets the constraints applicable to individual assets.

82. Several measures of the credit quality of the portfolio as a whole are important for this case.

83. Weighted Average Spread ("WAS") is a measurement of the average spread of the securities in the portfolio. It is calculated as the dollar-weighted average of the difference between the London Interbank Offered Rate ("LIBOR") and the interest rates payable on the assets in which the CDO invested. As noted above, a higher spread corresponds with a riskier investment and potential for a higher return. Because WAS expresses an average, one asset with a low spread can be grouped with another asset with a high spread. Conversely, the acquisition of a security with too high or low a spread can alter impermissibly the WAS for the entire portfolio.

84. Weighted Average Rating Factor ("WARF") is a measure of the risk of the aggregate securities in the portfolio. It is calculated as the quantitative average of the credit ratings assigned by Moody's to each of the securities in the portfolio. It assigns a score of 1 to a triple-A rated asset, 20 to a double-A asset, 360 to a triple-B asset and so forth. WARF corresponds to a 10-year probability of default. For example, a WARF of 610 implies a 10-year probability of default of 6.1%, whereas a lower WARF of 494 implies a 10-year probability of default of 4.94%. Accordingly, higher WARFs indicate greater risk and, all else being equal, a riskier portfolio of assets and riskier CDO investment.

85. The Moody's correlation factor is a measure of the correlation among the securities in the portfolio and, thus, the risk that all of the securities will fail at once. It uses correlations to stress-test portfolios, focusing on asset-sector variation.

86. All else being equal, a super-senior investor would prefer a more diversified, as opposed to less diversified, portfolio. An equity investor, conversely, would prefer a less diversified portfolio because that would yield a higher likelihood of yielding greater returns.

87. Numerous factors can affect the correlation of the assets in a portfolio. One of the most important factors is asset type. The collateral securitizing a CDO could comprise numerous different asset types, including RMBS, CMBS, CDOs, municipal bonds, corporate bonds, or other securities. RMBS, for example, were not necessarily correlated with CMBS because the borrowers on the underlying loans were different—residential borrowers as opposed to commercial borrowers. Correlation is impacted by industry sector—for example, RMBS subprime versus RMBS prime. That is because assets from the same sector are regarded as more likely to perform in line with each other than assets from different sectors. Accordingly, a portfolio that was invested in both RMBS and CMBS may be more diversified and have less correlation risk than a portfolio invested in RMBS or CMBS alone. A portfolio invested in both prime RMBS and subprime RMBS may be more diversified and have less correlation risk than a portfolio invested in prime RMBS or subprime RMBS alone. Likewise, a portfolio invested in seasoned and less seasoned RMBS may be more diversified and have less correlation risk than a portfolio invested exclusively in RMBS of more recent vintage.

88. The foregoing description about the formation of a CDO is general. In practice, the idea for the CDO and the initial steps taken to form it can be made by any number of the participants who will ultimately play a critical role in its success. Equity participants may come up with the idea for the CDO, identify a bank willing to structure the CDO, and lend their balance sheet to the financial product. Collateral managers can also identify the idea for the CDO and locate and identify the participants.

89. A managed portfolio CDO goes through several different phases (although not always in this precise order).

90. First, the early part of a CDO process is a collaborative one where the prospective sponsor, collateral manager, and investors all discuss the terms of the potential CDO, including the assets to be invested by the CDO, the structure of the CDO, and the relative roles of the different parties. During that process, the equity investors or mezzanine investors or other participants at different points in the capital structure may demand certain asset selection criteria as necessary for them to participate and invest in the CDO.

91. Next, the structuring bank, the collateral manager, and occasionally, certain relevant investors will negotiate and sign the agreements that will govern their relationship during the period before the closing of the CDO. Typically, these include a mandate letter, i.e., a commitment letter, which outlines the broad general terms of the relationship between the structuring bank and the collateral manager, and a warehouse agreement which outlines the process by which assets are acquired for ultimate transfer to and deposit in the CDO. The mandate letter will describe the services to be rendered, the respective obligations of each party, the release of funds from the transaction once it has been closed and settled, fees, expenses, and so forth. Both parties will also usually agree not to engage in transactions that might create a conflict of interest for the lifespan of the transaction.

92. Meanwhile, the warehouse agreement will describe the responsibility of the collateral manager to identify and recommend assets to be acquired for the benefit of the transaction and will set forth the responsibility of the warehouse provider (typically, the structuring bank) to finance the acquisition of the collateral and to take the first risk of loss on the collateral. The warehouse agreement may also identify other third parties (investors) who will finance the warehouse and take the first risk of loss on it. Typically, the warehouse agreement will set forth detailed criteria for the assets that can be purchased and will provide the party who takes the first risk of loss with a veto over the acquisition of assets. It will set forth procedures for the collateral manager to provide advanced notice to the warehouse provider, or to the party or parties with a veto, over the assets to be acquired and for the persons with a veto right the method and timing for the exercise of that veto.

93. During the pre-closing period, the structuring bank, the collateral manager, and the other relevant parties will conduct the activities for which they are responsible to ensure the successful closing of the transaction. The collateral manager will identify and acquire, using the most efficient means possible, the assets that will form the collateral for the CDO and inform the warehouse provider of those assets.

94. The structuring bank will coordinate the structuring process, identify potential investors at each point in the capital structure, and determine their appetite for the transaction with different levels of credit enhancement and entitlement to the cash flows. Based on different portfolio requirements, the structuring bank may negotiate a guaranty over the most senior tranche (if that has not earlier been done), prepare the necessary transaction documents including an offering memorandum and distribute them to potential deal participants, prepare the indenture with the indenture trustee who will protect the noteholders, and liaise with the ratings agencies, the portfolio manager and other interested participants to ensure that the notes issued will have the ratings necessary for them to be marketed while also ensuring that the portfolio has the characteristics necessary to satisfy all participants at each point in the capital structure.

95. The structuring bank will also frequently warehouse the securities used as collateral during the pre-closing period. It will review requests by the collateral manager to acquire assets, determine whether they fit within the asset constraints of the transaction and, if they do, purchase the assets and frequently take the first risk of loss.

96. At closing, the rating agencies will issue their credit ratings letters, the structuring bank will transfer the warehoused assets to the SPV and sell the CDO notes and preferred shares to investors, and any insurer will issue the policy.

97. Thereafter, the collateral manager will purchase any post-closing assets needed to complete the CDO's investments, make later purchases needed to reinvest principal amounts received on the CDO assets, and provide the trustee with information needed to prepare periodic investor reports.

98. The details of the investment arrangement are laid out in an offering memorandum, which is, essentially, a solicitation to potential investors to buy assets from a portfolio. It provides information about those assets, such as a detailed description of the notes and their respective agency ratings, preference shares, security for the notes, and so forth. The offering memorandum typically includes a detailed schedule of eligibility criteria to which securities purchased for the CDO must conform. Such eligibility criteria may specify collateral types, minimum credit ratings, minimum interest rates, and other features of the collateral. Also, an offering memorandum may specify criteria that the portfolio must meet as a whole.

99. Before the 2008 financial crisis, CDOs used and generated staggering sums of money in holding collateral. By one measure (Moody's rated volume), the CDO market grew from a handful of deals and $2-3 billion in rated issuance in 1994, to a peak of $80 billion in rated issuance in 2000. Estimates for 2004 counted roughly 160 transactions and just under $60 billion in issuance. These numbers also fail to reflect the growth in variety of assets securitized via CDOs.

100. Given the sums of money involved, and the complexity of the transaction, it is no surprise that the parties to CDOs are typically sophisticated. They have, by necessity, market experience as well as access to large amounts of money and intelligent advisors.

101. This case, like so many others surrounding the 2008 financial crisis, entails liability for investment losses. When the CDO heyday came crashing down in the 2008 financial crisis, the housing industry collapsed and MBS collapsed—taking CDOs with them.

102. Specifically, this case concerns the failure of one of those CDOs, Pyxis. The question is whether Pyxis's collateral manager defrauded the insurer of the structuring bank into issuing a guaranty against Pyxis, and, if so, whether the representations that induced such commitment caused the loss that the insurance company ultimately suffered.

5. The Participants in the Pyxis Transaction

103. Pyxis was a $1.5 billion CDO, created and sold in the summer and fall of 2006. It launched on October 3, 2006. Pyxis was a "hybrid" mezzanine asset backed securities CDO focusing on midprime and subprime RMBS securities. Its hybrid nature meant that its portfolio would be comprised of securities held in both "cash" and "synthetic" form, permitting it access to a wider range of exposures to assets at more efficient pricing. Because of the expanded ability to access assets, Pyxis was also a relatively large CDO. The mezzanine ABS nature of the Pyxis portfolio meant that the collateral backing the transaction, which would be purchased with the investments made in Pyxis by investors and which would fund the cash flow to pay the investors and provide a return to the equity, was comprised primarily of RMBS securities that were backed by pools of midprime or subprime mortgage loans.

104. The marketing for the Pyxis transaction launched in mid-July 2006 and the transaction closed on October 3, 2006.

105. Putnam, a Delaware limited liability company based in Boston, Massachusetts, was the collateral manager for the Pyxis transaction. At the time and today, Putnam was a significant global asset management company, registered with the Securities and Exchange Commission ("SEC") as an investment adviser since 1968. As of June 2006, Putnam was managing over $189 billion in assets for nearly 10 million shareholders and 180 institutions.

106. Putnam had particular sophistication in the CDO industry. By June 2006, Putnam was managing $3.5 billion in CDO assets across seven funds. Putnam began managing structured product CDOs in 2001, and was managing the collateral for three structured product CDOs—with $2.9 billion in assets under management—before the Pyxis transaction took place.

107. The Putnam team members on Pyxis all had significant experience in structured finance, including with CDOs and MBS.

108. Carl Bell, Senior Portfolio Manager and Team Leader for the CDO & Portfolio Credit Team at Putnam, received a Bachelor of Science degree in Applied Mathematics from Carnegie Mellon University in 1988 and a Master of Business Administration from the Fuqua School of Business at Duke University in 1992. At the time of the Pyxis transaction, Bell had served in "buy side" investment roles with institutional clients for 14 years continuously. He was also a Chartered Financial Analyst ("CFA") charterholder. As the head of Putnam's CDO group, he was Putnam's team leader for Pyxis.

109. John Van Tassel received a Bachelor of Science degree in Finance from the University of Missouri-Columbia in 1983, worked in the banking industry for five years, and then received a Master of Science degree in Finance from the University of Colorado Boulder in 1990. He worked as a Senior Vice President at Putnam and was responsible for analyzing and selecting RMBS assets for the Pyxis portfolio. Because more than 90% of the Pyxis assets were RMBS, Van Tassel was responsible for selecting the vast majority of Pyxis's portfolio assets. At the time of the transaction he had eighteen years of investment industry experience.

110. Michael Malm was a Senior Vice President and Portfolio Construction Specialist in Putnam's Investment Management Division. As a member of the CDO & Portfolio Credit Team, he supported Putnam's investment process by selecting CDO assets for the Pyxis portfolio. Malm was a CFA charterholder and, when he joined Putnam in 2001, brought five years of investment industry experience.

111. The members of the Calyon team working on the Pyxis transaction all had significant experience with structured finance, CDOs and MBS included.

112. Bertrand Delaunay was Calyon's Head of Structured Credit Sales in the Americas.

113. Alex Rekeda was Calyon's Director and Head of Cash CDOs from July 2004 to December 2006, and the deal lead for Pyxis. Before joining Calyon, Rekeda worked at Ambac, a monoline insurance company. In his career as a structurer, Rekeda worked on more than twenty CDOs. However, as of the time of trial, he had been out of the industry for ten years.

114. The analysts from Calyon on the transaction included Sachin Anand and Ben Lee.

115. Pyxis's equity investors and sponsors were Magnetar Capital LLC ("Magnetar") and Deutsche Bank. Magnetar and Deutsche Bank initiated the Pyxis transaction. They had the initial concept for the transaction, effectively designed the initial intent to move forward with the transaction, and provided the initial views and support for the transaction.

116. The teams responsible for Pyxis at those institutions were led by individuals with extensive experience in structured finance: Jim Prusko, the Co-Head of Structured Credit at Magnetar, and Michael Henriques, a Managing Director at Deutsche Bank. Prusko had previously worked at Putnam and was business friends with Bell.

117. Bell and Henriques were also business friends. They had previously worked on three prior CDOs managed by Putnam from 2001 to 2003 when Henriques was at Goldman Sachs & Co.

118. The debt investors in Pyxis included such sophisticated financial investors as IKB Deutsche Industriebank AG ("IKB"), ITC, and Merrill Lynch, all of which had substantial experience in the cash and synthetic RMBS and CDO markets. Indeed, TCW, Merrill Lynch, and another eight of the Pyxis debt investors were collateral managers for at least one CDO between 2005 and 2007. IKB was an early mezzanine investor in Pyxis and helped negotiate the portfolio constraints to ensure that the transaction met its needs and was one it could invest in.

119. In addition to being the initial purchaser for the Pyxis notes and preferred shares, Calyon was the super-senior investor in Pyxis, pursuant to a super-senior swap agreement that obligated Calyon to provide credit protection to Pyxis at the 28% attachment point in the form of payments if adverse credit events with respect to the underlying assets took place in exchange for the payment to Calyon of contractually specified fees.

120. Calyon protected itself against the risk of loss on Pyxis through transactions with two separate financial institutions at two separate attachment points. Banca Intesa provided protection to Calyon against the risk of loss from the 28% attachment point to the 40% attachment point. It was considered the junior super-senior insurer.

121. FGIC provided protection to Calyon against the risk that the losses in the Pyxis portfolio would exceed the 40% attachment point.

122. FGIC is a monoline credit protection insurance company organized under the laws of New York and established in 1983. "Monoline" insurers are so named because they operate only one line of business, which is to insure the timely interest and principal payments on bonds. FGIC was a sophisticated entity with the expertise necessary to evaluate RMBS and CDOs. By 2006, it was one of the four largest players in the global financial guaranty insurance industry and it dedicated a substantial portion of its business to insuring structured products, including RMBS and CDOs. In fact, structured finance comprised 14.2% of FGIC's insured portfolio in 2004, with over 90% of that amount written on mortgage-backed securities. By the end of 2005, structured finance represented 19.7% of FGIC's insured portfolio, and 22.7% (or roughly $68 billion in net par outstanding) by the end of 2006, with over half written on mortgage-backed products. By 2006, the company's top 15 structured finance exposures in the U.S. together accounted for over $16 billion in net par outstanding.

123. FGIC had an on-again, off-again relationship with Pyxis but ultimately, through a subsidiary, accepted Calyon's invitation to provide protection to Calyon against the Pyxis risk. FGIC's subsidiary, FGIC Credit Products LLC ("FGIC Sub"), agreed to reimburse Calyon for its losses under Calyon's super-senior swap agreement with Pyxis in exchange for regular premium payments. By separate agreement, FGIC guaranteed FGIC Sub's obligations under the Calyon/FGIC swap, which effectively assumed that Calyon's risk would be measured and calculated at Pyxis's termination date.

124. FGIC and Calyon had a working relationship both before and throughout the Pyxis transaction. Prior to the transaction, FGIC had participated in at least two other CDO transactions with Calyon as the structuring bank, and had underwritten collateralized loan obligations in which Calyon was involved. After Pyxis, FGIC continued to invest in CDOs structured by Calyon (such as the Havenrock II transaction, which will be described below).

125. FGIC and Putnam had no relationship prior to Pyxis. FGIC had not participated in any CDO transactions with Putnam as collateral manager. Putnam did not know how FGIC evaluated the credit risk of a CDO and, prior to FGIC's due diligence on Putnam, FGIC knew little about Putnam's investment strategy.

126. Every member of the FGIC team working on the Pyxis transaction had significant experience in structured finance, CDOs and RMBS included, although many had moved on to other careers by the time of trial.

127. At all relevant times, Elizabeth Menhenett was the Director of the Structured Finance Group at FGIC, and the Pyxis deal manager. Menhenett received a Bachelor of Science from the Massachusetts Institute of Technology in 1993 with a major in Computer Science and Engineering, and minor in Economics. Menhenett had worked in structured finance ever since. She became a CFA charterholder in 2000, and she remains certified. By the time she worked on Pyxis, Menhenett had accumulated extensive experience with CDOs, ABS, MBS, and RMBS.

128. Lynn Finkel, head of FGIC's CDO group, and Tom Adams, head of FGIC's MBS Group, were required to approve the company's participation in the Pyxis transaction before the deal could advance. Finkel received a Bachelor of Arts from the University of California, Berkeley in 1984 with a major in Economics, and a Masters in Business Administration from Harvard Business School in 1991. She worked at a number of institutions, including Merck and CentreRe, before joining FGIC as a Managing Director in 2004. The CDO Group's underwriters and analysts, including Menhenett, reported to her.

129. Finkel reported to Adams. Adams graduated from Fordham Law School in 1989 and proceeded to work on MBS transactions at the law firm Thacher Proffitt & Wood LLP. Adams accumulated additional experience in mortgage transaction ratings and MBS insurance before joining FGIC.

130. Dana Skelton was an analyst in FGIC's MBS group. She received a Bachelor of Arts in Mathematics and Statistics from Barnard College, and a Masters in Translation and Interpretation from Middlebury Institute of International Studies. Skelton joined Moody's Investor Service as analyst focusing on ABS and MBS. She joined FGIC in 2005 as a Vice President of the MBS team, where she was responsible for evaluating the creditworthiness of MBS—including that of the RMBS contained in CDOs. In 2005, she was appointed head of the MBS team (a subset of the group led by Adams) and became responsible for performing credit analyses on proposed RMBS transactions.

131. Finally, the CDO trustee for the Pyxis notes was LaSalle Bank National Association, a financial institution unaffiliated with Putnam.

6. Brief Chronology of Pyxis

132. The Pyxis transaction began with discussions in March 2006 among the equity investors (Magnetar and Deutsche Bank), Putnam, and Calyon about a synthetic or hybrid managed CDO focused primarily on midprime and subprime RMBS.

133. The transaction priced on September 8, 2006.

134. The transaction closed on October 3, 2006.

7. The Pre-Marketing Phase

135. In March 2006, Magnetar and Deutsche Bank had conversations with Putnam and Calyon about a synthetic or hybrid CDO focusing primarily on midprime and subprime RMBS.

136. The two equity investors were engaged in a long-short correlation strategy involving RMBS, of which the CDO investment was part. The strategy entailed going long on equity exposure to a pool of midprime and subprime RMBS and shorting the higher-rated tranches or single-named credit default swaps from the same sector (but not necessarily from the same CDO). In other words, investors would short RMBS securities that were the same or similar to the RMBS securities included in the CDO. The strategy was meant to, and did, perform well in (but not exclusively in) stressed economic environments. As the long side of the trade, in midprime and subprime RMBS, performed poorly with defaults in the home housing market, the short side of the transaction would make up for those losses (at least on a short-term basis) and generate at least a cash flow profit for the equity investors.

137. In approximately March 2006, Prusko of Magnetar and Henriques of Deutsche Bank approached Putnam about Putnam's interest in managing a CDO that would invest in mezzanine subprime RMBS and would have no cash triggers. Putnam historically had been reluctant, since October 2003, to manage such a CDO because it had been selling (rather than buying) mezzanine subprime securities in its general strategies for the funds it managed and could not get comfortable with the potential conflict of interest and otherwise of being long as well as short such securities and because it had believed that the market was risky. Putnam expressed interest, noting that it had started to add subprime credit to its other portfolios, that the market had developed to attract more sophisticated investors, and that it would be prepared to manage such a transaction if there were sophisticated investors at all levels of the capital structure who would be prepared to be exposed to the subprime risk.

138. In early April 2006, Prusko made an "official introduction" of Calyon to Bell in anticipation of a May 1, 2006 meeting the three were scheduled to have. PX67. The introduction email attached the term sheet for the precedent transaction the three would model Pyxis on, the Orion 2006-1 transaction ("Orion"). Id. Like Pyxis, Orion was a mezzanine hybrid CDO backed primarily by midprime and subprime RMBS with ratings of BBB to BBB-.

139. From approximately March through July 2006, the three parties (Calyon, Putnam, and the equity investors) worked on the pre-marketing phase of Pyxis. Magnetar and Deutsche Bank set forth the asset constraints they would require in order to sponsor and make the equity investment in the transaction. Putnam and Calyon modeled portfolio constraints, cash flows, and waterfalls to investors. All of them worked on, and to some extent collaborated on, a set of marketing materials. The parties also drafted the legal documents and contracts that would define their relationships and obligate them to one another and to the transaction.

140. The process of purchasing and assembling a portfolio of assets for a CDO is known as ramping. In the case of Pyxis, at least 85% of the assets would be ramped (or purchased) before the transaction closed, and the remainder could be purchased after the closing. The initial ramp-up of the portfolio and preparation of marketing materials proceeded in tandem, as Putnam explored the assets available in the market to satisfy the anticipated portfolio constraints and as the parties modified the asset constraints, based on the assets that were available or likely to be available. Put differently, Putnam revised the marketing materials' description of the targeted collateral while it ventured into the market and acquired securities for the transaction.

141. As would be typical before the marketing of a transaction, the initial parties reached agreement on certain defining features of the CDO including the percent of the portfolio (by notional value) that would be invested in midprime and subprime securities, the average spread the portfolio as a whole would enjoy (and thereby the cash flow on average the assets would enjoy), and the rating factor that the portfolio would need to satisfy (and thereby the average rating or risk the securities in the portfolio would have). As to the percent of assets invested in midprime and subprime RMBS, from the beginning, Prusko and Henriques made it clear that their institutions would only invest in a CDO with a highly correlated portfolio, at least 80% of which was invested in midprime and subprime RMBS. As negotiated and ultimately agreed, 80% of the Pyxis portfolio would be in midprime and subprime RMBS assets, and 90% of the Pyxis portfolio be invested in midprime and subprime assets and ABS CDOs. Those asset constraints were disclosed to other investors and participants.

142. The parties also ultimately agreed on a minimum WAS of 1.87% and a target WAS of 2.00%, which would imply that the CDO would be focused on higher risk, lower-rated assets.

143. The parties also agreed on a maximum WARF of 439 and target WARF of 511. The WARF constraint implied and conveyed that the CDO would have a focus on Baa2 and Baa3 assets.

144. The parties also discussed and negotiated the warehousing process for the pre-close period, i.e., who would bear the first loss on the assets that they planned to transfer and sell to Pyxis if the transaction closed. They also discussed what rights that party would have over the assets to be purchased.

145. As ultimately agreed, Calyon was the warehouse bank. It bore the first risk of loss on the warehoused assets (subject to whatever agreement it might have reached with the equity investors for them to absorb all or some of that risk). In exchange, Calyon had a veto over the warehoused assets. Thus, before any assets could be purchased for the warehouse, Putnam needed to give Calyon advance notice, and Calyon had the opportunity to review the assets and advise Putnam if they did not fit within the asset constraints.

146. Importantly, the evidence did not establish that the equity investors had a right to direct Putnam's selection of the collateral for Pyxis, that Putnam took their direction, or that Putnam managed the collateral in their interest and to the detriment of investors at other levels of the capital structure. As Bell testified credibly at trial, although the disclosed asset criteria demanded by the equity investors constrained what assets the collateral manager could purchase and required that Pyxis have a specified percentage of midprime and subprime assets, Putnam exercised independent judgment about which assets to purchase within those constraints. His testimony was not contradicted by Prusko, Henriques, or any representative of Deutsche Bank or Magnetar, none of whom FGIC called to testify in Phase 1. Nor was it inconsistent with any of the documentary evidence at trial. In particular, there was no documentary evidence of Deutsche Bank or Magnetar ever asking Putnam to purchase certain assets in the Pyxis transaction. FGIC argued that the equity investors used Pyxis as a "managed account" to take the long side of short investments that were part of their trading strategy. In fact, of the 186 assets in Pyxis, of which 174 were RMBS, Magnetar was the counterparty on three. It was the short counterparty on two. And it was the long counterparty on one. The suggestion that Magnetar was shorting all of the assets in the deal and that Putnam knew that to be true was entirely unsupported by the evidence.

There was vague testimony (but no documentary evidence) that Magnetar and Deutsche Bank had an agreement with Calyon to veto the assets that went into Pyxis, Tr. 1090, but there was no evidence that Putnam knew of that agreement. In any event, there is no evidence that the equity investors ever vetoed an asset or directed Putnam to include any specific assets in the deal.

The parties spent significant time at trial exploring what Putnam told FGIC about the involvement of Magnetar and Deutsche Bank in the Pyxis transaction. It was undisputed that Putnam did not identify Magnetar and Deutsche Bank to FGIC, and did not have a right to do so under confidentiality rules and conventions. Magnetar and Deutsche Bank had a contractual relationship with Calyon and not with Putnam. The Court also finds that FGIC knew Pyxis was a Magnetar transaction by at least September 2006—before FGIC was obligated to go forward with the deal. In any event, Menhenett admitted at trial that if the collateral manager independently selected the collateral, it would be irrelevant to FGIC who the equity investors were and why they chose to buy the equity. Tr. 450.

147. In addition, Bell had persuasive explanations for the few documents that FGIC suggested were consistent with Putnam taking direction from the equity investors or ceding independent judgment. First, he addressed the email in which Henriques (to several colleagues but not to anyone from Putnam) referenced Bell as stating that Putnam would view the transaction as "a highly structured separate account mandate for the equity investor." PX47. Although Bell did not remember the comment, he did not deny that he might have made a similar comment and explained it in a way that was consistent with Henriques's email. Putnam previously had been reluctant to invest in subprime or midprime RMBS but might have been willing to manage such a transaction if all investors wanted to be invested in subprime and midprime RMBS and the portfolio was thus "highly structured." Tr. 1381. In that circumstance, the CDO could be considered a "managed account," not just for the equity investors, but for all investors. Id. Second, he persuasively testified about several emails in which he and Prusko discussed the transactions to be purchased into the deal and Magnetar's shorting strategy. According to Bell, the emails concerned opportunities for Pyxis, presented by the equity investors, to be on the other side of transactions which the equity investors sourced for Pyxis (and thereby perhaps to get better liquidity and pricing) and not dictates as to which specific assets to buy.

148. Finally, Bell was cross-examined extensively on his understanding that the equity investors might have had an arrangement with Calyon, as the warehouse provider, to veto the assets acquired for the Pyxis portfolio. FGIC argued Putnam must have had that understanding and must have acted at the direction of the equity investors, because Pyxis's resulting portfolio (one comprised heavily of midprime and subprime assets) was perfectly suited to the equity investors' correlation trading strategy. Bell testified that he was indifferent as to whether the equity investors had a veto; rather, he was concerned that whatever veto rights existed in connection with the warehouse not impede Putnam's efforts to efficiently assemble assets for the portfolio. He was unaware of a veto held by the equity investors and, in any event, no veto was ever exercised either by Calyon or by the equity investors. The final portfolio, he explained, was dictated by asset constraints which were disclosed and which defined the transaction and which favored Baa2/Baa3 midprime and subprime assets, market conditions, and Putnam's own independent views (which too favored midprime and subprime over prime assets).

8. The Parties' Agreements with Respect to Pyxis

149. The relationships and duties among Calyon, Putnam, Deutsche Bank, and Magnetar during the pre-close phase were all governed by contract. It is important to understand these contracts to understand the parties' behavior thereafter and the legal implications of that behavior.

150. Prior to the closing of Pyxis, Putnam was party to two contracts. Both were with Calyon and Calyon alone. Those two, both effective June 21, 2006, were the "Mandate Letter" and "Warehouse Agreement."

151. The "Mandate Letter" between Calyon and Putnam, dated June 21, 2006, outlined Calyon's and Putnam's respective roles and responsibilities with respect to Pyxis.

152. The Mandate Letter dictated that Calyon's role in the transaction was to act as the sole arranger, placement agent, bookrunner and structuring advisor to Pyxis, to render financial advisory and investment banking services, and to underwrite 100% of the equity and each class of notes that Pyxis would issue to finance the $1.5 billion purchase of assets that would collateralize the securities.

153. To that end, Calyon agreed to "(a) work with [Putnam] to finalize (which include[d] structuring and modeling the [t]ransaction, preparing marketing materials and investor cash flow runs, coordinating the preparation of all documentation including offering documentation in relation to the [t]ransaction) a structure for issuance of the [s]ecurities which [was] mutually acceptable, (b) work with the rating agencies to obtain desired ratings for the [n]otes, (c) act as placement agent and underwriter for the private placement of the [s]ecurities, including underwriting the potion of [n]otes and [e]quity . . . and (d) warehouse up to $1.5 billion of [c]ollateral at the direction of the [Putnam], for [Pyxis]." PX 719.

154. Putnam agreed to cooperate with Calyon in connection with the services Calyon was to render. Specifically, Calyon and Putnam both "agree[d] to, in a timely manner, provide each other with all relevant information and feedback from external parties (including rating agencies, potential investors and regulators) that could be considered relevant for the [t]ransaction." Id. The two "further agree[d] that mutual agreement [was] required in order to start the marketing process in relation to the [n]otes (other than the [e]quity)" and "agree[d] to cooperate in good faith to create a mutually acceptable structure." Id.

155. The Mandate Letter described that Putnam's role in the transaction would be to manage the pool of securities that would collateralize the Pyxis notes. Putnam would, in other words, act as an "investment advisor" to Pyxis. Id.

156. The obligations of Calyon, to act as underwriter and placement agent, and of Putnam, to act as investment advisor to Pyxis, were subject to numerous conditions including: the execution of a warehouse agreement, Calyon purchasing at Putnam's direction "at least the amount" of original collateral specified in offering materials, the valid and binding transfer of such original collateral from Calyon to Pyxis, completion of due diligence by Calyon, written confirmation of the contemplated ratings of each class of notes, and completion of review of the collateral and certificate of conformance to the eligibility criteria specified by the parties. Id.

157. The Mandate Letter assumed that Calyon would be the party responsible for providing information to prospective investors and imposed obligations on Putnam "[i]n connection with [Calyon's] activities" under the agreement. Specifically, Putnam agreed as follows:

In connection with [Calyon]'s activities hereunder, [Putnam] will provide [Calyon] with reasonable access during reasonable business hours and upon reasonable notice, to all information concerning [Putnam] which [Calyon] and [Putnam] reasonably deem appropriate (the "Information") and will provide [Calyon] with reasonable access to [Putnam's] officers, directors, accountants, and counsel. [Putnam] agrees to make reasonable efforts to participate in conference calls or meetings (including "road shows") with potential investors during reasonable business hours and upon reasonable notice, to discuss such information as is reasonably appropriate. [Putnam] will also cooperate with [Calyon] so that [Calyon] can provide prospective investors with reasonable access to the Information.
Id. (emphasis added).

Bell understood the Mandate Letter to "ma[k]e clear that Calyon would be the primary communicator with deal participants." Bell Decl. ¶ 16.

158. Putnam did not have any other responsibilities to, or any formal relationship with investors or participants in Pyxis other than the obligation to make information and its officers and other employees available to investors in connection with Calyon's activities under the Mandate Letter.

159. The Mandate Letter also generally precluded Putnam from disclosing the identity of an equity or note investor. It provided:

The Investment Advisor [Putnam] shall not disclose to any person (other than affiliates of the Investment Advisor who have a need to know and legal counsel to the Investment Advisor in relation to the [t]ransaction), without the consent of CALYON, the name of any prospective or actual [e]quity or [n]ote investor with whom the Investment Advisor . . . does not have a pre-existing business relationship (except such disclosure as may be required by regulators of or
counsels to the Investment Advisor or its affiliates or to comply with federal or state securities laws) except where (i) the context of such disclosure is unrelated to the [t]ransaction, (ii) another person has a pre-existing relationship with such prospective or actual [e]quity or [n]ote investor and such person explicitly inquires of the Investment Advisor about such investor, (iii) required by law or regulation, or (iv) the receipt of such information is subject to an agreement of confidentiality reasonably acceptable to the Arranger [Calyon].

160. The Warehouse Agreement, also made effective on June 21, 2006, set forth Calyon's obligation to acquire and finance up to $1.5 billion of assets as collateral for securities acquired during the so-called "Accumulation Period" (which would be selected by Putnam). Id. The Accumulation Period was defined as the time period between the date of the Mandate Letter's execution and the earlier of: (a) the date that Pyxis closed or (b) the date the Warehouse Agreement terminated for another reason. The Warehouse Agreement provided that, if and when Pyxis closed, Putnam would direct Pyxis to purchase, and Calyon to sell to Pyxis, the warehoused securities. Under the Warehouse Agreement, Calyon would bear the risk of loss if the transaction did not close.

161. Under the Warehouse Agreement, and befitting the fact that Calyon would assume the risk of loss on the assets until their transfer to Pyxis, Putnam had to identify to Calyon each security it wished to recommend for inclusion in Pyxis's collateral. In turn, Calyon had to use commercially reasonable efforts to acquire such security subject to such security satisfying (in Calyon's discretion) eligibility criteria set forth in an Annex to the agreement. Annex 1 to the Warehouse Agreement set forth "General Eligibility Criteria" and "Asset Concentration Limits" that would constrain Putnam in its selection of the collateral (including the maximum percentage of securities with specified credit ratings or specified types) and the "Collateral Quality Tests" the collateral as a whole would have to satisfy, including that it needed to have a target WARF of 511, a target correlation factor of 22, and a target WAS of 2.03. Id. Each of those will be further explained later in this opinion.

162. Section 5(a) of the Warehouse Agreement addressed the procedure for Calyon selling warehoused securities in connection with certain triggering events. It provided that Calyon would "either arrange for the sale of such [security] in an arm's length transaction or obtain bid quotations from one or more third parties for such [security], and the sale price or the highest bid quotation obtained shall be the value of such [security]." Id.

163. Prior to Calyon signing the Mandate Letter and the Warehouse Agreement, and assuming the obligations under those agreements, it signed agreements with each of Magnetar and Deutsche Bank that imposed the irrevocable obligation on those institutions to purchase the equity that would be issued by Pyxis in amounts (for each) of not less than $35 million (and not more than $45 million). The equity was described as that which would be issued on terms substantially similar to those in an attached term sheet that described the CDO as a "Cash/Synethic Mezz ABS CDO" and that contained the portfolio constraints that would also appear in the warehouse agreement, including the figures for WARF, WAS, and the correlation factor and the General Eligibility Criteria. Id. Putnam was not a party to the agreements with the equity investors and was referenced therein only in its capacity as future collateral manager.

9. Preparation of the Draft Marketing Materials

164. During the period before Pyxis was marketed to investors, each of Putnam and Calyon performed the obligations they had agreed with one another to perform. Consistent with the Mandate Letter and the Warehouse Agreement, Putnam began to ramp up the portfolio of securities that would ultimately collateralize the transaction. Consistent with the Mandate Letter, Calyon began preparing marketing materials that, among other things, indicated the portfolio's composition by asset type and credit rating. The companies worked in tandem. As Putnam engaged in the market, and learned about the types and prices of assets available, it communicated with Calyon to refine the marketing materials. Putnam and Calyon, at times, consulted the equity investors about changes to the portfolio constraints and the indicative targets.

165. The earliest draft of a pitchbook introduced at trial was dated June 19, 2006. It was shared between Putnam, Calyon, and the equity investors and it relied on the Orion transaction as a model. Consistent with that transaction, it contained a pie chart with a preliminary target portfolio on an indicative basis. The preliminary target portfolio was consistent with the objectives of the parties to create a transaction that was primarily focused on mezzanine midprime and subprime RMBS and consistent with Calyon's agreements with the equity investors as well as with the Warehouse Agreement. It indicated that Pyxis would invest 16% of the collateral in RMBS A assets (3% cash and 13% synthetic), but that 62% of the assets would be RMBS B/C (11% cash and 51% synthetic). It also reflected a target WARF of "[511]" and maximum WARF of "[539]." PX 717. After Bell received the June 19, 2006 draft from Calyon, he returned various comments, including that the pitchbook should reflect an asset constraint of a minimum of 80 percent subprime and midprime RMBS and ABS CDOs.

The bracketed figures in the foregoing paragraph and other sections in the pitchbook represented items subject to negotiation and not yet finalized. For convenience, the remainder of this opinion omits brackets notwithstanding their appearance around certain numbers quoted.

166. Another draft of the pitchbook was created on June 21, 2006. The June 21, 2006 draft projected 16% RMBS A assets (3% cash and 13% synthetic) and 62% RMBS B/C (11% cash and 51% synthetic). The June 21, 2006 draft also specified that there would be a maximum of 20% RMBS prime assets and a minimum of 80% RMBS midprime and subprime RMBS assets. The target WARF remained 511 and the maximum WARF remained 539.

167. July 7, 2006 brought about yet another update to the pitchbook. In the July 7, 2006 version, there were, on an indicative basis, 6% RMBS A assets (3% cash and 3% synthetic) and 72% RMBS B/C (11% cash and 61% synthetic). The July 7, 2006 draft also specified that there would be a maximum of 20% RMBS prime assets and a minimum of 80% RMBS midprime and subprime assets. The target WARF remained 511 and the maximum WARF remained 539.

168. The evolution of the asset concentration charts in the pitchbook drafts reflected point-in-time expectations for Pyxis based upon market conditions, investment views, and estimates, as well as discussions among the parties. The charts, and other aspects of the pitchbook, changed as cash-flow modeling and discussions with the equity investors, and other investors, progressed.

169. For example, two early drafts of the pitchbook suggested the portfolio manager could invest up to 20% of the portfolio (the remainder after investing in mid and subprime RMBS) in CMBS, but a final version of the pitchbook, dated July 14, 2006 and to be discussed below, limited that amount to 10%. 10. The Launch and Marketing Phase, Generally

170. As noted above, by July 14, 2006, Calyon and Putnam had prepared the final version of the pitchbook. It bore the names of both organizations and was to be used with potential investors and participants.

171. The Pyxis launch email, drafted by Calyon and not Putnam, was sent to potential investors and participants on July 14, 2006. It described Pyxis as "a mezzanine ABS CDO transaction that takes advantage of synthetic technology to achieve better funding and access a ramped portfolio while utilizing a traditional cash CDO format." PX212. It made clear the nature of the transaction. It stated that the portfolio would be "invested primarily in real-estate related mezzanine ABS securities, with a particular focus on midprime and subprime residential criteria," and was "expected to be at least 85% ramped by the closing date." Id.

172. As finalized, the pitchbook's cover page introduced the proposed transaction as a "Mezzanine ABS CDO" transaction "Managed by The Putnam Advisory Company, LLC." Id.

173. The cover page was followed by cautionary disclosures labeled "Important Information," which warned investors that neither Calyon nor Putnam made "any representation or warranty, express or implied, as to the accuracy or completeness of the information contained [in the presentation]" and that "nothing contained herein shall be relied upon as a promise or representation whether as to the past or future performance." Id. The disclosures explained further:

The information set forth herein includes estimates and projections and involves significant elements of subjective judgment and analysis. No representations are made as to the accuracy of such estimates or projections or that all assumptions relating to such estimates or projections have been considered or stated or that such estimates or projections will be realized. The information contained herein does not purport to contain all of the information that may be required to evaluate the securities that may be issued and any recipient hereof is encouraged to read the Offering Memorandum and should conduct its own independent analysis of the data referred to herein. CALYON, Putnam and their respective affiliates . . . disclaim any and all liability as to the information set forth herein or omissions herefrom, including, without limitation any express or implied representation or warranty with respect to such information[.]

174. The disclaimer made clear: "None of CALYON, Putnam or any of their respective affiliates expect to update or otherwise review the information contained herein except by means of the Offering Memorandum with respect to the securities described herein. Additional information is available on request." Id.

175. The "Important Information" section also included a "forward-looking statement" disclaimer, advising that the forward-looking statements contained in the portfolio (including projections, forecasts, targets, and sample portfolio structures or compositions) were based on "certain assumptions." Id. The disclaimer warned, "Actual events are difficult to predict and will be beyond the issuer or Co-issuer's control. Actual events will differ from those assumed, perhaps significantly . . . and none of Calyon, Putnam or any of their respective affiliates assume any duty to update any forward-looking statement to reflect changes in underlying assumptions or factors, new information, future events or otherwise." Id.

176. As "important factors which could cause actual results or performance to differ materially from those expressed or implied in any forward-looking statements," the disclaimer listed: "the actual composition of the collateral, the price at which the collateral is actually purchased by the Issuer, any defaults on the collateral, the timing of any defaults and subsequent recoveries, changes in interest rates and any weakening of the specific credits included in the collateral, among others." Id. The disclaimer anticipated that "[o]ther risk factors" would be "described in the Offering Memorandum." Id. "Accordingly," the warning concluded, "there can be no assurance that estimated returns or projections can be realized, that forward-looking statements will materialize or that actual results will not be materially lower than those presented." Id.

Previous versions of the pitchbook all had the same or similar disclaimers.

177. The "Important Information" section further explained that neither Calyon, Putnam, nor any of their respective affiliates were acting as financial advisors or in any "fiduciary capacity [in] respect of the transaction." Id. The disclaimer advised that any decision to invest in Pyxis should be made after consulting the investor's "own legal, accounting, and tax advisors in order to make an independent determination of the suitability and consequences of an investment[.]" Id.

178. The July 14, 2006 pitchbook continued with an executive summary section (the "Executive Summary"). The Executive Summary presented a schematic diagram of the transaction structure (the "Transaction Diagram"). On the far right were the preferred shares of $82.5 million par value, which would comprise the first 5.5% of the capital structure. They would enjoy "certain excess" from the transaction. Id. Next in the capital structure were four classes of notes (ranging from Class A to Class D) with an aggregate par value of $337.5 million and that would comprise the next 22.5% of the capital structure. They would enjoy payments of principal and interest. Above the notes and the preferred shares sat the super senior of $1,080 million par value that would attach at the 28% level and enjoy super-senior payments in exchange for which it would provide not proceeds (as in the case of the notes) but protection.

179. The Executive Summary described Pyxis as follows:

Pyxis ABS CDO Ltd. is a $1.5 billion collateralized debt obligation (the "CDO") secured by WARF of 539 consisting primarily of U.S. dollar-denominated Residential Mortgage Backed Securities ("RMBS"), Commercial Mortgage Backed Securities ("CMBS"), ABS CDO securities and CDS on RMBS/CMBS/ABS.

180. The Executive Summary presented the terms that had been agreed by the parties, including the equity investors.

181. It described the transaction as one that would be "lightly managed," meaning that the securities in the collateral pool would not change dramatically over the duration of the transaction. Id. There was "a 5 year Reinvestment Period" and, "[s]ubject to certain conditions, up to 5% of the portfolio [could] be sold per year during the Reinvestment Period at the Collateral Manager's discretion" with the exception of Credit Risk and Credit Improved Assets that could be sold at any time. Id.

182. The Executive Summary advised: "It is anticipated that 85% of the collateral will be purchased by closing." Id. It allotted a ramp-up period of "3 months (post-close)." Id. It also reflected that there would be no triggers protecting the super senior tranche.

183. A section called "Composition of Assets" indicated Pyxis would purchase RMBS and other securities on a cash and synthetic basis, with a 28/72 split between cash investments and synthetic investments. Id.

184. The "Transaction Summary" section indicated that the transaction maturity would be 40 years. Id. It also contained a "Targeted Collateral Securities Profile," which included eligibility criteria and asset concentration limits for collateral assets, indicative collateral quality tests, a pie chart indicating asset type distribution, and a pie chart indicating S&P Rating Distribution (explained below). Id.

185. The Targeted Collateral Securities Profile specified that the collateral manager would be required to place at least 80% by par value of the portfolio in subprime and midprime RMBS assets and at least as much as 90% in RMBS and ABS CDOs and had discretion to put all of the portfolio in subprime and midprime RMBS and ABS CDOs.

186. It also contained a lengthy list of assets other than midprime and subprime RMBS and ABS CDOs that the collateral manager had the discretion but not the obligation to purchase for Pyxis. For example, it indicated that the collateral manager could invest up to 10% of the portfolio but no more in REIT debt securities, up to 10% of the portfolio but no more in CMBS securities, up to 5% of the portfolio in student loans guaranteed by the Department of Education, and up to 10% of the portfolio but no more in PIK bonds.

187. The "Portfolio Notes" accompanying the Targeted Collateral Securities Profile specified that "[e]ach asset must be rated at least investment grade by one rating agency at the time of acquisition" and mandated a "5% limit for split rated IG [investment grade] assets" (i.e., assets rated investment grade by only one rating agency and not another). Id.

188. The final pitchbook also announced the "Indicative Collateral Quality Tests," that would limit the discretion of the collateral manager and define the nature of the collateral pool and its quality, again as had been discussed among the parties. These provided for a target WARF of 500 and a maximum WARF of 539; a target Moody's correlation score of 24% and a maximum of 26%; and a target CDS Spread of 2% with a minimum of 1.87% (put differently, the projected minimum weighted average spread was 187).

189. As noted above, the final pitchbook (like the drafts that came before it) contained pie charts with "indicative" asset type and rating distributions. Id. These reflected, among other things, that the structuring bank and the collateral manager expected that the majority of the portfolio would be invested in midprime and subprime RMBS securities with only 4% invested in RMBS A securities. It also reflected that the majority of the portfolio would be invested in lower-rated assets:

Indicative Asset Type Distribution

Image materials not available for display.

Indicative S&P Rating Distribution

Image materials not available for display. Id.

190. The final pitchbook also contained a section describing Putnam's qualifications and philosophy. That section touted Putnam's nearly 70 years of experience managing money and that it managed $180 billion in assets for nearly ten million shareholders and approximately 170 institutions. The section also highlighted that Putnam was committed to ongoing monitoring and surveillance. It stated that, in addition to regularly monitoring rating actions, headlines, and key economic data that may impact the value or liquidity of its investments, Putnam tracked various other indicia (voluntary prepayment speeds, periodic net loss rates, cumulative losses, and more) for each credit-sensitive security.

191. Finally, the pitchbook concluded with a series of "Risk Factors." A section called "Collateral Risk: Collateral Impairment" warned that each collateral security would carry "credit, liquidity, interest rate and prepayment risk." Id. It explained that a "[d]ecline in credit quality of the collateral or defaults could result in losses which would adversely affect" securities. Id. And it explained that "impaired market liquidity following an event of default could impair the total return on the portfolio." Id. A section called "Hypothetical Illustrations" warned about "forward-looking statements." Id. It explained that "hypothetical illustrations are only estimates" and that "[a]ctual results will vary, and the variations may be material." Id.

11. The Marketing of Pyxis to FGIC

192. As early as June 20, 2006, prior to the Pyxis launch, Calyon approached FGIC—with whom it had a prior relationship—to explore whether FGIC would be interested in guaranteeing and insuring its entire exposure to the most senior tranche of Pyxis's capital structure. FGIC was the first, and for some time the only, financial institution Calyon approached to be its counterpart.

It was not the only available monoline insurer.

193. The interactions between Calyon and FGIC were characteristic of two large, sophisticated institutions negotiating a complex transaction at arms-length. At a lunch on June 20, 2006, Calyon's Delaunay discussed the deal with Menhenett and Skelton of FGIC.

194. Prior to the Pyxis launch on July 14, 2006, Calyon sent FGIC two versions of the pitchbook: the drafts of June 21, 2006 and July 7, 2006. Both pitchbook drafts anticipated that Putnam would invest the majority of the portfolio in midprime and subprime RMBS securities. As noted above, the June 21, 2006 draft pitchbook projected 62% RMBS B/C assets and 16% RMBS A assets. It specified that there would be a maximum of 20% RMBS prime assets and a minimum of 80% RMBS midprime and subprime RMBS assets. The July 7, 2006 draft projected the much-higher 72% RMBS B/C assets and much-lower 6% RMBS A assets. It also specified that there would be a maximum of 20% RMBS prime assets and a minimum of 80% RMBS midprime and subprime assets.

195. Calyon also sent FGIC the final (July 14, 2006) pitchbook, which projected the even-lower 4% RMBS A assets and the even-higher 81% RMBS B/C assets. The final pitchbook did not reflect any maximum for RMBS prime assets but it did reflect that the portfolio manager would have to invest at least 80% of the portfolio in midprime and subprime RMBS assets and at least 90% of the portfolio in RMBS and ABS CDO assets.

196. Calyon initially offered FGIC the opportunity to participate in the Pyxis CDO by guaranteeing Calyon's entire exposure to Pyxis, i.e., at a 28% attachment point. On July 14, 2006, Menhenett wrote to Rekeda expressing interest in participating in Pyxis at the 28% attachment point.

12. FGIC's Initial Requests for Information from Calyon

197. FGIC immediately began to request information about the ramped portfolio, and to consider and negotiate the deal's terms. As anticipated by the Mandate Letter and consistent with the fact that Calyon and FGIC (not Putnam and FGIC) would be counterparties, the correspondence was largely between Calyon and FGIC with Putnam playing a supporting role. On the afternoon of July 14, 2006, even before receiving the launch email, Menhenett asked Calyon for information about the assets that had been ramped into the portfolio so that FGIC's MBS group could begin to perform a credit analysis.

198. Calyon responded four days later, on July 18, 2006, with a portfolio spreadsheet of the ramped assets acquired for Pyxis to date. No employee of Putnam was copied on this communication.

199. The July 18, 2006 portfolio spreadsheet reflected that Putnam had, up to that point, included only assets with lower credit ratings. As of July 18, 2006, Putnam had acquired 82 collateral assets with a notional value exceeding $1 billion (approximately two-thirds of the final $1.5 billion portfolio), but none of the securities in the ramped portfolio were RMBS A or prime RMBS. All were post-2005H1 vintage and none were pre-2005H2 vintage. Collectively, the assets were riskier than would be permitted for the final Pyxis portfolio. The ramped portfolio had a WARF of 601 (compared to the final pitchbook's target WARF of 539).

200. The following day, July 19, 2006, Rekeda of Calyon sent Menhenett the same spreadsheet of the ramped portfolio and a target portfolio of assets to be acquired. No Putnam employee was copied on the email. Rekeda explained to Menhenett:

Note, that Putnam so far has been concentrating mainly on the BBB assets (almost done there) and only now is starting ramping up As and AAs. That's why I think once your RMBS guys get comfortable with the ramped portfolio, they will be happy with the rest of it, which will be much higher quality.

The spreadsheet cover page contained various disclaimers. Specifically, it clarified that the information contained was confidential and to be used only to evaluate potential interest, and the document did not constitute an offer or solicitation but was solely for informational purposes. Further, Calyon expressly made no representations or warranties about the accuracy of the information therein, and nothing contained could be relied upon as a promise of past or future performance.

201. The July 19, 2006 target portfolio was in the form of a matrix and contained the same breakdown of the target portfolio as appeared in the final July 14, 2006 pitchbook. It indicated that the vast majority of the portfolio would be in midprime and subprime RMBS and that only 4% of the portfolio was projected to be in RMBS A.

202. FGIC knew from the outset that the bonds in the pool would be comprised primarily of midprime and subprime assets of lower (Baa2 and Baa3) quality. The launch email sent to Menhenett (to which the July 14, 2006 final pitchbook was attached) described Pyxis as "a mezzanine ABS CDO transaction" and indicated that it would have "a particular focus on midprime and subprime residential criteria." PX212. The target for the Moody's WARF in the launch materials implied that the assets in the pool would be between Baa2 and Baa3. At trial, Menhenett also admitted that she understood that a minimum WAS spread of 187 basis points, with a target of 200 basis points, implied that the majority of the assets would be Baa3 in rating. Tr. 283.

203. The deal nonetheless was attractive to Menhenett and FGIC. In a July 20, 2006 email, Menhenett informed Finkel and Adams that she wanted an "initial read" of the asset names in the July 18, 2006 portfolio spreadsheet (the ramped portfolio). PX750. She then anticipated diving into FGIC's "re-rating process." Id. Menhenett also provided her two bosses with the July 14, 2006 final pitchbook, which contained the asset concentration criteria and the pie chart suggesting that Pyxis would invest most of the collateral in midprime and subprime RMBS and projected that it would invest 4% in what it characterized as prime RMBS. She further referred her supervisors to the "target" pool in the final pitchbook. Id. Menhenett's email identified the Pyxis WARF as correlating to "Baa2/Baa3" assets. Id.

204. On July 21, 2006, Menhenett asked Calyon for information about certain below-investment grade RMBS in the ramped portfolio, including the seasoning of those assets. She did not copy Putnam.

205. On July 23, 2006, Finkel emailed Menhenett and Adams that FGIC needed to decide "whether to soft circle, subject to credit, due diligence and negotiation of terms" the Pyxis transaction within "the next couple of days." PX1034. Finkel raised questions about a number of features of the transaction, including with respect to certain of the triggers and coverage tests and some of the permitted asset types. She did not raise questions either about the amount of subprime and midprime RMBS or about the absence of any requirement that any of the portfolio be invested in prime RMBS or seasoned RMBS. Menhenett responded: "I'll address these items in a separate email later. In short, the deal items are appr[o]priate for this type of transaction. You should not jump to quick conclusions." Id. No mention was made of the nature of the transaction being one focused on midprime and subprime RMBS of lesser quality.

206. On July 24, 2006, Menhenett emailed Delaunay, Lee, and Rekeda at Calyon to ask for further information. PX220. In part, she asked them to confirm her understanding that the maximum WARF would be 539, or for Calyon to "indicate the worst level." Id. She also asked for the "lowest level of WAS test." Id. She commented on the ramped portfolio's RMBS data saying, "the biggest missing piece is CLTV [combined loan-to-value ratio]. Would you be able to fill that in." Id.

207. Calyon replied the same day, stating the target WARF was 500, the maximum WARF was 539, and the then-present WARF was around 599 because "the manager ramped riskier assets first." Id. Delaunay advised Menhenett that the minimum WAS was 1.87 basis points and the target WAS was 1.96 basis points. Calyon advised that, "[o]n RMBS data, neither Putnam nor Calyon will be able to provide the CLTV. The only way would be to hire external auditors." Id. Putnam was not copied.

208. Negotiations continued between FGIC and Calyon regarding whether, and if so on what terms, FGIC could attach at 28%.

13. FGIC Interactions With Putnam

209. While FGIC reviewed the ramped portfolio with Calyon and negotiated their terms of business, FGIC also conducted due diligence on Putnam. That due diligence was important to FGIC for two reasons. First, FGIC understood that Putnam would have substantial discretion in its responsibility to manage the portfolio in the interest of all investors within the required constraints. Whatever Calyon or the pitchbook said in July or August about the projected composition of the portfolio might not necessarily be accurate in February after the transaction closed and when the ramping was to be completed or at any future time over the projected 40-year term of the transaction. Markets can change rapidly, and a target portfolio delivered on day one, which projects what the collateral manager intends to purchase weeks and months ahead, can become outdated on day two. FGIC wanted to understand Putnam's investment philosophy, the types of securities it preferred, its methodology for trading and conducting due diligence, and whether it had any conflicts of interest given the securities it held for its own account. Second, although Menhenett had some passing acquaintance with Bell from prior employment, FGIC and Putnam as institutions did not know one another.

210. The interactions initially proceeded through Calyon, as the Mandate Letter contemplated. On July 21, 2006, Menhenett emailed Delaunay asking certain questions she requested be relayed to Putnam. Her questions pertained to the deal structure, Putnam's investment philosophy, and whether Putnam had proprietary investments that could compromise its investment advice for others: "What is Putnam's view on the benefit of a hybrid structure?"; "Are RMBS names in this pool also held in [Putnam's] books?"; "Do they use single-name cds/abx to hedge their book?"; and "How do they see issuing this hybrid deal affect their other activities outside the deal?" PX751. Referencing a change in S&P's criteria for the amount of subordination required before a bond could be rated investment grade, Menhenett asked of Putnam, "Do you have trading data showing market's reaction, to mbs bonds issued b[e]f[ore] vs after July 1, due to SP's criteria change . . . ?" Id. Menhenett did not mention prime RMBS or seasoned bonds.

211. Delaunay relayed, in his words, the "questions that FGIC need[ed] answered" to Lee and Rekeda, but not to Putnam, adding, "I had a LONG conversation to accompany this email." PX751. As summarized by Delaunay, FGIC's questions focused on the ramping of the deal, not its specific asset composition or asset classes:

FGIC needs better understand[ing] about how we ramp these deals:
- Is Putnam hedging their own book as they are a[n] RMBS investor as well?
- Do we put out bid list and the street lifts us
- Does the manager ask for specific bonds and we shop around for best level
- Does the manager trade on its own and tells you 'you're done with dealer xyz'
Id. Delaunay never mentioned prime or seasoned RMBS.

212. Van Tassel and Bell of Putnam answered these questions. Their answers were relayed by Calyon to FGIC. Among other things, they answered how "Putnam . . . put[s] out the bid lists to ramp up the deal":

Reference bonds are identified through our investment process and then sent out to 12 or so dealers as a BWIC under standard ISDA ABS CDS confirm language for the following day. Each of our lists to date has been comprised of 20-25 bonds with a size of $12mm per bond. If the winning (wide spread) bid is acceptable, then we notify the dealer and Calyon to confirm the trade for the warehouse.

BWIC is an abbreviation for "Bid Wanted In Competition."

213. In addition to providing information about its investment philosophy and trading practices, Putnam (indirectly) addressed questions about the ramped portfolio. When Calyon received requests for information, it occasionally emailed Putnam to supply it that information. For instance, on July 21, 2006, Xavier Capdepon (a Calyon analyst) emailed Van Tassel that FGIC "would like more details" on the seasoning of the BB-rated assets in the ramped portfolio. PX769. Calyon collected some of this data and then asked Putnam to provide the rest, which Putnam did. Calyon forwarded this data to FGIC.

214. On July 25, 2006, Menhenett called Bell to ask if Putnam could provide the average CLTV data on the ramped portfolio. The prior day, Calyon had indicated neither Putnam nor Calyon would be able to provide that information. Putnam worked with Calyon to obtain the required information. On July 25, 2006, Bell emailed Calyon, "FGIC was clear that average CLTV [combined loan-to-ratio value] is a required statistic for their process," that Putnam did not "have this statistic in hand" and did "not know how to source it electronically." PX223. Bell asked for Calyon's help in providing that information to FGIC.

215. Putnam made itself available for two due diligence sessions with FGIC—on August 3, 2006, and August 7, 2006—as required by the Mandate Letter. Those sessions constituted the vast majority of Putnam's contact with FGIC prior to the closing of the transaction.

216. On August 3, 2006, FGIC conducted an in-person due diligence meeting at Putnam's Boston offices. As contemplated by the Mandate Letter, the meeting was chaperoned by Calyon.

217. FGIC had prepared a list of topics for discussion, focused on: (1) the deal's rationale; (2) the alignment of interests between Putnam and FGIC; (3) Putnam's views on different asset types, including prime RMBS; and (4) Putnam's collateral selection process.

218. Menhenett sent the list of discussion topics to Calyon in a two-page agenda. Calyon then forwarded it to Putnam. Bell and other Putnam employees received the agenda on August 2, 2006.

219. The agenda included the following questions:

• "Why issuing this structured finance CDO now?"
• "Why hybrid instead of a traditional cash flow structure?"
• "Any investment over-lap with your existing CDO transactions?"
• "Asset mix"
• "Rating quality: why wt. avg. Baa2/Baa3 instead of a 'high grade'?"
• "RMBS:
? Prime, Subprime? Agency MBS?"
• "Describe credit screening, use of tools and analytics"
• "Process for approval / rejection"

220. The agenda did not express an interest in or an aversion to any particular asset type. Nor did it ask about specific assets that were in the ramped portfolio or targeted by Putnam for purchase.

221. The August 3, 2006 meeting was attended by Menhenett, Finkel, and Katya Sverdlov of FGIC; Bell, his superior Rob Bloemker, Van Tassel, Malm, David Galvin, and Geoff Trawick of Putnam; and Ben Lee of Calyon.

Menhenett, Finkel, Bell and Van Tassel have all submitted testimony now, fourteen years later, about that meeting. Bell and Van Tassel do not remember the precise contents of the meeting but have testified based on pattern and practice to certain things that would not have been said. Menhenett and Finkel claim recollection of the meeting. Menhenett took handwritten notes during the meeting, which she preserved.

222. At the meeting, Putnam was transparent about its preference for subprime securities over prime RMBS securities. It explained that subprime had a better risk/reward ratio, offered express spread, had more spread than prime, and that "subprime [would] hold[] up to stress testing, while [p]rime [would] not." Tr. 309, 312. In essence, because the market preferred the risk of prime loans to subprime loans, prime RMBS had less subordination and were relatively overpriced and more difficult to source. As Putnam's trader Van Tassel explained at trial, "Prime deals were enhanced razor thin, to perfection, really. They couldn't really handle defaults and losses . . . [S]ubprime were designed to handle the worst previously observed default and loss scenarios that we had seen to the triple B level. So that sort of set the stage in terms of our investment thesis at that point in time in terms of why I favored subprime over prime." Tr. 1873. His views were similar to those conveyed to Menhenett by the collateral manager in a previous deal, RFC-IV, which was consummated just a few months earlier.

223. With respect to more seasoned RMBS, Putnam explained that by using CDS, Putnam was free from the new-issue market and able to look at issuance for the past year while searching for appropriate opportunities. In essence, it could get exposure without sourcing the physical bonds. There is no evidence that FGIC pushed for any firm commitment regarding exposure to more seasoned assets.

224. FGIC did ask how, given the amount of Baa3-rated RMBS that had already been acquired for the portfolio, Putnam would be able to balance the ratings distribution of the portfolio and acquire assets rated above Baa2. The question was directed to the credit ratings of the assets to be acquired and not their underlying collateral. Bell responded that Putnam's strategy had been to use the synthetic bucket in the deal to buy smaller dollar amounts of lower-rated bonds, which were easier to acquire synthetically, and that Putnam planned to subsequently use the cash bucket to acquire higher-rated bonds more readily available to purchase. Nothing was asked or said about acquiring prime RMBS or residential securities from industry categories other than subprime or midprime.

225. The remainder of the approximately three-hour-long meeting focused on asset constraints, Putnam's general investment philosophy, and Putnam's credit screening process, tools, and analytics—topics relevant to Putnam's future asset selection. Putnam explained that Pyxis would be exposed to sector risk from a concentrated RMBS subprime mezzanine portfolio.

226. Menhenett testified by declaration that, at the meeting, Putnam said it would diversify the portfolio by acquiring both assets of different collateral types (e.g., prime rather than subprime RMBS, and seasoned rather than unseasoned RMBS) and rating distributions (assets rated higher than Baa3 and Baa2). However, the Court finds Menhenett's contemporaneous notes and an internal memorandum that she prepared a week post-meeting (on August 10, 2006) more reliable. Those materials demonstrate that at the August 3, 2006 meeting, Putnam indicated the asset constraints permitted it to "diversify into non-RMBS asset[s]" and that such non-RMBS assets included CMBS and permitted ABS asset classes—not that Putnam indicated that it would purchase more RMBS A or prime RMBS or predicted how it would exercise discretion on behalf of all investors in the weeks and months ahead. PX829. Menhenett's notes and memorandum do not reflect any promise made by Putnam about what assets they intended to purchase, or any questions or answers about future plans—other than that the remainder of the pool would include more highly-rated securities. Indeed, Menhenett testified in her declaration that, while FGIC wanted "the remaining collateral to include a combination of more stable assets that were less correlated with Baa3 and Baa2 subprime RMBS," she did not "ask Bell to elect a minimum amount of any particular type of asset, whether prime, seasoned or otherwise." Menhenett Decl. ¶ 95.

227. Menhenett asked Bell how much overlap existed between the assets in Pyxis and in Putnam's own books, a question relevant to ensuring that Putnam did not have an adverse interest in any collateral it was selecting for the portfolio or the portfolio as a whole. Bell responded that Putnam was not using the deal "to hedge risk on its books." PX829.

228. Menhenett also asked Putnam why it was interested in doing the Pyxis deal. According to the August 10, 2006 memorandum, Menhenett understood Putnam's interest in Pyxis to have arisen from "internal studies" that "confirmed the strength of the Subprime product." Id. The August 10, 2006 memorandum reported that Pyxis was "an extension of Putnam's fundamental sector views." Id.

229. Another topic at the August 3, 2006 meeting was FGIC's tail-risk concern—tail risk being the risk of an event that would cause all or most of the collateral to default and thereby generate loss for the super senior tranche and thereby for FGIC as the insurer of Calyon with respect to that tranche. There is no convincing evidence to support FGIC's argument at trial that Putnam addressed the question of FGIC's tail-risk concern or even thought to address it by adding further diversification to the portfolio or by changing the portfolio's asset category mix. Rather, the conversation on August 3, 2006 (and the conversations that followed) centered on the subject of triggers and the fact that Putnam would be adding more highly-rated bonds as the portfolio ramped up.

230. On the subject of triggers, Menhenett inquired why the transaction was a triggerless CDO. Menhenett expressed interest in adding a cash diversion trigger that would shift cash to the super-senior tranche and thus improve its credit enhancement and risk profile when specified credit events occurred. Menhenett also asked about the pros and cons of having no cash diversion triggers during the reinvestment period. Bell responded that Putnam preferred additional credit enhancement to triggers because triggers did not have value in practice, distracted managers, and adversely affected their decision-making. Without triggers, he said, the manager was free to focus on what was best for the portfolio and the work was materially simplified.

231. It stands to reason that Putnam and FGIC did not discuss asset category diversification to reduce the tail risk of the transaction. In the first instance, and as to the portfolio asset constraints, the fact that Pyxis would be 80% invested in mezzanine subprime and midprime RMBS bonds was baked into the transaction. That was how it was defined. It was explicit in the asset constraints and implicit in the collateral quality tests. And if those portfolio constraints were to be changed, it would have been the responsibility of Calyon—not Putnam—to change them and ensure any changes would not undermine the ability to attract investors at other levels of the capital structure. Second, as Bell testified, adding a small amount of prime RMBS or other non-subprime or midprime RMBS would not have changed the essential riskiness of the transaction, as one based on a heavily concentrated portfolio with at least 80% invested in midprime and subprime securities. Third, as Bell reminded Menhenett, Putnam would be contractually and legally responsible to Pyxis, and would have to make future purchase decisions in a rapidly evolving market. Those responsibilities required Putnam to make investment decisions in the interest of the investors at all levels of the capital structure, not just to satisfy the insurer of the super-senior tranche (regardless how important that insurer was). As a practical and legal matter, that responsibility precluded Putnam from making commitments to asset selection in the future other than those reflected in the portfolio constraints. Bell and Menhenett were both sophisticated professionals. Bell understood that "there can be conflicts of interests between different investors in a CDO" but that "[t]he obligation for the collateral manager is to the CDO, as the CDO is the client." Tr. 1327. Menhenett understood that a collateral manager manages the different interests of investors by choosing assets in a balanced way that fit within the portfolio constraints; she acknowledged that the constraints give managers "flexibility." Tr. 474. It would have been foolhardy and uncharacteristic of a sophisticated professional like Bell to make a commitment he could not honor (or one which, if he attempted to honor it, could cause him to be called to task by investors at all other levels of the capital structure who also would have access to information about the portfolio's contents). It would have been equally foolhardy and uncharacteristic of a sophisticated professional like Menhenett to accept such a representation. The Court finds that Bell did not make any such representation, nor did Menhenett ask for one.

Menhenett testified that, as a matter of philosophy, FGIC did not want to impose asset constraints. Menhenett stated that FGIC did not have the same market flow of information as the collateral manager because "[o]n the corporate side, I can say that as a corporation FGIC acknowledged we weren't trading in the market. We didn't know as much as somebody like Putnam." Dkt. No. 317-9 at 91-92. "And I discussed earlier FGIC as a remote part of the debt stack was aware that for a good manager who knew what they were doing that we didn't already constrain them." Id. at 131.

232. At trial, and in argument, FGIC claimed that Putnam made certain material misrepresentations at the August 3, 2006 meeting. For example, FGIC asserted that Putnam failed to disclose that "one of the reasons Putnam was issuing Pyxis . . . was that the equity investors had told [Bell] they wanted to sponsor a primarily subprime mezzanine RMBS deal." Tr. 2053. FGIC also complained that Putnam failed to disclose: that "the equity investors sponsored Pyxis," that "the equity investors required the deal to be primarily subprime RMBS," that the "equity investors were pursuing a long-short correlation strategy based on subprime mezzanine RMBS," that "the equity investors would only participate if there were no triggers," and, in sum, "the critical role and motivation of the equity investors." Id.

233. These routes to fraud or negligence liability are plainly foreclosed by Judge Torres's rulings. But, even if they were not, the Court would consider them to be meritless. The evidence established that the occasion for Putnam to manage a primarily subprime mezzanine RMBS deal arose from the interest of the equity investors, but that such interest was not the reason why Putnam chose to manage Pyxis. The reason why Putnam chose to manage Pyxis, as reflected in the contemporaneous documents and in Bell's testimony, was that over the years from 2003 until 2006, Putnam had become comfortable with managing a subprime mezzanine RMBS deal, particularly if sophisticated investors at all levels of the capital structure were interested in such a transaction. Whereas previously Putnam had been inclined to short subprime RMBS assets, it now had some such assets in its portfolio.

234. Nor would it have been a surprise, or did FGIC establish that it was unaware, that at the time Pyxis was marketed to it, there were equity investors lined up for the transaction and that the portfolio constraints on the collateral manager and other structural features (including the absence of triggers) that defined the transaction were ones that the equity investors would need in order to make the substantial investment they made in the transaction. The evidence at trial established that the equity tranche of a CDO was among the most difficult to sell; it was the one that presented the most risk. Therefore, it was also common practice (and common sense besides) that before an equity investor committed, it would make sure that the transaction was one that would meet its risk appetite. There is no evidence that, at the time Pyxis was marketed to it, FGIC thought that the equity tranche was unaccounted for or remained to be sold; the transaction would not have been able to price if there were not equity investors. Nor did FGIC ask. Indeed, when after the August 3, 2006 meeting, FGIC asked both Calyon and Putnam further about triggers, Calyon and Putnam indicated that the equity investors would need to be consulted. Clearly, FGIC knew there were equity investors and that—regardless of who came up with the portfolio constraints in the first place—deal conditions needed to be such that would make equity investors willing to commit substantial capital to the transaction.

235. Finally, although FGIC also complains that Putnam failed to disclose to it that the equity investors made their investment as part of a long-short correlation trading strategy, there is also no evidence that FGIC asked or that such information would have been important to FGIC (or unknown by FGIC) at the time of the transaction. The evidence at trial established that institutions at all levels of a capital structure commit millions of dollars to a transaction only as part of a strategy. The strategy might involve laying off some of the exposure, as Calyon did with its super-senior exposure. Or sometimes it might be a different hedging strategy. It would not have been a surprise to FGIC that the equity investors too were making their commitment as part of a strategy. The conclusion might be different if the evidence had established that Putnam took direction from the equity investors contrary to Putnam's representations. But the evidence did not establish that Putnam's independence was in any way compromised. FGIC's complaints that it was defrauded by representations other than the PCS would fall flat even if they could establish an independent basis for liability after Judge Torres's decision. Indeed, when both before and after the closing of the Pyxis transaction, FGIC was informed that the equity investors had a trading strategy, it expressed no surprise or upset.

236. In addition, at bottom, and further elaborated below in the conclusions of law, FGIC's complaint rests on a faulty premise: that Putnam was obligated to volunteer information in the course of FGIC's dealings with Calyon (information that Putnam did not know that FGIC wanted). But Putnam was not a fiduciary to FGIC. It answered FGIC's questions completely and truthfully, and it was not required to go beyond that.

237. In any event, after the August 3, 2006, meeting, Putnam did convey to Calyon that FGIC had a tail-risk concern and left it to Calyon, as appropriate under the Mandate Letter, to address that concern if it could be addressed.

238. On August 4, 2006—one day after the in-person due diligence meeting—Bell and Menhenett had a follow-up conversation that pertained exclusively to triggers. They did not address the assets in the portion of the portfolio yet to be ramped. Rather, they spoke at length "about the possibility of adding cashflow triggers to the Pyxis deal structure [to] protect the senior debt classes." Menhenett Decl. ¶ 128. "Bell responded that this would upset the expectations of junior debt and equity investors." Id. Menhenett further averred that she made clear to Bell "that based both on the absence of triggers in the deal and, more fundamentally, the largely mezzanine hybrid portfolio, it was far from certain that FGIC would be able to get sufficiently comfortable with the deal to provide the requested insurance." Id. ¶ 129.

FGIC complains that Putnam did not limit its comments to the equity investors alone. The complaint is utterly meritless. There is no evidence that a trigger—which would have diverted funds from the equity and debt alike—was not of concern to the junior debt investors or that it would have made any difference whatsoever to FGIC if Bell had limited his reference to just the equity investors.

239. On the morning of August 4, 2006, Bell emailed Rekeda and Lee expressing his view that managing FGIC's tail risk concern "seems like the largest issue we need to clear to get this deal completed" and conferring responsibility to address those concerns onto Calyon: "Please confirm your plans and timetable to handle this." PX238 (emphasis added). As he testified at trial, the tail-risk issue was a matter to be handled by Calyon. Bell was asked how he was proposing to manage FGIC's tail risk concern on August 4, 2006 and he responded, "I don't believe I was proposing. I was just identifying the issue to Calyon, whose role it was to source and communicate and do analysis for investors." Tr. 1533.

240. In another email from Bell to Lee and Rekeda on August 4, 2006, Bell suggested a "potential solution" of "a cumulative loss trigger that ends the reinvestment period early and makes the deal pro rata." PX240. Again, Bell closed his email with an indication that the responsibility to address the trigger issue ultimately rested with Calyon: "Please let me know when we can discuss your strategy around this." Id. Rekeda proposed a call between Putnam and Calyon the following day. Id. Bell accepted the invitation but again noted: "[I]t is Calyon's role to structure and sell this transaction." Id.

241. Menhenett testified that, following Menhenett's August 4, 2006 conversation with Bell, and "prompted by his statement that triggers would upset the expectations of the equity investors," she called Delaunay to ask about the relationship between Putnam and the equity investors. Menhenett Decl. ¶ 130. Menhenett testified that Delaunay eased her concerns and confirmed that Putnam's relationship with the equity investors was arms-length, no different from a typical third party equity relationship, and Putnam was not being unduly influenced by the equity investors. No other witness testified to this conversation. In any event, the Court finds that Delaunay's statement was truthful.

14. FGIC's Continued Negotiations with Calyon and Its Request to Calyon for Additional Information About the Ramped Portfolio and the Target Assets

242. By August 7, 2006, FGIC had completed its preliminary analysis on Calyon's offer for FGIC to attach at the 28% attachment point. The initial feedback was not positive. At 12:38 p.m., Menhenett emailed Delaunay, Rekeda, and Lee of Calyon, with the results of FGIC's analysis of the July 18 portfolio spreadsheet (which contained mostly lower rated assets), reporting: "I have feedback on our RMBS, not good on the surface. But I'd like to see if there's been add'l ramp up at above Baaa3 rating." PX242.

243. In a bilateral conversation shortly thereafter, Menhenett told Delaunay that, based upon FGIC's analysis of the ramped portfolio, FGIC was not willing to attach at 28% and did not believe it could attach below 40%. FGIC was concerned about the ramped portfolio's credit composition, which consisted almost exclusively of Baa3-rated subprime RMBS. FGIC estimated high losses on such a portfolio. Menhenett further explained that, if better assets were to be added, FGIC would need to review and evaluate those assets to give them proper consideration.

244. Menhenett requested that Delaunay provide FGIC with two pieces of information: an updated warehouse (i.e., ramped portfolio) showing any additional assets acquired since July 18, 2006, and a detailed target portfolio showing the assets to be acquired for the rest of the portfolio. Menhenett wanted "not merely a pie chart like the target portfolios in the [pitchbooks] or a matrix target portfolio like the [one she] had been provided on July 18, 2006, but rather an itemized target portfolio filled in line by line with the names of the target bonds still to be acquired, with as much detailed metrics for those bonds, including credit ratings, spread, servicer, size of bond, whether the bond was funded or unfunded, and so forth, similar to the ramped data." Menhenett Decl. ¶ 133.

245. Delaunay summarized his conversation with Menhenett in a 1:32 p.m. email to colleagues at Calyon, asking them to provide an updated warehouse and a detailed target portfolio. His email reflects that, while FGIC was not willing to attach at 28% based on the ramped portfolio, it would likely be able to attach at 40% based on that portfolio:

Initial feedback is very bad indeed... AS it is: 40% attachment (FG or no FG)

The problem flows from the fact that there are only Baa3 ramped right now so portfolio quality is really bad. (I guess they extrapolate on the same portfolio quality). I guess the manager's strategy is to ramp lower quality assets 1st, but this has to be clarified.
1) Please send me an updated warehouse ASAP
2) Please send me a DETAILED target portfolio:
Not the guidelines with the buckets from the pitch book but itemized tentative portfolio, i.e., you fill in line by line with target bonds mentioning the rating, spread, Servicer, size, funded/unfunded...

Urgent... In case FGIC drops we need[] more time to find someone else.
PX242 (emphasis in original). Notably, while his request was urgent, Delaunay did not copy anyone from Putnam or even mention Putnam.

246. Eight minutes later, Lee replied to Delaunay (without copying anyone from Putnam) with a message that focused on the improved credit characteristics of the yet-to-be-acquired assets:

As Alex told Elizabeth very clearly on July 19, Putnam so far had been concentrating mainly on the BBB assets (almost done there) and only now is starting ramping up As and AAs. That's why the FGIC RMBS guys should not extrapolate the ramped portfolio to the rest of the collateral, which will be much higher quality. We have made this point clear to every investor including FGIC. Also, please see attached the target portfolio, an abbreviated version of the one Alex had already sent to FGIC on July 19. Rachel will send you the updated and ramped portfolio when she's back on the desk.

247. Delaunay followed up at 6:54 p.m. stating, "I need a detailed target portfolio." His email also did not copy anyone from Putnam. Id.

248. On August 7, 2006, at 12:57 p.m., Menhenett emailed Skelton, Perlman, and Finkel of FGIC. The email stated that Calyon indicated that the target pool would have a WARF of 494. She advised that, within RMBS, "we expect some portion to be rated higher than Triple-B." PX246. And she explained that she was waiting to hear back from Calyon regarding the specifics of additional ramp up to date. She promised to let them know what she heard.

249. At 2:58 p.m. on August 7, 2006, Delaunay wrote to Rekeda again: "Please make sure this is well done and if Putnam could give the names of servicers, etc... as much info as possible. I hear on the call that they want to buy CDO synthetically... what rating, spread, collateral . . . All of this must be reflected in a detailed manner." PX244.

250. A follow-up due diligence meeting—this time telephonic—was scheduled for 4:00 p.m. on August 7, 2006. The reason for the follow-up call was so that FGIC could better understand Putnam's collateral selection process, and because Skelton was unable to attend the earlier meeting. At 1:28 p.m. on August 7, 2006, Menhenett emailed Bell and asked, "Carl, at 4pm, shall we dial your # or John's?" D-J44. Bell responded, "My number." Id.

251. At 3:07 p.m., on August 7, 2006, Menhenett followed up with Delaunay asking if there was "[a]ny chance of getting an updated pool bf 4pm (mbs call w/Putnam)." PX245. The updated pool to which she was referring was the updated detailed target portfolio.

252. The detailed target list was not sent to FGIC before its 4:00 p.m. due diligence call.

253. On the due diligence call, Menhenett and Skelton represented FGIC, while Van Tassel and Bell represented Putnam.

254. To prepare for the call, FGIC had requested additional materials from Putnam. The requests centered on Putnam's RMBS investment process, including how it screened out and analyzed deals. FGIC did not ask Putnam for an updated target portfolio or anything about the specific assets or even general asset categories that Putnam intended for the remaining unramped portion of the portfolio. Van Tassel sent FGIC information about Putnam's investment process for subordinate asset-backed securities and, in particular, home equity subprime RMBS, as well as other information including the types of filters he used to filter out RMBS names (i.e., bonds) from the 500-name RMBS database he had told FGIC he used to select the Pyxis collateral.

255. During the August 7, 2006 due diligence call, Putnam described its collateral selection process and methodology for Pyxis. Significantly, although Van Tassel was the representative of Putnam responsible for selecting RMBS, Menhenett did not ask him for details of what Putnam intended to purchase for Pyxis—either by asset or by asset category. Nor did she ask him to supply a spreadsheet of the target portfolio. The other participants on the call do not purport to recall exactly what was said. Van Tassel, who took the lead for Putnam, stated, "I have participated in many such calls with potential investors during my career, and to the best of my knowledge, nothing unusual occurred during this call that would have caused me to remember what was said on it." Van Tassel Decl. ¶ 64. Bell took the call while on vacation and testified that, accordingly, and because it was a "routine business call" that "occurred 14 years ago" after which "no contemporaneous notes were exchanged," he did "not remember much of the details of what was discussed[.]" Bell Decl. ¶ 96. Menhenett recalls that Bell stepped away at some point during the call and was not present for the majority of it.

256. At 4:45 p.m., during the due diligence call, Delaunay replied directly to Menhenett's earlier email with an email that did not copy anyone at Putnam, stating, "We need more time to do a decent job and provide you with a detailed 'target portfolio,' i.e., with line by line target items. The only thing we can do tonight is re-send you the matrix of intended portfolio, which is attached. (This is a condensed version.)" PX245. Attached was the July 18, 2006 portfolio spreadsheet, showing a target of 3% RMBS prime in cash and 1% RMBS prime synthetically. With respect to the ramped assets and FGIC's initially negative view of the transaction based on the assets ramped to date, Delaunay reiterated that FGIC "in no way . . . could extrapolate [from] the initial purchases [because] the manager's intention was to ramp the lower quality assets first." Id. He explained, "The manager is only starting to buy A and AA as we speak and average portfolio quality will only increase from now on. If you extrapolate you will obviously hate the risk." Id. "This is a much better position for you to be in from an underwriting perspective," he advised, "because the riskier assets are already identified on day 1 and the unknown is on safer assets where you probably have less margin to make mistakes." Id. He added, "[I]f you think Putnam did a good job picking BBB relatively to what is available, then you can trust that they will do the same on higher rated collateral[.]" Id. He concluded, "If Putnam had ramped High quality first: you would love the deal today but would have a less-clear picture in terms of where your risk is going to come from. Also it leaves the deal exposed to the most volatile part of the collateral (high beta) and should the market go the wrong way, it would force the manager to go into collateral they would not have otherwise considered." Id.

257. A great deal of trial testimony was expended on an alleged phone call between Menhenett and Bell during or after the August 7, 2006 follow-up due diligence meeting, during which FGIC claims Bell directed Menhenett to obtain a detailed target portfolio from Calyon. The evidence (or, more properly, lack thereof) to support the content or existence of that call will be explained below.

258. At 6:13 p.m. on August 7, 2006, Menhenett wrote again to Skelton, Perlman, and Finkel. "Further to my initial email," she began, "Putnam's plan was to ramp the lower quality assets first (hence ramped pool data we have showing Baa3). They're just starting to buy A and AA. Calyon is putting together a more detailed target pool. Until then, attached is the best they have . . . ." PX246. She passed along what Delaunay had just re-sent to her: the July 18, 2006 portfolio spreadsheet. She never mentioned Putnam putting together a target pool.

259. Menhenett replied to Delaunay at 6:33 p.m. on August 7, 2006 in two paragraphs. The first responded to Delaunay's agreement to provide a detailed target portfolio. It stated: "thx. We'll assume the target pool mix until you send add'l detail. Specific issuance/vintage can help fine-tuning our assumptions, since we do bench-marking in terms of loss expectations." PX248. The second related to what was in the warehouse at that time, and the contents Calyon expected to add from its communications with Putnam: "fr speaking to Carl at Putnam, he indicates Calyon is managing the warehouse process, and that you have all the details of names that's bought vs ones that's likely but not bought yet. Let me know what you can share. Also, do you have the target ABS CDOs names?" Id.

260. Shortly thereafter Delaunay responded: "We will have something for you tomorrow morning: line by line, what is acquired, what is targeted." Id.

261. The next day—August 8, 2006—at 5:14 p.m., Menhenett emailed Delaunay, Lee, and Rekeda: "Pls send the target pool when you can." PX810. No mention made was made of Putnam.

262. Rekeda responded with a reference to Sachin Anand, a junior analyst at Calyon: "Sachin is working on it now. I told him to send it out tonight." Id. Again, no mention was made of Putnam.

263. At 11:10 p.m. on August 8, 2006, Anand finally emailed the detailed targeted portfolio (what the parties have called "the Peach Colored Spreadsheet" or "PCS") to Menhenett. The transmitted email copied Rekeda, Lee, and Delaunay of Calyon but no one from Putnam, and it stated:

I apologize for the delay in delivering this Portfolio, None-the-less please find it attached to this e-mail.

Just to highlight a few points, the assets highlighted in a peach colour are the target assets we anticipate in the portfolio, they are not yet traded as indicated in the file. The rest of the assets are currently in the portfolio.

I must also remind you that Alex Rekeda has not yet checked over this pool, so please be aware there might be some changes, he might wish to make as the deal ramps up further.
PX253 (emphasis added). The email made no mention of Putnam.

264. As noted, Putnam was not copied on any communications regarding the detailed target list.

265. The first page of the PCS included a disclaimer from Calyon. Although Menhenett testified that she did not read it because it was in "small font" and "not the focus of delivering deal-specific information," Tr. 399, the disclaimer stated in relevant part:

[T]he information contained herein has been prepared solely for informational purposes and is not an offer to buy or sell or a solicitation or any offer to buy or sell any security or to participate in any trading strategy. If any offer of securities is made, it shall be made pursuant to a definitive offering memorandum prepared by or on behalf of the issuer and the co-issuer which would contain material information not contained herein and which would supersede this information in its entirety. Any decision to invest in the securities described herein should be made after reviewing the Offering Memorandum, conducting such investigations as the investor deems necessary and consulting the investors own legal, accounting, and tax advisors in order to made an independent determination of the suit liability and consequences of an investment in securities.
. . . .
None of Calyon or any of its affiliates made any representation or warranty, express or implied, as to the accuracy of the information contained herein and nothing contained herein shall be relied upon as a promise or representation whether as to the past or future performance. No representations are made as to the accuracy of any estimates or projections, or that all assumptions relating to such estimates or projections have been considered or stated, or that such estimates or projections will be realized. CALYON and its affiliates and their directors, partners, officers, employees and representatives disclaim any and all liability as to the information set forth herein or omissions herefrom, including, without limitation, any express or implied representation or warranty with respect to such information.
Id. (emphasis added).

266. Unlike the target portfolio matrices previously distributed that displayed the names of both Putnam and Calyon, the PCS was on Calyon letterhead and contained no attribution to Putnam. While previous portfolios offered category-based descriptions and projections of Pyxis collateral assets, the PCS listed particular securities.

267. The PCS contained roughly 75 columns of information (running from A to Z, AA to AZ, and then BA to BN) regarding 151 assets. Of the 151 assets in the PCS, 98 were listed as already acquired: 82 of them were Baa3 assets from the July 18, 2006 portfolio spreadsheet, and 16 of them were assets that Putnam had arranged for Pyxis to acquire since that date. Generally, these 16 additional assets were higher-rated bonds. However, only one (with a par value of $2.75 million) was listed as RMBS A; that amounted to .28% of the notional value of the portfolio that had been acquired by July 18, 2006. The remainder were listed as RMBS B or C or ABS CDO. The notional value of these 98 assets constituted nearly 75% of the final Pyxis portfolio. None of them were prime RMBS or from 2005H1 or earlier. Of the 151 assets in the PCS, 138 were RMBS deals.

268. The PCS also included 53 "target" assets Putnam and Calyon had not purchased, highlighted in a peach color. The PCS went into a great amount of detail. As noted above, the PCS contained roughly 3,750 separate pieces of information for the specific targeted assets. The 53 target assets were identified by CUSIP, issuer, servicer, underwriter, and ratings (among other identifying information). Of the target assets, 26 were to be acquired synthetically and 27 were in the form of cash assets. Further, 24 assets (totaling $177.15 million) were labeled as "prime" RMBS (although many were not prime but rather Alt-A RMBS). Nine assets were RMBS B/C, ten were HEL, and ten were CDO (all 10 with the underlying CMBS and CDO listed as "Cash CDO"). In terms of ratings, six assets were AA+, 17 were AA, 14 were A, six were A-, six were BBB+, and six were BBB.

15. Evidence Regarding Putnam's Knowledge of or Involvement with the PCS

269. The evidence does not support that Putnam knew of Menhenett's request to Calyon for information about the remainder of the unramped portion of the portfolio before, during, or after it was made, or that Putnam had any role in preparing or reviewing the PCS. Nor does it establish that the PCS was drawn from information provided by Putnam to FGIC.

270. Until August 7, 2006, and thereafter, FGIC's questions to Putnam concerned assets in the ramped portfolio and Putnam's general investment philosophy, including its purchase of lower credit-rated assets first and higher credit-rated assets later. FGIC never asked Putnam questions about specific unacquired assets in the unramped portion of the portfolio or about the allocation among specific asset categories. Indeed, FGIC's questions assumed (prudently) that Putnam had not yet decided, and would not yet need to decide, the assets in the unramped portion. Menhenett conversed with Putnam on July 25, 2006 (about the CLTV on the ramped assets), on August 3, 2006 (to conduct due diligence on Putnam as an institution), on August 4, 2006 (to discuss triggers), and on August 7, 2006 (to conduct further due diligence on Putnam's investment philosophy). It is undisputed that FGIC did not ask Putnam to provide FGIC with information about the yet-to-be acquired assets for the remainder of the portfolio in any of those conversations. Nor did FGIC ask Putnam for that information after those conversations occurred.

271. There is no documentary evidence suggesting that Putnam was aware that FGIC asked Calyon for information about the yet-to-be acquired assets for the remainder of the portfolio or that Calyon had provided such information. Nor is there documentary evidence that the information Calyon provided to FGIC originated with Putnam, was shared with Putnam, or was even known to Putnam.

272. The emails regarding FGIC's need for detailed information regarding the remainder of the portfolio are all between FGIC and Calyon. They neither copy anyone at Putnam nor refer to Putnam (or anyone who works for Putnam). That is true both for the emails between FGIC and Calyon and for the emails within Calyon (on which FGIC is not copied). In Rekeda's 4:45 p.m. email on August 7, 2006 (the one sent to Menhenett during the diligence call she was on with Putnam), he wrote to his Calyon colleagues, "We need more time to do a decent job and provide you with a detailed 'target portfolio.'" PX245 (emphasis added). He continued, "the only thing we can do tonight is re-send you the matrix of intended portfolio." Id. (emphasis added). When Menhenett responded, she said that FGIC would assume that target pool "until you send add'l detail." PX248 (emphasis added). Shortly thereafter, Rekeda responded, "We will have something for you tomorrow morning: what is acquired, what is targeted." Id. (emphasis added). In Rekeda's email from the evening of August 8, 2006, responding to Menhenett's request for Calyon to send the "target pool when you can," PX810 (emphasis added), he stated "Sachin is working on it now. I told him to send it out tonight." Id. (emphasis added). His email referred to a Calyon analyst; no reference was made to Putnam. It is conspicuous that, notwithstanding the perceived urgency of FGIC's request and the urgency and importance with which it was treated within Calyon, there is not a single email from within Calyon indicating that Calyon was soliciting Putnam's help, nor is there an email or any documentary evidence from within Putnam indicating that Putnam's help was sought. Indeed, there would have been no reason for Sachin to work late into the night on the PCS if the PCS was simply being forwarded from Putnam to FGIC.

273. The August 8, 2006 email from Anand to FGIC transmitting the PCS did not include anyone from Putnam and made no mention of Putnam. It referred to target assets as those that "we," meaning on its face Calyon, "anticipated in the portfolio." PX253. It also noted that Rekeda had not checked over the pool, a step presumably unnecessary if Calyon was simply relaying a list assembled by Putnam and which was therefore Putnam's responsibility. And it noted that Rekeda—not anyone from Putnam—"might wish to make [changes] as the deal ramps up further." Id.

274. Likewise, with one significant exception weeks later, FGIC's contemporaneous internal and external emails demonstrate an understanding that Calyon generated the PCS. When Menhenett relayed the PCS to others at FGIC on August 9, 2006, she referred to the document as the attached target pool "Calyon provided," without making reference to Putnam. PX817. The day after receiving the PCS, on August 9, 2006, at 1:33 p.m., Menhenett emailed Lee of Calyon asking about certain asset constraints in the Pyxis term sheet including the maximum percentage of Baa3 assets allowed. Later that afternoon, Menhenett sent the PCS to Skelton and Perlman. Her email stated: "Calyon provided the attached target pool with more color on remaining target bonds. 11.8% Resi A by rating agency designation. Also a portion of the B/C & HEL are above Baa3. Assets highlighted in 'peach color' are the target assets expected to be bought but not yet traded. The rest of the assets are currently in the portfolio. I need to get your re-ratings soon!" PX262. No mention was made of Putnam.

The exception is a reference in the FGIC credit application for FGIC's participation in Pyxis. It states that the "FGIC CDO group was provided information from Putnam regarding the makeup of these new bonds." PX325. Menhenett testified, however, that the only source for that statement was her conversation with Bell. The Court has found that Menhenett's conversation with Bell does not establish that the information in the PCS came from Putnam. It follows that the statement in the credit application, based on that purported source, also cannot support that the information in the PCS came from Putnam.

275. FGIC's argument is also inconsistent with the due diligence call on August 7, 2006. The purpose of that meeting was for FGIC to understand Putnam's collateral selection process. The meeting, and the correspondence before it, centered on the type of detailed review Putnam conducted before deciding to purchase an asset. The diligence call lasted approximately one hour. If Putnam had already assembled a detailed targeted list of specific assets it intended to purchase for the remainder of the portfolio, it is inconceivable that Putnam would not have mentioned it during the call. If FGIC thought that Putnam had a list of assets reflecting its carefully considered view of what it intended to select for the Pyxis portfolio, it is also inconceivable that FGIC would not have asked Putnam for it. The content of that call strongly supports the conclusion that Putnam had not prepared a detailed targeted list of the assets it wished to purchase for the remainder of the portfolio. It also supports the fact that, when FGIC asked Calyon for such a list, FGIC did so because it knew that Calyon would be able to construct a list of assets available for purchase within the asset constraints—but not necessarily the assets that would or were intended to be purchased.

276. FGIC's argument is additionally inconsistent with its actions after receiving the PCS. FGIC asked follow-up questions of Calyon, but not of Putnam. FGIC never circled back to Putnam with respect to any of the yet-to-be-acquired peach-shaded assets.

277. FGIC did request certain information from Putnam, but not regarding the assets in the PCS. On August 8, 2006, for example, in his response to Menhenett's request for information about Putnam's management of three earlier CDOs, Bell sent Menhenett the "time series of portfolio characteristics for our 3ABS CDOs and the most recent collateral marks for each fund." PX254. FGIC did not ask for, and Putnam did not mention, any information regarding the PCS or any of the assets to-be-acquired. The following day, also in response to a request, Bell sent Menhenett "detail from the Trustee on writedowns and credit risk default/defaulted sales for our 3 existing ABS CDOs. The two primary reasons for classifying a sale as credit risk are a downgrade since the time of purchase or a material decline in the credit quality of a collateral pool." PX822. Again, Menhenett did not ask for any information regarding the PCS or the yet-to-be-acquired assets, and Putnam did not mention the PCS or any assets in it.

278. Menhenett was not shy about contacting Putnam. And FGIC and Calyon did not hesitate to include Putnam in emails when requests were being made of Putnam. If FGIC had asked Putnam, or even asked Calyon to ask Putnam, for information, it stands to reason there would be emails reflecting their requests. The fact that FGIC failed to unearth any such emails corroborates that Putnam was not involved in, and did not know of, FGIC's request for a detailed targeted list or Calyon's provision to FGIC of such a list.

279. At no point, before or after the delivery of the PCS, did Menhenett have a conversation with Bell stating that she expected that 11.8% of bonds in the portfolio to be prime RMBS. In fact, after August 7, 2006, FGIC did not communicate with any representative of Putnam until August 17, 2006 (when Menhenett called Bell to say that given the concentrated subprime risk in the portfolio, FGIC could not get comfortable attaching below 40%). There were no conversations about the PCS.

280. Calyon's internal emails further demonstrate that Calyon generated the PCS independently. Its communications do not mention Putnam at all—let alone do they indicate that Calyon received information from Putnam or that Putnam had any involvement in the processing of the information or the delivering of the information to FGIC. To the contrary: Calyon's internal emails reveal that Calyon prepared the PCS without Putnam's involvement or knowledge. On the morning of August 8, 2006, Anand sent Rekeda a spreadsheet titled "Pyxis Warehouse v006 - Dummy portfolio.xls." PX252. Anand stated:

Please review it we can go through it together, when I get back from lunch. But basically I have added the 'dummy assets' and the end of the portfolio list in a peach colour, it should correspond to the test portfolio, where I have adjusted the targets and added assets to fit with what was already bought. I tried to keep it granular and diversified, we might just have to play with some spreads, before we send [sic] only concern is if we do 'adjust spreads, [sic] they can easily cross reference inflated spread with Bloomberg seeing as they asked for specific assets!

281. At 2:50 p.m. on August 8, 2006, Anand emailed Rekeda a revised draft.

282. At 6:35 p.m. on August 8, 2006, Anand sent Delaunay a further revised draft, then titled "Pyxis - Full Warehouse v006 - 20060808.xls." PX251. Anand told Delaunay: "Here is a file for FGIC. We have taken assets from other Managers deal's [sic] that will fit criteria of portfolio, I have highlighted all the new assets in Peach and have indicated in the trade column an 'N' for not traded. Alex has looked it over and thinks it should suffice for their purposes." Id.

283. Further documentary evidence supports Anand's statement that the PCS was drawn from prior deals that did not involve Putnam. Before Pyxis, the last CDO that Putnam managed closed in 2004. The assets in the PCS were from 2005 and 2006. By simple logic, the reference to other managers' deals was not a reference to Putnam deals.

284. FGIC chose not to call Delaunay or Anand as witnesses at trial. If Anand had been called, he could have been asked the simple, dispositive question in this case: where did you get the information that you put into the PCS? Delaunay too presumably could have provided direct evidence regarding Putnam's involvement with or knowledge of the PCS, if Putnam had any knowledge or involvement.

285. FGIC did call Rekeda as a witness, but his testimony undermined FGIC. His most probative evidence, if it could be called evidence and not speculation, was his deposition testimony that Anand would have "engage[d] in some exercise to figure out what is it that Putnam likes" and "theoretically could buy," and while that exercise could include contacting Putnam, it could also be limited to speaking with Calyon's traders and asking what they thought Putnam liked. Tr. 1222. Thus, his conjecture was consistent with the notion that the spreadsheet was entirely Calyon-generated. Rekeda remembered nothing about the actual preparation of the PCS. He testified that he did not recall learning that FGIC wanted a detailed target portfolio with line-by-line target bonds and their credit characteristics. He did not recall Delaunay's email conveying that request. The contents of the email did not refresh his recollection. He had no recollection of a call with Putnam concerning the types of assets they wanted to buy for Pyxis at that time. He did not recall being on a phone call with Putnam at any time around the date of August 7, 2006 or August 8, 2006. His recollection could not be refreshed that Calyon was preparing a detailed target portfolio for FGIC. He did not recall Anand's email sending him a draft of the PCS, or the draft PCS itself. On the critical question, he did not recall how Anand selected the target assets for the PCS because he did not remember the document at all. He did not recall reviewing the PCS, and had "no idea" whether he did so or discussed it with Anand. Tr. 1203-05. He also had "no idea" whether it was correct, per Anand's email, that Rekeda had looked over the PCS. Tr. 1207. He stated generally, and without regard to the PCS, that "target assets are the assets the manager would consider buying into portfolio and consider a good fit for the portfolio," and that while he did not remember the PCS, he believed there were "some reasons why those assets were selected" and "some justification why those specific assets would have been a good fit." Tr. 1208. However, he did not know those reasons in the case of the PCS, had "no idea . . . how these assets were selected," Tr. 1211, and did not know whether it was accurate that the assets in the PCS were "target assets we anticipate in the portfolio," Tr. 1213. When asked for his best understanding of what basis he had in 2006 to believe that the target assets were assets the manager considered a good fit for the portfolio, he responded: "I have no idea what was the basis, how these assets were selected. I don't remember this e-mail, and I don't remember me checking this portfolio or not checking. I don't even see the portfolio at this point. So it's hard for me to comment on something that I have no idea." Tr. 1211. When asked who the "we" referenced in Anand's statement that the PCS reflected the assets "we anticipate in the portfolio," Rekeda answered, "Could be anyone. Could be he and the structurer. Could be he and a trading desk. Could be he and all of us together. It could be he and Putnam. I have no idea." Tr. 1221. He stated, without contradiction, that he had no knowledge of the source of the information that Anand used to prepare the PCS and that he was not aware of any input by Putnam. Rekeda did not know whether Calyon even had, for its own purposes, a list from Putnam of assets to be purchased into the portfolio. He stated that, while the warehouse provider cares about the assets in the portfolio because it bears the risk of loss, it protects itself by understanding the constraints on which assets the manager can purchase. Rekeda was not even asked whether he told Putnam, before or after the fact, about the PCS or the request for it. In short, Rekeda did not provide any evidence to support FGIC's claim.

286. For his part, Bell testified credibly that his team could not have known in advance what assets they would have the opportunity to buy (or which assets they would buy) because there was a good deal of uncertainty in the buying process. As Bell explained, it takes months to construct the full portfolio for a CDO like Pyxis, and market conditions can change materially over that time period, transforming the attractiveness of a particular asset or asset class. Bell testified in his declaration, "[I]t makes sense that I would direct Ms. Menhenett to Calyon for the details of . . . what was in the warehouse [because] Calyon was funding the warehouse and was the official record keeper for all transactions." Bell Decl. ¶ 98. But, he stated, "I would not have directed Ms. Menhenett to Calyon to obtain a target portfolio of specific targeted assets because I had no reason to believe that such a portfolio existed." Id. ¶ 102. He explained, "I, and to my knowledge, members of my team, never created such a target portfolio of each specific collateral asset that we might seek to acquire in the Pyxis transaction." Id. ¶ 99. On August 7, 2006, he "did not know what specific assets Putnam would buy, or exactly what percentage of prime or earlier vintage RMBS assets Putnam would buy, because that would depend on market conditions." Id. ¶ 100. "Such a portfolio would have taken significant effort to prepare, and it would have been based largely on speculation, because Putnam could only acquire in the marketplace assets that an independent third party wanted to sell, at a price we thought was appropriate." Id. ¶ 99. He stated that the only target portfolios sent to potential Pyxis participants (of which Putnam was aware and commented on) were those listing the percentage and notional amount of various types of assets or assets of different credit ratings.

287. Bell also testified credibly that he never saw the PCS before this litigation and does not know where Calyon obtained such a list. He asserted that it "was not provided by Putnam." Bell Decl. ¶ 103. In his declaration, he explained:

As the Senior Portfolio Manager for the Pyxis CDO and the point person at Putnam for communications with Calyon, it would not make sense for someone at Putnam to create such a portfolio without involving me in the process. I do not recall any conversations with anyone at Calyon in any way related to such a target portfolio that was created at FGIC's request, and I do not believe any such conversation occurred, given the creation of such a portfolio would have been highly irregular and outside of what I would have thought made sense or could be readily achieved.
Id. ¶ 103.

288. Bell also testified that no person at FGIC ever asked him or, to his knowledge, any members of his team any questions about the PCS. He further testified that he did not have any conversation with anyone at Calyon about Pyxis's targets in the days leading up to August 8, 2006. Bell, according to his testimony, never expressed a view to Calyon that any Pyxis targets should include 10% prime or any particular amount of earlier vintage RMBS assets. He does not recall any discussions with Calyon over targeted amounts of seasoning for any targets and does not believe any such discussions occurred.

289. Bell was thoroughly cross-examined about his declaration. He maintained that Putnam "did not have a detailed list of all of the assets" it was going to select for the Pyxis portfolio, and could not have prepared such a list. Tr. 1470, 1600. He explained that Putnam could have assembled a partial list of assets it was actively seeking to acquire at a point in time, but that "it was an iterative process where we would see how we're doing versus weighted average spread, all the different characteristics that we talked about before." Tr. 1599. "I don't think [Van Tassel] could construct a line-by-line target portfolio that would forecast what he could purchase over six months at what spread and comply with all the myriad of tests," Bell explained. Tr. 1600-01.

290. Although Bell admitted that it was possible for Menhenett to have asked him for a target portfolio, he was definitive and persuasive that he did not promise her a detailed targeted list of remaining assets and would not have told her to obtain such a list from Calyon. If he had told Menhenett to obtain anything from Calyon, Bell testified, he would have expected it to be the target portfolio from the marketing materials.

291. Bell was also cross-examined about an email that he wrote to Lee on July 31, 2006, in which he asked to receive "a post call/meeting summary for each investor conversation" including the "name, title, [and] contact info" for "each person that was on the call (including the Calyon salesperson)." PX230. The email and Bell's testimony were telling with respect to Putnam's role in the Pyxis transaction and the limits to its role: in essence, Putnam had an interest in the meetings it participated in and in keeping a record of them. It had no role with, or professional responsibility with respect to, meetings Calyon had with investors without Putnam about whether and how those investors should participate in the Pyxis transaction. Thus, Bell made clear that, in his July 31, 2006 email, he was referring to conversations in which he was on the phone, and for which he wanted "help in recording the details of those conversations" so that he could keep a record of people he "had contact with." Tr. 1475-76. He did not recall asking for any post-call or post-meeting summaries for any investor conversations that Calyon had without Putnam, and the record reflects no evidence that he did. Based on his regular practice, he would have made at most for informational purposes "general inquiries [to Calyon] about how marketing [was] going" but not "specific information" in the form of post-call or post-meeting summaries of Calyon's meetings with investors in which Putnam did not participate. Id.

292. Van Tassel similarly testified by declaration that he was unaware of anyone at Putnam providing the information in the PCS to Calyon, and that he did not recall seeing or knowing about the PCS at any time prior to this litigation. He testified that he did not know "where or how Calyon would have obtained a list of specific, yet-to-be-acquired assets for Pyxis." Van Tassel Decl. ¶ 67.

293. Both Van Tassel and Bell testified compellingly to Putnam's preference for subprime over prime securities (and that FGIC did not express to them a preference for prime securities). Their testimony was detailed and credible. In short, in 2006, there was more credit enhancement available for subprime securities than prime, and thus Putnam believed subprime securities were a safer and better investment than prime securities. It is implausible that Putnam was the source of a document reflecting that the Pyxis portfolio would have 11% prime securities when, as both Bell and Van Tassel testified persuasively, they would never have thought that it was in Pyxis's interest to accumulate such a large share of prime RMBS securities at the expense of other securities including midprime and subprime that were better credits for the transaction. Indeed, if FGIC thought it had or was going to get a detailed target list of all securities to be added to the Pyxis portfolio, it would have had little reason to spend hours asking Putnam how the collateral manager planned to select securities in the future and ample reason to request, simply, that very list (and to question Putnam about that list).

294. No copy of the PCS, or anything resembling it, was introduced in evidence as having come from Putnam's files.

295. Putnam did not maintain a target list, bond-by-bond, of all assets to be acquired for the transaction. As Bell described, the asset acquisition process was iterative and dynamic. Putnam looked for market opportunities to fill certain buckets in the transaction (in the case of Pyxis, starting with lower rated assets to be acquired in smaller sizes synthetically). The initial acquisitions determined what Putnam was able to purchase with the remaining, unfilled portion of the portfolio within the constraints of spread, correlation, and the pool's cumulative rating. At the same time, there would be changes in the market as to the assets available and at what price. It thus would have been impossible to predict the securities Putnam would have purchased for Pyxis with any reliability, and Putnam had no reason to do so. Rather, Putnam created bid lists that reflected what it wished to acquire during a particular week for a portion of the transaction. Putnam provided those bid lists to Calyon. In short, Putnam did not have a detailed targeted list of assets for the portfolio as a whole to give Calyon and could not have generated such a list without significant work.

296. FGIC offered no convincing evidence or argument in opposition to these points. Its argument that Putnam was aware of the request for information about the yet-to-be-acquired assets for the remainder of the portfolio hinges entirely upon a single email from FGIC to Calyon, the deeply-flawed testimony of a single witness (Menhenett), and alleged circumstantial evidence including that of motive and opportunity. None of that evidence supports FGIC's claim that Putnam was aware before or after the fact of FGIC's request to Calyon or that Putnam delivered any information about the PCS, much less that Putnam had any involvement in creating it.

297. The single email was sent by Menhenett to Delaunay on August 7, 2006 at 6:33 p.m. As noted above, that email was in two parts and had two paragraphs. The first paragraph responded to Delaunay's agreement to provide a detailed target portfolio. It stated: "thx. We'll assume the target pool mix until you send add'l detail. Specific issuance/vintage can help fine-tuning our assumptions, since we do bench-marking in terms of loss expectations." PX248. The second paragraph referred to Putnam. It stated: "fr speaking to Carl at Putnam, he indicates Calyon is managing the warehouse process, and that you have all the details of names that's bought vs ones that's likely but not bought yet. Let me know what you can share. Also, do you have the target ABS CDOs names?" Id.

298. After hearing all of the evidence, the Court draws the most powerful and logical inference from the email's text and structure: that the two paragraphs refer to two separate and distinct requests. The first paragraph pertains to FGIC's request to Calyon earlier that day for detailed information about the yet-to-be-acquired assets for the remainder of the portfolio, and Calyon's promise to provide that information. As noted, Putnam was not copied on that correspondence or even mentioned in it. It would not have had such a list. The second paragraph involves something entirely different—the assets Putnam had purchased for Pyxis since July 18, 2006, and any other specific ones that Putnam intended to but had not yet purchased. The two were different, and the second was not even a subset of the first. The first necessarily would have to reflect hypothetical assets—assets Putnam would be able to purchase consistent with its investment thesis and the portfolio constraints. (It would not have to even begin purchasing a full 15% of the portfolio until two months after the October 2006 closing.) The second related to real assets, which Putnam knew were available in the market and could be purchased within the portfolio constraints.

299. That interpretation is consistent with the language in Menhenett's email. In particular, Menhenett would have had no reason to ask what Calyon could "share" regarding assets "likely but not bought yet," if—as Menhenett also testified and as was established by the documentary evidence—Calyon had already promised to provide FGIC a detailed targeted portfolio listing the "target" assets for the entire remainder of the portfolio. In addition, Menhenett would have had no reason to ask for the "target ABS CDOs names" if those were the same assets referred to in the first paragraph regarding the information Calyon had already agreed to provide. Nor, in referring to the "target ABS CDOs names," did Menhenett indicate Bell had told her Calyon had those names from Putnam, the inference one would have to draw if—as Menhenett claimed—Bell told her Calyon had the target portfolio for the rest of the portfolio.

300. The inference is inescapable that Menhenett was referring to two separate (but related) categories of information: (1) a set of "target" assets that could be purchased within the asset constraints and would satisfy the cash flows necessary to pay down the debt that Pyxis was expected to incur and that could be used by FGIC for modeling purposes; and (2) the more limited (and different) collection of actual assets already purchased and those assets (not comprising the remainder of the portfolio) Putnam was in the process of trying or expecting to purchase for the portfolio.

301. FGIC also relies upon the trial declaration of Menhenett. No other witnesses testified to any knowledge of Putnam's involvement in or knowledge of the PCS.

302. Menhenett testified in her declaration that after reviewing Delaunay's mid-call email on August 7, 2006, she left the due diligence meeting with Van Tassel (in which Bell had participated) and called Bell. Menhenett Decl. ¶ 140. She stated that she did so in order to procure the detailed target portfolio that Calyon had stated in its email to Menhenett that it was working on. Id. ¶ 141. She further testified by declaration that, when she reached Bell, she told him that FGIC "needed to know the actual bonds Putnam was planning to purchase, just as [they] did for the ramped assets, and [that they] did not merely need a Matrix like the one Calyon had given [them] back in mid-July." Id. ¶ 142. Menhenett's declaration states that Bell told her that she "should get the detailed target portfolio from Calyon." Id. ¶ 143. Menhenett's declaration further states that Bell told her that Calyon "was managing the warehouse process and that Calyon had all the details of the names that had already been bought for the portfolio as well as the names that were likely but had not been bought yet." Id. Menhenett testified by declaration, "I am confident that those were Bell's words almost verbatim because, immediately after I got off the phone with him, I emailed Delaunay to let him know precisely what Bell had told me, hoping that would prompt him to respond more quickly." Id.

303. As the Court advised at the outset of this opinion, Menhenett was a deeply flawed witness and unreliable reporter. At trial, she contradicted the answers she gave in her deposition several years earlier and she disavowed portions of her declaration that she had signed only weeks earlier (and that was submitted as her direct testimony). Her answers were frequently evasive and unresponsive, and her demeanor differed dramatically depending on whether she was on cross-examination or re-direct examination. The cross-examination made clear that, notwithstanding the language of her declaration, she was testifying to significant events based on a reconstruction of what she believed must have happened based on her review of emails (which was selective) rather than an actual recollection of what happened. Even when she did not directly contradict her declaration, cross-examination indicated that her declaration was selectively drafted to highlight events that could be favorably construed to Plaintiff's case while entirely ignoring and obscuring events and conversations that could be construed to be unfavorable.

For example, her declaration fails to mention the crucial events from 2007, when FGIC discovered the amount of prime RMBS in the actual Pyxis portfolio, that will be described below. It is also noteworthy that, in 2013, Menhenett submitted an affidavit in a Massachusetts securities enforcement proceeding involving Putnam. The affidavit summarizes Menhenett's experience with the Pyxis transaction. At no point does it mention FGIC having been misled or defrauded in connection with a false target portfolio. It does state, however, that Pyxis "was designed to be a CDO whose underlying collateral would be managed by Putnam, as a result of which the actual collateral at any time could change substantially." D-B81.

304. Specifically, her account of the August 7, 2006 events at trial differed materially from her account three years earlier at her deposition. In 2017, Menhenett testified repeatedly that her conversation with Bell occurred during and at the end of the August 7, 2007 diligence call. For example, she stated that "on the August 7 telephone due diligence . . . toward the end I did speak to Carl." Dkt. No. 317-9 at 150. She stated, "toward the end of the call I did ask Carl about - I'm not using the same word 'target.' But I mean, what other names do you intend to put in the deal?" Dkt. No. 317-10 at 222. She was asked, "And it was during this conversation that you recall him talking about the target portfolio?" Dkt. No. 317-10 at 414. And she responded, "It was toward the end of it." Id. When asked why this conversation with Bell was not memorialized in her notes from the August 7, 2006 call, she testified, "I recall"—recall—"feeling that, oh, you know, I wasn't really taking down other details because it's [Skelton's] call." Id. That testimony left an important vulnerability in FGIC's case. No one else on the due diligence call remembers FGIC asking for or Putnam promising a detailed target list of the rest of the Pyxis portfolio.

305. In her declaration authored three years later, and 14 years after the conversation, Menhenett changed course. She testified in her declaration that she called Bell "following [the] due diligence meeting with Putnam[.]" Menhenett Decl. ¶ 140. At first, when Menhenett was asked about the discrepancy, she tried to deflect: "I think there is a little bit of vagueness in that previous deposition that is incorrect because it doesn't match with some of the other memory I have." Tr. 333. Then she rejected the inconsistency: "I think at the original deposition I did mention that the conversation with Carl regarding the additional names happened after the . . . due diligence call had ended, so it was after the hour-long call had ended." Id. Eventually, she acknowledged the inconsistency, but she immediately attributed her misrecollection at the deposition to the fact that "at the time," she "did not get an opportunity to review the written email exchanges." Tr. 337. When defense counsel clarified, "Is it your testimony that you weren't shown [the relevant exhibit] during your deposition?" she conceded, "It's not." Tr. 337.

306. Yet newer details apparently surfaced in her memory during cross-examination, even since the time she authored her declaration (a few weeks earlier). She testified that she saw the 4:45 p.m. email on her BlackBerry during the due diligence meeting, then "left the conference room briefly," and then "went back to it," but by that time, "the call had already ended." Tr. 334. She explained that when she left the room, she called Bell but she "must have left him a voice message" because it was "later on" when they eventually spoke. Tr. 335, 357. Admitting she could not have left a voice message for Bell at his office number during the due diligence call (because that number was being used for the due diligence call), she was unable to answer where she would have called him if not at his office, only speculating that she must have called his cell phone. Tr. 354-55.

307. In Menhenett's declaration, she states that she told Bell that she was "still waiting on Calyon for the detailed target portfolio" and that Bell told her she "should get the detailed target portfolio from Calyon." Menhenett Decl. ¶ 142. The declaration continues, "[Bell] told me that Calyon was managing the warehouse process and that Calyon had all the details of the names that had already been bought for the portfolio as well as the names that were likely but had not been bought yet." Id. ¶ 143. It insists, "I am confident that those were Bell's words almost verbatim." Id. But under cross-examination, Menhenett conceded that she did not use the word "target" when asking Bell for the names that Putnam was intending to put in the deal, and indeed that the words "detailed target portfolio" did not come out of her mouth. Tr. 363.

308. On cross-examination, Menhenett disavowed other significant portions of her declaration. In her declaration, Menhenett stated she told Bell FGIC "needed to know the actual bonds Putnam was planning to purchase, just as [they] did for the ramped assets, and [that they] did not merely need a Matrix like the one Calyon had given [them] back in mid-July," and Bell told her to "get the detailed target portfolio from Calyon." Id. ¶ 143. On cross-examination, however, she conceded that she did not use the words "actual bonds," as she testified on direct by declaration. Tr. 365. She admitted, consistent with her deposition testimony, the words she used instead were "what other names do you intend to put in the deal." Tr. 364. She also admitted that she did not tell Bell that FGIC "did not merely need a matrix like the one Calyon had given us back in July." Tr. 365, 368. She conceded she did not reference the July 19, 2006 Putnam target portfolio or a matrix during their conversation. Tr. 368. Indeed, at trial, she recalled differently, explaining that she actually stated, "[T]o the extent you've added more names or identified names you are going to buy, we need to know that." Tr. 368. As she phrased it on redirect examination, "[I]f they purchased more bonds since the initial 82 we saw, then please let us know." Tr. 556. According to her trial testimony, it was "a very quick call that we needed more names." Tr. 557.

309. Thus, even crediting her testimony on cross-examination and redirect, there is no swearing context. Menhenett's version of the conversation with Bell—at least how she put it on cross-examination—is entirely consistent with Putnam's argument and inconsistent with FGIC's claim. When Menhenett spoke to Bell on August 7, 2006, she asked for an updated ramped portfolio including assets that had been purchased since July 18, 2006 and ones that Putnam was in the process of trying to purchase. She did not ask for a complete detailed list of all of the remaining assets in the $1.5 billion portfolio, whether they were in the process of being acquired or not. She asked for the names of assets that Putnam had identified that it was going to buy and not for all of the names that it would buy eventually. Not only is that what Menhenett testified to and what was in the email but it is also the most sensible way to construe the evidence. Significantly, however, it is entirely different from a request for a list of target assets for the remainder of the portfolio. As indicated, updates to the ramped portfolio were a matter of keen interest and importance both to FGIC and Calyon. FGIC had last received a list of the assets in the ramped portfolio on July 18, 2006, in which 79 of the 82 assets were rated Baa3. FGIC analyzed that list and began to develop a negative view of the transaction. However, once new assets were acquired, FGIC requested to see those assets since Calyon had indicated they were more highly rated. Therefore, it was important to FGIC to confirm Calyon's representation and the quality of those assets currently in or being added to the warehouse. It also follows that FGIC would make the request to Putnam. Not only did Putnam know best what assets it had purchased and was about to purchase, but FGIC had also requested such information from Putnam in the past and received it. But, to state it again, the statement that FGIC should go to Calyon if it wanted to know what Putnam had purchased or was in the process of purchasing is entirely different from a statement about a list of target assets.

310. The remainder of FGIC's argument rests upon speculation and inference. On August 5, 2006, Rekeda proposed to have a call with Bell at 9:00 a.m. on August 7, 2006 regarding FGIC's tail-risk concern. PX241. Bell agreed. The call did not take place as scheduled—or even, reasonably likely, at all. At 10:47 a.m. on August 7, 2006, Bell emailed Lee and Rekeda, "I was expecting to hear from you at 9:00 a.m. on FGIC." PX796. Bell's email noted that Putnam was simultaneously managing "IKB going to Committee," "Other Investor Calls," and "Asset Aggregation." Id. The lynchpin of FGIC's theory that Putnam gave Calyon the information in the PCS is that this call took place at some point later on August 7, 2006 and that on the call, Putnam gave Calyon the information contained in the PCS. Rekeda, for his part, testified that he had no memory of a conversation with Bell about managing FGIC's tail risk-concern. FGIC did not call Lee to testify. Bell testified, credibly, that he did not remember whether any conversation about managing FGIC's tail-risk concern ended up taking place over the phone or whether, instead, the matter was worked out over emails. He did not recall discussing the issue with Rekeda or Lee on August 7, 2006.

311. As evidence that this call did take place, FGIC relies on an email sent on August 7, 2006 at 12:06 p.m. from Lee to his Calyon colleague Ralf Otzen, CC'ing Rekeda, Anand, and two other Calyon employees. PX792. The reliance is curious. The subject line of the email is "Pyxis - IKB eligibility criteria." Id. It does not refer to FGIC at all. It refers to IKB, the mezzanine debt investor, who had expressed interest early in the transaction but who had also negotiated certain features of the transaction. The body of the email refers to the fact that IKB was going to its credit committee, that it was asking for revisions of the transaction, and that IKB needed to get an answer from Putnam because "[f]or IKB it [was] essential to have . . . this commitment from Putnam." Id. Lee responded to the email from his Calyon colleague about when IKB could receive an answer: "We're on the phone with Putnam." Id. FGIC draws attention to the fact that, a couple hours later (at 2:58 p.m. on August 7, 2006), in Delaunay's email to Rekeda responding to the request for (1) an updated warehouse and (2) a detailed target portfolio, he stated: "Please make sure this is well done and if Putnam could give the names of servicers, etc... as much info as possible. I hear on the call that they want to buy CDO synthetically... what rating, spread, collateral, they also want to buy AA CRE CDO . . . All of this must be reflected in a detailed manner." PX244 (emphasis added). From these two documents, FGIC would have the Court conclude that Putnam had a call with Calyon at 12:06 p.m. on August 7, 2006 and that, on that call, Putnam gave FGIC the information that comprised the PCS. There are several fatal flaws in that logic.

312. First, there is no evidence that the call referenced in the 12:06 p.m. email about IKB was at all related to any communications from Putnam about "the names of servicers, etc..." Id. From all that appears, it was about IKB—another significant investor as to which there was an issue of urgency. Second, FGIC's argument has an Alice-in-Wonderland type quality to it. Menhenett's first request for the detailed target list did not come until after August 7, 2006, at 12:38 p.m., when Menhenett asked Rekeda to give her a call in light of "not good on the surface" feedback from the RMBS team. Id. (Although Menhenett claimed at trial that she made a previous oral request for such a list, no evidence supports that contention.) But, if the request was not made until after 12:38 p.m., the 12:07 p.m. call could not have been one in which Calyon requested and Putnam delivered the information regarding target assets that FGIC was seeking. Third, the PCS contained approximately 75 columns of information for each of the 53 targeted assets, totaling roughly 3,975 pieces of information. The notion that such information could have been supplied by Putnam on a call with Delaunay does not stand to reason. Nor does it stand to reason that someone from Putnam would recite 53 assets over the phone as opposed to sending a spreadsheet or email. Any one of these points would be fatal to FGIC's argument. In short, not only is it speculation, but it is speculation undermined by the evidence.

313. FGIC also argues that, because Putnam allegedly had motive to deceive FGIC and an opportunity to do so, the Court should infer that Putnam was involved with, or at least knew about, the PCS. FGIC did not prove this theory at trial.

314. FGIC has not cited any law for the proposition that authorship of a misleading statement can be established by evidence that the defendant had the motive to make the misstatement. Even if it had, the evidence does not establish either motive or opportunity. First, there is no evidence Putnam had a motive to deceive FGIC. As laid out above, under the Warehouse Agreement and the Mandate Letter, Putnam was entitled to a fee even if the transaction did not close. Although undoubtedly it earned additional fees from the transaction closing, that would not give it a motive other than the ordinary one a collateral manager would always have in a transaction; certainly, it would not give Putnam a motive to make a misstatement. Moreover, from the evidence at trial, Putnam would have had no reason to believe that if FGIC did not participate as Calyon's counterparty, another guarantor would not emerge to take FGIC's place. After all, when FGIC declined the offer to attach at 28%, a new guarantor appeared. Nor, even if Pyxis failed, is there reason to believe that such failure would affect Putnam unusually or redound unfavorably on Putnam or make it more likely that Putnam was involved in the PCS as opposed to Calyon unilaterally. Putnam was the collateral manager for Pyxis, not the structurer. It was not responsible for obtaining investors. Putnam did not have transparency into that process. The evidence establishes that, in 2006, there were entities interested in participating at a senior level in midprime and subprime transactions. There is no evidence that Putnam thought, or was told, that if FGIC did not participate, no other monoline would come forward. Nor was Putnam the one seeking a counterparty—that was Calyon. Certainly, as Bell acknowledged, if no counterparty emerged for Calyon, Pyxis might not close, which would result in one less structured finance transaction; however, this would not distinguish Putnam's interest in closing Pyxis from its interest in another transaction.

315. Second, FGIC did not prove anything close to opportunity. In fact, the evidence proves the opposite. The evidence was uncontradicted that Putnam did not have a detailed targeted list of assets to give FGIC, because assembling such a list would have been speculative and time-consuming. As importantly, Bell testified credibly that he never considered the possibility of changing the portfolio's composition to keep FGIC in the transaction. The reason is obvious. The parameters of the Pyxis portfolio were contracted by way of indenture. Putnam's ability to select assets within the asset constraints were also under contract, first through its agreements with Calyon and later, as will be explained, through its agreement with Pyxis. Putnam was bound by law to make decisions in compliance with its fiduciary duty to Pyxis; it did not owe fiduciary duties to the underlying investors or participants. Favoritism towards FGIC, at the expense of those other investors and participants, would have been a violation of Putnam's duties both to Pyxis and, derivatively, to all investors at other points in the capital structure. And such a violation would have been easily discovered: the portfolio's composition was available to all investors, on request prior to closing and as a matter of right after the closing.

FGIC also asked the Court to infer from the fact that Putnam closely reviewed and edited the pitchbooks prior to the launch of Pyxis that it must necessarily have reviewed and provided comments on the PCS. That conclusion does not follow. As indicated above, the Mandate Letter required Putnam and Calyon to cooperate on the marketing materials and prevented Calyon from launching the transaction until it received Putnam's consent. It thus makes sense that Putnam commented on the pitchbooks. It does not follow, and the evidence does not support, that Putnam was involved in all (or even most) other communications that Calyon had with investors and participants after the launch.

16. Communications with Rating Agencies

316. As noted above, Calyon was required under the terms of the Mandate Letter, to "work with the rating agencies to obtain desired ratings for the [n]otes." PX719. To that end, on August 11, 2006, Calyon sent modeling assumption packages to Moody's, S&P, and Fitch.

317. Before Calyon provided its cash flow modeling assumptions to the rating agencies, Bell took certain steps to ensure that the rating agencies would receive correct information regarding the Pyxis cash flows, including providing comments on the Pyxis cash flow waterfall. Bell noted to Lee, Rekeda, and others at Calyon that it was "critical" for that to be "documented in an extremely clear fashion" because "an unambiguous waterfall is clearly a pre-req[uisite] for pricing." PX265.

318. After Calyon sent the modeling assumption packages to the ratings agencies, on August 14, 2006, Bell asked Calyon to send them to Putnam. The next day, Calyon sent the packages to Putnam.

319. FGIC argued at trial that the rating agency portfolios were materially false in that they showed inflated amounts of prime RMBS assets, pre-2005H2 RMBS assets, and assets rated above Baa2 and BBB. FGIC asked the Court to draw three inferences from the portfolios: (1) that Putnam had the theoretical ability to put together a detailed target portfolio like the PCS; (2) that the rating agency portfolios, like the PCS, reflected Putnam's intent to acquire the specific assets therein; and (3) that because Putnam knew the rating agency portfolios were false, Putnam either knew or should have known about the PCS and that it was false (and therefore had a duty to correct it).

320. In particular, FGIC focused on the draft Fitch presale report for Pyxis, which Calyon sent to Bell on September 13, 2006 along with a request for his comments. The Fitch draft presale report set forth a breakdown of the Pyxis portfolio that was based on the August 11, 2006 rating agency portfolio Calyon had provided to Fitch. The breakdown displayed 10.4% prime RMBS and 24.01% assets rated above BBB. FGIC faults Bell for failing to point out the allegedly false information about the quantity of prime and higher-rated assets. It also reasons from Calyon's provision of this information to Fitch, and Putnam's non-correction of it, that Putnam also was aware that incorrect information was provided to FGIC, that Putnam knew of it, and that Putnam disregarded a duty to correct it.

321. The evidence does not support FGIC's argument either in its specifics or more generally. Bell testified that he does not recall receiving the draft presale report. But, after reviewing the evidence, he also testified credibly and consistently that he was not surprised he did not raise the issue that the communication to Fitch reported more prime RMBS than communications to other rating agencies did. Bell explained that it was "well known that there were inconsistent RMBS bond classifications among the rating agencies," that Fitch was particularly likely to be different from other rating agencies because it "had only rated approximately 50% of the underlying bonds in the Pyxis portfolio," and that, unlike Moody's, Fitch did not have a midprime category (such that Fitch, or Calyon in reporting to Fitch, would have had to determine whether each midprime bond belonged in the prime or subprime category). Bell Decl. ¶ 159. O'Driscoll, FGIC's own expert, confirmed that Fitch did not use a midprime RMBS category in 2006. It was reasonable for Bell to have assumed that the higher-rated midprime assets would have been reallocated to the prime RMBS bucket for the Fitch draft presale report and not jump to the conclusion that Fitch's report was mistaken.

322. The evidence also does not support the notion that Calyon, or Putnam, was trying to deceive Fitch. In particular, on September 12, 2006, Calyon emailed Fitch, "Can you please revise[] the presale report according to the attached materials[?]" PX959. The attached materials were the preliminary offering memorandum and a September 7, 2006 term sheet—also sent to FGIC (as will be discussed later) which contained a revised target portfolio showing 2% prime RMBS. If Putnam had been directing Calyon to supply inflated representations of prime assets to Fitch, it had been doing so quite unsuccessfully. Nothing about the presale report—or Bell's interaction with it—supports the conclusion that Putnam intended to mislead FGIC via the PCS.

323. More generally, each of the proposed inferences FGIC would have the Court draw fails. The first proposed inference is immaterial; Putnam has not argued that it would have been impossible to put together a detailed target portfolio like the PCS. To the contrary, Putnam has embraced Bell's acknowledgement that it would have been "feasible" to put together a detailed target portfolio. Tr. 1599. But, as Putnam emphasizes and as Bell explained, because asset acquisition was an "iterative process" that depended on market availability and fluctuations, Putnam would have to "iterate again" after any such portfolio was constructed. Id. Therefore, Putnam did not generate a detailed target portfolio like the PCS. Van Tassel testified consistently that it would have been theoretically "possible" to put together a list of assets Putnam "would like to acquire for Pyxis" but that limitations on doing so "had to do with market appetite and not flooding the market." Tr. 1936. The modeling assumption packages theory establishes the opposite inference from that which FGIC would have the Court adopt: that Calyon, as the structurer, was equally capable as Putnam in generating a list of target or dummy assets.

324. Turning to the second and third inferences, the evidence did not establish either that the rating agency presentations were intended to be a reflection of Putnam's projections of the final Pyxis portfolio or that Putnam should have drawn the conclusion, from the fact that the ratings agencies received "detailed" information that did not correspond to Putnam's intent for the portfolio, that FGIC would also have received such information—much less a duty on Putnam's part to correct any such information. The modeling assumption packages served a purpose different from what FGIC claims the PCS was intended to serve. Numerous witnesses testified about the purpose of the information provided to the rating agencies. The rating agencies understood that no projection of what a collateral manager intended to purchase, if delivered mid-marketing of the transaction, could substitute for what the manager actually purchased and that the best way to stress test a transaction's structure was to look at the portfolio constraints and construct a hypothetical portfolio that would be within those constraints. Dolan explained that rating agency portfolios were not designed to convey the actual assets that Putnam intended to buy. Rather, the dummy assets in the rating agency portfolios were representative assets that were provided to the rating agencies so the rating agencies could run their models and assign preliminary ratings. Dolan testified that market participants would have been aware, at the time, that ratings would need to be confirmed by the rating agencies after the deal closed based on the final portfolio. In the case of Pyxis, this process was laid out in the indenture and it transpired accordingly.

325. Rekeda described how the structuring bank provided "placeholder" assets for the rating agencies—not assets that were "likely to be acquired for the portfolio," but rather "assets that potentially could be acquired." Tr. 1030-32. He explained, "[I]n general, no deal is ramped before the closing. So typically the deals are ramped 60, 70 percent prior to the closing. And after closing the manager continues to buy whatever the assets that fit into the box that's allowed by the indenture." Tr. 1031. Rekeda also explained that the rating agencies were most interested in the asset constraints for the portfolio (which he called the "box") and the waterfall (i.e., how cash would proceed from the bonds to the administrator). Tr. 1028-29.

326. Bell too testified that his "understanding [was] that Calyon's purpose in providing the representative assets to the rating agencies was for those assets to serve as placeholders, and was not to convey that Putnam intended to purchase those specific assets for the Pyxis portfolio." Bell Decl. ¶ 154. Simply put, the assets in rating agency portfolios were not considered to be statements of actual, targeted, yet-to-be-acquired assets. They were intended to be dummy assets consistent with the portfolio constraints that could be used to model the transaction but which were not necessarily the assets (or in the asset categories) that would be in the final portfolio.

327. Far from supporting FGIC's theory, the rating agency evidence tends to support the notion that when the PCS was sent by Calyon to FGIC, it was understood by both Calyon and FGIC to be analogous to the dummy assets sent to the ratings agencies: assets that could be used to model a transaction but not necessarily reflective of the collateral manager's intent as to which assets or asset categories it would purchase. The pie charts in the final pitchbook—which Calyon and Putnam represented would not be updated until the offering memorandum—were the reflections of the collateral manager's point-in-time projections, not the PCS.

328. There also was no evidence introduced at trial from which the Court could infer that, because dummy assets that did not correspond to Putnam's intended purchases were provided to the ratings agencies, Putnam should have known that similar lists of assets would have been provided to FGIC. First, the evidence at trial did not establish that Putnam in fact reviewed the assumption packages to see whether they corresponded to the assets Putnam intended to purchase. As noted, Putnam did not have a master plan as to what it intended to purchase and did not understand the rating agency material to purport to represent specific target assets. Dolan explained that, to the extent Putnam received and reviewed the assumption packages sent to the rating agencies, he would not have expected Putnam to undertake a "detailed line-by-line review of placeholder assets." Dolan Decl. ¶ 15. Bell confirmed that he did not review the rating agency portfolios with an eye to whether the specific representative assets were ones that Putnam actually intended to acquire. Rather, as he explained, his review would have been "focused on ensuring that the overall blend of assets contained therein met the general portfolio constraints such that those assets theoretically could have been placed in the Pyxis portfolio—i.e., assets that in the aggregate satisfied the general eligibility criteria and asset concentration limits." Bell Decl. ¶ 154. Bell's understanding was that "the rating agencies [were] less focused on target assets than on actual assets and the constraints of the deal (the weakest portfolio consistent with the deal's quality and profile tests, stressed by ratings)." Id. ¶ 156. When asked on cross-examination how the placeholder assets in the rating agency portfolios were selected, he described his understanding that the assets "were put together in whatever manner was required consistent with the criteria for each rating agency based upon Calyon's understanding of that process given their role in working with the rating agency." Tr. 1679. He later explained, "I don't believe we provided specific information to Calyon that we had an expectation or knowledge they would incorporate into the rating agency portfolios. We were provided with what they gave to the rating agencies several days after it was completed, and I've seen substantial documents that support we were very focused on the cash flow tie outs, just making sure the modeling of the waterfall, the structure, that all of the math worked. I've not seen any indication that we were very focused on the representative assets that were provided to each rating agency." Tr. 1734. The documentary evidence confirms Bell's summary. While Bell and others at Putnam provided input to Calyon about the cash flow waterfall and making sure the math tied out, there is a distinct lack of documentary evidence suggesting that Putnam was focused on which representative assets Calyon included in the rating agency portfolios.

329. Second, there was no reason to believe from the fact that Calyon provided dummy assets to the rating agencies that it would have provided any detailed list of future purchases to FGIC, much less one that was similar to what was provided to the rating agencies. There was no evidence that any investor other than FGIC had ever requested from Putnam a detailed list of target assets. Nor was there evidence that investors generally in the market asked for detailed lists of target assets; they relied on the pie charts and the rating agency reports. Putnam had never received any such similar request from FGIC or any other super-senior participant. And the evidence does not establish that Putnam knew of FGIC's request. In short, when Calyon sent FGIC the modeling assumptions package on August 14, 2006 that it had sent to the rating agencies on August 11, 2006, there would have been no reason for Putnam to suspect the same information had been provided to FGIC.

Moreover, when asked the question of whether there would be a reason for the information in the PCS and information sent to the rating agency to differ, Rekeda offered an answer that, though speculative as to the facts, speaks to the dynamic nature of building a portfolio: "It could be, you know, new purchases, more than the spreads, manager changed their view[.]" Tr. 1236-37. Bell's and Van Tassel's above-quoted testimony, about why it would have been possible but wasteful to construct a comprehensive detailed target portfolio corroborates that understanding.

330. FGIC's principal witness with respect to ratings agency communications was its expert, O'Driscoll. Although O'Driscoll testified on direct examination that the "target portfolios sent to the rating agencies were reformatted to meet the rating agencies' requirements for ratings submissions and included some inconsequential changes and corrections, but in substance they were the same as the PCS," O'Driscoll Decl. ¶ 157, that testimony was undermined on cross-examination. Cross-examination revealed significant differences between the unramped portion of the portfolio sent to Moody's and the PCS. The unramped portion of the portfolio sent to Moody's listed generic names, issuers, and identification numbers for each transaction instead of actual names, issuers, and CUSIP numbers (as the PCS did). For 26 assets, the Moody's ratings were at least three notches below the PCS asset ratings. For 39 out of the 53 unramped assets, the Moody's asset was rated below the PCS asset. The PCS had a WARF of 449. The Moody's portfolio had a WARF of 498. For 17 of the 53 assets on the chart, the issuer date was different from the deal closing date on the PCS. In the 14 instances where information was provided about the master servicer or servicer for the PCS asset, there was a discrepancy between the servicer listed in the Moody's column and that in the PCS. In seven instances, the notional amount of the bond in the Moody's portfolio was different from that in the PCS. Indeed, even the figures for the percentage of the capital structure that any particular security represented differed between the Moody's portfolio and the PCS. O'Driscoll appears to have drawn his figures from what he eventually characterized as the "wrong version" of the PCS—a draft of the PCS never delivered to FGIC. Tr. 1008, 1014.

331. As dictated by the terms of the indenture, each of the rating agencies received the final set of Pyxis assets post-closing. In February 2007, all three rating agencies supplied letters confirming their initial ratings for the notes on the basis of the final assets, not the modeling assumptions.

17. FGIC Declines Calyon's Offer to Participate at the 28% Attachment Point

332. After August 7, 2006, when FGIC held its follow-up due diligence call with Putnam, FGIC and Putnam next communicated on August 18, 2006. Menhenett called Bell and told him that FGIC would not attach at the proposed 28% attachment point or below 40% given the concentrated subprime risk in the portfolio. In an email to others at Putnam on August 18, 2006, Bell reported: "I spoke with Elizabeth at FGIC yesterday. FGIC has never been involved in a deal like this before. As I mentioned a few weeks ago, they chose this deal w/ Putnam as a 'test case'. When they finished their work on the collateral (concentrated sub-prime risk) they wanted 40% subordination. Given deal economics Calyon could not offer more than 35%. They said they were impressed with Putnam and hope we can work together in the future." PX298.

333. FGIC had engaged in three general activities, in parallel, leading up to this conclusion: (1) it analyzed the information available about the transaction's structure and Putnam's investment philosophy and experience; (2) it collected further information from Calyon about the "target" assets in the unramped portfolio and finished analyzing the collateral pool; and (3) it continued to negotiate deal terms with Calyon. These activities inform the Court's analysis of FGIC's reliance and loss causation claims and confirm its findings with respect to FGIC's claim that Putnam had an involvement with the PCS.

18. FGIC's Analysis of Information from Calyon

334. On August 10, 2006, Menhenett sent an email to various FGIC colleagues with the subject line: "Summary of Pyxis ABS CDO 2006-1." PX 829. It attached her memorandum of August 10, 2006.

335. The August 10, 2006 memorandum summarized FGIC's August 3, 2006 meeting with Putnam, sharing that Putnam had positive views about subprime, but not prime, assets. It stated, "In 2005 the group held a slightly defensive view in Subprime (see below) reducing acquisitions of BBB-bonds. In 2006, Putnam reaffirmed its interest in Subprime based on improved outlook on economic scenarios, plus structural improvements in Subprime (and stress testing). Currently, Putnam likes Subprime at IG and BIG levels (BBB & BB)." Id. The message was similar to that which Putnam had conveyed to the equity investors back in March 2006 when they asked whether Putnam would be interested in managing a CDO focusing on midprime and subprime assets.

336. The August 10, 2006 memorandum gave a detailed description of the basis for Putnam's view that subprime RMBS was a better, safer, and more credit-worthy investment than prime:

Putnam's interest to issue this hybrid CDO came indirectly, as a result of an internal effort (in 2005) to create a short-credit strategy, to short the Subprime RMBS market. Contrary to its sense on the sector at that time (uncertain interest rate environment, uncertain effect on housing price & borrower performance), Putnam's internal studies confirmed the strength of the Subprime product, rather than the opposite. The Subprime sector proved too robust to be shorted under stress testing.

Putnam likes MBS for its huge opportunity set, highly liquid asset class with wide participants, and quantifiable risk. Within RMBS, Putnam likes Subprime over Prime: Subprime holds up to stress testing, while Prime does not (according to Putnam's analysis, even AAAs can be hit under historical stress). Putnam prefers Subprime because the investment community expects Subprime to not perform hence deals have better return for risk (higher spread, higher subordination). Prime on the other hand cannot withstand marking of error in expectations; plus recent performance has been stellar / rating agencies do not apply harsh stresses; therefore investors have no pricing power or ability to ask for higher subordination. We have heard similar sentiment from RFC, where RFC worries about "the good assets not performing as expected."
Id. The memorandum also outlined how Putnam's view informed its collateral selection process:
For CDS, Putnam selects single-name RMBS from a universe of credits it tracks (appx. 500 names). Internal criteria are used to help the group further weed out certain bad characteristics, while leaving a sufficient pool size (e.g. 300 names) for Putnam to create bid lists, and recycle the bid list over a ramp-up period.

337. Menhenett's August 10, 2006 memorandum added that Putnam had initially bought bonds with "lower ratings . . . [and] small $ issuance, which tend to have strong CDS interest," and that going forward, Putnam would use the cash bucket to buy "higher rated bonds (AA, A), those issued at larger $ sizes where there was enough for buy-and-hold investors." Menhenett Decl. ¶ 125 (quoting PX829).

338. The memorandum identified a key risk to the transaction: "Exposure to mezzanine collateral, min. rating of Ba3/BBB-, mostly subprime." PX829. It also discussed risk mitigants; however, it did not identify investments in prime or seasoned RMBS as a risk mitigant to the investment in subprime and midprime RMBS. Rather, it highlighted that the portfolio constraints permitted the collateral manager to offset the risk of subprime and midprime RMBS by "diversify[ing] into non-RMBS asset, including CMBS and permitted ABS asset classes." Id. As Menhenett testified at trial, her reference to "permitted ABS asset classes" in the memorandum specified a subset of non-RMBS assets. Tr. 608, 637. With it, she intended to communicate that Pyxis's ability to invest in non-RMBS assets was a positive, and to make a "bigger point" that it was good that the manager could invest in other asset classes originating from non-consumer risk, including those such as container leases and corporate borrowers. Tr. 609. Recognizing there was a minimum amount of subprime and midprime RMBS in the portfolio, and understanding that an increase in CMBS and other asset classes would come at the expense of prime RMBS, Menhenett explained her point:

Menhenett also testified to the uncontroversial proposition that, with respect to the preliminary offering memorandum, the language "other permitted asset types" also included prime RMBS. Tr. 613, 633. But that does not change the language of her August 10, 2006 memorandum, or her testimony about what "other permitted asset types" meant in that document she authored.

We wanted to give good managers selectability. I think where FGIC was coming from, our style was to work with the manager. We appreciated the added value for the manager. We weren't micromanaging, very restrictive type of shop. There were other investors who took the position of, you must give me this, you must give me this. Where we saw in the case of somebody who is having experience, we didn't want to box them into a bod. Instead - and this strategy was not just with ABX CDOs. With other managers we carried a process-working relationship. We acknowledged and we wanted to give the good managers room to be able to do their job, where if we confined the box too strictly, market change for a reinvested portfolio of five years - the market definitely changes within five years. Preferred asset type change, the best asset type change, risk and return basis. So it's for those reasons that when we did the due diligence - we wanted to pick experienced-enough manager who can decide, as the market changed, what would be the best way to put together a balanced
pool. And because we acknowledged the market did change, we didn't want to say, you must have this, you must have this because the job of investing was in the hands of managers. With a good manager, we wanted to allow them to do their job.
Tr. 611-12. She later added: "[W]e liked the fact that prime was used. And . . . also . . . [i]f for some reason one asset sector would go out of favor or not be able to be selected, there are other asset sectors that the deal can benefit from to achieve diversification." Tr. 613.

339. The August 10, 2006 memorandum further relayed that FGIC had negotiated structural features to mitigate risk: "To mitigate the above features, we are negotiating performance triggers to address: a) loss of par during reinvestment period, b) stop discretionary trading/reinvestment if mgr or market not performing. We expect to have a cumulative par loss trigger to cover the initial 5 years, which will be set at the AAA stress level." PX829.

340. The references to asset allocation from the PCS information Calyon sent were limited to a pie chart in the section on preliminary portfolio characteristics. The pie chart reflected information from the PCS indicating the target portfolio would comprise 12% RMBS A assets and 50% RMBS B/C.

Image materials not available for display. Id.

341. The memorandum did not otherwise mention adding RMBS A to the transaction, and in particular, did not mention that FGIC believed such an addition would lend any value to the transaction.

342. Menhenett sent her August 10, 2006 memorandum to Finkel, Adams, Perlman, Skelton, Hurst, and five others at FGIC with a note that FGIC would consider attaching at the proposed 28% point. Her email advised that the CDO team was preparing to bring the issue to committee the following week. Finkel replied that the group would schedule a meeting "to discuss these synthetic deals and make sure everyone is comfortable before we bring anything to SCC." PX840.

19. FGIC's Analysis of the Collateral Pool

343. The day after Menhenett received the PCS, she emailed Lee of Calyon on August 9, 2006, at 1:33 p.m. asking for additional items:

min Moody's rating: Ba3 is allowed ? What is the max %?
min SP's rating: BBB-
min Fitch rating: BBB-

Agg. CDO max 15% ? as discussed a few wks ago, this need to be in documents, pls note this for later drafting.

344. On Friday evening, August 11, 2006, the RMBS team gave "preliminary feedback" on the Pyxis collateral. PX279. Its feedback was based on the Loss Coverage Requirement ("LCR") analysis, one of several analyses FGIC would perform on a pool of collateral before deciding whether to proceed with a transaction. This particular analysis (which was built in-house at FGIC) gauged RMBS transactions. Specifically, LCR modeled the risk of the RMBS portion of the collateral. FGIC first calculated expected defaults and the severity of loss of the underlying loans in each (or a representative sample of each) of the bonds comprising the portfolio at different confidence levels. It then compared those loss estimates to the bond's level of credit enhancement (including both subordination and excess spread) to determine expected losses on the bond itself. The outputs of the LCR model, which represented loss estimates at the loan pool level, also served as the inputs of the MBS Analysis Spreadsheet, which estimated aggregate losses for the RMBS portfolio as a whole. The LCR model assumed correlation. It ran 10,000 simulations and, by calculating how many loans would default and with what severity in each of those 10,000 scenarios, it attempted to predict the severity of loss on the assumption that all or substantially all of the bonds would default at once. The model then predicted the outermost loss the portfolio would suffer at the 97% confidence level (i.e., 3 of 100 times).

345. Importantly, the LCR model had limits. It ignored the structural protections in the CDO, such as excess spread (the increase in premium paid to FGIC over a market benchmark that could be used to cover a loss in the portfolio and still generate a profit or a zero loss for FGIC on the transaction). It also ignored triggers. It assumed—counter to fact and more conservatively than fact—instant loss (meaning loss was assumed to occur immediately, rather than at some date in the future) and no step down (meaning that it was assumed there would be no provisions in any RMBS whereby, after 36 months, the cash flow would convert from sequential pay to pro rata pay). It used FGIC's conservative rating assumptions, not those of credit rating agencies. It also ignored the non-RMBS portion of the portfolio.

346. The analysis that led to FGIC's "preliminary feedback" focused on expected loss at the 97% confidence level of two separate hypothetical pools of bonds. Id. First, assuming a pool comprised 72% Baa3 subprime and 28% Baa2 subprime at the 97% confidence level, the analysis concluded that the pool would sustain a 39.99% loss, meaning that, if FGIC attached above that level, it would not be at risk of loss. Second, assuming a pool comprised assets from the ramped portfolio and dummy assets from the July 18, 2006 portfolio spreadsheet (which included 4% prime RMBS but a greater percentage of Baa3 subprime bonds) at the 97% confidence level, the analysis concluded that the pool would sustain a 49.67% loss, meaning FGIC would have to attach above that level to avoid a loss.

347. Menhenett contextualized those results in a reply email to the RMBS team at 12:36 a.m. on August 12, 2006. She stated: "49.67% in RMBS is loss of 42.22% in the cdo (RMBS being 85pct of the CDO). Or, 34% (based on you [sic] 39.99 figure)." PX275. Those figures reflected that, based on the pool of entirely subprime assets, FGIC could attach at the 35% attachment point or above without being at risk of loss. She observed: "This is essentially a gutt [sic] check since your loss results assume instant loss, and no step down, correct?" Id. She added, "I'll be working over the weekend as well. Pls feel free to send me stuff or if you need anything." Id. Clearly, these results did not deter her from continuing on the path towards credit approval.

348. Early the next morning, in response to Delaunay's email requesting feedback on the 28% attachment point then being discussed, Menhenett told Calyon that the "results [were] still not looking good," and advocated for an attachment point of at least 34%. PX274. She summarized, "[B]ased on our RMBS re-rating & running through CDOs (they're rechecking #'s over the weeknd): we need brk-even df of 45%/SS shows 40% results (rr of 30%), so it's not enough." Id. Referring (without stating it) to the internal email of that night reflected above, she stated that "another angle" was "we need attachment of 34% or 42% depending on rating mix in the pool. This is where our RMBS grp chks tranche thickness vs our loss expectations on the pool. This is a static # assumes losses occur right away, based on current enhancement. 42% is based on target pool mix w/ some AA and A's 34% if assuming pool has WARF of 539 and max out Baa3 rest in Baa2." Id. She added: "Our mbs guys are rechking #s over the weekend" and "Touch base w/you on Monday." Id.

349. That Monday, August 14, 2006, Menhenett asked Calyon for additional data on the prime RMBS in the PCS. She needed to assess the credit profile of the prime RMBS in order to generate loss their profiles.

350. Menhenett explained in an email to the MBS credit team that same day: "I'm working on getting you CE [credit enhancement], OC (orig. curr), XS, on as many Prime bonds as possible in the Target pool, a la Calyon. ... As discussed, this will allow you to add a few lines above the current 82 bonds that's Baaa3." PX289. (The then-current 82 bonds were those in the ramped portfolio.)

351. Early the following morning, on August 15, 2006, Anand of Calyon responded to Menhenett's request for information: "Please find a spreadsheet with the 'Prime RMBS A' assets you highlighted in peach with subordination and additional straits, as much that I could find." PX284. Anand did not copy anyone from Putnam. Menhenett relayed his response to the MBS credit team.

352. On August 15, 2006, Menhenett also asked Calyon (but not Putnam) for additional data about every asset in the PCS.

20. Negotiation of Deal Terms

353. On August 8, 2006, at 12:06 p.m., Delaunay emailed Menhenett: "As per your request, we have talked to the Equity investor and came up with the following potential additions to address your concerns during the reinvestment period." PX 816. His email listed several proposed triggers for Pyxis during the reinvestment period.

354. FGIC understood that Calyon, as the agent placing multiple parts of the deal, needed to clear FGIC's desired performance triggers (that is, triggers which would have redirected the payment of principal and interest to the super-senior tranche in the event that Pyxis failed to meet certain performance metrics) with the equity investors first.

355. Ultimately, FGIC negotiated a performance trigger to be added to Pyxis.

21. FGIC's Decision to Reject the 28% Attachment Point

356. By August 16, 2006, FGIC completed its analysis of the collateral pool. Menhenett emailed Finkel and Adams that afternoon setting forth the completed analyses, the analyses in progress, and the deal options. She described the deal options accordingly:

Deal options: 28% for 11bps
35% for 10bps
Attach above 35% is possible, pricing won't be very attractive/Calyon may need us to attach very high (to line up sub interest) which puts us out of range for attractive pricing.
PX290. The last option referenced a scenario in which Calyon contracted with a different insurance company to insure the tranche from 28% to a level above 35%.

357. Menhenett stated it would work for FGIC to attach at "a 35% attachment . . . fr risk perspective." Id. Her email outlined four analyses completed to date and those in progress that supported her conclusion. The first test was based on an analysis of the CDO. It concluded that, based on a hypothetical pool that was re-rated by FGIC's own analysis—assuming a 25% recovery—FGIC would break even at a 35% attachment point if the CDO experienced losses of 42.5%. Without RMBS re-rating (using the higher ratings given by the ratings agencies instead) and using the pool maximum WARF of 539, the CDO could experience losses of 42.7%, and FGIC would still break even at a 35% attachment point.

358. The second test looked at FGIC's expectations of default of the bonds in the Pyxis portfolio and the coverage it would have for those defaults at a 35% attachment point. Using the RMBS re-ratings, FGIC expected 22.9% of the bonds to default, which would generate a coverage ratio of 1.86% at a 35% attachment point, meaning that FGIC could attach at the 35% attachment point without risk of loss.

359. The third test was the RMBS loss estimate; it yielded a result at the 97% confidence level of 41.15%. But, Menhenett wrote, "this assumes no step-down/aggr loss." Id. She added: "This is where RMBS vs CDO results differ. Attachment of 35% is below this loss check of 41.15%. We haven't used this as a hard rule in the past due to this is point in time assessment assuming no step-down . . etc." Id.

360. The fourth analysis was the FGIC spread model. Menhenett reported the results of that analysis based on work that had been performed several weeks earlier, before FGIC received the PCS. The analysis looked at the bonds FGIC expected to be in the Pyxis portfolio and calculated the cash flows the bonds would yield (over a specified benchmark) to determine whether those yields would be sufficient to cover any incremental defaults. Menhenett reported that the "[r]esults were within [the] range of deals we've done to date." Id. The fourth analysis thus indicated that the transaction would pass the FGIC spread model at a 35% attachment point.

361. Menhenett's final point related to the secondary pricing comps analysis. That analysis was not yet completed but it related to a pricing decision and not whether to do the transaction or not. The current information suggested a wide range of pricing from 12-18bps.

362. Menhenett summarized the "points" for the team to consider in making a business decision on the Pyxis transaction in an email of August 16, 2006. Id. She described the deal options as "28% for [11]bps," "35% for 10bps," and "Attach above 35% is possible, pricing won't be attractive/Calyon may need us to attach very high (to line up sub interest) which puts us out of range for attractive pricing." Id. As noted above, her conclusion after reviewing all of the above analyses was: "35% works fr risk perspective." Id.

363. There was no mention in the August 16, 2006 email of prime RMBS, seasoned RMBS, or the allocation of the Pyxis portfolio among different asset classes.

364. Finkel responded to Menhenett's request for approval to proceed at a 35% attachment level by advising that she "want[ed] to review final mbs analysis and alex's input before going back." Id. Even though the MBS loss estimate was 41.15%, Finkel did not see that as a show-stopper even to attaching at a 35% attachment point. She later responded to Menhenett's request for a response to give to Calyon that day: "Need to see the MBS info and mkt pricing data before deciding. Not a slam dunk, so if I have to give an answer now, my vote would be no. I am hoping that once we look at info tomorrow, we can formulate a response (price, attachment pt, size). That gets to yes." Id.

365. Menhenett countered Finkel's assessment and stated, "[M]y vote would be yes. market comp is hard to pin down, and illusive / hard to pin down if its hard trading level. and whether it is sans liquidity. but i understand it would be much easier had it been higher attachment." Id. Menhenett expanded on her communications with Calyon: "[I] told Calyon 35% is not a clear easy answer, 40% would be easier. Pricing aside. Calyon said they can't even quote us a price there because it makes placing the add'l pieces difficult, meaning attachment will need to be higher / pricing they said would be quite low." Id. Again, there was no mention of prime RMBS or seasoned RMBS anywhere in the email chain.

366. As noted above, FGIC eventually made the decision not to accept Calyon's offer to attach at the 28% attachment point. On August 17, 2006, Finkel and Menhenett called Rekeda, Delaunay, and Lee to tell them that FGIC could not attach below 40%. Rekeda responded, apparently with some frustration, that Calyon could not offer a higher attachment point than 35% and negotiations broke off.

367. Lee reported his conversation with Menhenett to Bell. His email stated: "FGIC could not get comfortable with the collateral. We're approaching other monolines." PX298.

368. As noted above, Menhenett called Bell and told him that, given the concentrated subprime risk in the portfolio, FGIC could not attach below 40%. As Bell explained it to Van Tassel: "When [FGIC] finished there [sic] work on the collateral (concentrated sub-prime risk) they wanted 40% subordination. Given deal economics Calyon could not offer more than 35%." Id.

369. The deal between Calyon and FGIC was effectively over and Calyon began approaching other monolines.

22. The Revised Transaction

370. It did not take long for Calyon to find a counterparty willing to attach at the 28% to 40% attachment point. By August 22, 2006, Calyon informed FGIC of that counterparty (which was Banca Intesa) and asked if FGIC could obtain approval to attach at 40% for 10bps, 45% for 9bps, or 50% for 8bps. Their following correspondence reveals whether and to what extent FGIC relied on information in the PCS, and whether FGIC ever had the opportunity to attach above 40%.

371. On the evening of August 22, 2006, Delaunay wrote Menhenett: "Just to summarize our discussion, we need to understand two things clearly: 1) that the 40% attachment works; 2) if a pricing of 9bp works." PX874. He added, "There are two separate issues and at this stage, we would appreciate color on each as we need to be very clear with the Junior Super Senior investor." Id. The following morning, Menhenett responded: "1. At 40%, we would need 10bps; 2. At [50-53%], we would need 8bps." Id. She added, "Either option works. The latter is to give you flexibility [with this junior supersenior investor, i.e.,] if offering them larger size/thicker tranche gives elbow room for lower pricing for their piece." Id.

372. On August 23, 2006, Rekeda wrote to Menhenett, apologizing for statements he had made to her on the phone the previous week. "What drove me crazy is not the deal rejection itself, but . . . comments about 'few counterparties available' and the 'attachment, given the price', which revealed not very good understanding of the subject," he said. PX875. He added that Calyon had "managed to find several SS counterparties . . . willing to take either a piece or the entire SS tranche," so FGIC need not feel pressured to transact. Id. He advised that one counterparty could take the 28% to 40% tranche, "so FGIC still has a good chance to participate on this deal if you get back to us soon." Id.

373. The next day, Menhenett relayed their conversation to Finkel and Adams: "[C]alyon has two parties interested at 28-40%. They're checking if either party prefer a thicker tranche e.g. 45%. Calyon indicated [50-53%] at 8bps is too much to pay for that risk layer. They are not doing [sic] down this path right now. Calyon asks if we can get approval for either of 3 options below as they complete their jr piece: 40% at 10bps, 45% at 9bps, 50% at 8bps." PX307. Adams responded, "Sounds interesting - what are the par amounts of these tranches? Are they saying the attachment/pricing mix is up to us?" Id. At 1:43 p.m., Menhenett responded with the par amounts at different attachment points (noting 40% par amount would be $900 million, 45% par amount would be $825 million, and 50% par amount would be $750 million). Id. She continued, "[N]o, calyon is not saying up to us. They are asking if FGIC can agree to either of these 3 options." Id. She concluded, "I would pick 45% at 9bps, if calyon can get the jr's to do 28-45%. But then, we don't know at this pt." Id.

374. In a separate email thread, Finkel replied, "Total deal size is 1.5 bil, so $600-750 mm. The price/subordination matriz [sic] is what we had quoted them. They are . . . saying they will decide which price/ subordination level to take based on where they have demand. Sounds more likely they will hit our bid at 10/40 or 9/45, than at 8/50. I like 8/50 marginally best, but they think that price is too high for the attachment point. I am also very happy with the other attachment/ pricing levels." PX 306. Finkel then proposed that their group "talk to Calyon and if we are confident that they will hit our bid at one of the levels in the grid, then we go to SCC on that basis." Id. Adams affirmed, "I agree. I am ok with all of the bids with a moderate preference for the 50 percent attachment which I think is the best value." Id.

375. Finkel followed up at 2:15 p.m.: "We spoke to Calyon. They will narrow down the choices to 1 by Wednesday. We agreed to go to SCC by September 6th. Much as I like the 50%, sounds like that is the least likely." Id.

The following Wednesday was August 30, 2006.

376. Shortly thereafter, Finkel wrote other members of the FGIC team: "Due to Elizabeth's excellent negotiations, Pyxis is back on. We and Calyon have agreed to a pricing/subordination matrix, which they have agreed to narrow down to one option by Wednesday. Matrix at this point is : 50% subordination at 8 bp; 45 9 bp; 40 10bp." She noted, "We have further agreed to go to SCC on September 6th, which means we need to distribute the SCC memo before Labor Day." PX308.

377. Thus, as Menhenett and Finkel both testified, FGIC understood that Calyon would decide what attachment point to offer FGIC. It was not up to FGIC to decide the attachment point that Calyon would accept.

378. On August 30, 2006, Calyon narrowed the choices to just one: the 40% attachment point. Menhenett emailed Delaunay, Lee, and Rekeda that she wanted to "confirm 40% attachment today (or if the other options we discussed that are less likely)." PX900. By the other options that are "less likely," Menhenett was referring to potentially attaching at 45% or 50%. Calyon confirmed that attaching at 40% was the option available.

379. As Menhenett testified, Calyon never offered FGIC the opportunity to attach at 50%. Finkel corroborated that fact, testifying that she could not recall Calyon ever offering FGIC the opportunity to attach at 50% and that she understood Calyon was primarily looking for someone to attach at 40%. The evidence does not support that FGIC was offered an opportunity to participate in Pyxis at any different attachment point, including one of 50%.

There was evidence that Calyon had a potential counterparty who was willing to go up to the 55% attachment point (which would have permitted Calyon to offer FGIC an opportunity to attach at the 50% attachment point), PX 296, but there is no evidence that Calyon pursued a transaction with that counterparty or that it could have done so at pricing that was attractive to Calyon.

23. FGIC Credit Approval for Pyxis

380. By September 6, 2006, FGIC's CDO team was prepared to seek company-wide approval to engage in Pyxis at a 40% attachment point as it had committed to Calyon.

381. On August 29, 2006, Calyon emailed Menhenett and attached the latest investment criteria as well as the most recent draft of the offering memorandum and indenture, reflecting that the collateral manager would have to invest at least 80% of the portfolio in midprime and subprime RMBS and 90% in RMBS and ABS CDO securities and need not invest anything in prime RMBS. Even at this point, FGIC did not inquire as to whether the targets changed and did not ask Calyon to change the investment criteria to require Putnam to acquire prime RMBS securities.

382. Menhenett responded that day with changes that she wanted to the criteria—but not the addition of prime RMBS or seasoned RMBS. She "noticed bespoke ABS CDO has been added to the criteria." PX311. She asked, "What prompted this addition? What bespoke transactions is Putnam planning to add for initial ramp-up?" Id. Menhenett also asked for other changes to the eligible criteria.

383. Before agreeing to the Pyxis transaction, FGIC had to review and approve it at five different levels.

384. The first level of review was conducted by the deal underwriter, who was Menhenett in this case. The deal underwriter screened the transaction's proposed structure, manager, and collateral, to determine if further review was warranted. If so, she performed due diligence on the collateral manager and prepared an analysis of the manager's strengths and weaknesses. In addition, the deal underwriter coordinated and facilitated the collateral pool's credit analysis by the MBS group and CDO analysis by the CDO group. Based on that information, if she wished the transaction to proceed, the deal underwriter would then propose an attachment point at which FGIC should participate.

385. Thereafter, to accomplish the remaining steps, FGIC's participation needed approval from: (2) the head of its CDO Group, Finkel; (3) the head of its MBS Group, Adams; (4) its Credit Group, headed by Alessandra ("Sandy") D'Imperio; and (5) its Structured Credit Committee ("SCC"), comprising members of FGIC's senior management. If a proposed CDO transaction was rejected at any of these levels, FGIC could not participate in the deal. By this time, all but the fifth step had been completed.

386. Before deciding whether to participate in the revised transaction, FGIC did not conduct any further due diligence, ask Putnam or Calyon for an updated ramped portfolio, or ask Putnam or Calyon whether the targeted portfolio as represented in the PCS had changed.

387. On September 6, 2006, FGIC's CDO team presented the SCC with its proposal to provide Pyxis's $900 million super-senior tranche credit protection at the 40% attachment point. Menhenett sent the SCC a written proposal in the form of a credit application, before the SCC convened for an in-person meeting. The credit application opined that FGIC's "opportunity to earn 10 bps at super senior attachment of 40% . . . represents strong risk/return compared to previous mezz[anine] ABS opportunities we participated/reviewed." PX325.

388. In several places, the credit application highlighted that the transaction would allow FGIC to take exposure to the subprime RMBS sector, and that such exposure was a key feature of the transaction.

389. The "Recommendation" section at the beginning of the application pitched the transaction as "an attractive opportunity for FGIC at 40% attachment (2x natural AAA) covering risk layer up to 100% with Putnam acting as asset manager" and made clear that the collateral pool would "consist of primarily RMBS securities with average rating quality of triple-B" with "[a]dditional asset classes . . . permitted to provide sector diversification, including ABS/CRE CDOs, CMBS, other permitted commercial and consumer ABS to be selected by Putnam (meeting investment criteria) . . . This deal allows FGIC to take exposure to the subprime RMBS sector, and have the benefit of an experienced CDO manager who is permitted to defensively manage the deal through the CDO structure." Id.

390. The "Asset Manager Assessment" section noted that "Putnam has held a general preference for Subprime RMBS since 2001" which is "continually tested by Putnam combined with its macro economic outlook." Id.

391. The "Collateral Description/Assessment" section stated that all 82 bonds in the ramped portfolio were subprime. Id.

392. The credit application identified the "exposure to Subprime RMBS" as a "key risk." Id. Notably, however, as with the August 10 memorandum, it did not list exposure to prime or more seasoned subprime RMBS under "mitigants." Id. Rather, the credit application compared Pyxis's subprime exposure to FGIC's subprime book and noted their quality seemed comparable. It stated, "CDO's exposure is at Investment-grade. Each bond in the CDO portfolio must be rated IG by at least one agency[]" and observed "[g]ood diversification across originator/issuer/servicer, which provides protection against event risk." Id. It mentioned that Pyxis's structure permitted diversification into non-RMBS including credit card, student loan, and automotive. The section concluded the transaction "is tested under FGIC's RMBS review process where FGIC holds very conservative assumptions towards the Subprime product." Id.

393. With respect to the ramped portfolio, the credit application reported, "FGIC's MBS team performed a review of all 82 RMBS bonds in the ramped portfolio and found the quality of issuers/originators and servicers to be representative of the market with no adverse selection." Id.

394. Three pages of the credit application contained FGIC's modeling of the CDO's sensitivity to the possibility of losses in the portfolio and its analysis of the CDO's cash flow. The sensitivity and cash flow analyses modeled the portfolio based solely on the CDO's portfolio constraints and not on the target assets from the PCS. Specifically, they assumed—consistent with the portfolio constraints—that the portfolio would enjoy a WAS of 2.00%, a WARF of 539, and a Moody's Correlation Factor of 26.00%. FGIC also tested the structure using more conservative assumptions than the public ratings assigned by the ratings agencies based on its analysis of the assets in the ramped portfolio. Running a transaction of those characteristics through its model, FGIC forecast that the point at which the losses would equal the break-even (i.e., the lowest point at which FGIC could attach without incurring losses) was 35% at a 99.5% confidence level. The 40% attachment point thus was safely at a range where FGIC did not expect it would incur losses at a high level of confidence.

395. The information that was conveyed in the PCS was mentioned in two separate places in the credit application. The sixth page of the credit application, containing FGIC's "Collateral Description/Analysis," after mentioning that all 82 bonds in the ramped portfolio were subprime, added: "The above captures the initial ramped portfolio. Additional bonds will be added to the portfolio and is expected to include Prime bonds rated A2 and above." Id. A bar chart on that same page reflected the asset mix as represented in the PCS including that the overwhelming majority of the portfolio would comprise RMBS B/C and that approximately 11% would be RMBS A. Id. Those were the only mentions of the PCS in the body of the credit application.

396. Appendix 8 to the credit application contained the MBS team's review along with their MBS analysis. The bulk of that analysis examined the bonds in the ramped portfolio, as the MBS team admitted that it had "limited information" about the additional bonds. Id. It concluded as to the bonds in the ramped portfolio that they consisted entirely of subprime bonds, that the portfolio of subprime bonds did not exhibit any adverse selection, that it did not include any seasoned bonds, that the mix of issuers between originator-issuers and dealer-issuers "result[ed] in a less correlated portfolio and offer[ed] us better protection from event risk," that the portfolio was "well diversified with respect to originator," that "[w]ith respect to issuers, again there [was] a strong mixture," that "[t]he portfolio show[ed] a very strong mix of servicer names," and that the credit profile and credit enhancement of the ramped portfolio was "about average in comparison to our subprime book." Id.

397. The MBS analysis also contained an "Investing" section, reporting that Putnam's "MBS/ABS team believes that the current [s]ubprime RMBS transactions are very well enhanced compared to worst historical and projected future closes." Id. It explained, "[A]fter initially studying the market for the purpose of shorting it, [Putnam] concluded that on the contrary, the RMBS market is currently a good opportunity, and began to purchase securities in late 2005." Id.

398. With respect to the remainder of the portfolio which had not yet been ramped, Appendix 8 stated:

Additional bonds will be added to the pool during the ramp up period. The FGIC CDO group was provided information from Putnam regarding the makeup of these new bonds. The CDO group then confirmed the characteristics were logical relative to the targeted WARF. The original pool will comprise ~75% of the ramped up pool. The complete MBS pool will look as follows . . .
Id. (emphasis added).

399. The chart below followed:

Issuers Classification




FGIC Classification

Subprime (Original Pool)











































400. The latter portion of the MBS section of the credit memorandum contained "Credit Analysis." Id. For the 82 bonds in the ramped portfolio, FGIC ran LCR for 25, or 30% of the total transactions. For deals where FGIC did not have actual LCR results, it used LCRs from deals that were most similar based upon originator and pool characteristics. Id. For the additional bonds that would be added to the CDO, the MBS team made an assumption different from that applied to the bonds in the ramped portfolio. The MBS team assumed that they would be representative assets and not real bonds. For the subprime bonds, it assumed that if new bonds came from subprime transactions, they would perform, on average, similar to the subprime market, without treating seasoned bonds differently from unseasoned ones. For the prime bonds, FGIC selected as proxies four prime bonds from the FGIC portfolio that it believed would be most similar to those in the fully ramped portfolio. Two of those involved Alt-A bonds (not prime RMBS by FGIC's own internal characterization) and one of which was secured by home equity loans.

401. Based on those assumptions, the Credit Analysis generated results. It assumed, without taking credit enhancement into account, that, at the 97% confidence level, there would be average losses of 8.81% on subprime RMBS and 2.22% on prime RMBS.

402. After taking into account the number and percentage of fails, and the credit enhancement of the underlying bonds in the RMBS portion of the Pyxis portfolio, the MBS team generated an aggregate loss figure of 41.15% at the 97% confidence level. The Appendix concluded: "Applying this analysis to the current transaction, 58 of the underlying [approximate] 110 RMBS transactions fail at the 97% confidence level. This represents a 41.15% loss as a percent of the RMBS collateral. Since we are attaching at a super senior level for this deal, we have 40% of subordination. This implies that the other 1.15% of credit enhancement must come from the excess spread from the CDO." Id.

403. At the conclusion of the SCC's meeting, the SCC approved FGIC's insurance of Pyxis at a 40% attachment point, for a premium of 10bps. There was no testimony about the discussion that transpired during the meeting or what factors the SCC found to be important. No witness testified that Appendix 8 or the analysis therein was discussed at the meeting, or even that the witness reviewed it and thought it was important prior to the meeting.

404. Although the first page of the credit application contained an asterisk followed by the statement: "Please note that we are also asking for approval to attach at >=50% for premium of 8bps, in the unlikely case where the 40% opportunity is no longer available," PX325, Menhenett, the person at FGIC who sought SCC approval for Pyxis, confirmed that the credit committee was not asked to approve an attachment point of 50%.

405. After the meeting, Menhenett emailed Lee, Delaunay, and Rekeda to tell them the news: "We have credit approval on this deal . . . Pls include us in draft distribution of: Indenture, OM, CMA, SS documents." PX930.

406. Menhenett also emailed Bell: "I wanted to let you know that we received credit approval on this transaction and expect to work through doc review to complete the ss swap facing Calyon outside the deal." PX336. She continued: "Thanks for sending performance information which helped our credit process. In the next couple of wks, in preparation for closing, I'd like to coordinate with your team and our surveillance group to obtain web access, and discuss getting trustee to report class OC/IC ratios and the cum of variation in this deal. Have a good pricing tomorrow!" Id.

407. Bell responded: "Thank you very much. That's excellent news! We are pleased to have an investor like FGIC as a client. We will work with you to get the reporting that you need." Id. Bell testified that, with this response, he intended to communicate that Putnam would treat FGIC as a client in terms of making additional reporting available. That testimony is credible. As noted above, and as Bell confirmed multiple times throughout his testimony, Putnam considered Pyxis—the CDO—to be its client as a legal matter.

24. The Pricing of Pyxis

408. The pricing date for Pyxis was September 8, 2006. That was the date the debt investors would commit to Pyxis, provided that Pyxis satisfied the closing conditions. Prior to pricing, Calyon provided two key documents to deal participants: a term sheet and a draft of the offering memorandum (the "preliminary offering memorandum" or, in industry jargon, "the red").

409. As described above, the development of a CDO is shaped by input from various investors, who seek to ensure the transaction will satisfy their needs by imposing terms or constraints into the deal. As Bell testified, "We [had] received comments from debt investors. We [had] comments from preferred shares equity investors. There were new triggers being proposed by FGIC and others." Tr. 1503. Before the pricing, Bell wanted to "make sure that there was a careful scrub of all these different constraints." Id. He wanted to "get them into a term sheet that had them all in there." Id. He "wanted to make sure that . . . they were all clearly communicated out to investors at the time of pricing," so he "push[ed] for that." Tr. 1504. Among other things, Putnam wanted to make sure investors and participants received an updated target portfolio. Bell was not in charge of drafting the term sheet, but he did provide feedback to Calyon as the drafts of the term sheet evolved.

410. Throughout the entire Pyxis transactional history, Calyon—who was in charge of the transaction and who was the would-be counterparty to FGIC—provided drafts of the term sheet to FGIC. On August 9, 2006, Lee emailed Menhenett (copying Rekeda and Delaunay) a draft term sheet. No person from Putnam was on that email. On August 29, 2006, Lee emailed Menhenett (copying Delaunay) an updated draft of the term sheet. No person from Putnam was on that email.

411. On September 8, 2006, moreover, before the pricing and three weeks before the closing, Delaunay emailed Menhenett attaching the preliminary offering memorandum for Pyxis, and another updated term sheet for the transaction. The two were attached as separate documents. The preliminary offering memorandum was 251 pages long. The term sheet was seven pages long.

412. The term sheet contained a "Summary of Key Terms" for the final set of asset constraints. PX953. It stated the portfolio manager would be required to invest at least 80% of the portfolio in RMBS midprime and subprime assets and was permitted to invest as much as 100% of the portfolio in such assets. It also contained a list of other assets the portfolio manager could purchase but was not required to purchase, with caps for each category of assets.

413. The term sheet also contained a "Target Portfolio" section, which displayed two pie charts with indicative asset allocations based on asset category and asset ratings. Id. The pie chart indicated that 2% of the Pyxis portfolio, on an indicative basis, would be invested in RMBS Prime:

Image materials not available for display. Id.

414. Turning to the preliminary offering memorandum, it disclaimed: "The information contained herein supersedes any previous such information delivered to any prospective investor and may be superseded by information delivered to such prospective investor prior to the time of sale." Id. Menhenett testified that she was familiar with such disclaimers.

415. The preliminary offering memorandum too contained a listing of the permitted asset categories but—unlike the term sheet and reflecting the preliminary nature of the draft preliminary offering memorandum—the figures for the maximum and minimum investments were not filled out.

416. Menhenett testified she did not review the term sheet Delaunay sent on September 8, 2006. She said she did not need to because she had already received two previous versions of the Pyxis term sheet—the one on August 9, 2006 and the other one on August 29, 2006. She stated that she would defer to the preliminary offering memorandum and indenture for the final deal terms instead. She also stated that the earlier term sheets did not include a target portfolio, such that seeing one on September 8, 2006 would have been atypical and unexpected.

417. Menhenett's testimony that she did not review the term sheet was evasive and not credible. Menhenett stated that she asked for the preliminary offering memorandum because, as the deal manager, she needed to review the details of that document and confirm that the items she negotiated with Calyon, such as triggers and portfolio constraints (including an aggregate CDO limit), were reflected therein.

418. The preliminary offering memorandum could not have confirmed the portfolio constraints, as they were left blank—which was common for early drafts of this type of document. These numbers were not filled into the indenture and offering memorandum until approximately October 3, 2006, when the transaction closed. Only the term sheet would have had the numbers for the portfolio constraints, a critical issue for the transaction.

419. The earlier term sheets clearly did not answer Menhenett's concerns, as she now claims. Indeed, Menhenett continued to ask about final investment criteria after she received the August 29, 2006, term sheet. Specifically, on August 29, 2006, she asked Calyon about the aggregate CDO limit of 15% and about a bespoke bucket, and she did not receive a response. When Menhenett received the term sheet on September 8, 2006, her inquiry was still unanswered.

420. At the time FGIC received the term sheet, FGIC was considering—but not yet committed to—insuring a nearly $1 billion portfolio. The term sheet was seven pages long and was attached as a distinct attachment to Calyon's September 8, 2006 email. There is no other document Menhenett received before it entered into the Pyxis transaction that she testified to ignoring. It is inconceivable and not credible that Menhenett would not have taken the time to look at such a short and important document given the commitment FGIC was making.

25. The Closing of Pyxis

421. By October 3, 2006, the Pyxis transaction was ready to close. The equity investors had negotiated and signed the necessary agreements, the note holders had agreed to the deal, there was an indenture in place between Pyxis and the indenture trustee to protect the note holders, and Putnam had negotiated and was prepared to sign the collateral management agreement prescribing its duties going forward.

422. By October 3, 2006, FGIC received the final offering memorandum (the "Offering Memorandum"). The Offering Memorandum made various disclaimers. For example, it warned: "In making an investment decision, investors must rely on their own examination of the co-issuers and the terms of the offering, including the merits and risks involved." PX361. It also warned, as had the preliminary offering memorandum: "The information contained herein supersedes any previous such information delivered to any prospective investor and may be superseded by information delivered to such prospective investor prior to the time of sale." Id.

423. The Offering Memorandum directed the reader's attention to the fact that Putnam would be subject to those constraints: "The Indenture and the Collateral Management Agreement place significant restrictions on [Putnam's] ability to advise [Pyxis] to buy and sell securities for inclusion in the Collateral as Collateral Debt Securities, and [Putnam] is subject to compliance with such restrictions." Id.

424. The Offering Memorandum made clear that Putnam's ability to purchase securities for the portfolio was limited by the restrictions contained in the collateral management agreement and the indenture, including the eligibility criteria and that, even within those constraints, Putnam would not have unfettered discretion:

Pursuant to the Collateral Management Agreement and in accordance with the Indenture, the Collateral Manager will manage the selection, acquisition and disposition of the Collateral Debt Securities on behalf of [Pyxis] ... based on the restrictions set forth in the Indenture (including the Eligibility Criteria described herein) and on the Collateral Manager's research, credit analysis and judgment (in each such case, subject to the standard of care set forth in the Collateral Management Agreement).

425. The Offering Memorandum stated that, in purchasing securities for the portfolio, the collateral manager would have to ensure that the portfolio as a whole would have a maximum WARF of 539 and a minimum WAS of 188.

426. It also provided that at least 80% of the portfolio would have to be invested in RMBS midprime or subprime securities and 90% of the portfolio would have to be invested in RMBS or ABS. It did not require the portfolio manager to purchase any other types of securities but did impose limits on the maximum amount that could be invested in those other types of securities: no more than 20% of the portfolio could be invested in prime RMBS, no more than 5% of the portfolio could be invested in asset backed securities backed by equipment leases, no more than 10% could be invested in CMBS, no more than 10% could be invested in automobile securities, no more than 10% could be invested in REIT debt securities, no more than 10% could be invested in credit card securities, no more than 5% could be invested in Small Business Loan securities, and no more than 5% could be invested in student loan securities.

427. The collateral management agreement between Pyxis and Putnam, to which the Offering Memorandum made reference, was signed effective October 3, 2006, the date of the closing. It was also provided to FGIC. Going forward, and subject to the terms of the indenture, which trumped any inconsistent provision of the collateral management agreement, Putnam was charged with "supervis[ing] and direct[ing] the investment and reinvestment of the Collateral," and "performing on behalf of the Issuer, those investment related duties and functions assigned to the Issuer under the Indenture." D-C29 (emphasis added). Those responsibilities included, of course, the responsibilities to "(i) select the Collateral Debt Securities, U.S. Agency Securities and Eligible Investments to be acquired by the issuer; [and to] (ii) invest and reinvest the Collateral and facilitate the acquisition and settlement of Collateral Debt Securities by the Issuer." Id.

428. Under the collateral management agreement, Putnam was specifically charged with the obligation "[i]n performing its duties [t]hereunder, [to] manage the Collateral with the objectives that there be sufficient funds available on each Distribution Date in accordance with the Priority of Payments (1) to pay interest on the Offered Notes in a timely manner, (2) to repay principal on the Offered Notes in full on or prior to the stated maturity date and to make distributions on the Preference Shares [the equity] in an amount necessary to ensure that the Preference Shareholders have received the Target Return by their stated maturity date; (3) to pay expenses; and (4) to provide for additional return to the Preference Shares." Id. Putnam did not undertake, and was not charged with, any responsibility to the super-senior tranche directly.

429. Putnam also agreed to, and was charged with the obligation to generally "perform its obligations . . . with reasonable care and in good faith using a degree of skill and attention no less than that which the Collateral Manager exercise[d] with respect to comparable assets that it manage[d] for others with similar objectives and policies." Id.

430. Section 9 of the collateral management agreement further set out for whose benefit Putnam was to act:

[Putnam] shall perform its obligations hereunder in accordance with the terms of this Agreement and the terms of the Indenture applicable to it and in the course of such performance shall at all times act in good faith in accordance with its (and its Affiliates') fiduciary duties to clients, and, consistent with such fiduciary duties, shall use all reasonable endeavors in the course of carrying out such obligations to protect the interests of the Noteholders and to produce the greatest likelihood of preserving the ability of the Issuer to retire the Offered Securities in accordance with their terms. The Collateral Manager agrees that such obligations shall be enforceable at the direction of the Administrator, on behalf of the Issuer and the Trustee on behalf of the Secured Parties.

431. The collateral management agreement tied Putnam's fees to the performance of the portfolio. The monthly "Senior Collateral Management Fee" of 15bps it received was calculated on the basis of the non-defaulted assets in the portfolio. Id. The "Subordinated Collateral Management Fee" also built in a performance component. Id.

432. Two FGIC entities signed agreements with Calyon effective October 3, 2006. First, FGIC Sub signed a confirmation for a credit swap agreement with Calyon (the "Credit Swap Agreement"). Under the terms of the Credit Swap Agreement, FGIC Sub as seller or Floating Rate Payer agreed to pay Calyon a floating amount based on the performance of the super-senior swap agreement, in exchange for the payment by Calyon to FGIC of a fixed annual fee. Second, effective also October 3, 2006, FGIC signed a surety bond for the benefit of Calyon insuring FGIC Sub's obligations to make payments under its swap agreement with Calyon, referencing the Pyxis super-senior swap agreement. The surety bond contained a waiver by FGIC of "all defenses of any kind (including, without limitation, the defense of fraud in the inducement or fact)." D-F31. The Credit Swap Agreement, creating the obligation that FGIC was insuring, contained no similar waiver by FGIC Sub.

433. FGIC was not required to close and to sign the deal. Menhenett testified that FGIC would not have participated in the transaction if the preliminary offering memorandum did not reflect the changes that she had negotiated. She understood that, even after the transaction priced, FGIC would have time to review the transaction documents and confirm they reflected FGIC's desired transaction. She informed Calyon that FGIC did not believe the September 8, 2006 pricing date created a binding obligation on FGIC and that FGIC would continue to review the preliminary offering memorandum thereafter to confirm that it contained the deal terms FGIC had negotiated. FGIC's expert, O'Driscoll, also testified that FGIC could have reneged on its commitment and would not have been required to close if it no longer wanted to participate on the basis of the September 8, 2006 target portfolio. Bell too testified that, based on his industry experience, because FGIC was not a Pyxis investor but merely a counterparty to Calyon, it was not contractually committed until closing. FGIC could have withdrawn, leaving Calyon with the exposure.

434. The fact that FGIC understood that it could walk away from the deal post-SCC approval but pre-closing is underscored by its conduct in a transaction that occurred a year later. In July 2007, FGIC explored providing insurance for the super senior tranche of a CDO deal called Grand Avenue III. Menhenett was the underwriter of Grand Avenue III. On July 9, 2007, the SCC approved the transaction. In the two days afterwards—July 10, 2007 and July 11, 2007—Moody's and S&P downgraded certain collateral assets in Grand Avenue III. Thereafter, and despite the fact that FGIC had received credit approval and indicated it was prepared to move forward with the transaction, it took the position that it was not committed to the transaction until it signed and told its counterparty that it would not sign the agreement unless the portfolio was changed and three specific bonds were removed from it.

26. The Ramp-Up

435. Following the closing, and as required by the collateral management agreement and the indenture, Putnam continued to purchase securities in the interest of Pyxis and consistent with the portfolio constraints. There is no allegation or evidence that Putnam violated its duties under the collateral management agreement.

436. As noted above, Putnam had begun ramping the Pyxis portfolio by the time it signed the Mandate Letter and Warehouse Agreement in June 2006.

437. By July 18, 2006, and as reflected in the July 18, 2006 portfolio spreadsheet, Putnam had acquired 82 assets for Pyxis, with a notional value of $972 million. Over $500 million in assets remained to be ramped. Those 82 assets comprised about two-thirds of the portfolio, and all but three of them were rated Baa3. None of the securities were prime RMBS. The WARF of the ramped portfolio as of that date was 601, reflecting a risk significantly greater than that which would be permitted of the final portfolio.

438. Calyon had represented to FGIC that the credit characteristics of the assets purchased after July 18, 2006 would be better than those purchased prior to July 18, 2006. They were. From July 19, 2006, to the date of closing on September 8, 2006, Putnam purchased assets of an additional $301 million notional value. The WARF of the additional assets purchased between July 19, 2006 and pricing was 313, reflecting significantly less risk. Sixteen of the additional assets purchased after July 18, 2006 (those listed in the PCS) had a total notional value of $102.1 million. The remainder were purchased after the PCS. Of the assets purchased prior to pricing, only one (with a par value of $2.75 million) was listed as RMBS A. As noted above, FGIC did not ask either Calyon or Putnam about the assets acquired for the Pyxis portfolio after receiving the PCS and prior to pricing.

439. From pricing until closing, Putnam purchased assets with an additional notional value of $106 million for the Pyxis portfolio. None of those assets were prime RMBS. The assets purchased from September 9, 2006 to October 3, 2006, had an aggregate WARF of 306, reflecting even better credit quality. FGIC did not ask Calyon or Putnam about the assets acquired from pricing to closing.

440. At closing on October 3, 2006, $1.379 billion worth of the $1.5 billion in assets had been ramped.

441. After the closing and until the full ramp-up was completed before the deadline of February 2007, Putnam acquired the remaining $121 million notional amount of assets for the Pyxis portfolio. The remaining $121 million of assets had a WARF of 365, also well below the WARF for the period prior to July 18, 2006.

442. By February 2007, the Pyxis portfolio was fully ramped. The fully-ramped portfolio complied with all of the requirements set forth in the indenture, the collateral management agreement, and the Offering Memorandum. It had a WARF of 503, well within the maximum permitted under those agreements. However, it included only miniscule amounts of assets that could be characterized as prime RMBS (approximately 0.22%) prime RMBS assets and pre-2005H2 RMBS assets (approximately 2.3%) and had 15.47% assets rated above Baa2.

443. Thereafter, from February 2007 until at least August 2007, the composition of the Pyxis portfolio did not change.

444. The Court finds that, in selecting assets, Putnam engaged in a detailed analysis of RMBS and CDO assets available in the marketplace and applied independent judgment to determine the assets to include in Pyxis. It acquired some securities through a bilateral negotiation between a potential buyer and seller, and others by sending a bid list to the market. The vast majority of assets in the Pyxis portfolio were acquired through bid lists.

27. FGIC's Post-Closing Surveillance of Pyxis

445. FGIC's involvement with Pyxis and with Putnam did not end with the closing. After closing, FGIC completed post-closing certifications. It also received monthly trustee reports regarding the assets in the Pyxis portfolio and it had access to Putnam's Pyxis website, which recited the rating and vintage of each collateral asset in Pyxis, as well as the exact percentage of assets that were prime RMBS and subprime RMBS. FGIC also had access, at least by telephone, to Putnam and spoke to Putnam on several occasions. Had FGIC asked, it would have learned that, at closing, Putnam had invested less than 1% of the Pyxis portfolio in prime RMBS. Indeed, by February 2007, when the ramping was complete, FGIC would have known that less than 1% of the Pyxis portfolio was invested in prime RMBS, that there was 2.3% pre-2005H2 RMBS, and that there were 15.47% assets rated above Baa2. It also would have known of the credit characteristics and quality of the portfolio as a whole. FGIC also would have had the ability to recommend changes in the composition of the portfolio had it cared about the percentage of prime RMBS and seasoned RMBS (and had such assets been available). The collateral management agreement and the indenture gave Putnam a 5% discretionary trading authority each year after the closing. Nothing prevented FGIC from asking Putnam to exercise that authority in favor of the kind of assets it wanted. It also had the right, if it believed it had been materially defrauded, to walk away from the transaction. Nothing in the guaranty that FGIC Sub issued to Calyon made FGIC Sub (and, derivatively through the surety bond, FGIC) irrevocably bound to the Pyxis transaction.

446. Yet, FGIC did not inquire or complain. The tale of FGIC's post-closing conduct is the nail in the coffin of its case for reliance.

447. In late October 2006, almost one month after Pyxis closed, Menhenett was presented the "Closing Memorandum" for Pyxis and asked to attest: "There were no material differences between the closed transaction and the transaction approved by the SCC." D-B55. The question recognized that portfolio selection is a dynamic process and there are differences between the portfolio presented to the SCC and the portfolio at the time of closing. By the time of closing, at least $1.379 billion worth of the Pyxis portfolio (approximately 92%) had been acquired. Its composition would have reflected what FGIC in this litigation claims is a material difference from what was presented to the SCC. Yet, Menhenett signed the attestation, conveying that there were no material differences. When Menhenett was asked about the reason behind FGIC's contrary attestation at trial, she answered unconvincingly and conclusorily: "Based on all the information we had received up to the closing of the transaction." Tr. 438. When Menhenett was asked if she treated the Closing Memorandum as a "pro forma document" that she signed without any additional examination of the closed transaction, she again responded evasively: "It's based on the reasoning I stated . . . I deemed there was no material difference between the characterization we presented to the credit committee . . . I think our purpose here is the overall quality - qualitative reviews, credit reviews encompassing the transactions. This statement was not an accounting statement." Tr. 438-39.

448. In late 2006, FGIC (with Menhenett as the lead underwriter) insured the super-senior tranche in a CDO transaction called Raffles Place II. She testified that she performed due diligence on the assets in that deal. One of those assets was Pyxis. The credit memorandum for Raffles II, which was dated November 10, 2006, stated that, in reviewing the underlying CDO assets, the team focused "on collateral quality as represented by ratings distribution, investment guidelines and current holdings." PX1038. By that time, Pyxis's current holdings consisted of approximately $1.35 billion of securities (the majority of the portfolio) and only a miniscule portion was in prime RMBS—what FGIC now claims reflected the results of a misrepresentation. FGIC nonetheless went forward with the transaction with no complaint about Pyxis's then-current holdings.

449. FGIC's Structured Finance & International Underwriting Risk Management Credit Guidelines & Procedures, dated December 1, 2004, required: "Each transaction is to be monitored on an ongoing basis to ensure that assets are performing as projected and that the counterparties are in compliance with the relevant covenants, triggers, reps & warranties and other factors." D-C12.

450. Thus, in February 2007, and every month thereafter, an independent trustee prepared reports for FGIC and others regarding the composition of the Pyxis portfolio (the "Trustee Reports"). The reports were sent to FGIC. The February 2007 Trustee Report listed each portfolio asset, whether it was prime or subprime RMBS, and confirmed that it met all of the eligibility criteria for the deal. It reflected that the fully-ramped portfolio included only 0.22% prime RMBS assets, 2.3% pre-2005H2 RMBS assets, and 15.47% assets rated above Baa2. Nonetheless, FGIC did not complain.

451. On April 6, 2007, Finkel completed an analysis of "Competitor Subprime RMBS Exposure in ABS CDOS" and emailed it to senior management (including FGIC's Chief Risk Officer, Tom Adams, its head of RMBS and CDOs, and the head of the Surveillance Group) with that subject line. D-A14. She then forwarded her email to Menhenett, Skelton, and three other analysts with the message "fyi." Id. Her original email stated: "We have compared FGIC's ABS CDO sub-prime MBS exposure to that of our competitors. We appear to be in line with our competitors. The results are set forth below and the attached charts provide more detail[.]" Id.

452. The attached chart showed the subprime exposure in Pyxis. It reflected that Pyxis contained just 2% prime assets compared to the 11.8% reflected in the PCS. Although Menhenett—who apparently recalled verbatim comments from her telephone conversations in August 2006—claimed that she could not recall receiving Finkel's email and its attachment. But she acknowledged that, since Finkel was her direct supervisor, she likely would have read emails and attachments that came from her. She also testified that, had she noticed the chart, the difference reflected therein would have constituted a "drastic change." Tr. 199. In 2007, however, Menhenett did not react. Instead, she testified at trial that she did not recall "digging into the numbers." Tr. 200.

Discrepancies in percentages of prime RMBS are explained by the fact that there is no single common definition of prime RMBS.

On redirect examination, Menhenett retreated from her cross-examination testimony and stated that she very likely reviewed just the body of the email. The Court credits her testimony on cross-examination.

453. On April 27, 2007, Calyon requested FGIC's consent to the sale of Putnam's indirect parent company (Putnam Investments Trust) by Putnam's parent company (Marsh & McLennan Companies, Inc.) to Great-West Lifeco Inc. FGIC's consent was sought in its capacity as a Pyxis noteholder. Bell sent a letter accompanying the consent form stating that consent was sought "in order to ensure that [Putnam] may continue to manage the [Pyxis collateral] under the Collateral Management Agreement[.]" D-A16.

454. By this time, Menhenett and FGIC had been informed: (1) of the discrepancies between the PCS and the fully ramped Pyxis portfolio, including that Pyxis had a miniscule amount of prime RMBS; and (2) that the actual Pyxis portfolio complied with the investment criteria and quality guidelines. FGIC had also already received at least two Trustee Reports for Pyxis.

455. FGIC consented. The consent form stated: "[T]he deal [Pyxis] is performing as expected." Id. The form bore signatures from Menhenett, two surveillance analysts, and Mathias Pulster, who was the Head of Credit Risk Management. Their consent is inconsistent with FGIC's claim it was deceived by Putnam in August 2006.

456. A month later, in May 2007, FGIC insured its largest CDO-type transaction ever—a $2.5 billion deal called Havenrock II. Approximately $70 million of the Havenrock II initial reference portfolio consisted of Pyxis notes. FGIC performed diligence on the Havenrock II collateral assets, one of which was Pyxis. Nonetheless, and despite the fact that the Pyxis portfolio contained only one prime RMBS asset compared to what had been reflected in the PCS, FGIC went forward with the Havenrock II transaction and did not complain about the Pyxis portfolio to either Putnam or Calyon.

457. As will be discussed below, the 2007-08 financial crisis forced price declines, rating downgrades, and losses in the assets that collateralized Pyxis. Although Menhenett testified that she did not review the February 2007 Pyxis Trustee Report, she acknowledged that she and her team did review the Trustee Report that they received in July 2007.

458. By August 2007, a substantial number of RMBS assets had been downgraded as a result of the credit crisis.

459. On August 15, 2007, Sverdlov emailed the MBS group (including Menhenett) with an agenda for a meeting the following day. The second item on the agenda was: "report for Pyxis CDO transaction (prepared by Dana Skelton)." D-A20. When asked if she recalled whether there were meetings scheduled among members of the CDO and MBS groups to discuss FGIC's CDO deals in the aftermath of the July rating downgrades, Menhenett testified with characteristic inconsistency: "I don't recall that. I believe this was a committee-style meeting, where the committee would have . . . presented . . . to senior management." Tr. 223. When asked if she attended those meetings, Menhenett said, "I'm vague on that. I don't recall . . . I may have." Tr. 224.

460. Sverdlov's August 15, 2007 email attached a memorandum about Pyxis prepared by Skelton. Her memorandum discussed the then-current state of Pyxis and the analysis comparing it to the original Pyxis portfolio. Specifically, it reported: "The portfolio is concentrated heavily in subprime sector with 91% of bonds coming from subprime MBS deals and 7% from Mezz ABS CDO, which, in turn, are comprised primarily of mezzanine bonds backed by subprime MBS transactions. The Alt-A and CMBS CDO concentrations are negligible." D-A20. A chart showed that the portfolio included 91.1% subprime, 7.02% mezzanine ABS CDOs, .9% CMBS CDO, and .98% of Alt-A. There was no mention of any prime assets.

During cross examination, Menhenett attempted to counter defense counsel's procurement of her admission that the chart contained no prime assets by advising that "some marketplaces may have categorized Alt-A as prime." Tr. 225. Defense counsel promptly elicited testimony that FGIC did not categorize Alt-A as prime, that FGIC recognized that Alt-A was of a lower credit quality than prime, and that there are understood differences between Alt-A products and prime products.

461. Skelton's memorandum also charted the then-current analysis compared to the original analysis reflected in the Credit Memorandum, reflecting the difference of 11% prime in the original analysis and only 1% prime in the current analysis:


Current Analysis

Original Analysis


100% ramped:• 91% Subprime• 1% Prime• 7% ABS CDO• 1% MBS CDO

63% ramped:• 89% Subprime• 11% Prime• 0% ABS CDO• 0% MBS CDO

Sample Size

87% sample

30% sample

Loan Population

~ 12 - 24 month seasoned loans

0 - 12 months seasoned loans


462. Menhenett testified that she read Skelton's memorandum. She testified that she would have read the chart contained within it too. She noticed the difference between the amount of prime in the August final portfolio and that reflected in the PCS.

463. In spite of the so-called "dramatic difference" between the actual Pyxis portfolio and the PCS, and the fact that she could have easily contacted them, Menhenett did not raise any issues with Bell or Putnam or have any conversations with them about Pyxis. Tr. 228. Nor did she or anyone else from FGIC complain about the final Pyxis portfolio, state that it was different from what was originally represented, or request that changes be made and assets be added. Although she testified that she was "shocked" by the difference between the prime in the original and current analyses, that testimony is undercut by the fact that she did not say anything of the sort to her colleagues contemporaneously or raise any issues with Putnam. Tr. 228.

464. Skelton's memorandum reflects that, on August 15, 2007, FGIC conducted a "second" diligence call with Putnam to determine the manager's expectation of performance and strategy for handling the transaction going forward. D-A20. Her memorandum states that Menhenett, Skelton, Carson, and Sverdlov were on the call, and it explains that Putnam had reported that it considered the transaction distressed and was open to exploring loss mitigation strategies. There was no discussion of the amount of prime or seasoned RMBS in the Pyxis portfolio or the differences between the final Pyxis portfolio and what FGIC claims that it expected as of the time it approved entering the transaction with Calyon with respect to Pyxis.

465. On August 15, 2007, the same day that Menhenett received the analysis showing that less than 1% of the Pyxis portfolio had been invested in prime RMBS assets, she had a conversation with Bell. The fact of their call is memorialized in Menhenett's August 20, 2007, email to Bell that opens with the line: "I have a few items to follow up (listed below) since our call last wk. Please let me know if you have a few minutes this afternoon so we can briefly discuss." D-B95. She continued: "Separately, our team thought the 8/15 call was productive so we'll move the on-site visit to later in the year . . . ." Id.

466. Menhenett's August 20, 2007 email requested six categories of information about Pyxis's performance. None of them related to Pyxis's asset composition, the amount of prime or seasoned RMBS therein, or any discrepancy between Pyxis's final contents and those in the PCS. Although Menhenett offered various rationalizations on redirect as to why she would not have spoken to Bell about the differences between the actual Pyxis portfolio and the allocations reflected in the PCS, the Court finds it inconceivable that she would not have said something in August 2007 if the issue had been essential to FGIC in deciding to underwrite Pyxis. The implication is apparent that FGIC did not consider the allocations reflected in the PCS to be material.

467. On August 16, 2007, Skelton emailed two FGIC analysts, copying Menhenett, with the subject line "Pyxis." D-J13. Skelton's email stated: "Since both of you worked on this deal I want to keep you in the loop. The work you did most recently went into the August update on the deal. I also included the July surveillance memo and the original SCC memo for your reference." Id. The July surveillance reflected that Pyxis was "concentrated heavily in subprime sector with 91% of bonds coming from subprime MBS deals and 7% - from Mezz ABS CDO, which, in turn, are comprised primarily of mezzanine bonds backed by subprime MBS transactions." Id. It added, "The Alt-A and CMBS CDO concentrations are negligible." Id.

468. On September 10, 2007, Bell responded to some of Menhenett's requests for information from August 20, 2007. Among other things, Bell provided a list of all the RMBS assets in the deal. The list reflected only one asset that could be characterized as prime RMBS. Menhenett testified she could not remember a call with Bell after she received this information. There is no evidence that she, or anyone else from FGIC, raised any concern either internally or with Calyon or Putnam about the absence of prime or seasoned RMBS assets from the final Pyxis portfolio or any discrepancies between the PCS and the final Pyxis portfolio or asked that prime or seasoned RMBS be added to the portfolio.

469. On August 27, 2007, Menhenett emailed FGIC's MBS and CDO groups, saying: "Attached are memos for today's meeting." D-A60. Her email attached a memorandum (the "LCR Memorandum") about Pyxis. Id. Menhenett testified that she believes she played a role in preparing the CDO analysis contained in the LCR Memorandum.

470. Menhenett testified by declaration that, had FGIC known of the composition of the final Pyxis portfolio and not based its analysis on the allocations and information reflected in the PCS, it would have relied on an MBS loss estimate at a 99% confidence level rather than a 97% confidence level, which would have generated a higher loss estimate. FGIC made that argument at trial. There is no evidence to support it besides FGIC's after-the-fact assertions. (FGIC apparently did apply a higher confidence level when analyzing single RMBS.)

The Court discounts Adams's testimony as discussed below and does not find it to be credible.

471. The LCR Memorandum sent by Menhenett on August 27, 2007 was based on the final Pyxis portfolio. At the time, FGIC knew the correlation among all the final assets. The analysis was based on "the most current portfolio composition, including RMBS, ABS CDOs and CMBS CDOs." Id. Yet, when FGIC calculated the MBS loss estimate, it used a 97% confidence level, not a 99% confidence level, as illustrated by the chart below that was included in the memorandum. The chart reflected that, even after all of the credit losses of July and August 2007, and even knowing the final Pyxis portfolio, FGIC calculated a loss estimate of 45% as compared to the 41.15% calculation in the credit application:

The "New method" referenced in the top left corner was a method that FGIC developed in August 2007. It took into account delinquencies in underlying loan payments as well as ratings downgrades for the RMBS in the Pyxis collateral pool.


RatingOf our position

LCR Deficiencyfor the Portfolio(Wt. Avg.)

Loss to FGIC (%Portfolio)

Loss to FGIC(% Exposure)

Loss lo FGIC(SMM)

New methodMean LCR97%LCRMax LCR






Original method97%LCR






Prior Results (7/16/07)DGs - 3DGs - 3 + 2xCorrIndex-based analysis






Actual Current Ratings


At time of UW97%LCR(FGIC ratings)





472. On August 30, 2007, FGIC prepared a set of slides for its board of directors entitled: "Heightened Surveillance BoD [Board of Directors] slides." D-A41. The slides were sent by an FGIC analyst to Finkel, Menhenett, and Hurst with a note that the draft "incorporate[d][ [Menhenett]'s edits." Id. One of the slides concerned Pyxis (the "Pyxis Slide"). Id. By that time, as noted in the Pyxis Slide, 25% of the underlying assets in Pyxis had been downgraded and the transaction had been placed on "heightened surveillance" the prior month. (Pyxis was not alone. The presentation also included draft slides for three other CDO transactions that had been placed on heightened surveillance in July 2007, after downgrades in their portfolios of 5%, 21%, and 22%, respectively.) The Pyxis Slide summarized for the board of directors that "92% and 8% of the collateral pool is nonprime and CDO, respectively" and indicated that none of the RMBS had a vintage earlier than 2005. Id.

473. The Pyxis Slide did not say anything about any discrepancy between FGIC's original understandings in approving the transaction and the final portfolio, or that Putnam had misled FGIC with the PCS (or otherwise). About Putnam's performance as a manager, the slide only said: "Putnam is a large and financially stable manager with average RMBS credit skills." Id.

474. The Pyxis Slide also stated that FGIC had conducted three calls with Putnam: on July 12, 2007; July 15, 2007; and August 15, 2007. It advised that future calls were scheduled to occur monthly. On none of those calls did FGIC say anything about the discrepancy between the PCS or what FGIC claims it expected about the final Pyxis portfolio at the time it approved the transaction with Calyon.

475. Menhenett was asked whether, as of August 2007, she had reason to believe the portfolio "might" be materially different from the PCS portfolio. Tr. 630. She affirmed, "Not only yes, I would remove the word 'might.'" Id. She testified FGIC "recognized that it was materially different." Id. But if Menhenett or any other FGIC employee believed that Putnam had made a misrepresentation while underwriting Pyxis and it was material, she never raised the issue with the Board. That communication further undercuts FGIC's current argument that the PCS constituted a representation of what would be in the Pyxis portfolio, and that the difference between the assets in the PCS and the final Pyxis portfolio was critical to its underwriting decision.

476. Finally, in February 2008, Menhenett emailed her colleagues with loss projections for a number of CDOs including Pyxis. The loss projection for Pyxis was 43% at a 97% confidence level—the same confidence level applied to the portfolio in the Pyxis credit application, even though 25% of it had been downgraded by February 2008. The document stated that Pyxis had 0% prime. No complaint was lodged with Putnam (or anyone else).

477. Menhenett testified via declaration that it was only during the course of the litigation that she became aware the actual Pyxis portfolio was materially different from the PCS portfolio insofar as it included materially less prime RMBS, pre-2005H2 RMBS, and assets rated above Baa2. That testimony is contradicted by the numerous communications to Menhenett showing that she knew as early as February 2007 and certainly by August 2007 the composition of the fully-ramped portfolio (and by her own testimony on cross-examination admitting the same). Although Menhenett tried to explain away her declaration, none of the rationalizations were compelling.

478. Menhenett expressly acknowledged at trial that she understood market conditions could change between August 2006 (when FGIC was sent the PCS) and February 2007 (when the assembly of the final Pyxis portfolio was to be completed), and she testified that that, indeed, they could change at "any time" and that "each month," "the market introduces different factors." Tr. 661. However, neither Menhenett nor any other FGIC employee checked with Putnam or Calyon to determine if the target portfolio had changed.

479. The Court concludes, first, that FGIC did not understand the information conveyed in the PCS to constitute a promise or a representation of the percentage of different types of securities that would be purchased by Pyxis and, second, that FGIC considered any disparity in the amount of securities in the different buckets unimportant.

28. Additional Evidence Regarding Reliance

480. At trial, FGIC contended that it would not attach below the 41.15% attachment point where the MBS team's LCR review expected there would be losses. Both Finkel and Adams (but not Menhenett) testified that to that effect. The testimony was offered in support of the argument that, had FGIC known that Putnam would invest the Pyxis portfolio in the assets reflected in the final Pyxis portfolio, rather than those in the PCS, the LCR loss estimate would have been significantly above 41.15% and FGIC would have rejected the transaction at an attachment point of 40% and demanded that it attach at a 50% attachment point.

481. There were several flaws in the argument. First, the contemporaneous evidence from the Pyxis transaction does not support the contention that FGIC would not attach below the LCR loss estimate. On August 16, 2006, the RMBS loss estimate Menhenett presented to Finkel and Adams, and ultimately to the SCC, based on LCR analysis, yielded a result at the 97% confidence level of 41.15%. That figure of 41.15% did not prevent FGIC from making the decision to attach at the 40% attachment point. If FGIC was willing to attach at a level one point below the MBS loss estimate (thus exposing it to risk of loss based on that estimate alone), FGIC offered no evidence to support the notion that a higher loss estimate than 41.15% would have prevented it from attaching at the 40% attachment point.

482. Second, the contemporaneous language that the relevant decision-makers used to describe the significance of the LCR loss estimate makes clear that they did not believe that there was a hard-and-fast rule that FGIC would not attach below it. They understood, and expressed, that the LCR loss estimate was one factor among many to consider in deciding whether to participate in a transaction. In her August 16, 2006 email, thus, Menhenett proposed an attachment of 35%, well "below this loss check of 41.15%." PX920. She explicitly stated, "We haven't used this as a hard rule in the past[.]" Id. And she concluded that 35% attachment "work[ed] fr risk perspective." Id. Finkel responded to Menhenett's request for approval to proceed at a 35% attachment level by advising that she "want[ed] to review final mbs analysis and alex's input before going back"; as she acknowledged in her testimony, the difference between the 35% attachment point and the 41.15% loss estimate did not cause her to give a "hard no." Tr. 787. Clearly, from her contemporaneous perspective also, the LCR loss estimate was only one factor among many and the loss estimate did not need to be below the attachment point for FGIC to be willing to engage in a transaction.

483. The credit application for Pyxis too relied substantially on the sensitivity and cash flow analysis and not on the LCR loss estimate. The sensitivity and cash flow analysis was presented in the body of the credit application. It showed that, looking at the portfolio as a whole, and considering its structural features, FGIC was not exposed to loss even though the LCR loss estimate was above the attachment point. The sensitivity and cash flow analysis was not based on the PCS or the information in the PCS but on the hypothetical portfolio that could be constructed based on the portfolio constraints. In other words, it looked at the risk to FGIC given the wide discretion Putnam had to make investment decisions within the portfolio constraints and Putnam's legal obligation to make those decisions in the interest of Pyxis itself, and not necessarily in the interest of the most risk-averse tranche of Pyxis investors. It concluded that, regardless of the investments made within those contractual constraints, FGIC should not be exposed to risk of loss. The LCR loss estimate was buried in an appendix in the back of the credit application and was not referenced in the body of it.

484. Furthermore, as noted above, there were weaknesses in the LCR model (which FGIC acknowledged). It ignored the structural protections in the CDO, which would result in cash flow being diverted away from the other tranches of securities and that would improve the super-senior's credit position, such as excess spread and triggers. It assumed that all losses that would ever occur in the portfolio would occur immediately, rather than at some date in the future, and thus generated a loss estimate more conservative than the analysis of the underlying portfolios suggested (i.e., they looked at the risk of loss over time). Also, it used FGIC's conservative rating assumptions rather than those of credit rating agencies. And it ignored the non-RMBS portion of the portfolio. For those reasons, the contemporaneous evidence that it was not critical for the LCR loss estimate to be below the attachment point makes sense. By applying such a rule, FGIC would have left transactions—and money—off the table, eschewing to engage in transactions and to receive the premiums from them when an analysis of the transaction as a whole would have suggested that it would not generate a loss and would instead be profitable for FGIC. Indeed, for those reasons, it is not surprising that the LCR analysis was relegated to Appendix 8 of the credit application, far behind the sensitivity and cash flow analyses.

485. FGIC's witnesses retreated from their testimony regarding the importance of the LCR analysis on cross-examination. Finkel admitted that FGIC would attach at a level below where the MBS group expected losses would fall when there were structural features that protected FGIC. As noted above, during her deposition, she stared at the LCR analysis for seven minutes before admitting she did not recognize it.

486. Moreover, the evidence reveals there were other transactions, such as the BFC Ajax transaction, where FGIC attached at a point lower than the MBS loss estimate. In the BFC Ajax transaction, FGIC attached at 38%, far lower than the MBS loss estimate of 85%. Although FGIC sought to dismiss these precedents on the theory that they were unique transactions, FGIC also admitted that every transaction has its unique aspects.

487. FGIC's principal witness on the question of reliance was Adams. Adams testified on direct and by declaration that he "would not approve a CDO transaction where the MBS group estimated aggregate transaction level losses that exceeded the level of FGIC's credit enhancement for such CDO." Adams Decl. ¶ 41. Adams added that, based on what he implied was a comparable transaction involving Deutsche Bank, FGIC would use the 99% confidence level for MBS analysis if the CDO portfolio was "very highly correlated" with even higher loss estimates. Id. He stated that he "took the position that based on MBS group's loss estimate of 41.15%, at 97% confidence level [in the Pyxis transaction], FGIC should not attach below 40 percent expected excess spread to provide credit enhancement." Id. ¶ 55. However, had he "understood that the portfolio consisted entirely of mezzanine subprime RMBS with no prime RMBS and virtually no RMBS issued before 2005H2[,] [he] would have required that FGIC attach at no lower than the loss estimate generated at the 99% confidence level rather than that generated at the 97% confidence level." Id. ¶ 59. Further, he testified that had he "known that the reason the Pyxis portfolio would be less diversified and riskier than represented in the PCS was that the collateral mix was dictated by equity investors who were using Pyxis to facilitate a long-short correlation trading strategy, [he] would have required substantial additional scrutiny of the deal before approving FGIC's participation and likely not have been able to approve FGIC's participation at all." Id. ¶ 61.

Menhenett was an unreliable witness and the other witnesses had no probative testimony with respect to decisions made in 2006.

488. Adams's direct testimony, for which he was paid at an hourly rate by FGIC, was thoroughly discredited at trial. His statement on direct that he would not have approved of FGIC attaching at a point lower than the RMBS loss estimate was founded on an allegedly comparable bespoke transaction with Deutsche Bank which he claimed was not approved because the loss estimate was above the attachment point. However, the evidence established that he did not object to the Deutsche Bank transaction being presented to the SCC, which Adams omitted from his declaration. Without being refreshed on cross-examination, he did not recall the most basic facts about the supposedly comparable transaction, including the tranche that FGIC was being asked to insure (unlike in Pyxis, the junior super-senior) or the attachment point. After being refreshed, he admitted the transaction was one of "first impression" for FGIC, which was why it was ultimately not presented to the SCC. Tr. 936. FGIC was not being asked to insure the most senior tranche. The pool was static and not managed; the same institution was serving as sponsor, structuring bank, and collateral agent; and the party selecting the collateral was compromised because it sought to use the CDO to move risk from its books to FGIC's. Finally, at the time, FGIC did not obtain loan tapes for any of the proposed portfolio assets. All of those facts distinguished the Deutsche Bank transaction from Pyxis and make the evidence of what was done with respect to Deutsche Bank unreliable as a gauge of what would have been done with respect to Pyxis had FGIC assumed the asset composition of the final Pyxis portfolio rather than that reflected in the PCS.

489. Adams's testimony that he would never approve FGIC attaching at a point below RMBS loss estimate was undermined by the fact that, at his recommendation, FGIC was prepared to attach in the Deutsche Bank transaction at a 25% point when it projected a 40% loss at a 99% confidence level, and 28% loss at a 97% confidence level. Adam's testimony was further belied by testimony that in the BFC Ajax transaction, FGIC agreed to participate at a roughly 30% attachment point even though the loss estimate generated by the LCR analysis was 85%.

490. Adams's additional testimony that, based on the Deutsche Bank transaction's analysis of losses at the 99% confidence level, FGIC would have conducted loss estimates at a 99% confidence level if FGIC had it known of the final portfolio, was undercut in numerous respects. FGIC always reported loss deficiency at a 99% confidence level as well as 97%. When Adams expressed acceptance of the 40% attachment point for Pyxis on August 24, 2006, well before the credit application, he would not have known anything more about the contemplated composition of the portfolio other than what was reflected in the pitchbooks (and certainly not what was in the PCS). Neither Menhenett's August 16, 2006 email nor any of the other documents Adams was on referred to the PCS or the allocation reflected in it. He only learned of the allocations that were in the PCS (without knowing about the PCS) from the credit application that arrived weeks after he agreed to the transaction. Adams's testimony is further undercut by the fact that, in August 2007, FGIC knew exactly what assets were in the Pyxis portfolio and analyzed the risk only at the 97% confidence level. Although he was shown language on redirect examination from the Pyxis credit application to the effect that "a moderately diversified pool such as one that contained 50 or more underlying RMBS collateral pools, which had a break-even LCR level of 97 percent would equate to a single collateral pool with a break-even level of 99.99 percent," PX325, his explanation of that language was singularly unhelpful to FGIC: he testified that the comparison being made was between a standalone MBS transaction that FGIC would insure only at the 99% confidence level, and a CDO transaction that had "not just one single mortgage deal but dozens of them." Tr. 918-20. In other words, it was the number of transactions in the CDO (thus reflecting diversification) and not their asset sectors that determined whether a 97% or a 99% confidence level would be used. Finally, his testimony on direct is undercut by the fact that, whether at a 97% or a 99% confidence level, FGIC was measuring how the bonds would perform in a tail-risk scenario that assumed the assets would perform similarly in a market downturn. In short, there is no convincing evidence to support Adams's proposition.

Adams also contradicted his declaration testimony that FGIC preferred underlying assets with higher credit ratings over assets with lower credit ratings, all else being equal, by admitting that FGIC insured a number of CDOs that were backed by assets with an average rating of BBB or BBB-. Adams's testimony that FGIC preferred RMBS with a meaningful amount of seasoning was undercut by the transactions he approved (several of which he did not remember) that did not have seasoned assets. His testimony that FGIC generally sought to insure tranches with at least two times the deal's initial AAA attachment point was undercut by transactions FGIC had insured that were once AAA. His testimony that he would have been "sensitive to whether the CDO began as a reverse inquiry," Adams Decl. ¶ 32, was undercut by the fact that, even after he was informed that Pyxis was started as a result of an inquiry from an equity investor, he could recall no steps he took to follow up on that information.

491. The second weakness in Plaintiff's argument is that there was no evidence that the LCR loss estimate would have been higher if FGIC had analyzed the final portfolio. FGIC attempted to introduce such evidence through Skelton, whose declaration contained testimony about how the LCR analysis would apply to the actual Pyxis portfolio if FGIC had been given the actual Pyxis portfolio in advance of its decision to insure. The Court struck that testimony at trial as impermissible expert testimony and testimony based on improper speculation. Skelton did not have personal knowledge of how FGIC would have calculated the LCR loss estimate or what results such analysis would yield had FGIC had the actual final Pyxis portfolio in August 2006. She testified to what she understood from the documents and from her background as to what FGIC did with the hypothetical Pyxis portfolio it assumed based on the notion that, as to the assets Putnam had not purchased, it would purchase not those specific assets but only assets of that same asset type. Thus, while FGIC analyzed a subset of the actual assets in the ramped portfolio based on low-level detail of the actual bonds in the underlying pool for those assets (or a good proxy), FGIC operated on the assumption that the unpurchased assets in the PCS were "dummy assets or placeholders." Tr. 744. The MBS team understood that they were "just representative assets," "not specific," and that there were "no inferences to be made [on] how they would perform exactly." Tr. 744-45. But there was no evidence comparing the actual assets purchased for the Pyxis portfolio (those which FGIC's process would have analyzed as to their specifics) with the assumptions based upon which FGIC analyzed the unpurchased PCS assets considering them to be "dummy assets." Id. That would have been the work of an expert, and not of a lay witness, particularly of a lay witness who disclaimed memory of any events from 2006. Moreover, even had the evidence been admissible, the Court did consider it and found it unpersuasive. It would not have moved the needle. As the Court noted, Skelton's analysis was apples to oranges—assuming that FGIC knew the actual assets in August 2006 but applying an analysis based on the notion that the assets were not actual assets but merely dummy assets.

Putnam made a motion to strike this testimony before trial began. After considering the arguments on both sides, the Court determined that it would "be useful to have voir dire of Skelton on the point before rendering a decision." Dkt. No. 344 at 12. The Court's decision to strike Skelton's testimony was informed by that voir dire.

29. The Downfall of Pyxis and FGIC's Commutation Agreement

492. The Court likely need not remind readers who have made it this far into the opinion that the housing market collapsed in the summer of 2007 and the country fell into a credit crisis, resulting in the value of RMBS generally plummeting across the board. It catapulted the American economy into crisis. And it produced catastrophic losses for CDOs that had been issued in the second half of 2006 and the first half of 2007 regardless of the specific assets chosen for the portfolio—losses that reached even the highest tranches of those CDOs' collateral structures.

493. As of August 30, 2007—within 11 months of Pyxis's closing—25% of Pyxis assets had suffered credit rating downgrades. On December 8, 2008, the Pyxis trustee declared that an event of default for Pyxis had occurred, which indicated that the collateral assets and reserves of Pyxis were insufficient to meet its expected obligations. By January 2017—nine years after closing—Pyxis had only repaid a few tens of millions of dollars of the $1.5 billion invested in it.

494. FGIC's liabilities on various CDOs like Pyxis led it into financial peril. By the end of 2007, FGIC was reporting large losses. As of March 2008, S&P had downgraded FGIC's credit rating by six notches—from BBB to B. It continued to drop until April 2009, when S&P lowered its rating from CCC to CC, indicating that the company was near default. Then S&P withdrew its ratings, citing the company's inability to provide sufficient information.

495. On July 2, 2009, FGIC negotiated a commutation agreement with Calyon by which it agreed to pay $100 million to discharge its liabilities to Calyon associated with the Pyxis Guaranty and two other CDOs which had also suffered losses and as to which it had issued Calyon a guarantee. The commutation agreement did not specify what portion of the total payment was attributable to which CDO. FGIC argues that $74.5 million of that amount was allocated to Pyxis. The Court need not resolve what portion of the commutation payment was attributable to Pyxis because the question of damages was reserved for Phase 2. But for the commutation agreement, FGIC would have had to pay Calyon $782 million (the amount of payments Calyon eventually made to Pyxis) beginning in September 2016.

496. On October 1, 2012, FGIC filed its Complaint against Putnam. As described (far) above, the Complaint focused primarily on allegations that Putnam had misrepresented that it would "independently" select collateral for the Pyxis portfolio (when in fact selection was controlled by equity investors). Dkt. No. 1 ¶ 1. The Complaint also alleged that the final Pyxis portfolio lacked the amount of prime RMBS and seasoned RMBS that had been promised in the PCS. This federal court complaint, filed six years after Pyxis closed, and in the aftermath of a devastating financial crisis, was the first time FGIC alerted Putnam—or anyone—to this apparently game-changing bait-and-switch.


497. FGIC brings claims for fraud, negligent misrepresentation, and negligence.

498. "Under New York law, to state a claim for fraud, a plaintiff must demonstrate: (1) a misrepresentation or omission of material fact; (2) which the defendant knew to be false; (3) which the defendant made with the intention of inducing reliance; (4) upon which the plaintiff reasonably relied; and (5) which caused injury to the plaintiff." Wynn v. AC Rochester, 273 F.3d 153, 156 (2d Cir. 2001); Pasternack v. Laboratory Corp. of America Holdings, 59 N.E.3d 485, 491 (2016) ("The elements of a fraud cause of action consist of a misrepresentation or a material omission of fact which was false and known to be false by [the] defendant, made for the purpose of inducing the other party to rely upon it, justifiable reliance of the other party on the misrepresentation or material omission, and injury.") (internal citations and quotation marks omitted). "To establish causation, the victim of fraud must show both transaction causation—that the fraud caused the victim to engage in the transaction in question—and loss causation—that the misrepresentations or omissions caused the economic harm." Cont'l Grain Co. v. Meridien Int'l Bank, Ltd., 894 F. Supp. 654, 661 (S.D.N.Y. 1995).

499. "Under New York law, the elements for a negligent misrepresentation claim are that (1) the defendant had a duty, as a result of a special relationship, to give correct information; (2) the defendant made a false representation that he or she should have known was incorrect; (3) the information supplied in the representation was known by the defendant to be desired by the plaintiff for a serious purpose; (4) the plaintiff intended to rely and act upon it; and (5) the plaintiff reasonably relied on it to his or her detriment." Hydro Inv'rs, Inc. v. Trafalgar Power Inc., 227 F.3d 8, 20 (2d Cir. 2000).

500. As the plaintiff, FGIC bears the burden of proving every element of every claim. FGIC must prove each element of its fraud claim by "clear and convincing evidence." Merrill Lynch & Co. Inc. v. Allegheny Energy, Inc., 500 F.3d 171, 181 (2d Cir. 2007). FGIC must prove its negligent misrepresentation claim by a "preponderance of the evidence." Kortright Capital Partners LP v. Investcorp Inv. Advisers Ltd., 392 F. Supp. 3d 382, 402 (S.D.N.Y. 2019); see Tuckett v. Slade Indus., Inc., 2018 WL 3910821, at *5 (S.D.N.Y. Aug. 14, 2018).

501. As noted above, the Court bifurcated trial into two phases pursuant to Fed. R. Civ. P. 42(b). During Phase 1, the Court tried the following issues: (1) whether Putnam is responsible for the PCS or had a duty to correct it; (2) whether Putnam had a duty to correct the PCS under the "special facts" doctrine; (3) actual reliance or transaction causation; (4) reasonable reliance; and (5) loss causation. On those matters, the Court has reached the following conclusions of law.

1. Fraud

A. Whether Putnam is Responsible for the PCS

502. The Court concludes that FGIC has failed to prove its fraud case by clear and convincing evidence. The evidence does not support the arguments that Putnam made a false or misleading statement or was involved in the statement made to FGIC or that it had any duty to disclose; that FGIC relied on the PCS; that FGIC reasonably relied on the PCS; or that the alleged falsity of the PCS caused any loss on the part of FGIC. The Court would reach the same conclusions on a preponderance of the evidence standard.

503. Courts applying New York law have held defendants liable as principals for fraud (who take on corresponding duties to correct), even when third parties make the alleged misstatements to putative victims under two circumstances. Those circumstances are: (1) where the defendant makes a false or misleading statement to an intermediary, intending it to be conveyed to the plaintiff and it is so conveyed, see Peerless Mills, Inc. v. Am. Tel. & Tel. Co., 527 F.2d 445, 451 (2d Cir. 1975); Pasternack, 59 N.E.3d at 492-93; and (2) where the defendant has "so 'entangled itself' with the issuance of" a document that false statements therein "may, in effect, be attributed to the [defendant]," In re ICN/Viratek Sec. Litig., 1996 WL 164732, at *3 (S.D.N.Y. Apr. 9, 1996) (citation omitted), or constitute "an implied representation [from the defendant] that the information" being conveyed "is true or at least in accordance with the [defendant's] views." Elkind v. Liggett & Myers, Inc., 635 F.2d 156, 163 (2d Cir. 1980); see In re N. Telecom Ltd. Sec. Litig., 42 F. Supp. 2d 234, 249 (S.D.N.Y. 1998) ("Indicia of entanglement may include a corporation's review, correction and approval of a final version the analyst's report."); see also Aris Multi-Strategy Offshore Fund, Ltd. v. Devaney, 907 N.Y.S.2d 98 (Sup. Ct. 2009) (citing Elkind as applicable to common law fraud claim); Prime Mover Capital Partners L.P. v. Elixir Gaming Techs., Inc., 898 F. Supp. 2d 673, 696 (S.D.N.Y. 2012), aff'd, 548 F. App'x 16 (2d Cir. 2013) ("The elements of common law fraud are substantially identical to the elements of a Section 10(b) claim."). FGIC has not established a basis for Putnam's liability under either theory.

504. In order to establish a claim under the line of cases identified in Pasternack, it is critical that (1) the defendant be the source of the false and misleading statement or information conveyed to the plaintiff; and (2) the defendant intends and knows that the information will be conveyed to the plaintiff. 59 N.E.3d at 492-93 (citing Eaton, Cole & Burnham Co. v. Avery, 83 N.Y. 31 (1880); Rice v. Manley, 66 N.Y. 82 (1876); Bruff v. Mali, 36 N.Y. 200 (1867)). The intermediary acts as a conduit. In other words, "if A. makes the statement to B. for the purpose of being communicated to C. or intending that it shall reach and influence him, he can be . . . held [liable]." Eaton, 83 N.Y. at 34-35.

505. The facts of Eaton are instructive. In that case, the defendant firm gave false information about its capital to a "mercantile agency" (a present-day credit reporting agency) knowing that the agency would convey the information to its customers seeking to sell goods to the defendant firm on credit. Id. The mercantile agency then reduced that information to a writing and, when the plaintiff later asked the mercantile agency for information about the defendant firm, the mercantile agency conveyed the false information to the plaintiff, on which the plaintiff firm relied to its detriment. Id.

506. The New York Court of Appeals concluded that it was "clear that the defendant, when making the [false statements], knew that they were to be used by a mercantile agency" and "understood the effect of the representations, and desired to have a high [credit] rating," which established "a fraudulent intent." Id. at 38. "[T]he alleged false statement [about] the capital . . . was made for the purpose of giving [the defendant firm] a credit to which it was not entitled, and of defrauding any person who might inquire of the agency, or consult its lists." Id. (noting that the defendant firm "knew that these [mercantile] agencies issued such lists to merchants who paid for them"). As the Eaton court summarized, "[a] person furnishing information to such an agency in relation to his own circumstances, means and pecuniary responsibility, can have no other motive in so doing than to enable the agency to communicate such information to persons who may be interested in obtaining it, for their guidance in giving credit to the party." Id. at 34 (emphasis added). Crucially, the false information conveyed by the mercantile agency in Eaton found its source in the defendant firm. The mercantile agency had not developed the information on its own. Moreover, the actionably false information that the mercantile agency conveyed to the plaintiff was the precise information that the defendant had given to the mercantile agency.

Even assuming Putnam provided the information in the PCS to Calyon, Putnam would have had many reasons to do so other than that Putnam expected Calyon to communicate that information to FGIC (or even investors generally). Calyon was actively working on structuring the transaction. Calyon and Putnam were, between the two of them, balancing various investor concerns. Calyon had an ultimate veto right over the assets. For this reason, Calyon's role in Pyxis—vis-à-vis Putnam—is not comparable to the role that a credit reporting agency has with respect to the individuals from whom it receives information for distribution to others.

507. Bruff stands for the same proposition. The defendants there issued or caused to be issued false stock certificates attesting them to be genuine and then "thr[ew] them on the market . . . 'with a view to well known and established commercial usages' . . . with intent to defraud any and all purchasers, well knowing that every person to whose hands these false certificates should come by fair purchase might be injured." 36 N.Y. at 200. The court held that, regardless of the absence of privity, "having authenticated and issued these certificates for the purpose of defrauding, the defendants should be held liable to any one sustaining damage by purchasing on the faith of their genuineness." Id. "[T]he defendants having issued the false certificates of stock authenticated by them as genuine, and cast them upon the market with fraudulent intent, are liable to every holder to whose hands they may come by fair purchase." Id.

508. FGIC has failed to establish liability under Pasternack, Eaton, and Bruff. Unlike the false information in Eaton and Bruff, which originated with the defendant and was merely conveyed to the victim through an innocent intermediary, the allegedly false information here did not originate with Putnam. The information was generated by Calyon, for Calyon's own purposes, from Calyon's records, based on transactions of other portfolio managers and not transactions of Putnam. The information did not originate with Putnam. Putnam did not provide Calyon with security-by-security information regarding a target portfolio. The record was uncontradicted that Putnam did not have such information. Nor did Putnam convey information to Calyon about Putnam's intended purchases of specific securities for the Pyxis portfolio such that the information conveyed by Calyon could be said to have originated from Putnam in any sense. The last communications Putnam had with Calyon about the amount of prime RMBS Putnam expected to be included in the Pyxis portfolio were from mid-July 2006. Those communications reflected that Putnam intended the Pyxis portfolio to include no more than 4% prime RMBS, down from higher numbers in earlier target portfolios. And Putnam conveyed consistently negative views about the value of prime RMBS, as compared to midprime or subprime RMBS. If, as happened, Calyon conveyed figures higher than the 4% prime RMBS to FGIC, that was because Calyon—on its own—chose to do so. It was not because Putnam conveyed any information to Calyon that would have given Calyon the impression that Putnam would purchase that quantity of prime RMBS or any of the other securities in the PCS.

509. Second, even if the information in the PCS had been based on information Putnam had given Calyon (which it was not), there has been no showing that Putnam knew that Calyon was providing detailed target ratings information to FGIC, much less that Putnam provided such information to Calyon for purposes of dissemination to FGIC. Putnam communicated with Calyon about the ratings of target assets for two specific purposes. First, Putnam and Calyon updated the target portfolio for drafts of key marketing documents such as the pitchbook. Those target portfolios were intended to be viewed by all investors and were intended to be distributed on two occasions—the launch and the pricing. Second, Putnam would occasionally communicate with Calyon about specific future asset purchases Putnam would recommend for Calyon's own purposes and not for external consummation—because Calyon would need to measure assets that Putnam had purchased or proposed to purchase relative to deal targets. Putnam had no reason to suspect that Calyon would convey target information to FGIC on other than those two occasions, and no reason to suspect it would convey detailed asset-by-asset information to FGIC on any occasion whatsoever. Putnam had never done a transaction with FGIC before. There was no evidence it had ever received a request before from an investor or participant for an asset-by-asset list of target securities. There was no evidence it was even aware that any investor or participant ever used such a target list of securities in its credit analysis or for any other purpose or that, in general, investors or participants in the marketplace ask for such information from any collateral manager or structuring bank.

510. To establish liability under the Elkind line of cases, the plaintiff need not establish that the defendant alone is the source of the conveyed false and misleading information but it is essential to show that the defendant had a "level of involvement" with the allegedly fraudulent communication that is sufficient to "indicate entanglement." S.E.C. v. Wellshire Sec., Inc., 773 F. Supp. 569, 573 (S.D.N.Y. 1991). In such circumstance, false statements "may, in effect, be attributed to the [defendant]." ICN, 1996 WL 164732, at *3. In ICN, the standard was met because the defendant officers had given the third-party analyst "access to non-public information," had "reviewed the full text of" the analyst report, and had "made several changes and additions to [the] report to ensure its accuracy." Id. By contrast, in Elkind itself, the plaintiffs could not establish liability because the defendant had not "place[d] its imprimatur, expressly or impliedly, on the analysts' projections." 635 F.2d at 163. Such was the ruling notwithstanding the fact that the company had examined and commented on a number of draft reports (just not the earnings forecasts). Id. The defendant in Elkind even "made suggestions as to factual and descriptive matters in a number of the reports it reviewed," but still the court concluded that "the record [did] not compel the conclusion that this conduct carried a suggestion that the analysts' projections were consistent with [the company's] internal estimates." Id. "Nor," the court found, had "plaintiff demonstrated that [the company] left uncorrected any factual statements which it knew or believed to be erroneous." Id.

511. FGIC has not established the facts necessary to hold Putnam liable for the alleged misrepresentations in the PCS conveyed by Calyon to FGIC under the Elkind line of cases for several reasons. First, the information in the PCS was not in any way attributed to Putnam. It was delivered to FGIC in a document on Calyon letterhead. Unlike other communications that were prepared jointly by Calyon and Putnam, the PCS did not share the letterhead of Putnam. Moreover, Putnam did not review any part of the PCS, let alone its "full text." ICN, 1996 WL 164732, at *3. Nor did it check any of Calyon's work on the PCS, let alone all of it. Putnam did not know that FGIC had requested the information that was conveyed in the PCS, did not know that Calyon was preparing the information in the PCS, was not asked by Calyon (or by FGIC) about the PCS, and did not know that Calyon had prepared and delivered the PCS. The PCS was prepared entirely by Calyon, without any involvement or knowledge by Putnam.

FGIC has argued that the Court should infer that Calyon and Putnam worked with one another on the PCS because they worked with one another on target portfolios that appeared in pervious marketing materials. But the difference is that Calyon was required to cooperate with Putnam on the pitch materials, pursuant to the Mandate Letter. Calyon could not start marketing the transaction until it got Putnam's agreement. The same cannot be said with respect to the PCS.

512. The cases cited by FGIC are not to the contrary; indeed, they underscore why liability would not extend to Putnam.

513. Tindle v. Birkett cites Eaton as "controlling in the case at bar" because "the two cases are almost identical in their facts." 171 N.Y. 520, 524 (1902). Tindle merely reinforces the proposition that, "when a member of a firm makes statements to a commercial agency, which he knows to be false as to the financial condition of the firm, with the intent that the statements shall be communicated to persons interested in ascertaining the pecuniary responsibility of the firm, intending thus to procure credit and to defraud such persons, and such statements are communicated to one who in reliance thereon sells goods to the firm on credit, an action for deceit may be maintained against the buyer of the goods in favor of the seller who has suffered by the fraud." Id. In other words, it is critical that the false statements conveyed to the plaintiff have found their source in a statement made by the defendant.

514. Koch v. Greenberg, 14 F. Supp. 3d 247 (S.D.N.Y. 2014), aff'd, 626 F. App'x 335 (2d Cir. 2015) is similarly unavailing. In Koch, following a jury trial, the court rejected the defendant's argument that defendants cannot be liable in fraud for misstatements that they do not make directly to plaintiffs. Id. at 257. As the court explained, "even if [the defendant's] statements were made to [a third party] or its owner, rather than to [the plaintiff] himself, so long as [the defendant] reasonably expected auction attendees to rely on those statements—a conclusion the jury was free to reach in light of the record—they may be actionable representations under New York law." Id. at 257-58. Again, there, the defendant's statements were made to the third party and the defendant reasonably expected a "class of persons" that included plaintiff to rely on those statements. Id. at 257 (quoting Rest. (Second) of Torts § 533 (1976)). Those elements are lacking here.

515. FGIC fares no better with Nagelberg v. Meli, 299 F. Supp. 3d 409 (S.D.N.Y. 2017) and Ostano Commerzanstalt v. Telewide Sys., Inc., 794 F.2d 763 (2d Cir. 1986). In Nagelberg, the court denied a motion to dismiss because the plaintiffs "adequately allege[d] that [a defendant] knew that [one plaintiff] was considering [an] investment, knew that [another plaintiff] would relay [the defendant's] statements to [the first plaintiff], and knew that [the plaintiff who received the information from the other plaintiff] would rely on those statements" in making an investment decision. 200 F. Supp. 3d at 415; see id. ("One can reasonably infer that [the defendant] intended his statements to be conveyed to [the plaintiff who received the information from the other plaintiff]."). Ostano Commerzanstalt rests on the proposition that "a fraudulent misrepresentation made with 'notice in the circumstances of its making' that the person to whom it was made would communicate it to third parties subjects the person making the misrepresentation to liability to the third party." 794 F.2d at 766 (quoting Ultramares Corp. v. Touche, 174 N.E. 441, 444 (1931) (Cardozo, J.)). The cases are inapposite because Putnam did not make the statement at issue, and was not even the source for it.

516. Departing from the fraud-by-third-party cases, FGIC seeks refuge under Fresh Meadow Food Servs., LLC v. RB 175 Corp., 2013 WL 527199 (E.D.N.Y. Feb. 11, 2013), aff'd, 549 F. App'x 34 (2d Cir. 2014). In Fresh Meadow, the plaintiff-lessees brought a fraud claim against the defendant-lessors premised on the allegation that they (the plaintiffs) had executed a lease in reliance on the defendants' representations that "all underground petroleum bulk storage tanks [had been] removed from the [s]ite and that [the defendants] had no knowledge of additional storage tanks buried on the property at the time of the [l]ease's execution." Id. at *3. Indeed, the lease proclaimed that the defendants did not even have "constructive" knowledge of any conditions that would have interfered with the tenancy. Id. But the evidence at trial revealed that an individual defendant who made such a representation "was not in a position to verify that all of the tanks that existed at the [s]ite were removed because he [had] left the [s]ite during the removal before the job was completed." Id. at *4. "[T]he record contain[ed] no evidence that [this individual] made any attempt to follow up with the tank removal company after the day he saw them at the [s]ite to ensure the job was complete." Id. (emphasis added). And the court cited various other evidence that would have "put him on notice of the incomplete nature . . . removal job." Id. at *6. Based on all of these facts, the court concluded that "the plaintiffs [had] proved, by clear and convincing evidence, that [this individual defendant] acted, at a bare minimum, in reckless disregard of the truth in his representations to plaintiffs that all the underground storage tanks had been removed and that he had no knowledge of any hazardous materials under the ground, and that he therefore committed a fraud on the plaintiffs under New York state law." Id. at *9. In other words, the defendant had made a false statement and acted with the requisite scienter. The case is inapposite to the present one. It is well-established that a plaintiff can prove fraud in New York by establishing that a defendant "recklessly misrepresented a material fact" to the plaintiff. Merrill Lynch & Co., 500 F.3d at 181. But it is not established that a plaintiff can prove fraud by a defendant when the defendant was not the source of the false or misleading information provided to the plaintiff and did not know and was not on notice that such information was being provided to the plaintiff.

Already, the case is distinguishable from this one. The defendant here never saw the PCS and did not know of its existence.

B. Failure to Correct PCS as Culpable Omission of Material Fact

517. FGIC argues that, even if Putnam was not the maker of a fraudulent statement and the statement originated exclusively from Calyon, Putnam is liable in fraud for a culpable omission of material fact. But this theory fails because FGIC failed to establish any duty to disclose under any theory. "If a plaintiff is proceeding under a material omission theory, it must further allege that the 'defendant had a duty to disclose material information.'" First Hill Partners, LLC v. BlueCrest Capital Mgmt. Ltd., 52 F. Supp. 3d 625, 637 (S.D.N.Y. 2014) (quoting Nealy v. U.S. Surgical Corp., 587 F. Supp. 2d 579, 585 (S.D.N.Y. 2008)). A duty to disclose arises in the following circumstances:

[1] the parties are in a fiduciary relationship; [2] under the special facts doctrine, where one party possesses superior knowledge, not readily available to the other, and knows that the other is acting on the basis of mistaken knowledge; or [3] where a party has made a partial or ambiguous statement, whose full meaning will only be made clear after complete disclosure.
Aetna Cas. & Sur. Co. v. Aniero Concrete Co., 404 F.3d 566, 582 (2d Cir. 2005) (citations omitted); see also Miele v. Am. Tobacco Co., 770 N.Y.S.2d 386, 391 (2d Dep't 2003) ("New York recognizes a cause of action to recover damages for fraud based on concealment, where the party to be charged has superior knowledge or means of knowledge, such that the transaction without disclosure is rendered inherently unfair.").

518. The first and third circumstances clearly are not applicable here. FGIC and Putnam were not in a fiduciary relationship. Prior to the closing of the Pyxis transaction, Putnam was a true third party with respect to FGIC. Calyon was negotiating a contract with FGIC for FGIC to provide insurance with respect to an asset of Calyon's—Calyon's super-senior contract with Pyxis. Putnam was not a party to that transaction. Its sole role, for which it contracted with Calyon, was to be the collateral manager of Pyxis (an entity with which FGIC would have no relationship) if the transaction closed. Putnam had no relationship with FGIC, fiduciary or otherwise. Indeed, both Calyon and Putnam expressly disclaimed any fiduciary relationship with investors and participants, including FGIC, in pitchbooks provided to FGIC.

519. Nor is the third circumstance applicable. Putnam did not make any partial or ambiguous statements. It had no role with respect to the PCS or its allegedly misleading content. It made no statement whatsoever to FGIC about the PCS. It did not know of its existence, that it was requested by FGIC, or that it had been delivered to FGIC by Calyon. There was no statement it made that required it to make any additional disclosure. Given that it did not know of the PCS or what was in it, there would have been no way for it to know what to disclose.

520. FGIC argues nevertheless that a duty to disclose arises under the special facts doctrine. Liability could arise under the "special facts doctrine" if Putnam "possess[ed] knowledge, not readily available to [FGIC], and kn[e]w that [FGIC was] acting on the basis of mistaken knowledge." Aetna, 404 at 582.

521. FGIC's argument fails for several reasons. First, the special facts doctrine applies only when the defendant is a party to a transaction with the plaintiff; it does not provide a license, much less a duty, for a third party to intervene in a relationship between buyer and seller and provide its opinion regarding "special facts" the buyer should have before contracting with seller. See Merkin v. Berman, 1 N.Y.S.3d 21, 23 (1st Dep't 2014) ("[T]he special facts doctrine did not require petitioner to disclose the information at issue, since it applies only in 'business dealings' between parties to a prospective transaction.") (citation omitted); Tatintsian v. Pryor Cashman LLP, 2018 WL 6505401, at *6 (N.Y. Sup. Ct. Dec. 10, 2018) ("Defendants were not parties to the transaction that was entered into between Plaintiff and [a third party] . . . , and Plaintiff does not allege, as a [an] investor [in the third party], the existence of a fiduciary relationship with Defendants. Thus, the 'special facts' doctrine is inapplicable."). Thus, the special facts doctrine exists to protect against unfair dealing between parties to commercial transactions. See Newbro v. Freed, 409 F. Supp. 2d 386, 401 (S.D.N.Y. 2006), aff'd, 2007 WL 642941 (2d Cir. Feb. 27, 2007) ("The superior knowledge held by one party renders its transaction without disclosure with another party 'inherently unfair' and thus the transaction may constitute fraud.") (emphasis added); see also id. (rejecting plaintiff's assertion of duty to disclose because plaintiff "failed to show evidence that he and defendants engaged in any sort of transaction that could give rise to a fraud claim"). That is consistent with well-established federal securities law holding that "[a] party has no duty to correct statements not attributable to it." In re Bristol-Myers Squibb Sec. Litig., 312 F. Supp. 2d 549, 560 (S.D.N.Y. 2004) (citing Elkind, 635 F.2d at 162-63)).

522. Here, of course, Putnam was not a party or a prospective party to a transaction with FGIC. As FGIC's own credit application stated, FGIC's transaction with Calyon was "outside the [Pyxis] transaction." PX325. Putnam was not negotiating a transaction with FGIC and would not have a relationship with FGIC after the transaction closed. Putnam would be the collateral manager to Pyxis, which was the reference entity in a swap between Calyon and FGIC. FGIC was contracting with Calyon and would have no relationship direct or indirect with Putnam.

523. Further, the "special facts doctrine" requires the tortfeasor to "know[] that the [victim] is acting on the basis of mistaken knowledge." Aetna, 404 F.3d at 582 (quoting Aaron Ferer & Sons Ltd. v. Chase Manhattan Bank, 731 F.2d 112, 123 (2d Cir. 1984)). The evidence here does not establish that Putnam knew that FGIC was acting on the basis of mistaken knowledge. The evidence does not establish that Putnam knew that FGIC had requested detailed information for yet-to-be-acquired assets in the portfolio or that Calyon had provided such information, much less that FGIC would be relying on such information as a representation of what would actually be in the final Pyxis portfolio. Putnam had no knowledge of the request for, or the provision of, the detailed information regarding the remaining assets to be purchased for the portfolio. That was handled exclusively by the potential counterparties to the swap—FGIC and Calyon. FGIC did not share with Putnam that it would analyze or had analyzed a detailed line-by-line credit profile of a portfolio. Nor did FGIC provide Putnam with information from which Putnam could discern how FGIC would understand and apply the information in any such portfolio. There was no basis for Putnam to know how FGIC would use the information in the PCS—even if it had known that such information was provided by Calyon to FGIC.

524. Finally, the special facts doctrine "requires satisfaction of a two-prong test: that the material fact was information 'peculiarly within [the] knowledge' of [the alleged tortfeasor], and that the information was not such that could have been discovered by [the victim] through the 'exercise of ordinary intelligence.'" Jana L. v. W. 129th St. Realty Corp., 802 N.Y.S.3d 132, 135 (1st Dep't 2005) (quoting Black v. Chittenden, 503 N.E.2d 1370, 1372 (1986)); see Schumaker v. Mather, 30 N.E. 755, 757 (1892) ("[I]f . . . the other party has the means available to him of knowing, by the exercise of ordinary intelligence, the truth or the real quality of the subject of the representation, he must make use of those means, or he will not be heard to complain that he was induced to enter into the transaction by misrepresentations."). Put slightly differently, a duty to disclose under the special facts doctrine can only arise when the information known to the non-disclosing party "is not readily available to the other party." Banque Arabe et Internationale D'Investissement v. Maryland Nat. Bank, 57 F.3d 146, 155 (2d Cir. 1995). For reasons articulated below in the section addressing reasonable reliance, the truth of the information within the PCS "could have been discovered by [FGIC] through the exercise of ordinary intelligence," Jana L., 802 N.Y.S.3d at 135, and was "readily available to [FGIC]," Banque Arabe, 57 F.3d at 155.

525. Unable to satisfy the standards under Aetna, FGIC tries a different tack. It argues first, that Putnam is liable because Putnam neither asked Calyon to show it the portfolio information it had sent to FGIC, nor notified FGIC that the information in the PCS might be incorrect, and second, that Putnam was reckless in not knowing that Calyon had supplied the PCS to FGIC. FGIC thus would amend the test so as to require only a showing that Putnam "knew or must have known Calyon was providing information to FGIC and that Putnam knew or must have known that such information was materially false." Dkt. No. 238 at 7. FGIC further posits that, "[u]nder these circumstances, Putnam's failure to inquire into the accuracy of the PCS amounted to willful blindness, which, under New York law, is equivalent to actual knowledge of the PCS's inaccuracy." Id. FGIC's argument is flawed no matter how it is framed.

526. First, Putnam did not know of the existence of the PCS or that detailed information about yet-to-be-acquired assets had been provided by Calyon to FGIC. It also had no reason to suspect such information was being provided or had been provided by Calyon to FGIC. As the Court has found, Menhenett did not ask Putnam where she could obtain detailed information about target assets to be added to the portfolio. She did not mention to Putnam that she was even seeking such information. Nor did Calyon inform Putnam of the request.

527. Second, Putnam owed no duties to FGIC to ensure that the information it received was accurate. Putnam's duty was to make sure that it did not make a false or misleading statement and was not involved in one being made to FGIC. It discharged its duties.

528. Third, and independent of any knowledge of the existence of the PCS, Putnam did notify FGIC that all of the prior target projections (and not just the PCS) were outdated. Specifically, before the pricing (and well before FGIC was committed to the transaction), Putnam had Calyon provide to all participants and investors (not just to FGIC) updated target information in the same form as that which had been previously provided to FGIC: a term sheet. That term sheet clearly disclosed, contrary to what had been reflected in the PCS, that as of the beginning of September 2006, the Pyxis portfolio was anticipated to have only approximately 2% prime RMBS. Thus, even assuming that Putnam had a duty to correct the misleading information supplied to FGIC by Calyon, Putnam satisfied that duty.

529. That fact is fatal to FGIC's claim. Although Menhenett claimed (incredibly) that she did not open the seven-page document, a defendant's duty under the special facts doctrine would be discharged when it advises that the prior information upon which the plaintiff might otherwise have relied is incorrect. The "special facts doctrine" punishes "concealment"—when one party does not disclose information as to which it has superior knowledge to another party. See Miele v. Am. Tobacco Co., 770 N.Y.S.2d 386, 391 (2d Dep't 2003) ("New York recognizes a cause of action to recover damages for fraud based on concealment, where the party to be charged has superior knowledge or means of knowledge, such that the transaction without disclosure is rendered inherently unfair."). Here, Putnam did not conceal anything. It is not incumbent on a defendant to make sure that a plaintiff has read the information it should read before entering a massive transaction. And a plaintiff's assertion that it ignored the defendant's subsequent disclosure cannot resurrect its cause of action.

530. FGIC's argument that liability arises for Putnam because Putnam was willfully ignorant that FGIC would request, and Calyon would provide, a list of specific yet-to-be-acquired assets is one step further removed. From the information in Putnam's possession, there was no reason for it to know that FGIC would be requesting such particularized target information. Putnam knew or had reason to know that FGIC had requested detailed information about ramped assets from Calyon and that Calyon had responded. That information would have comprised the bulk of the portfolio that Putnam was to assemble. FGIC also knew— and Putnam knew that FGIC knew—the portfolio constraints for the transaction and the duties that Putnam would have under the indenture and the collateral management agreement. The published documents informed FGIC (and all Pyxis investors) that Putnam was required to invest 80% of the portfolio in subprime and midprime securities and 90% of the portfolio in RMBS or ABS CDOs and that Putnam did not need to invest in prime RMBS or seasoned RMBS at all. That information also informed FGIC that Putnam would have the responsibility at all times to invest the remainder of the portfolio in the assets that would bring the most value to Pyxis and insure that Pyxis was able to satisfy its obligations to the noteholders and that Putnam could not and would not make decisions solely in the interest of the super-senior tranche. Those decisions would be made over a period of six months from August 2006 until the ramping was required to be complete in February 2007, and they would be made in rapidly fluctuating markets. FGIC did not provide evidence suggesting Putnam had any reason to suspect that FGIC, in August 2006, would ask for a detailed list of the specific assets that Putnam or Calyon then expected would be added to the portfolio when any representation in August 2006 as to what would be purchased over the ensuing half-year would necessarily be tentative, preliminary, and subject to change.

531. The cases FGIC relies upon do not support it. During summation, FGIC's counsel stated that in Koch, 14 F. Supp. 3d 247, the court held that the defendant with superior knowledge was obligated to disclose certain authenticity defects in a wine purchase "despite the fact that the court found the plaintiff could have actually inspected each bottle for fraud itself," and specifically despite the fact that the plaintiff "could have found some signs of fraud." Tr. 2116-17. Koch involved a dispute between buyer and seller in which the buyer alleged that the seller had superior knowledge of the facts. For that reason alone, it is inapposite. Moreover, the case says exactly the opposite of what FGIC represents. "Throughout the trial," Judge Oetken explains in that opinion, "[the defendant's] counsel emphasized that [the plaintiff] had the opportunity to inspect the bottles at issue, and, had he done so, he would have seen the indicia of inauthenticity that served as readily apparent indicators of their counterfeit status." Koch, 14 F. Supp. 3d at 258. "The jury was free to credit that argument and chose not to do so—a determination that was within the province of the jury." Id. (emphasis added). "The jury chose to agree with [the plaintiff's] position that '[n]o amount of inspection would have revealed what [the defendant] knew' and the Court concludes that that choice was sufficiently supported by the evidence at trial." Id. (emphasis added) (internal citation omitted); see id. at 259-60 ("[I]t was reasonable for the jury to conclude that, in light of all the circumstances, and despite [the plaintiff's] sophistication and his right to inspect the bottles, it was unreasonably difficult or impossible for him to have discovered what [the defendant] knew.") (emphasis added). Accordingly, Koch stands for precisely the converse of FGIC's assertion. Fraud liability obtains when "[n]o amount of inspection" would reveal what a defendant knows, or at the very least when it would be "unreasonably difficult" to acquire that knowledge. Id. No such circumstance was present here.

532. FGIC's citation to Heineman v. S & S Mach. Corp., 750 F. Supp. 1179 (E.D.N.Y. 1990) is similarly perplexing. Heineman once again involved a dispute between parties to a transaction; it is distinguishable for that reason alone. Moreover, Heineman confirms that the "special facts" doctrine does not apply when the "information not disclosed could have been easily obtained through 'duly diligent inquiry,'" id. at 1187 (quoting Klamberg v. Roth, 473 F. Supp. 544, 552 (S.D.N.Y. 1979)), or "by the exercise of ordinary intelligence," id. (quoting Kurz v. Nicolo, 510 N.Y.S.2d 390, 391 (4th Dep't 1986)). In Heineman, and unlike in this case, the court determined that "independent verification" would have been "difficult, if not impossible." Id.; see id. at 1186 (citing "[c]ase law holding that fraud cannot be based on a failure to make simple inquiries").

533. FGIC failed to show that Putnam was responsible for the PCS.

C. Actual Reliance or Transaction Causation

534. "In a fraud action, the plaintiff must show a belief in the truth of the representation and a change of position in reliance on that belief." Hecht v. Components Int'l, Inc., 867 N.Y.S.2d 889, 895 (Sup. Ct. 2008) (citing CBS Inc. v. Ziff-Davis Pub. Co., 553 N.E.2d 997, 1000 (1990)). Put another way, to establish "transaction causation," the plaintiff must show that the "defendant's misrepresentation induced plaintiff to engage in the transaction in question[.]" Laub v. Faessel, 745 N.Y.S.2d 534, 536 (1st Dep't 2002). Transaction causation requires proof "that 'but for the claimed misrepresentations or omissions, the plaintiff would not have entered into the detrimental securities transaction.'" Basis PAC-Rim Opportunity Fund (Master) v. TCW Asset Mgmt. Co., 48 N.Y.S.3d 654, 656 (1st Dep't 2017) (quoting Lentell v. Merrill Lynch & Co., 396 F.3d 161, 172 (2d Cir. 2005)).

535. At trial, FGIC espoused alternative theories for how it relied on the PCS. It argued first that it relied on the MBS loss estimate that the PCS generated. In other words, it urged that had Calyon or Putnam disclosed what would be in the final Pyxis portfolio as opposed to providing the PCS itself, FGIC would not have entered the transaction with Calyon. FGIC also argued that it relied on the asset allocation in the PCS, and specifically, on the quantity of prime RMBS and seasoned RMBS reflected in the PCS. Whichever theory FGIC chooses, it has not proven that the PCS was a but-for cause of FGIC's decision to insure Pyxis at the 40% attachment point.

536. The evidence does not establish that FGIC relied on the 41.15% MBS loss estimate generated by FGIC's analysis of the PCS portfolio in entering the transaction and that it would not have entered the transaction at the 40% attachment point but for that loss estimate. On August 7, 2006, even before FGIC received the PCS (much less the analysis generated from it), FGIC's employees came to the conclusion that it would attach at the 40% attachment point and expressed that to Calyon. That conclusion was based on FGIC's understanding of the ramped portfolio and the portfolio constraints in the transaction, including the target WARF, as well as of the other structural features in the transaction, including the absence of triggers that would provide additional protection to the super-senior tranche. After FGIC received the PCS, but before the LCR analysis of that target portfolio was completed, FGIC conducted a LCR analysis of two other hypothetical portfolios—one that was comprised entirely of subprime assets and the other that assumed a pool comprised of assets from the ramped portfolio and the July 18, 2006 portfolio spreadsheet. They generated loss estimates for the MBS portion of the CDO of 39.99% and 49.67% respectively. Notwithstanding those results, Menhenett advocated and believed that FGIC should attach at the 35% attachment point, noting that the LCR analysis was just a "gutt [sic] check." PX275. After FGIC received the results of the LCR analysis of the PCS showing the 41.15% loss estimate, Menhenett still concluded that FGIC could and should attach at the 35% attachment point, stating to her superiors that such a transaction "works fr risk perspective." PX290. The analysis showed that FGIC would break even at a 35% attachment point even if the CDO experienced losses of 42.5%. The contemporaneous evidence shows that Menhenett did not believe it was important that the attachment point be above the MBS loss estimate. She stated that the loss check (which generated the 41.15%) had not been "used . . . as a hard rule in the past" due to the weaknesses of the model. Id. The evidence also does not show that Finkel believed that the difference between the proposed 35% attachment point and the 41.15% MBS loss estimate would be a show-stopper either. Her reaction to Menhenett's recommendation of the 35% attachment point was simply that the transaction was "[n]ot a slam dunk," and that she wanted more information. Id. When FGIC rejected the proposed transaction at a 28% attachment point and told Calyon that "35% is not a clear easy answer, 40% would be easier," id., and then that FGIC could not attach below 40%, it was not because such figure was above the MBS loss estimate. It was based on the entirety of the transaction, including, importantly, the portfolio constraints and the structural features.

537. That conclusion is further confirmed by the credit application that the SCC considered before it authorized the transaction with Calyon at the 40% attachment point. The credit application reflected the MBS loss estimate and the 41.15% figure it generated as a datum in Appendix 8, which simply noted that because FGIC would have 40% of subordination, the other 1.15% would have to come from the excess spread of the CDO. Finkel, who testified for FGIC to the extraordinary significance of the LCR analysis, spent seven minutes staring at it during her deposition only to confess that she did not recognize it. The bulk of FGIC's analysis—and that reflected in the body of the credit application—was the sensitivity and cash flow analysis showing that, based on the portfolio constraints and the structural features of the CDO, FGIC could attach at a 35% attachment point without incurring losses and that the 40% attachment point was thus safely within the range where FGIC would not expect to be incurring losses.

538. Moreover, although not critical to the Court's legal conclusions, FGIC failed to offer evidence that, had it been provided a portfolio in August 2006 that replicated the final Pyxis portfolio as completed in February 2007, the results of the MBS loss estimate would have been any higher than those reflected in the credit application. Remember that FGIC received an MBS loss estimate of two hypothetical portfolios on August 11, 2006—one of which included prime RMBS and the other which did not. The portfolio with only subprime performed better than that with prime RMBS—the result which would have been anticipated by Putnam based on its statements to FGIC at the August 3, 2006 and August 7, 2006 meetings. Subprime RMBS was not inherently more risky than prime RMBS or more likely to generate losses. FGIC analyzed the assets that had not been purchased as of the time of the credit application as dummy assets. It understood, and built into its analysis, that Putnam might not purchase those assets in the ramping process. There is no evidence as to how FGIC would have treated the assets in the PCS if, as it now claims, it believed them to be the actual assets that Putnam would purchase or what loss estimate they would have generated.

539. Finally, as will be further elaborated below, FGIC's lack of reliance on the loss estimates generated by the PCS is confirmed by the evidence of its conduct after the transaction closed. By February 2007, FGIC was on actual notice of the true composition of the final Pyxis portfolio. FGIC prepared detailed analyses of that portfolio beginning in July 2007. It never raised issues about the discrepancies between that portfolio and the PCS until after it paid the commutation amount to Calyon and was looking for someone and some reason to sue.

540. The evidence also is clear that FGIC did not rely on the asset allocation in the PCS in deciding to enter the transaction at the 40% attachment point. The contemporaneous evidence, both before and after FGIC signed the agreement with Calyon, establishes that FGIC well knew that Putnam was required to invest 80% of the Pyxis portfolio in midprime and subprime RMBS (and that consequently the portfolio would be heavily concentrated in those securities). FGIC knew that Putnam's discretion to make investment decisions in the future would be severely constrained both by the portfolio constraints and by the obligation to make decisions in the best interest of Pyxis. FGIC did not rely on the portfolio having any particular composition so long as it complied with the portfolio constraints, and it did not rely on the particular composition set forth in the PCS. As Menhenett acknowledged, FGIC "wanted to give the good managers room to be able to do their job" because "the market definitely changes." Tr. 611-12.

541. As will be detailed further below in the section on reasonable reliance, FGIC was warned and knew that any predictions about what would be in the final Pyxis portfolio and its asset allocation was necessarily tentative and should not be relied upon. Markets change. What was in the best interest of Pyxis on day one might not be in the best interest of Pyxis on day two.

542. Before FGIC agreed to participate in the Pyxis transaction, the evidence shows that it was largely indifferent to the asset composition in the PCS and, in particular, to the amount of prime RMBS and seasoned RMBS reflected in it. FGIC had two meetings with Putnam to discuss Putnam's asset selection process: in person on August 3, 2006 and by telephone on August 7, 2006. Putnam was transparent about its preference for subprime over prime RMBS securities and it explained why: prime RMBS had less subordination, were relatively overpriced, and were more difficult to source. FGIC did not question that thesis or voice objection to it. Putnam also disclosed that Pyxis would be exposed to sector risk from a concentrated RMBS subprime mezzanine portfolio. Putnam expressly said that on August 3, 2006. It stated that the asset constraints would permit it to invest in non-RMBS assets such as CMBS and permitted ABS classes, not prime RMBS. FGIC did not ask about increasing the amount of prime RMBS or any other particular type of asset. On the August 7, 2006 call, when Putnam described its collateral selection process and methodology, FGIC never asked Van Tassel—the person responsible for selecting RMBS—for the details of what Putnam intended to purchase for Pyxis, let alone for any representation concerning the amount of prime RMBS.

543. FGIC received the PCS on August 8, 2006. The asset allocation in it was reflected in a pie chart in Menhenett's August 10, 2006 memorandum summarizing the transaction for her bosses. The memorandum made no comment, positive or negative, about that asset mix. When, in the memorandum, Menhenett summarized the risk mitigants for the amount of midprime and subprime RMBS in the portfolio, she focused on the ability of the manager to invest in non-RMBS assets (assets other than prime RMBS) and the performance triggers that FGIC was negotiating. She did not say anything about prime RMBS or seasoned RMBS as a risk mitigant. In fact, she stated, without any red flag or warning, that Putnam "like[d] Subprime over Prime." PX829. FGIC was simply not focused on the amount of prime RMBS or seasoned RMBS in the Pyxis portfolio.

544. Nor, in any of the internal emails either preceding FGIC's rejection of the proposed transaction at the 28% attachment point or its acceptance of a transaction at a 40% attachment point, is there discussion of the amount of prime RMBS or seasoned RMBS in the proposed portfolio. That was not a consideration in FGIC's communication to Calyon that it was prepared to go forward with a transaction at a 40% attachment point.

545. Indeed, there are only two mentions of the amount of prime RMBS in the credit application—a notation that additional bonds would be added to the portfolio and those would be "expected to include Prime bonds rated A2 and above" and a bar chart reflecting the asset mix as represented in the PCS. PX325. There is no evidence that any member of the SCC gave those mentions any notice in approving the transaction. There is also no evidence that, but for the asset allocation in the PCS, FGIC would not have entered the transaction.

546. Finally, FGIC's silence in 2007 even after it was informed repeatedly of the asset allocation in the Pyxis portfolio is telling evidence that such asset allocation was not important to FGIC just months earlier. FGIC, like any sophisticated insurer, dedicated personnel to monitoring the transactions it underwrote. Its internal policies required it to do so. For that reason, shortly before pricing, Menhenett told Bell that she wanted to coordinate with him and FGIC's surveillance group to obtain access to website updates on Pyxis and regular trustee reports. FGIC did not present any witness who could testify to receipt of the first Pyxis trustee report in February 2007. But FGIC's institutional behavior, and the words of its employees, in late 2006 and early 2007, reflect the unimportance to FGIC of the amount of prime RMBS or seasoned RMBS. FGIC was repeatedly put on notice of the small amounts of prime RMBS and seasoned RMBS in the Pyxis portfolio and the discrepancies between the final Pyxis portfolio and that reflected in the PCS—in late 2006, in February 2007, in April 2007, in May 2007, in July 2007, and in August 2007. It did not ask that more prime or seasoned RMBS be added to the transaction or raise or identify any issues regarding the differences between what was in the PCS and the final Pyxis portfolio. That asset allocation simply had not been important to FGIC when it approved the transaction. Like the dog that did not bark, FGIC's silence about the asset allocation before and after the approval of the FGIC transaction is a powerful clue that FGIC did not rely upon the percentage of prime RMBS or seasoned RMBS in the PCS, but upon other factors, in deciding to provide insurance to Calyon in connection with the Pyxis transaction. See Arthur Conan Doyle, Silver Blaze, in The Complete Sherlock Holmes 335 (1930); In re Chateaugay Corp., 89 F.3d 942, 954 n.1 (2d Cir. 1996) ("[T]he failure of the dog to bark—its silence when it would ordinarily be heard—was a clue the legendary detective considered in solving the crime.").

547. FGIC thus has failed to show that it relied on the PCS as a but-for cause in connection with entering the Pyxis transaction.

D. Reasonable Reliance

548. The Court also concludes that, even if FGIC had been able to show that it relied on the PCS, such reliance would not have been reasonable. See Wynn, 273 F.3d at 156. "In assessing the reasonableness of a plaintiff's alleged reliance, [the Court] considers the entire context of the transaction, including factors such as its complexity and magnitude, the sophistication of the parties, and the content of any agreements between them." Meisel v. Grunberg, 651 F. Supp. 2d 98, 118 (S.D.N.Y. 2009) (quoting Glidepath Holding B.V. v. Spherion Corp., 590 F. Supp. 2d 435, 459 (S.D.N.Y. 2007)). "In applying the test of reasonableness, the court must determine whether the plaintiff availed itself of the means available to it given its size and sophistication." Ambac Assur. Corp. v. Countrywide Home Loans, Inc., 2016 WL 7374210, at *12 (N.Y. Sup. Ct. Dec. 19, 2016) (internal quotation marks and citation omitted). "Factors to be considered in making this determination may include the particular parties' prior course of dealing, the roles of other participants in the transaction, the time and expense of the investigation, the feasibility of practical[ity] of the plaintiff's performance of a particular type of investigation before entering into the transaction, the foreseeable risk to the plaintiff of failing to undertake the investigation, and responsible industry custom." Id. (internal quotation marks and citation omitted).

549. Before reasonably relying on a representation regarding information material to participation in a significant transaction, a sophisticated party "has a duty to conduct an independent appraisal of the risk it is assuming and a duty to investigate the nature of its business transactions." Granite Partners, L.P. v. Bear, Stearns & Co. Inc., 58 F. Supp. 2d 228, 259 (S.D.N.Y. 1999); see id. (noting that they "cannot be heard to complain when they fail to make diligent inquiries"). "Where sophisticated businessmen engaged in major transactions enjoy access to critical information but fail to take advantage of that access, New York courts are particularly disinclined to entertain claims of justifiable reliance." Crigger v. Fahnestock & Co., Inc., 443 F.3d 230, 235 (2d Cir. 2006).

550. "Where . . . a party has been put on notice of the existence of material facts which have not been documented and he nevertheless proceeds with a transaction without securing the available documentation or inserting appropriate language in the agreement for his protection, he may truly be said to have willingly assumed the business risk that the facts may not be as represented." Lazard Freres & Co. v. Protective Life Ins. Co., 108 F.3d 1531, 1543 (2d Cir. 1997) (quoting Rodas v. Manitaras, 552 N.Y.S.2d 618, 620 (1st Dep't 1990)). "Succinctly put, a party will not be heard to complain that he has been defrauded when it is his own evident lack of due care which is responsible for his predicament." Id. (quoting Rodas, 552 N.Y.S.2d at 620).

551. FGIC was a sophisticated market participant with particular expertise in structured financial products. It was one of the largest players in the global financial guaranty insurance industry and dedicated a substantial portion of its business to insuring structured financial products. Its teams included underwriters with prestigious academic and industry credentials. FGIC's teams were both specialized and collaborative; analysts with honed expertise in CDOs and MBS contributed jointly to FGIC's study of Pyxis. See Lazard, 108 F.3d at 1543 (noting that defendant was a "substantial and sophisticated" player in the bank debt market); Emergent Capital Inv. Mgmt., LLC v. Stonepath Grp., Inc., 343 F.3d 189, 196 (2d Cir. 2003) (highlighting plaintiff's self-professed "knowledge and experience in financial and business matters" and its ability to "readily evaluate the risks of the transaction"); Marini v. Adamo, 995 F. Supp. 2d 155, 177 (E.D.N.Y. 2014), aff'd, 644 F. App'x 33 (2d Cir. 2016) ("[T]he relevant inquiry is not whether a plaintiff has money or has made investments generally, but whether a plaintiff is sophisticated regarding this specific investment.").

552. The transaction that FGIC was considering was large and complex. See Emergent Capital, 343 F.3d at 195 ("In assessing the reasonableness of a plaintiff's alleged reliance, we consider the entire context of the transaction, including factors such as its complexity and magnitude[.]"). The exposure at stake was massive. FGIC was being asked to extend $900 million of insurance on a structured financial product with a maturity of 40 years. The structure of the deal was intricate; it created relationships between several different types of investors, all of which had different incentives and input into the transaction. Not only was the structure of Pyxis complex, but many of the underlying assets in the Pyxis portfolio were themselves complex financial vehicles, leading to several additional layers of complexity. See, e.g., Deutsche Bank Tr. Co. v. Am. Gen. Life Ins. Co., 2016 WL 5719783, at *9 (S.D.N.Y. Sept. 30, 2016), aff'd sub nom., 707 F. App'x 741 (2d Cir. 2017) ("CDOs are complex . . . transactions."); Assured Guar. Mun. Corp. v. DLJ Mortg. Capital, Inc., 2014 WL 3288335, at *5 (N.Y. Sup. Ct. July 3, 2014) (noting the "complexities of RMBS").

553. The prior course of dealing among FGIC, Putnam, and Calyon, and the nature of the risk that FGIC was being asked to insure, put FGIC on notice that any communications it received in early August 2006 about target assets for a portfolio that was not to be complete until February 2007 would be necessarily incomplete. As early as July 14, 2006, Putnam and Calyon warned FGIC (and other participants and investors) that actual events would "differ from those assumed, perhaps significantly," that among the "important factors" that "could cause actual results or performance to differ materially from those expressed or implied in any forward-looking statements" were "the actual composition of the collateral," and that none of Calyon or Putnam expected to update the information contained in the [pitchbook] "except by means of the Offering Memorandum." PX715; PX212.

554. When the PCS was delivered by Calyon to FGIC, it too came with a disclaimer warning "[n]one of Calyon or any of its affiliates make any representation or warranty, express or implied, as to the accuracy of the information contained herein and nothing contained herein shall be relied upon as a promise or representation whether as to past or future performance. No representations are made as to the accuracy of any estimate or projections . . . ." PX253. It added that the Offering Memorandum—which would be forthcoming if Pyxis issued securities—"would supersede this information [in the PCS] in its entirety" and that "[a]ny decision to invest in the securities described herein should be made after reviewing the Offering Memorandum, conducting such investigations as the investor deems necessary and consulting the investors own legal, accounting, and tax advisors . . . ." Id. The body of Anand's email tracked the warnings attached to the spreadsheet. His email first identified the "assets highlighted in a peach colour" and then warned, with respect to "this pool": "[P]lease be aware there might be some changes[.]" Id.; see DynCorp v. GTE Corp., 215 F. Supp. 2d 308, 319 (S.D.N.Y. 2002) (stating that "particularized disclaimers" can "make it impossible" for a plaintiff "to prove . . . that it reasonably relied on . . . representations that it alleges were made to induce it to enter into [agreements]").

555. FGIC understood that the process of assembling a portfolio was dynamic and iterative and, therefore, subject to change. It was foreseeable to FGIC that the asset allocation might change. Ambac Assur. Corp., 2016 WL 8254810, at *12 (one factor to consider in assessing reasonable reliance is "the foreseeable risk to the plaintiff of failing to undertake the investigation"). On June 21, 2006, Calyon sent to FGIC a draft pitchbook projecting that the Pyxis portfolio would contain 16% RMBS A assets. On July 7, 2006, an updated draft projected 6% RMBS A assets. Just seven days later, FGIC received a new, and final, pitchbook that reflected an indicative asset allocation of 4% prime RMBS. One day before receiving the PCS, FGIC received an "intended portfolio" matrix again indicating that the amount of prime RMBS would be 4%. FGIC had no reason to believe that, if the intended or indicative allocation had fluctuated so much between July and August 2006, it would not fluctuate equally as much or more in the days and months between August 2006 and February 2007.

556. FGIC also knew and understood from the documents it had been sent the responsibilities Putnam had prior to the closing and those that Putnam would have after the closing, including that Putnam would need not acquire more than 85% of the portfolio prior to closing, that it would have to devote 80% of the portfolio to midprime and subprime RMBS, and that, thereafter, and with the remaining part of the portfolio, it had a wide range of options including CMBS, student loan debt, credit card securities, REIT debt securities, and small business loan securities (as well as more subprime and midprime RMBS and ABS CDOs) among which to choose. FGIC understood that Putnam would have to make the best decisions for Pyxis as a whole based on what was available in the marketplace and consistent with the constraints for the collateral pool as a whole, regardless of what any particular investor at any particular point in the capital structure might have wanted or expected based on its particular risk profile.

557. The role of other participants in the transaction also imposed a duty on FGIC to inquire. The participants and investors in Pyxis included numerous other substantial financial institutions who were committing substantial sums of capital. Each of those participants and investors had their own motivations, strategies, and requirements for the transaction based on where they stood in the capital structure and on their own risk appetites. The investments that Putnam would be able to make with the portfolio—expressed in the portfolio constraints—were a negotiated amalgam of the needs of the investors at those different points. The constraints, and the requirements of the investors, changed virtually up until the day of pricing. For example, the evidence showed that IKB, a large debt investor, had its own requests with respect to the portfolio constraints that Calyon had to try to satisfy. The fact that other investors in Pyxis all would make requests with respect to the contractual portfolio constraints, and FGIC's knowledge that the other investors would make such requests in order to participate and in order for Pyxis to be a success, all make it unreasonable for FGIC to rely on a private, subject-to-change representation of target assets in August 2006 as a definitive indication of what Putnam actually would purchase over the ensuing months. FGIC knew, or should have known, that the other sophisticated investors would all have their own requirements and that up until and after the closing Putnam would have to operate within the publicly-disclosed and negotiated portfolio constraints in making investment decisions rather than hew blindly to what any individual investor might have expected or wanted months earlier in the negotiation of the transaction.

558. The time and expense it would have taken for FGIC to conduct an investigation of what Putnam would have purchased for the portfolio and its feasibility was not great. In the first instance, FGIC knew that the PCS was represented as coming solely from Calyon; that is what the transmittal indicated and the earlier pitchbook had put FGIC on notice that it could not expect an update from Putnam until the Offering Memorandum. It also knew that the asset allocations within the PCS, and in particular the allocations to prime RMBS, were inconsistent with what Putnam was telling FGIC simultaneously—that it did not like prime RMBS. That alone should have put FGIC on notice that the PCS might not have represented Putnam's views. If FGIC wanted to know whether the PCS reflected Putnam's views, it would have been easy enough to call up and ask. FGIC had Putnam's contact information and the two had been in touch on other issues, including those such as triggers that were clearly outside of Putnam's area of responsibility.

559. Moreover, and importantly, FGIC had warning notices that anything it was told in August 2006 about the Pyxis portfolio might not be true in September 2006 or October 2006. If FGIC wanted to know whether the PCS continued to represent Putnam's views, it was easy enough for it to find out. It could have called. The evidence at trial was that other investors in CDOs would, from time-to-time, call and ask what the portfolio manager's current views were with respect to particular asset classes. FGIC itself did that with Putnam in August 2006 when it inquired about the bespoke CDO bucket. It also asked Putnam's views about prime RMBS and was told they were not favorable. There was nothing to prevent it from calling at any subsequent time and asking questions about the amount of prime RMBS or whether Putnam's views had changed. FGIC complained at trial that Putnam did not call it and say that the asset allocation had changed. Even if that were true, the obligation of reasonable reliance does not put a duty on the defendant to volunteer information that would dispel reliance. FGIC "was under a . . . duty to protect itself from misrepresentation" and "easily could have done so" by asking for an update on Putnam's investment views prior to pricing or closing. Lazard, 108 F.3d at 1543.

560. In addition, if the asset allocation in the PCS were important to FGIC, FGIC had the means to protect itself. See, e.g., Emergent Capital, 343 F.3d at 196 ("[H]aving been told that NETV's largest investment was its $14 million purchase of an equity interest in Brightstreet, Emergent should have protected itself by insisting that this representation be included in the stock purchase agreement."); Lazard, 108 F.3d at 1533 (no reasonable reliance where sophisticated buyer should have and easily could have protected itself); Rodas, 552 N.Y.S.2d at 620 (same). FGIC understood the difference between a binding contract and an informal communication (from a junior analyst, with the caveat that his supervisor had not approved it, and with the warning that it was subject to change). The binding contracts that structured the Pyxis transaction were made available to all investors and participants in Pyxis and contained the terms by which Putnam would have to abide in selecting assets for the portfolio in the future. All investors and participants would be entitled to rely upon them in conducting their own risk analyses and to assume Putnam would not depart from them. Any "target" or "indicative" assets were just that: indicative assets that a potential deal participant could use to run analyses but that would not necessarily be reflective of what was in the final portfolio.

561. When investors wanted to limit the discretion of Putnam and ensure that it purchased a certain amount of a particular asset class (or did not purchase a particular asset class), they knew how to do so—by asking for changes in the contractual portfolio constraints. See Ambac Assur. Corp., 2016 WL 8254810, at *12 (industry custom is a relevant factor in considering reasonable reliance). The equity investors did that at the beginning of the transaction. The trial evidence established that IKB did that. The evidence also established that FGIC did so, asking for changes to some of the portfolio constraints but not those with respect to prime RMBS or seasoned RMBS. Had FGIC wanted to ensure that Putnam buy a certain amount of prime RMBS or seasoned RMBS, it could have negotiated for such an obligation. It deliberately eschewed doing so, not wanting to tie the hands of the collateral manager. Having made that choice, it was required to live by it. Its failure to insist that Putnam invest a particular percentage of the portfolio in the assets it now (in litigation) says it wanted Putnam to invest in alone precludes it from claiming that it reasonably relied on supposed oral representations that Putnam would invest the portfolio in that percentage of prime RMBS and seasoned RMBS assets. See Emergent Capital, 343 F.3d at 196.

562. Finally, FGIC failed to establish reasonable reliance because the evidence at trial established that, prior to pricing and well before closing, FGIC was sent an updated term sheet with a target portfolio indicating that—whatever Calyon might have thought in August 2006 about a target portfolio—by September 2006, Putnam intended to acquire only 2% of prime RMBS in the Pyxis portfolio. That information was contained in a seven-page attachment to an email sent to Menhenett on September 8, 2006. Menhenett testified at trial that she did not open the attachment and did not see the target portfolio because she did not expect such a document to be attached to a term sheet. The Court finds Menhenett's testimony to be incredible. It is simply not believable that Menhenett, as the underwriter for a transaction that committed FGIC to nearly a billion dollars of insurance, when presented with an email attachment with an updated term sheet that would have addressed open items that she identified with the prior version of the term sheet, would not have opened the email attachment. And if she had opened the email attachment she would have seen the new target portfolio. The Court does not credit her testimony. Even if the Court did credit that testimony, however, Menhenett's evident lack of care with respect to a target portfolio would preclude FGIC's claim of reasonable reliance.

E. Loss Causation

563. In order to make out a claim for fraud under New York law, a plaintiff must prove "both that the 'defendant's misrepresentation induced plaintiff to engage in the transaction in question (transaction causation) and that the misrepresentation directly caused the loss about which plaintiff complains (loss causation).'" Putnam III, 783 F.3d at 402 (quoting Laub, 745 N.Y.S.2d at 536). Loss causation "is the causal link between the alleged misconduct and the economic harm ultimately suffered by the plaintiff." Id. (quoting Lentell, 396 F.3d at 172). FGIC must prove that the "subject of the fraudulent statement or omission was the cause of the actual loss suffered." Id. (quoting Lentell, 396 F.3d at 173). If a false representation caused the plaintiff to engage in the transaction, but the plaintiff would have suffered the identical loss regardless of whether the representation was accurate or not, the plaintiff has not established loss causation. There is no causal link between the alleged misconduct and the economic harm ultimately suffered by the plaintiff. The plaintiff must show that "it was [the] misrepresentations, rather than market forces [or some other factor], that caused the investment losses." Basis PAC-Rim Opportunity Fund, 48 N.Y.S.3d at 657. If "the lie" did not "increase[] the chances of the actual damage," then there was no loss causation. Loreley, 797 F.3d at 183-84.

In its decision sustaining FGIC's loss causation allegations against Putnam's argument that they were legally insufficient, the Second Circuit held that FGIC had "plausibly allege[d] that Putnam's alleged misrepresentations and omissions caused at least some of the economic harm suffered." Putnam III, 783 F.3d at 403. The Second Circuit relied on FGIC's allegations that "[h]ad Putnam selected the Pyxis collateral itself, as it represented it would do, and had it not acquiesced in Magnetar's control of collateral selection, Pyxis would not have defaulted as quickly as it did, and may well not have defaulted at all" and that "[m]any of the assets selected for the Pyxis portfolio by Magnetar, on their face, were more liable to default than the assets Putnam would have selected had it acted independently," including that $145 million of prime RMBS was replaced by $145 million of subprime RMBS. Id. FGIC offered no evidence to support any of those contentions at trial.

564. FGIC failed to prove loss causation. The loss FGIC complains of is the unexpected decrease in the value of its credit swap agreement with Calyon as the securities in the Pyxis portfolio started to default beginning in the summer of 2007. The exposure through the floating rate payments it would owe at the end of the swap would greatly exceed the fixed rate premiums it earned. It ended up with a losing bet. FGIC measures the decrease in value by the commutation payment it made to Calyon, which relieved FGIC of its obligations under the Pyxis swap and two other agreements with Calyon that also turned unprofitable.

565. FGIC failed to prove a causal link, however, between those losses and the misconduct of which it complained—the alleged misstatement in the PCS and the failure of Putnam to invest the Pyxis portfolio as represented in the PCS. In particular, FGIC failed to prove, or indeed offer any evidence, that the Pyxis portfolio performed worse as a result of Putnam's investments than it would have performed had Putnam invested in the securities identified in the PCS (assuming it would have been able to do so). The defaults in the Pyxis portfolio, and the decline in value of the swap as a result, coincided with a market-wide financial crisis which impacted all CDOs backed by subprime RMBS issued during the same period. The slides prepared for FGIC's board of directors in August 2007 reflected that four CDO transactions—not just Pyxis but also three others—had been placed on heightened surveillance as a result of the credit crisis beginning in July 2007. By October 2010, every CDO insured by FGIC from January 2006 to July 2007 had been downgraded from AAA to below investment grade and by March 2012 nearly every one of these CDOs had defaulted or suffered substantial losses.

566. In the face of such evidence, it was incumbent on FGIC to prove "that its loss was caused by the alleged misstatements as opposed to intervening events." Basis PAC-Rim Opportunity Fund, 48 N.Y.S.3d at 656-57 (quoting Lentell, 396 F.3d at 174). FGIC failed to do so.

567. In the absence of evidence that FGIC performed worse as a result of Putnam's investment decisions, and in the face of the possibility that it might have performed worse had Putnam invested as represented in the PCS, FGIC pursued an alternative theory of loss causation at trial. It claimed that had the target portfolio not been misrepresented to it and had it known the investment decisions Putnam would make, it would have engaged in an alternative transaction. It would have taken a lower fee to attach at a higher attachment point. And had it attached at a higher attachment point, it would have extended less insurance to Calyon and suffered fewer losses.

568. There is a substantial question whether, under New York law, a defrauded plaintiff can obtain damages based on the value of a foregone opportunity. Under New York law, the measure of damages for claims based on fraud (and negligent misrepresentation) is the "out of pocket" rule: loss is computed by ascertaining the "difference between the value of the bargain which a plaintiff was induced by fraud to make and the amount of value of the consideration exacted as the price of the bargain." Sager v. Friedman, 1 N.E.2d 971, 974 (1936). Fraud damages are intended to compensate a plaintiff for what was lost because of the fraud, not to compensate for what might have been gained. See Lama Holding Co. v. Smith Barney Inc., 668 N.E.2d 1370, 1373 (1996). Applying those principles here, the loss (if any) would be measured by the difference in value between a CDO that invested in the final Pyxis portfolio and the CDO that FGIC contends it thought it was paying for that invested in the securities referenced in the PCS. But there is no evidence that there was a difference in value that disfavored FGIC.

The sole case FGIC cites for the proposition that courts have permitted recovery based on loss of an alternative transaction, In re MarketXT Holdings Corp., 2006 WL 2864963 (Bankr. S.D.N.Y. Sept 29, 2006), merely held that the plaintiff had stated a claim for loss causation. That case is also easily distinguishable on its facts. The plaintiff there was induced by a fraud to forego a job the plaintiff actually held. It thus did not involve a lost opportunity at all.

569. Thus, for example, in Lama Holding Co., the New York Court of Appeals squarely held that in a fraud case, "[d]amages are to be calculated to compensate plaintiffs for what they lost because of the fraud, not to compensate them for what they might have gained." 668 N.E.2d at 1373. It added, "Under the out-of-pocket rule, there can be no recovery of profits which would have been realized in the absence of fraud." Id. at 1374. There, the plaintiff, the largest single owner of Smith Barney shares, alleged that it was fraudulently induced to vote its shares in favor of Smith Barney's merger with Primerica, which resulted in sizable unanticipated tax consequences for the plaintiff. The plaintiff argued that, but for the misrepresentation, it would have withheld its consent and Primerica would have been forced to engage in an alternative transaction with plaintiff that did not have the same adverse tax consequences. The court rejected the plaintiff's argument, holding that plaintiff suffered no recoverable losses because it received "more than twice the fair market value for its shares." Id. As to the claim that it was deprived of the opportunity to gain even more profits through the alleged alternative contractual bargain, the court held that "the loss of the bargain was 'undeterminable and speculative.'" Id. (citation omitted). "Primerica's refusal to go forward with [plaintiff's] proposal [of the alternative transaction] was [not alleged to be] because of defendant's fraud rather than an independent business decision made by Primerica." Id.

570. FGIC relies on the proposition that "[i]n the rare case, fraud damages based on the relinquishment of an alternative bargain have been recognized, but only because the fraud caused the plaintiff to give up some quantifiable, concrete alternative opportunity." Spithogianis v. Haj-Darwish, 2008 WL 82188, at *7 (S.D.N.Y. Jan. 7, 2008). It cites Dress Shirt Sales, Inc. v. Hotel Martinique Assocs. in which the New York Court of Appeals restated its "policy of our consistent refusal to allow damages for fraud based on the loss of a contractual bargain, the extent, and indeed, in this very case, the very existence of which is completely undeterminable and speculative." 190 N.E.2d 10, 12 (1963). FGIC argues that the lost opportunity of a contract with Calyon with a 50% attachment point was not undeterminable and speculative.

571. However, in Dress Sales, the New York Court of Appeals denied recovery based on an alternative transaction theory. The case is hardly good authority for FGIC's claim for such damages here. The decision is illustrative of principles of loss causation as applied by the New York courts and is not helpful to FGIC here. The plaintiffs in Dress Sales held a lease to occupy space in defendants' hotel for a period of ten years. Based on the defendants' false representation that they would not permit plaintiffs to sublease their space to a particular type of inexpensive restaurant, the defendants induced plaintiffs to cancel the lease. After the lease was canceled, the defendants leased the property to the restaurant. Plaintiffs sued for fraud, claiming that they were fraudulently induced to cancel the lease and suffered loss as a result of not being able to lease the space itself to the restaurant. There was no question of transaction causation or reliance. Plaintiffs clearly were induced by the false representation to cancel the lease; they would not have cancelled the lease but for the defendants' lie. The New York Court of Appeals nonetheless held that loss causation was not established. The only loss plaintiffs would have been entitled to recover was that "measured by the difference between the actual value of the remainder of the term and the price plaintiffs paid for it by reason of the lessors' deceit." Id. at 12. Because there was no assurance that plaintiffs would have been able to lease the space to the restaurant (the defendants had an unqualified right to refuse a sublease), the value of that lost opportunity or alternative transaction did not establish loss causation.

572. The Court need not here decide whether fraud damages are ever available for loss of a contractual bargain. Here, the existence of that alternative contractual transaction is entirely undeterminable and speculative. The evidence at trial was undisputed, and the Court has found that FGIC was never offered an alternative transaction to insure Calyon at the 50% attachment point (or any other attachment points above 40%). FGIC was not induced by a fraud to give up an asset it actually held or to forego an opportunity that was available to it, and Calyon's decision not to offer an attachment point above 40% is not alleged to be "because of defendants' fraud rather than an independent business decision made by [Calyon]." Lama Holding Co., 668 N.E.2d at 1374. Whether, if FGIC had turned down the transaction at the 40% attachment point, Calyon would have reconsidered and offered FGIC an opportunity to insure it at a 50% attachment point is entirely speculative. It would have depended on the decision of either the insurer of the junior super-senior tranche or some other group of investors to cover Calyon's risk of loss from the 28% attachment point to the 50% attachment point, and to do so at a price that would have been attractive both to Calyon and to the insurer. It also would have required Calyon to offer, and FGIC to be willing to accept, a fee at the 50% attachment point that the two entities both found acceptable. There was no evidence, however, that an insurer was readily available to take the tranche from 28% to 50% at a price Calyon found attractive. (If there were such an insurer, Calyon presumably would have taken that offer and offered FGIC only the opportunity to attach at the 50% attachment point.) Nor was there evidence that Calyon would have been willing to pay for insurance at the 50% attachment point a fee that FGIC believed would be acceptable. The two entities previously had walked away from one another during the course of the negotiations over the Pyxis transaction. There is no reason to believe that they would not have walked away from one another had FGIC demanded a transaction at the 50% attachment point.

573. FGIC failed to prove loss causation. 2. Negligent Misrepresentation

FGIC's pretrial brief asserted a theory of professional negligence in addition to negligent misrepresentation. The theory was that Putnam "deviated from the standard of care expected of collateral managers" because (1) it knew "that FGIC, like all Pyxis investors, was relying on it to provide accurate information about the Pyxis collateral" and (2) it "knew that Calyon could not be relied on to communicate accurately with investors" yet nonetheless "referred FGIC to Calyon" and then "failed to follow up to ensure Calyon had provided accurate information, even though it knew Calyon had sent materially inaccurate target portfolios to the rating agencies." Dkt. No. 238 at 19. For the reasons explained above and below, that theory cannot hold water. FGIC was not a potential investor in Pyxis, and Putnam did not owe a duty of care to FGIC either before or after the Pyxis transaction closed. Before the transaction closed, Putnam's duties were contractual and were owed to Calyon. It disclaimed any fiduciary duties to investors. After the transaction closed, its duties were owed to Pyxis, not to an insurer on a swap that used Pyxis as a reference entity. See Dkt. No. 238 at 18 (citing CMFF, LLC v. J.P. Morgan Inv. Mgmt., Inc., 2009 WL 6408599, at *5 (N.Y. Sup. Ct. Dec. 10, 2009) ("It is settled law that an investment manager of an investment account owes a fiduciary duty to its customer.")); see also Anunziatta v. Orkin Exterminating Co. Inc., 180 F. Supp. 2d 353, 358 (N.D.N.Y. 2001) ("Professionals, common carriers and bailees are examples of those who are subject to tort liability for negligent performance of a contractual obligation."). Even if Putnam owed FGIC duties, there is no evidence it failed to follow the applicable standard of care. There is no evidence that Putnam provided inaccurate information to FGIC or to investors about the Pyxis collateral. Moreover, Putnam did not refer FGIC to Calyon for a list of line-by-line target assets that were intended to be acquired for Pyxis. Nor is there evidence that Calyon could not be trusted to communicate accurately with investors, let alone that Putnam knew so. While FGIC offered some evidence that Putnam experienced frustration with Calyon at times, particularly in the drafting of the pitch materials as to which the two shared responsibility, that evidence came nowhere near demonstrating that Putnam knew Calyon would be unreliable in communicating with FGIC. Putnam had no duty to "follow up" on a communication that it did not know existed. Dkt. No. 238 at 18. Finally, although FGIC asserts that it, "like all Pyxis investors, was relying on Putnam to perform its duties independently in good faith in the interests of all investors," id. at 19, the Court has found that Putnam discharged that duty.

A. Bayerische Landesbank Test

574. FGIC also failed to establish the elements of negligent misrepresentation.

575. "Under New York law, the elements for a negligent misrepresentation claim are that (1) the defendant had a duty, as a result of a special relationship, to give correct information; (2) the defendant made a false representation that he or she should have known was incorrect; (3) the information supplied in the representation was known by the defendant to be desired by the plaintiff for a serious purpose; (4) the plaintiff intended to rely and act upon it; and (5) the plaintiff reasonably relied on it to his or her detriment." Hydro Inv'rs, 227 F.3d at 20.

576. A special relationship to give correct information sufficient to give rise to a claim for negligent misrepresentation exists when "(1) the defendant had awareness that its work was to be used for a particular purpose; (2) there was reliance by a third party known to the defendant in furtherance of that purpose; and (3) there existed some conduct by the defendant linking it to that known third party evincing the defendant's understanding of the third party's reliance." Bayerische Landesbank, 692 F.3d at 59 (citing Credit Alliance Corp. v. Arthur Andersen & Co., 483 N.E.2d 110, 118 (1985)). "The[se] indicia, while distinct, are interrelated and collectively require a third party claiming harm to demonstrate a relationship or bond with the once-removed [defendant] 'sufficiently approaching privity' based on 'some conduct on the part of the [defendant.].'" Sec. Pac. Bus. Credit v. Peat Marwick Main & Co., 597 N.E.2d 1080, 1083-84 (1992) (citation omitted).

577. New York courts have determined that such a rule is necessary "in order to provide fair and manageable bounds as to what otherwise could prove to be limitless liability." Prudential Ins. Co. of Am. v. Dewey, Ballantine, Bushby, Palmer & Wood, 605 N.E.2d 318, 382 (1992); see Ultramares Corp., 174 N.E. at 445-46.

578. Moreover, for a defendant to be liable for negligent misrepresentation, there must be some actionable misstatement or omission that is attributable to the defendant. See Woori Bank v. Citigroup Inc., 2013 WL 1235648, at *5 (S.D.N.Y. Mar. 27, 2013) (dismissing a negligent misrepresentation claim because plaintiff "failed to allege that there were any specific actionable misstatements or omissions that are attributable to [the defendants]").

579. FGIC failed to prove a case of negligent misrepresentation for multiple reasons. First, Putnam did not make or even contribute to any misrepresentation made to FGIC or, as explained above have a duty to speak to FGIC. Putnam did not author the PCS, it was not attributed to Putnam, Putnam did not contribute to the PCS or convey information to Calyon upon which it based the PCS, and it did not know that the PCS was going to be or was delivered to FGIC. It also did not have a fiduciary or other duty to speak to FGIC and did not have any duty to FGIC arising out of any statement it made to FGIC.

580. FGIC's proof at trial also failed all three elements of the Bayerische Landesbank test. Even if Putnam had had some awareness of the PCS and the fact that it was delivered to FGIC, the evidence did not show that it had an idea of the use to which FGIC would put it—as a measure not just of what Putnam could purchase within certain parameters, but of what it would in fact purchase in executing the Pyxis transaction. The evidence at trial showed that the rating agencies used portfolios as dummy assets to run their models but understood that the dummy assets did not represent real assets and that the manager retained discretion to purchase any assets that fell within the portfolio constraints and was in the interest of the CDO. There would have been no reason for Putnam, by contrast, to construe the PCS target assets as reliable promises of assets that would actually be acquired. See Parrott v. Coopers & Lybrand, L.L.P., 741 N.E.2d 506, 509 (2000) (finding no special relationship where defendant did not know its reports would be used specifically in connection with plaintiff's stock purchase agreement).

581. The evidence also is lacking that reliance by FGIC on the PCS was known to Putnam. Between the time of the delivery of the PCS by Calyon to FGIC and the closing of the Pyxis transaction, there was negligible contact between FGIC and Putnam and none regarding the assets in the portfolio or the analysis to which FGIC was putting the PCS. It thus would have been sheer speculation by Putnam to assume that FGIC was relying on the PCS in deciding whether to enter the transaction with Calyon.

582. Finally, there is no evidence of "conduct by the defendant linking it to [a] known third party evincing the defendant's understanding of the third party's reliance." Bayerische Landesbank, 692 F.3d at 59. It is not sufficient to impose liability for negligent misrepresentation under New York law unless the defendant "knew, should have known or was on notice" that the plaintiff was being shown reports prepared by the defendant in order to induce the plaintiff's reliance on those reports or that the defendant knew that the plaintiff, in fact, was relying upon defendant's reports. Credit Alliance Corp., 483 N.E.2d at 112. A defendant who gives false information to another knowing merely that it will be disseminated to third parties "according to the needs of the occasion" does not thereby take on a duty to all members of the indeterminate class to whom the information is shown. Id. at 116 (quoting Ultramares Corp., 174 N.E. at 442). There must be "some conduct on the part of the [defendant] linking them to that party or parties" and the conduct must evince the defendant's "understanding of that party or parties' reliance," id. at 118, as where conveyance of the information to the third party was the "end and aim" of the transaction, id. at 117 (quoting Ultramares Corp., 174 N.E. at 445). See also Sec. Pac. Bus. Credit, 597 N.E.2d at 1085 (finding no special relationship where the only contact between the parties was a "single unsolicited phone call").

583. The evidence fails the third element of the Bayerische Landesbank test. Even assuming that the limited information that Putnam provided to Calyon regarding Putnam's plans for the Pyxis portfolio in any way formed the basis for Calyon's subsequent transmission of detailed target assets to FGIC, there is no evidence that the "end and aim" of the transmission of the information from Putnam to Calyon was for it to be conveyed to FGIC. Outside of the preparation of the target portfolios in the marketing presentations, Putnam transmitted information to Calyon on limited occasions and for limited purposes. Putnam transmitted portfolio information to Calyon in Calyon's capacity as warehouse bank. Calyon had an independent interest and right to know the assets—the purchase of which it was financing and the risk of loss it would assume—if the transaction did not close. The evidence also admitted the possibility that Putnam and Calyon had communications, at least after the fact, about the information Calyon had transmitted to the ratings agencies to satisfy Calyon's obligations under the Mandate Letter to secure credit ratings for the notes. The evidence failed to show, however, that Putnam transmitted information about specific target assets to Calyon for the purpose, or with the "end and aim," of Calyon providing that information to FGIC.

FGIC's reliance on Kimmel v. Schaefer, 675 N.E.2d 450 (1996), is misplaced. There, the defendant "testified at trial that he expected plaintiffs to rely on [misleading] projections," that he "urged plaintiffs to review and rely on the projections," and that he "informed [one of the plaintiffs] that he could provide 'hot comfort' should plaintiff entertain any reservations about investing." Id. at 454-55. These facts are patently distinct from the facts at hand. As explained above, Putnam made no representations regarding the PCS to FGIC, was not aware of the contents of the PCS or that it was being transmitted to FGIC, and certainly did not encourage FGIC to enter the Pyxis transaction on the basis of the PCS. Plaintiff has adverted to no case holding that a special relationship exists in similar circumstances. --------

584. Finally, if Putnam cannot be held liable in negligent misrepresentation based on the theory that Putnam provided information to Calyon that Calyon—with Putnam's knowledge—then conveyed to FGIC for the purpose of inducing FGIC's reliance, Putnam cannot be held liable for a negligent omission for failure to alert FGIC to the falsity of information provided by Calyon to FGIC as to which Putnam was neither the source nor the author or co-author. Putnam did not stand in a confidential fiduciary relationship with FGIC, it did not make a partial or ambiguous statement giving rise to a duty to disclose, and it was not a party or potential party to a transaction with FGIC as to which it possessed superior knowledge, not readily available to FGIC. See Ellington Credit Fund, Ltd. v. Select Portfolio Servicing, Inc., 837 F. Supp. 2d 162, 201 (S.D.N.Y. 2011) (Sullivan, J.) (citing Lerner v. Fleet Bank, N.A., 459 F.3d 273, 292 (2d Cir. 2006); Brass v. Am. Film Techs., Inc., 987 F.2d 142, 150 (2d Cir. 1993)). It therefore had no duty to disclose to FGIC.

585. That result is sensible. Although FGIC repeatedly suggests otherwise, Putnam was not FGIC's counterparty in the guaranty or any other transaction. It was a third party to the FGIC-Calyon relationship. FGIC provided insurance to Calyon with respect to an asset (Calyon's interest in Pyxis) as to which Putnam was collateral manager. Putnam had a duty to Calyon, pursuant to the Mandate Letter, to make itself available to potential participants in Pyxis for due diligence sessions. It would have been in breach of that contract if it failed to make itself reasonably available. After Pyxis closed, Putnam had a duty to Pyxis, and derivatively to all of the investors in Pyxis as a whole, to manage the collateral of Pyxis in Pyxis's best interests. But each of these duties flowed from Putnam's contract with Calyon. It certainly owed no fiduciary or other duties in tort to Calyon's counterparty.

586. In a related context, the Supreme Court has instructed that the courts should hesitate and exercise care before creating new duties "in 'an area that demands certainty and predictability.'" Central Bank of Denver, N.A. v. First Interstate Bank of Denver, N.A., 511 U.S. 164, 188 (1994) (quoting Pinter v. Dahl, 486 U.S. 622, 652 (1988)). Those words of caution have resonance here. CDOs are complex transactions where the roles and responsibilities of the different parties are carefully delineated by contract and established law. Parties are entitled to rely upon contract and established law, to know their duties and when and how they are required to act. The contract, and established law, here placed responsibility on Calyon for ensuring that its statements to FGIC in connection with a Calyon-FGIC transaction were made with care. They did not impose a duty on Putnam—a third party to the Calyon-FGIC transaction who was not even privy to the details of conversations between the two—to investigate or ensure that Calyon acted with care. Nor did Putnam have a free-standing duty to FGIC to make sure that FGIC's decision was well-informed and that it reviewed all of the documents it should have. That responsibility is on FGIC. Indeed, by creating a new rule of law that would require a third party to a contract—with no duties to the contractual counterparty—to speak up and volunteer information it thinks (but does not know) the counterparty might want, the Court would risk putting the third party in an impossible position. On the one hand, the third party might be required to intervene in the relationship between the two parties and hazard a guess about whether each is acting with complete information to avoid later liability for a transaction gone bad. On the other hand, if it does hazard a guess and that guess turns out to be wrong, it could risk liability for defamation and tortious interference when the parties thereafter decide not to contract with one another. See Carvel Corp. v. Noonan, 818 N.E.2d 1100, 1103 (2004) (recognizing that New York law treats "interfering with a nonbinding 'economic relation'" as a tort); Restatement (Second) of Torts § 766B (1979) (noting that, as early as 1410, "it was said that 'if the comers to my market are disturbed or beaten, by which I lose my toll, I shall have a good action of trespass on the case'") (quoting 11 Hen. IV 47); Liberman v. Gelstein, 605 N.E.2d 344, 347 (1992) (holding an interloper's statements to the party (potentially slanderous per se because "tend[ing] to injure [the counter-party] in his or her trade, business or profession") could subject the interloper to liability for tortious interference if the party declined to proceed with the transaction on that basis); Amaranth LLC v. J.P. Morgan Chase & Co., 888 N.Y.S.2d 489, 494 (1st Dep't 2009) ("Defamation is a predicate wrongful act for a tortious interference claim."). In the end, however, the Court leaves to another day whether the creation of such a duty would be consistent with New York law. On the facts here, FGIC has not proved that Putnam had any reason to suspect that Calyon had given it incomplete or misleading information and thus FGIC's claim would fail even if there were such a duty.

B. Reliance and Reasonable Reliance

587. The same actual and reasonable reliance requirements apply to Plaintiff's fraud and negligent misrepresentation claims. See PHL Variable Ins. Co. v. Town of Oyster Bay, 929 F.3d 79, 94 (2d Cir. 2019); Vasquez v. Soto, 877 N.Y.S.2d 467, 468 (2d Dep't 2009). For the reasons stated above with respect to fraud, Plaintiff has failed to satisfy this element with respect to negligent misrepresentation.

C. Loss Causation

588. The standard for loss causation that applies to fraud claims also applies to negligent misrepresentation claims. See Laub, 745 N.Y.S.2d at 537. For the reasons stated above with respect to FGIC's fraud claim, it has not established loss causation for negligent misrepresentation.


The parties are directed to appear for a telephonic status conference on September 4, 2020 at 1:30 p.m. They are directed to call (888) 251-2909 and use Access Code 2123101.

The Clerk of Court is respectfully directed to terminate all pending motions.

SO ORDERED. Dated: September 14, 2020

New York, New York



United States District Judge

Summaries of

Fin. Guar. Ins. Co. v. Putnam Advisory Co.

Sep 14, 2020
12-cv-7372 (LJL) (S.D.N.Y. Sep. 14, 2020)
Case details for

Fin. Guar. Ins. Co. v. Putnam Advisory Co.

Case Details



Date published: Sep 14, 2020


12-cv-7372 (LJL) (S.D.N.Y. Sep. 14, 2020)