Federal Housing Finance Agency as Conservator for the Federal National Mortgage Association et al v. Goldman, Sachs & Co. et alMEMORANDUM OF LAW in Opposition re: 729 MOTION for Partial Summary Judgment on Defendants' Due Diligence and Reasonable Care Defenses. . DocumentS.D.N.Y.August 1, 2014UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF NEW YORK -------------------------------------------------------------------- x FEDERAL HOUSING FINANCE AGENCY, etc., Plaintiff, v. GOLDMAN, SACHS & CO., et al., Defendants. : : : : : : : : : : : 11 Civ. 6198 (DLC) -------------------------------------------------------------------- x GOLDMAN SACHS’ MEMORANDUM OF LAW IN OPPOSITION TO PLAINTIFF’S MOTION FOR SUMMARY JUDGMENT ON DUE DILIGENCE AND REASONABLE CARE DEFENSES July 18, 2014 Richard H. Klapper Theodore Edelman Michael T. Tomaino, Jr. Tracy Richelle High SULLIVAN & CROMWELL LLP 125 Broad Street New York, NY 10004-2498 Telephone: 212-558-4000 Facsimile: 212-558-3588 Attorneys for Goldman, Sachs & Co., Goldman Sachs Mortgage Company, GS Mortgage Securities Corp., The Goldman Sachs Group, Inc., Goldman Sachs Real Estate Funding Corp., Howard S. Altarescu, Kevin Gasvoda, Michelle Gill, David J. Rosenblum, Jonathan S. Sobel, Daniel L. Sparks and Mark Weiss Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 1 of 73 -ii- TABLE OF CONTENTS Page PRELIMINARY STATEMENT .....................................................................................................1 STATEMENT OF FACTS ..............................................................................................................9 I. GOLDMAN SACHS’ ROBUST DUE DILIGENCE PROCESS. ....................................10 A. Goldman Sachs Reviewed Counterparties Prior to Purchasing Loans, and Monitored Counterparties on an Ongoing Basis. ...................................................10 B. Goldman Sachs Conducted In-depth Loan-Level Pre-Purchase Diligence, Including Valuation, Credit, Compliance, Collateral and Fraud Reviews. ...........12 C. Goldman Sachs Continuously Enhanced Its Due Diligence Practices. .................15 D. E-mails and Witness Testimony Show that Goldman Sachs’ Diligence Was Stringent. ........................................................................................................16 E. Goldman Sachs Did Not Compromise Its Due Diligence Standards in Response to Originator Feedback. .........................................................................18 II. GOLDMAN SACHS WORKED CLOSELY WITH HIGH QUALITY THIRD- PARTY VENDORS TO CONDUCT DUE DILIGENCE. ...............................................18 III. GOLDMAN SACHS HAD STRONG INCENTIVES TO CONDUCT ROBUST DUE DILIGENCE. ............................................................................................................23 IV. GOLDMAN SACHS MAINTAINED A REASONABLE PROCESS TO PRODUCE MATERIALLY ACCURATE OFFERING MATERIALS. ..........................23 ARGUMENT .................................................................................................................................24 I. TO PROVE ITS REASONABLE CARE AND DUE DILIGENCE DEFENSES UNDER THE SECURITIES LAWS, GOLDMAN SACHS MUST SHOW IT EXERCISED REASONABLE CARE AND CONDUCTED A REASONABLE INVESTIGATION, RESPECTIVELY..............................................................................26 A. The Reasonableness of the Care or Investigation Undertaken Pursuant to Sections 11 and 12 and the Blue Sky Laws Is a Question of Fact. ........................26 B. There Is No Heightened Standard for a Section 11 Due Diligence Defense For Affiliated Underwriters in an Asset-Backed Securities Transaction. ..............29 Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 2 of 73 -iii- II. GOLDMAN SACHS’ DUE DILIGENCE PROVIDED A REASONABLE BASIS FOR ITS DISCLOSURES, WHICH GOLDMAN SACHS VERIFIED FOR MATERIAL ACCURACY. ...............................................................................................35 A. Whether Due Diligence at Time of Whole Loan Purchase Provided Reasonable Grounds for the Statements in the Offering Materials Is a Material Issue of Fact for the Jury. ........................................................................35 B. Goldman Sachs Verified the Material Accuracy of Representations in the Offering Materials. .................................................................................................39 III. GOLDMAN SACHS’ DUE DILIGENCE POLICIES WERE SOUND AND MATERIAL ISSUES OF FACT EXIST AS TO THEIR REASONABLENESS. ...........42 A. The Reasonableness of Goldman Sachs’ Use of Sampling. ..................................42 B. The Reasonableness of Adverse Selection. ...........................................................44 C. Whether Originators Limited Goldman Sachs’ Sample Sizes. ..............................45 D. Goldman Sachs’ Management of Third-Party Firms Was Effective. ....................47 E. The Reasonableness of Goldman Sachs’ Due Diligence When It Acted Only as Underwriter. ..............................................................................................49 F. The Securitization of Loans Graded as “3s”. .........................................................50 IV. GOLDMAN SACHS FOLLOWED ITS DUE DILIGENCE POLICIES. ........................54 A. Goldman Sachs Accurately Disclosed the Values of Properties. ..........................54 B. Goldman Sachs’ Representations Regarding Diligence of Loans Acquired Through Its Conduit Program Were Accurate In All Material Respects. ..............57 C. Information Obtained by Goldman Sachs From LownHome or Senderra Was the Same Information Known To the Public, Including the GSEs, and Goldman Sachs Did Not Conceal That Information To Gain Profits. ...................58 D. Goldman Sachs Properly Applied Its Policies Concerning Originators on “Watch” and “Suspend” Lists. ...............................................................................62 V. SUMMARY JUDGMENT IS IMPROPER BECAUSE GOLDMAN SACHS HAS NOT BEEN AFFORDED A FULL OPPORTUNITY FOR DISCOVERY. ....................64 CONCLUSION ..............................................................................................................................65 Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 3 of 73 -iv- TABLE OF AUTHORITIES Page(s) CASES Brody v. Village of Port Chester, No. Civ. 781 (HB), 2007 WL 735022 (S.D.N.Y. Mar. 12, 2007) ...........................................25 In re Countrywide Financial Corp. Mortgage-Backed Securities Litigation, 932 F. Supp. 2d 1095 (C.D. Cal. 2013) ...................................................................................28 FDIC v. Chase Mortgage Finance Corp., No. 12 Civ. 6166 (LLS), 2013 WL 5434633 (S.D.N.Y. Sept. 27, 2013) ...............................29 Federal Housing Finance Agency v. Bank of America Corp., No. 11 Civ. 6195(DLC), 2012 WL 6592251 (S.D.N.Y. Dec. 18, 2012) .................................26 Feinberg v. Katz, No. 01-cv-2739 (CSH), 2007 WL 4562930 (S.D.N.Y. Dec. 21, 2007) ...................................26 Feit v. Leasco Data Processing Equipment Corp., 332 F. Supp. 544 (E.D.N.Y. 1971) ................................................................................8, 33, 34 Foresta v. Centerlight Capital Mgmt., LLC, 379 F. App’x 44 (2d Cir. 2010) ...............................................................................................64 Franklin Savings Bank of New York v. Levy, 551 F.2d 521 (2d Cir. 1977).....................................................................................................28 In re Fuwei Films Securities Litigation, 634 F. Supp. 2d 419 (S.D.N.Y. 2009) ................................................................................26, 28 Ideal Steel Supply Corp. v. Anza, 652 F.3d 310 (2d Cir. 2011).........................................................................................25, 39, 42 Meloff v. New York Life Insurance Co., 51 F.3d 372 (2d Cir. 1995).......................................................................................................64 In re Metlife Demutualization Litigation, 262 F.R.D. 217 (E.D.N.Y. 2009) .............................................................................................27 NetJets Aviation, Inc. v. LHC Communications, LLC, 537 F.3d 168 (2d Cir. 2008), ...................................................................................................29 Quincy Co-op. Bank v. A.G. Edwards & Sons, Inc., 655 F. Supp. 78 (D. Mass. 1986) .............................................................................................26 Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 4 of 73 -v- Sanders v. John Nuveen & Co., 619 F.2d 1222 (7th Cir. 1980) .................................................................................................28 In re Software Toolworks Inc., 50 F.3d 615 (9th Cir. 1995) .....................................................................................................28 University Hill Foundation v. Goldman, Sachs & Co., 422 F. Supp. 879 (S.D.N.Y. 1976)...........................................................................................28 Vivenzio v. City of Syracuse, 611 F.3d 98 (2d Cir. 2010).................................................................................................25, 51 Weinberger v. Jackson, No. C-89-2301-CAL, 1990 WL 260676 (N.D. Cal. Oct. 11, 1990) ........................................28 In re WorldCom, Inc. Securities Litigation, 346 F. Supp. 2d 628 (S.D.N.Y. 2004) .............................................................................. passim In re WorldCom, Inc. Securities Litigation, No. 02 Civ. 3288 (DLC), 2005 WL 638268 (S.D.N.Y. Mar. 21, 2005) .......................... passim STATUTES, RULES AND REGULATIONS 15 U.S.C. § 77b(a) .........................................................................................................................30 15 U.S.C. § 77k(b) ...................................................................................................................29, 30 15 U.S.C. § 77k(c) ...................................................................................................................23, 27 15 U.S.C. § 77g ..............................................................................................................................43 17 C.F.R. § 229.1101(c)(2)(ii) .................................................................................................32, 33 17 C.F.R. § 230.191 .......................................................................................................................30 17 C.F.R. § 240.3b-19 ....................................................................................................................30 12 C.F.R. § 1026.43 ......................................................................................................................55 FED. R. CIV. P. 56(a) ......................................................................................................................24 FED. R. EVID. 407 ...........................................................................................................................44 S.D.N.Y. Local Rule 56.1(d) ...........................................................................................................9 Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 5 of 73 -vi- OTHER AUTHORITIES Gregory A. Markel & Gregory G. Ballard, Defending Claims Against Underwriters of Asset-Backed Securities: The Due Diligence Defense, Andrews Sec. Litig. and Reg. Rep. (2006) ..............................................................................................................................32 Harold S. Bloomenthal & Samuel Wolff, Chapter 13A: Asset-Backed Securities, the Financial Crisis and the Dodd-Frank Act, Going Public and the Public Corporation § 13A:1 (2014) .........................................................................................................................32 Hearing Transcript, 70:19-71:1, Dec. 17, 2012 (11 Civ. 6198 (DLC)) ..................................36, 42 Merritt B. Fox, Due Diligence with Residential Mortgage Backed Securities, Columbia Law & Econ. Working Paper No. 462 (2013) .........................................................................33 SEC Release Nos. 33-8518, 34-50905 (Jan. 7, 2005) ........................................................... passim SEC Release No. 8591 (Aug. 3, 2005) .........................................................................................28 SEC Release Nos. 33-9176, 34-63742 (Jan. 25, 2011) ..............................................................4, 43 Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 6 of 73 -vii- TABLE OF ABBREVIATIONS Action Federal Housing Finance Agency v. Goldman, Sachs & Co., et al., No. 11 Civ. 6198 (DLC) AVM Automated valuation model Certificates The specific RMBS that Fannie Mae and Freddie Mac allegedly purchased from the Securitizations at issue in this Action DTI Debt-to-income ratio Due Diligence Group Goldman Sachs due diligence personnel EPDs Early payment defaults Fannie Mae Federal National Mortgage Association Freddie Mac Federal Home Loan Mortgage Corporation FHFA Federal Housing Finance Agency, the regulator of Fannie Mae and Freddie Mac, which brings this Action in its capacity as their conservator Goldman Sachs Defendants Goldman, Sachs & Co., GS Mortgage Securities Corp., Goldman Sachs Mortgage Company, The Goldman Sachs Group, Inc., Goldman Sachs Real Estate Funding Corp., Howard S. Altarescu, Kevin Gasvoda, Michelle Gill, David J. Rosenblum, Jonathan S. Sobel, Daniel L. Sparks, and Mark Weiss GSMSC GS Mortgage Securities Corp. GSMC Goldman Sachs Mortgage Company Grice Database Database of all available credit, compliance, and valuation due diligence findings for loans in the Securitizations, compiled by Goldman Sachs’ expert Charles Grice Grice Report Expert Report of Charles Grice, dated March 27, 2014 GSE Government-Sponsored Entity (here, Fannie Mae and/or Freddie Mac) Harsch Declaration or Harsch Decl. Rule 56(d) Declaration of Bradley A. Harsch, dated July 18, 2014 Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 7 of 73 -viii- LTV Loan-to-value MCC Goldman Sachs Mortgage Capital Committee Motion FHFA’s Motion for Partial Summary Judgment on Defendants’ Due Diligence and Reasonable Care Defenses, dated June 10, 2014. Offering Materials The offering materials for the Securitizations in this Action, including documents filed with the Securities and Exchange Commission, such as registration statements or base prospectuses, prospectus supplements and any supplements, and any other free writing prospectuses or other documents pertaining to that Securitization, as well as any preliminary term sheets not filed with the Securities and Exchange Commission Originator(s) Any originator that originated loans underlying the Securitizations Pl. Br. or Brief Plaintiff’s Memorandum of Law in Support of Its Motion, dated June 10, 2014 PLS Private label securities RMBS Residential mortgage-backed securities Securitizations The PLS offerings from which the GSEs purchased the Certificates at issue in this Action Stmt. Goldman Sachs’ Local Rule 56.1 Counterstatement of Material Facts and Reply to Plaintiff’s Statement of Undisputed Material Facts SLGs The supporting loan groups for the Certificates WHAM Whole Loan Asset Miner Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 8 of 73 PRELIMINARY STATEMENT The many genuine issues of material fact concerning Goldman Sachs’ due diligence and reasonable care defenses preclude entry of summary judgment for FHFA. Whether the evidence as a whole supports those inherently fact-based defenses is a question for the jury. To succeed on its Motion, FHFA bears the heavy burden of showing a “complete failure of proof” with respect to the care and investigation Goldman Sachs utilized to conduct due diligence in connection with the forty Securitizations. See infra at 25-26. As this Court recognized, a determination of the sustainability of a defendant’s due diligence defense “requires an exquisitely fact intensive inquiry into all of the circumstances surrounding the facts upon which [the securities claim] is premised,” and “such fact-intensive inquiries do not lend themselves easily to resolution on summary judgment.” In re WorldCom, Inc. Sec. Litig., No. 02 Civ. 3288 (DLC), 2005 WL 638268, at *11 (S.D.N.Y. Mar. 21, 2005). FHFA has failed to cite a single case in which summary judgment was granted against a defendant where the defendant did more than rely on the statements of the issuer’s management. The details concerning Goldman Sachs’ due diligence on the Securitizations—which included a thorough review of both the counterparties involved and the loans underlying those Securitizations—as well as a respected expert’s opinion as to the reasonableness of Goldman Sachs’ conduct and both parties’ statements of undisputed fact, alone show Goldman Sachs did far more than rely on statements of management here, before even considering the vast additional evidence including countless pages of due diligence-related documents produced in this Action. These material issues of fact are for the jury to resolve in deciding whether Goldman Sachs’ due diligence was reasonable. Instead of showing the absence of material disputed facts, FHFA makes a series of largely rhetorical arguments based on its own disputed characterizations of documents and e- mails, while ignoring a vast amount of issue-creating evidence, including: Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 9 of 73 -2- None of the ten Goldman Sachs employees involved in due diligence on the Securitizations that FHFA deposed testified that Goldman Sachs’ due diligence practices were unreasonable, haphazard or ineffective, as FHFA contends—all testified to the contrary. See, e.g., Stmt. ¶¶ 542-546. Testimony of non-party witnesses confirmed that the Goldman Sachs personnel in charge of due diligence were “very serious,” “very competent,” “hands on,” and “knowledgeable.” Stmt. ¶¶ 622-626. Goldman Sachs’ management consistently refused to compromise the firm’s due diligence standards, as evidenced by contemporaneous documents, including e-mails stating: “I don’t see us reducing DD now,” “our due diligence standards should not be weakened,” “it does NOT mean that we should relax our standards,” “we aren’t going to alter our standards based upon the noise level from the client,” and “we don’t want to drop standards and compromise on these important things as a precedent.” Stmt. ¶¶ 629-631, 633. Goldman Sachs’ expert, Charles Grice, analyzed Goldman Sachs’ due diligence practices both generally and for the loans underlying the Securitizations, and found that those practices were consistent with, or exceeded, industry practices. Stmt. ¶ 531. The four experts FHFA intends to offer in response have not shown otherwise. Although FHFA now calls Goldman Sachs’ due diligence “shoddy” and “ramshackle,” Pl. Br. at 47-50, in 2006 Freddie Mac rated it “Above Average,” leading one Freddie Mac employee to comment that “No one ever gets a ‘superior,’ so these guys are at the top of their class.” Stmt. ¶¶ 640-642. Because a fair and accurate review of the record demonstrates the existence of myriad material issues of disputed fact, FHFA resorts to a brief riddled with inaccuracies, unwarranted generalizations, mischaracterizations of isolated e-mails taken out of context, and scores of impermissible inferences against Goldman Sachs, the non-movant. Each of the arguments FHFA raises creates issues of fact that must be decided by a jury. First, FHFA claims Goldman Sachs should have, as a matter of law, conducted a duplicative round of due diligence at the time of securitization, in addition to the due diligence that Goldman Sachs performed when it purchased the loans. Pl. Br. at 8-10, 46-48. FHFA fails, however, to show any consequences of performing due diligence at the acquisition stage rather than immediately before securitization. Goldman Sachs’ due diligence at acquisition evaluated Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 10 of 73 -3- the loans for all the criteria disclosed in the Offering Materials, including whether loans generally complied with underwriting guidelines, the reasonableness of appraisals, and owner occupancy status. Most of the information tested in due diligence at acquisition—such as compliance with origination guidelines, reasonableness of appraisals, loan terms, and other attributes of the borrower and the property—either by definition could not change (e.g., compliance with guidelines at origination) or was unlikely to change in the months after purchase of the loans (e.g., mortgage history). Although FHFA speculates that “additional information could become available” between acquisition and securitization, Pl. Br. at 9, it fails to show that any material information did change before the “cut-off date” as of which the Offering Materials represented the information was accurate. Moreover, nearly 80% of the loan pools were securitized within five months of acquisition. See infra Fact Sec. II.A.1. The jury must decide whether the extensive due diligence Goldman Sachs performed was unreasonable. Second, FHFA claims that, because of what it calls “commingling” between loan pools and SLGs, “the percentages of loans in the final SLGs that had been diligenced were often substantially smaller than the percentages of loans in the acquisition pools that Goldman’s own policies required.” Pl. Br. at 16. Overwhelming evidence refutes that assertion: about 36% of loans in the SLGs were subject to credit and compliance due diligence and about 84% received valuation due diligence, essentially the same as for the acquisition pools as a whole. Stmt. ¶¶ 529-530. The jury must determine the significance or reasonableness of these statistics. Third, FHFA asserts that, because neither the “internal diligence team” nor “third party vendors” reviewed prospectus supplements, Pl. Br. at 1-2, Goldman Sachs’ due diligence was not reasonable as a matter of law. FHFA ignores that the work of the due diligence team and third-party vendors provided the basis for other teams at Goldman Sachs to conclude that the Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 11 of 73 -4- loans generally complied with underwriting standards and that the information on loan tapes was accurate. Stmt. ¶¶ 386-392. Relying on the due diligence process, the Mortgage Finance Group, which had “oversight responsibility for the offering materials at large,” worked with accountants, outside attorneys and other Goldman Sachs teams to provide “materially accurate” Offering Materials. Id. Whether Goldman Sachs’ process for verifying the information in the Offering Materials for material accuracy was reasonable is a factual determination that cannot be resolved on a motion for summary judgment. Fourth, FHFA accuses Goldman Sachs of sampling without statistical validity, yet relies on a litigation sample of only 2% of the SLG loans. In any event, industry practice and SEC rules regarding reasonable due diligence for RMBS offerings do not require statistically valid samples. See, e.g., SEC Release Nos. 33-9176, 34-63742, at 4235 (Jan. 25, 2011), Stmt. ¶ 465. Because of the different characteristics of loans in each pool, Goldman Sachs relied on the expertise of the personnel responsible for conducting due diligence to select both an adverse and a random sample for each pool, and focused greater attention on loans perceived as having higher potential risk. The reasonableness of Goldman Sachs’ sampling methodology in the circumstances is an issue for the jury. Fifth, despite documentation and testimony to the contrary, FHFA claims that Goldman Sachs succumbed to pressure from originators to change its due diligence practices, and failed to upsize samples when warranted. Pl. Br. at 16-19. The contemporaneous evidence and deposition testimony shows that, although Goldman Sachs’ stringent due diligence standards placed it at a competitive disadvantage, Goldman Sachs consistently refused to compromise those standards. A jury must weigh the evidence and determine the facts. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 12 of 73 -5- Sixth, FHFA cites to cherry-picked documents to contend that Goldman Sachs’ due diligence was unreasonable as a matter of law because Goldman Sachs at times had concerns about third-party due diligence vendors—which were the same vendors used by the GSEs—on which it relied. See Pl. Br. at 24-28, 53-54. The testimony and contemporaneous e-mails show Goldman Sachs did not blindly accept the findings of third-party due diligence firms, or abdicate diligence review to them. Linda Peterson, the manager of the Due Diligence Group (who now works at Freddie Mac), testified that in dealing with diligence vendors on behalf of Goldman Sachs, she was “like a dog with a bone, they’re not getting away from me.” Stmt. ¶ 607. This Court said that questions of what constitutes “red flags” are “exquisitely fact intensive inquiries that depend on the circumstances” and are “not appropriate for determination at summary judgment.” In re WorldCom, Inc. Sec. Litig., 346 F. Supp. 2d 628, 679-80 (S.D.N.Y. 2004). Seventh, FHFA claims that Goldman Sachs’ due diligence was unreasonable as a matter of law because it securitized loans that due diligence vendors graded “3.” FHFA erroneously equates an event grade “3” with a defective loan and ignores that the meaning of a “3” could vary by deal. In fact, as the GSEs were aware from waiving loans graded a “3” into their own securities, an event grade “3” could be cured (e.g., through identification of a previously missing document), offset by compensating factors (e.g., good job stability and mortgage payment history for a slightly lower FICO score), mistakenly graded “3” (for example, by applying the wrong underwriting guideline) or graded a “3” only because it did not comply with the more stringent “overlays” Goldman Sachs instructed the diligence provider to apply during its review (although it complied with the originator’s underwriting guidelines). FHFA also glosses over the fact that the Securitizations included very few loans with a final grade of “3,” and a review of those loans by Goldman Sachs’ expert Charles Grice indicated that the loans Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 13 of 73 -6- were included in the Securitizations for valid reasons. Resolution of FHFA’s incorrect claim requires factual determinations by the jury, including the meaning and significance of an event grade “3” and whether the small number of loans graded “3” included in the Securitizations could support a finding that Goldman Sachs’ due diligence was unreasonable. Eighth, FHFA contends, based on nine loans in the SLGs that were discounted by less than 2% of the original purchase price—out of almost 74,000 loans in the SLGs—that Goldman Sachs purportedly determined, but failed to disclose, the “true value” of SLG loans. This argument ignores the plain language of the Offering Materials, which clearly state that LTV ratios are based on origination appraisals. FHFA misconstrues the meaning of Goldman Sachs’ estimates of value based on conflicting data points such as AVMs and BPOs. In order to resolve this issue, a jury must make several fact determinations, including the meaning of Goldman Sachs’ “final values,” whether re-pricing nine loans constitutes per se unreasonable due diligence, and whether Goldman Sachs’ reconciled values somehow rendered misleading the disclosed LTV formulations based on original appraisals. Ninth, FHFA argues that Goldman Sachs failed to disclose accurately which underwriting guidelines were used for loans acquired through the Goldman Sachs Mortgage Conduit Program, claiming that the Offering Materials erroneously stated that those loans had been underwritten according to the Goldman Sachs Mortgage Conduit Underwriting Guidelines. This contention is incorrect. Even if an investor interpreted the Offering Materials as FHFA suggests, Goldman Sachs either reviewed conduit loans on a loan-level basis, or performed reviews of either the Goldman Sachs Conduit Guidelines and originators’ guidelines to identify and address any differences between the two sets of guidelines. Thus, there are material issues of fact as to how a reasonable investor would interpret the Offering Materials and whether the Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 14 of 73 -7- “gap analyses” Goldman Sachs performed ensured that the conduit program loans generally complied with the underwriting criteria described in the Offering Materials. Tenth, FHFA claims that Goldman Sachs obtained supposedly “inside” information about fraud in the subprime market from its affiliates LownHome and Senderra, and used that information to benefit financially by shorting the market. No evidence links the speculation about causes of poor loan performance forwarded by Senderra or LownHome to Goldman Sachs’ later risk reduction efforts, and the purportedly “inside” facts were well known to the public and the GSEs. Further, Goldman Sachs’ effort to reduce risk during an increasingly uncertain time in the housing market (including reduction of both long and short risk) was wholly unrelated to the reasonableness of Goldman Sachs’ earlier due diligence or the accuracy of disclosures in the Offering Materials. Eleventh, FHFA argues that Goldman Sachs’ due diligence was unreasonable as a matter of law because it purportedly ignored issues it identified regarding originators from which it purchased loans. Far from ignoring such issues, Goldman Sachs created “watch” and “suspend” lists, based on a number of factors (many unrelated to loan quality), reflecting a decision that Goldman Sachs should not continue acquiring loans from certain originators until the reason for their suspension was resolved. At a minimum, the reasonableness of Goldman Sachs’ securitization of loans previously acquired from originators on a “watch” or “suspend” list is a question for the jury. Because FHFA realizes that its inherently fact-based challenges are inappropriate for determination on summary judgment, it alternatively seeks to strip away Goldman Sachs’ due diligence defense altogether by inventing a heightened legal standard for the “new world of RMBS transactions,” under which “underwriters will almost never be able to meet their burden Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 15 of 73 -8- of proof for a due diligence defense” when they are affiliated with issuers. Pl. Br. at 44. However, FHFA’s arguments for applying “near absolute” liability to underwriters affiliated with issuers have no basis in the statutory text and are particularly inappropriate in the context of asset-backed securities. Nothing on which FHFA relies suggests that an underwriter in an RMBS transaction cannot avail itself of Section 11’s due diligence defense merely because it is affiliated with a “solely passive entity” whose financial information is “not . . . useful information to investors.” SEC Release Nos. 33-8518, 34-50905 at 1508, 1510-11 (Jan. 7, 2005). Nor does FHFA justify its proposed heightened standard with respect to Goldman Sachs’ reasonable care defenses under Section 12 and the blue sky laws, which are available to the issuer itself, and a fortiori to the underwriter. Indeed, as this Court has observed, the Section 12 reasonable care defense is “less demanding than the duty of due diligence imposed under Section 11.” In re WorldCom, 346 F. Supp. 2d at 663. The cases FHFA cites to the contrary, Pl. Br. at 59, either are not binding precedent or do not support FHFA’s contention. Furthermore, FHFA mis-cites cases such as Feit v. Leasco Data Processing Equip. Corp., 332 F. Supp. 544 (E.D.N.Y. 1971), and In re WorldCom, 346 F. Supp. 2d at 628, to suggest that underwriters must be completely independent of issuers. Those decisions require underwriters of corporate offerings to make an “independent” investigation beyond merely accepting at face value the statements of management. Here, Goldman Sachs made an “independent” investigation by conducting a thorough investigation of the almost 74,000 SLG loans it acquired from unaffiliated originators (as well as the over 650 SLG loans originated by its affiliates), and did not accept originators’ representations at face value. The reasonable care and due diligence defenses both require a review of the entire evidentiary record, and hence are inappropriate for summary judgment. See infra at Argument I.A. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 16 of 73 -9- Finally, although the evidence Goldman Sachs presents in response to FHFA’s Motion demonstrates the existence of genuine issues of material fact for the jury, entry of summary judgment is also unjustified because of the restrictions placed on Goldman Sachs’ ability to pursue discovery on those issues. As set forth in the accompanying Harsch Declaration, the discovery limitations in this case have denied Goldman Sachs crucial discovery—particularly into the GSEs’ own due diligence practices—which provide a relevant benchmark for assessing the reasonableness of Goldman Sachs’ due diligence. STATEMENT OF FACTS Between September 7, 2005 and October 29, 2007, the GSEs purchased more than $11.1 billion in Certificates in 40 Securitizations for which Goldman Sachs served as the sponsor, depositor, or lead underwriter. Stmt. ¶ 374.1 GSMSC filed four shelf registration statements relating to 35 of the 40 Securitizations and acted as a depositor for those 35 Securitizations. Stmt. ¶ 379. GSMC, an indirect wholly-owned subsidiary of The Goldman Sachs Group, Inc., sponsored 36 Securitizations. Stmt. ¶¶ 376, 378. Goldman Sachs’ sole involvement in four of the Securitizations was as an underwriter. Stmt. ¶ 380. Goldman Sachs offered and sold each of the Certificates pursuant to a shelf registration statement and a prospectus supplement. The prospectus supplements for the Securitizations included information about the Securitizations and the underlying loans, including statements that the loans were underwritten “generally in accordance with the underwriting guidelines of the original loan sellers,” and that exceptions to those guidelines were “reasonable.” Stmt. ¶ 375. 1 Goldman Sachs maintains that deposition transcripts in regulatory matters are inadmissible and, thus, improper evidence in support of FHFA’s Motion. See S.D.N.Y. Local Rule 56.1(d). Goldman Sachs cites such transcripts to demonstrate Goldman Sachs’ anticipated trial testimony. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 17 of 73 -10- I. GOLDMAN SACHS’ ROBUST DUE DILIGENCE PROCESS.2 The extensive evidence demonstrating that Goldman Sachs’ due diligence was more than reasonable—all of which FHFA largely ignores—highlights the fact-intensive inquiry the jury must conduct, which is inappropriate for determination on summary judgment. See In re WorldCom, Inc., 2005 WL 638268, at *11. Goldman Sachs’ due diligence included practices that Freddie Mac witness Ron Feigles characterized as “robust”: on-site supervision of Clayton, use of AVMs to confirm appraisal values, fraud reviews and reliance on adverse sampling methodology. Stmt. ¶ 673. According to Goldman Sachs’ due diligence expert, Charles Grice, who has more than 30 years of experience in the industry, Goldman Sachs’ due diligence for the loans underlying the Securitizations was “well-documented, reasonably designed, . . . ensured a high standard of care, . . . consistent with industry practices and, in some respects, w[as] among the most conservative and best documented in the industry.” Stmt. ¶¶ 605, 608. Goldman Sachs’ due diligence on the loans underlying the Securitizations incorporated an in-depth counterparty review process prior to Goldman Sachs engaging with a counterparty, and continued monitoring of counterparties. At the loan level, Goldman Sachs also performed credit, compliance, valuation and fraud diligence—similar to the diligence conducted by the GSEs. See Stmt. ¶¶ 679-687. Consistent with industry practice and as the GSEs knew, third-party diligence vendors assisted Goldman Sachs with its diligence. See infra Facts Sec. II. A. Goldman Sachs Reviewed Counterparties Prior to Purchasing Loans, and Monitored Counterparties on an Ongoing Basis. Before Goldman Sachs purchased loans from an originator, it reviewed the originator’s finances, underwriting guidelines, policies and procedures, and management. Stmt. 2 For a more detailed description and analysis of Goldman Sachs’ due diligence policies, procedures and practices, see Krugman Decl. ¶ 94. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 18 of 73 -11- ¶¶ 407-408. These counterparty reviews included meetings with Goldman Sachs representatives, including outside counsel who inquired about the originator’s compliance with state and local anti-predatory lending laws, and an experienced Chief Underwriter who reviewed the originator’s underwriting guidelines. See Stmt. ¶ 408. Goldman Sachs also conducted ongoing monitoring of originators through its Asset Management Group, Stmt. ¶¶ 412-415, as well as periodic reviews of approved originators by members of the Steering Committee. Stmt. ¶ 412. In conducting those reviews, members of the Steering Committee received seller status reports, which provided monthly updates from Goldman Sachs’ due diligence team on risk issues concerning various originators. See Stmt. ¶ 416. An originator included in the report would receive an overall recommended status of “performing,” “monitor,” “watch,” or “suspend” based on a series of criteria, including EPD and repurchase claims, review by Goldman Sachs’ Business Intelligence Group, pull-through rates, financial results, liquidity, and credit exposure limits.3 Id. The determination of the scope of due diligence of loans from an originator was based, in part, on the seller’s classification in the report. Id. Contrary to FHFA’s claim, assignment of a “watch” or “suspend” status to an originator did not mean that all loans provided by that originator were of poor quality. See Pl. Br. at 3, 30-31, 47, 54-55; Stmt. ¶ 417. Typically, an originator’s assigned status was based on the lowest status of any of the reviewed criteria. For example, if an originator received a “watch” rating for pull-through rate, but “performing” for every other assessed category, the counterparty still would have a recommended status of “watch.” Stmt. ¶ 417. 3 A status of “performing” dictated that “[n]ormal trading and due diligence as outlined in policy and procedures” should continue. Stmt. ¶ 416. For originators assigned “monitor” status, “[a]pproval of [the] Conduit Steering Committee [was] required to continue normal trading and due diligence.” Id. A status of “watch” indicated “[n]o sampling, [and] 100% BPO review.” Id. Lastly, “suspend” meant Goldman Sachs “should not be doing a transaction” with the originator until resolution of the reason for suspension. Id. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 19 of 73 -12- B. Goldman Sachs Conducted In-depth Loan-Level Pre-Purchase Diligence, Including Valuation, Credit, Compliance, Collateral and Fraud Reviews. 1. Valuation Due Diligence. With the assistance of third-party due diligence vendors, the Due Diligence Group conducted extensive due diligence on whole loan pools that Goldman Sachs intended to purchase. Stmt. ¶ 383. This review included valuation due diligence at varying rates depending on the type of pool. Goldman Sachs procedures recommended conducting AVMs on 85-100% of subprime pools and 100% of conduit flow pools, while allowing AVMs on less of the pool for jumbo and Alt-A pools. Stmt. ¶ 453. Goldman Sachs also ordered BPOs as needed, such as when the AVM value varied more than 15% from the appraisal value, or when the AVM did not provide an estimate. Stmt. ¶ 470. Goldman Sachs engaged in a process to reconcile BPOs that were out of tolerance by more than 15%, by reviewing, or requiring the BPO vendor to review, all available data, including the AVM and BPO, and determining the best supported value of the property. Stmt. ¶ 483. Goldman Sachs then typically discussed with the seller whether the reconciled value was reasonable. Stmt. ¶ 486. If Goldman Sachs concluded that the value was not reasonable, it typically would drop the loan. Stmt. ¶¶ 486-487. Disputed issues of fact exist regarding FHFA’s claims that Goldman Sachs failed to obtain a reconciled value for certain loans in the SLGs, and “routinely”—an argument based on nine of the almost 74,000 loans in the SLGs—determined that the “actual values” of the mortgaged properties underlying the Securitizations were lower than reported by originators, but concealed these “true” values to negotiate lower prices for itself on the loans. Pl. Br. at 34-36. After a detailed analysis, Goldman Sachs’ expert Charles Grice found Goldman Sachs generally followed its own diligence procedures for the pools underlying the Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 20 of 73 -13- Securitizations. Mr. Grice determined that Goldman Sachs conducted valuation due diligence on 87.33% of loans in pools that contributed loans to the securitizations at issue, and 84.06% of loans in the SLGs. See Grice Database;4 Stmt. ¶ 530. 2. Credit and Compliance Due Diligence. Goldman Sachs also conducted credit and compliance due diligence on loans it considered purchasing. Credit due diligence reviewed loans for compliance with the applicable underwriting guidelines, including whether sufficient compensating factors existed to overcome any exceptions to guidelines. Stmt. ¶ 489. Similar to the GSEs, Goldman Sachs required that loans be reviewed to additional criteria or “overlays” more stringent than the applicable originator guidelines.5 Stmt. ¶¶ 491-492. Compliance due diligence reviewed loans for compliance with local, state, and federal laws. Stmt. ¶ 489. Goldman Sachs conducted credit and compliance due diligence on either all loans in a pool or a sample of adversely and randomly selected loans. Stmt. ¶¶ 449, 455. Factors considered in identifying the adverse sample included high LTV or CLTV ratio, Stmt. ¶ 458, low borrower credit score, id., loans subject to specific state or local compliance requirements, Stmt. ¶ 458, and loans identified during vendor fraud review, Stmt. ¶ 456. Goldman Sachs typically checked loans in a new pool against records of loans that had been dropped from previous pools, to ensure that any such loans were re-reviewed. Stmt. ¶ 458. Although FHFA discounts adverse sampling, Pl. Br. at 19-21, 51-52, 54, adverse sampling enabled Goldman Sachs to enhance its 4 In preparing his expert report, Mr. Grice compiled and analyzed databases of all available credit, compliance, and valuation due diligence findings for loans in the Securitizations. See Grice Report at 36; Grice Database. The Grice Database for credit and compliance due diligence includes approximately 220,000 loans from pools that contributed to the 40 Securitizations. 5 Items flagged by Goldman Sachs’ overlays included condominium loans, loans for purchase of manufactured housing, and loans for which the DTI exceeded a certain value, all of which met the originators’ guidelines. Stmt. ¶ 495-498. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 21 of 73 -14- due diligence by focusing on loans perceived as having higher potential risk. The internally recommended sample size for credit and compliance due diligence varied depending on the type of loan pool, and increased in late 2006. Stmt. ¶ 451. Because each pool had different characteristics, prior to determining the extent of, and procedures for, sample selection and diligence, Goldman Sachs reviewed several factors, including originator and borrower information, internal pre-due diligence reports, and geographic concentration. Stmt. ¶ 458. Mr. Grice determined that Goldman Sachs selected more than 85,000 (38%) of the approximately 220,000 loans included in the Grice Database for credit due diligence and more than 86,000 (39%) of the loans in that database for compliance due diligence. See Stmt. ¶ 531. The percentages were essentially the same for the SLGs. See Stmt. ¶ 532. 3. Collateral Due Diligence and Data Integrity Review. Goldman Sachs reviewed the collateral for all loans for potential purchase by confirming that the loan file contained certain key documents, including original notes, mortgages, title policies, assignments and insurance certificates. Stmt. ¶ 499-506. Goldman Sachs also compared the information in the loan files to loan tape data to assess the accuracy of key data fields, including the loan number, appraisal value, purchase price, property type, and occupancy type. Stmt. ¶ 499. Goldman Sachs also hired accounting firms to confirm the accuracy of loan tape data. Stmt. ¶¶ 507-511. 4. Fraud Due Diligence. FHFA’s incorrect—and at the very least disputed—assertion that “Goldman had no diligence process focused on fraud review and investigations until mid-2007,” Pl. Br. at 12, is contradicted by countless documents in the record, many of which Mr. Grice identified in his report. Stmt. ¶¶ 512-523. Starting in mid-2005, Goldman Sachs regularly performed fraud due diligence on whole loan pool purchases to identify loans with a higher risk of fraud. Stmt. ¶ 512. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 22 of 73 -15- To flag potential fraud, Goldman Sachs utilized fraud detection products including AppIntell’s Pre-Funding Fraud Filter System (“Preffis Plus”) or an Interthinx AVM add-on product (ClearValue), on all loans. Stmt. ¶ 512; see also Stmt. ¶ 514. 5. The WHAM Database. FHFA’s mistaken claim that Goldman Sachs used a database called “WHAM” as an “integral” part of the diligence process despite concerns regarding the accuracy of information included in WHAM creates further disputed issues of fact. Pl. Br. at 13. Goldman Sachs generally did not use WHAM in connection with the due diligence process other than to select samples in certain cases and as a source for identifying borrowers for review by Goldman Sachs’ Fraud Prevention Department. Stmt. ¶ 525. These inquiries were based, in part, on information in loan tapes uploaded to WHAM. Stmt. ¶ 526. To the extent Goldman Sachs used WHAM during the due diligence process, transaction managers would assess the accuracy of the data in WHAM before relying on it. Stmt. ¶¶ 527-528. C. Goldman Sachs Continuously Enhanced Its Due Diligence Practices. Goldman Sachs enhanced its due diligence practices based on continuing experience and feedback. Stmt. ¶¶ 608-616. For example, in January 2006, after Goldman Sachs became aware that mortgage notes for loans originated by Acoustic Home Loans as “40- due-in-30 ARMS” had language specifying an incorrect amortized term, Linda Peterson, the Due Diligence Manager, implemented a “note library” review for each transaction involving that loan type. Stmt. ¶ 609. The note library contained various documents that Goldman Sachs transaction managers reviewed to check the work of due diligence vendors. Id. Goldman Sachs expanded the recommended sampling rate for valuation diligence of prime and Alt-A loans from 5-15% to 5-25% in 2006. Stmt. ¶ 610. Goldman Sachs also introduced new tools and additional personnel to aid the detection of potential fraud. See Stmt. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 23 of 73 -16- ¶¶ 611-612. By December 2006, Goldman Sachs required new borrower credit scores for all loans in a pool. Stmt. ¶ 613. Goldman Sachs compared those scores to the scores already in the loan file to search for possible identity fraud, and included loans with a credit score variance greater than 80 points in the credit and compliance diligence sample. Id. In March 2007, Goldman Sachs established a separate fraud investigation team. Stmt. ¶ 614. D. E-mails and Witness Testimony Show that Goldman Sachs’ Diligence Was Stringent. Contemporaneous e-mails, which FHFA largely ignores, show that Goldman Sachs’ due diligence process was even more stringent than its competitors’ processes. In July 2006, a Goldman Sachs employee reported that “[c]ustomers expressed concern over Goldman’s perceived tougher stance on appraisals, which have resulted in a higher [percentage] of kicks than seen from [Goldman Sachs’] competitors.” Stmt. ¶ 617. In January 2007, a Goldman Sachs employee reported frustration from National City regarding higher drops and a more stringent due diligence process from Goldman Sachs as compared to its competition. Stmt. ¶ 618. In February 2007, Goldman Sachs employees reported that “[l]arge sellers[, as well as Conduit sellers were] penalizing [Goldman Sachs] for [its] due diligence kicks” by increasing the purchase price for loans. Stmt. ¶ 619. Goldman Sachs traders also testified that Goldman Sachs received complaints from originators that Goldman Sachs’ due diligence process was more stringent than those of its competitors. See Stmt. ¶¶ 547-548. Goldman Sachs’ due diligence was considered more onerous because it “tended to have lower pull through rates than some other banks that were buying loans” and a higher “absolute amount of due diligence that was performed.” Stmt. ¶ 548. Goldman Sachs witnesses—many of whom no longer work at Goldman Sachs—consistently reaffirmed their belief in the reasonableness of Goldman Sachs’ due Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 24 of 73 -17- diligence processes. Chris Gething, who oversaw the Due Diligence Group, testified that Goldman Sachs’ “very robust process” always started with a counterparty review because Goldman Sachs had to be “comfortable with [the originator’s] standard of care in all of those departments before [Goldman Sachs would] undertake any bid.” Stmt. ¶ 542. Mr. Gething also explained that transaction managers would “bore in” to potentially problematic loans and “keep pulling until [they were] comfortable” that there were no systematic issues. Stmt. ¶ 543. Linda Peterson testified that Goldman Sachs would drop loans with unresolved questions or concerns, even if the loan technically complied with underwriting guidelines. Stmt. ¶ 544. Bill Shuey, head of Goldman Sachs’ valuation diligence, testified that even when an AVM was within tolerance, Goldman Sachs would sometimes order BPOs on loans flagged as potentially risky. Stmt. ¶ 545. Mr. Shuey testified that although sellers “expressed their opinion that they thought our valuation process was too stringent or the number of drops potentially high,” Goldman Sachs “had continual conversations with AVM vendors” to identify potential improvements and achieve “comfort in the results.” Stmt. ¶¶ 488, 546. Non-party witnesses familiar with Goldman Sachs’ due diligence processes also consistently testified that Goldman Sachs took the due diligence process seriously, and actively engaged with third-party diligence vendors to ensure high quality and reliable results. They characterized Goldman Sachs’ diligence as “hands on” and providing regular feedback, and testified that Goldman Sachs was “knowledgeable, serious about [diligence] and did a good job checking [the third-party vendor’s] processes and work.” See Stmt. ¶¶ 622-623; see also Stmt. ¶ 626. They testified that Goldman Sachs’ due diligence processes were “rigorous” and its “requirements [we]re generally higher than other clients we worked with at the time.” See Stmt. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 25 of 73 -18- ¶¶ 624-625. For example, Joe Murray, the Clayton client service manager for Goldman Sachs (and currently a Freddie Mac employee), testified that Goldman Sachs was an “extremely” demanding client and “was very critical of the work we produced, in that they scrutinized every report we sent to them. They questioned every anomaly and expected us to report valid and accurate results.” Stmt. ¶ 626. Goldman Sachs had “no tolerance for errors.” Id. E. Goldman Sachs Did Not Compromise Its Due Diligence Standards in Response to Originator Feedback. Despite FHFA’s claims to the contrary, see Pl. Br. at 16-19, 55-56, all of which create disputed issues of fact, the leadership at Goldman Sachs consistently refused to compromise its due diligence standards, irrespective of whether that stance placed Goldman Sachs at a competitive disadvantage in bidding for mortgage loans. During e-mail correspondence in February 2007, Daniel Sparks, the head of the Mortgage Department, wrote: “I don’t see us reducing DD now.” Stmt. ¶ 629. Michelle Gill, a Managing Director who headed the Residential Whole Loan Finance desk, wrote that “it does NOT mean that we should relax our standards.” Stmt. ¶ 633. In June 2007, Chris Gething, head of the St. Petersburg office, which conducted due diligence activities, wrote “[w]e need the client contact side to understand and support the fact that we aren’t going to alter our standards based upon the noise level from the client.” Stmt. ¶ 630. Kevin Gasvoda, head of the mortgage loan trading desk, wrote: “we’re doing better fraud and property valuation due diligence than ever . . . we don’t want to drop standards and compromise on these important things as a precedent.” Stmt. ¶ 631. II. GOLDMAN SACHS WORKED CLOSELY WITH HIGH QUALITY THIRD- PARTY VENDORS TO CONDUCT DUE DILIGENCE. Goldman Sachs worked with third-party due diligence vendors to conduct due diligence on loans it was considering for purchase. See Stmt. ¶¶ 421-435. The GSEs, Goldman Sachs and countless others in the industry hired many of the same due diligence vendors to Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 26 of 73 -19- conduct credit, compliance and valuation due diligence reviews. See Stmt. ¶ 674, 677. Throughout the diligence process, the vendors provided Goldman Sachs with reports on their findings, which Goldman Sachs transaction managers reviewed on a daily basis. Stmt. ¶ 421. For credit and compliance due diligence, vendors such as Clayton included grades for the sample loans that were typically on a numerical scale of 1-3.6 Stmt. ¶ 552. Although the grades could vary by originator and even loan pool, a vendor generally would grade a loan as a “1” to indicate that the loan fully complied with the applicable underwriting guidelines as well as any Goldman Sachs overlay. Stmt. ¶ 556. A grade of “2” indicated that a loan was consistent with applicable underwriting guidelines and any supplemental criteria, such as Goldman Sachs’ overlays, because a deviation was either immaterial or offset by compensating factors. Stmt. ¶ 557. A grade of “3” indicated that the loan: (i) deviated in some way from the originator underwriting guidelines, including because of missing documents or a material exception that could be offset by compensating factors; (ii) could not be fully evaluated against the underwriting guidelines because of missing documents; or (iii) was flagged in accordance with a Goldman Sachs overlay. Stmt. ¶ 558. Vendors like Clayton also included slight variations on these numerical scales, such as “2Ws”, which indicated that the loan was initially graded a “3”, but the exception was waived by Goldman Sachs after further analysis. Stmt. ¶ 559. Like the GSEs, Goldman Sachs reviewed the due diligence vendors’ findings and applied compensating factors, including compensating factors beyond those identified by the vendor. Stmt. ¶ 565. In addition to relying on the strong professional reputation of the vendors it hired, Goldman Sachs generally evaluated each vendor prior to engaging them, which included 6 Though vendors typically applied the 1-3 grading scale, other vendors such as MDMC utilized a differing metric. MDMC often categorized loans as “Meets/Acceptable,” “Fails/Acceptable,” and “Fails/Unacceptable,” which corresponded to the 1-3 scale. Stmt. ¶ 299. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 27 of 73 -20- interviews with senior management, a review of the resumes of key personnel, a review of the vendor’s data and reporting capabilities, and discussion of Goldman Sachs specific recommendations and requests. Stmt. ¶¶ 427-428. After retaining the vendor, Goldman Sachs continued to review the vendor on an ongoing basis “to ensure conformity with [Goldman Sachs’] standards and philosophy.” Stmt. ¶¶ 429-432. Freddie Mac noted in its operational review of Goldman Sachs in 2006 that the firm’s “counter party/vendor approval process” was among “the solid controls that Goldman Sachs ha[d] in place.” Stmt. ¶ 433. The third-party vendors that conducted diligence on the loans at issue did their jobs well and needed to ensure that they provided reliable results to get repeat business. For example, Doug Lackey, founder of diligence firm MDMC, testified that MDMC was “a company that was in it for the long haul[.] [W]e wanted to be around, we were trying to grow the company, make it a mainstay and use proper personnel training and systems to make sure that we delivered a proper product.” Stmt. ¶ 549. The third-party vendors also had robust quality control processes to ensure the reliability of the diligence provided to Goldman Sachs. For example, MDMC had four layers of review for each loan—including those graded as a “1” or “2”, contrary to FHFA’s assertions, Pl. Br. at 28—before the due diligence results went to the client. First the loan was reviewed by an underwriter, then by a quality control person, then by the lead underwriter, and finally by the MDMC transaction manager assigned to the deal. Stmt. ¶¶ 585-586. Clayton also utilized a thorough quality control process, which subjected all loans—including those graded a “1” or “2”—to a quality control process that involved both system checks as well as manual checks by more experienced underwriters. Stmt. ¶ 588; see also Stmt. ¶ 588. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 28 of 73 -21- Goldman Sachs worked closely with third-party diligence vendors to ensure they met Goldman Sachs’ high standards. Stmt. ¶¶ 589-591. For example, in May 2006, Brian O’Brien, Goldman Sachs’ Chief Underwriter, coordinated with Clayton to review the standard diligence scripts Clayton was applying to Goldman Sachs’ deals “to make sure [Clayton is] capturing everything Goldman requires.” Stmt. ¶ 592. In July 2006, Linda Peterson and Joe Murray again reviewed and discussed the standard Goldman Sachs due diligence scripts, and explored whether a modified system might be more effective for Goldman Sachs’ data analysis. Id. In an e-mail to Peterson, Murray indicated that Goldman Sachs and Clayton often discussed improvements to Clayton’s diligence process: “I have discussed possible reworking, rebuilding or ‘tweaking’ the system on several occasions with Brian, yourself and Lauren to ensure Clayton is providing the absolute best quality and service we possibly can.” Id. Third-party due diligence vendors confirmed Goldman Sachs’ continuous engagement and involvement in the due diligence process. See generally Stmt. ¶ 591. For example, Doug Lackey, founder of MDMC, testified that several key Goldman Sachs employees met with MDMC to make sure Goldman Sachs understood exactly how MDMC’s due diligence process worked, and followed up with questions to ensure the due diligence conducted for Goldman Sachs was up to Goldman Sachs’ high standards. Id. The 30(b)(6) witness for JCIII & Associates testified that Goldman Sachs regularly provided feedback and asked questions when JCIII provided Goldman Sachs with exception reports for the loans on which it was providing due diligence. Stmt. ¶¶ 591, 622. Goldman Sachs also raised loan- or deal-specific issues with due diligence vendors on a day-to-day basis. In July 2006, Goldman Sachs transaction manager Joseph Ozment reviewed Clayton’s daily reports, and e-mailed Clayton after observing a compliance Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 29 of 73 -22- issue to remind Clayton that such issues should be escalated to Goldman Sachs immediately. Stmt. ¶ 601. As another example, in January 2007, Goldman Sachs transaction manager Tanner Donnizetti contacted Clayton regarding errors an underwriter made with regard to a loan that had been flagged under an AppIntell fraud report, and emphasized the issues Goldman Sachs expected Clayton to raise with respect to such loans. Stmt. ¶ 603. Goldman Sachs also frequently went on-site to work directly with the diligence vendors’ employees, and imposed strict requirements for which third-party diligence employees could work on Goldman Sachs reviews. Stmt. ¶ 605. These requirements included minimum years of experience for both underwriters and quality control personnel, as well as a review by Goldman Sachs of the resumes and underwriting test scores for third-party staff. Stmt. ¶ 430. The efficacy of the due diligence performed by Goldman Sachs and its third-party vendors is shown by the expert report of Zoe Goss, who found that the percent of sample loans underwritten by FHFA’s expert that potentially contained material defects was only 3.63%. Stmt. ¶ 627.7 Contrary to FHFA’s assertions, see Pl. Br. at 24-29, 53-54—which at best create 7 The high defect rate asserted by FHFA’s expert is demonstrably inflated. Almost half of his asserted defects rely on the assumption that missing documents in currently available loan files necessarily indicate a defect, despite the fact that record evidence establishes that missing documents typically could be, and were, found. FHFA’s expert bases a large proportion of his purported defects on supposed “minimum industry standards” that the record has shown did not exist. Many of his other defects result from his own errors, such as the misapplication of a guideline or failure to find a document in a loan file. See generally Report of Zoe Goss, dated June 3, 2014. The much lower defect rates found in contemporaneous diligence reviews by Goldman Sachs and the GSEs contradict FHFA’s made-for-litigation defect rates. For example, Doug Lackey testified that the numbers alleged regarding MDMC’s due diligence of loans at issue for Goldman Sachs were “unrealistic” because the 80% defect rate alleged by FHFA “would be an extremely high amount of differentiation between what [MDMC] found after going through all those QC checks and reviews.” Stmt. ¶ 628. Carolyn Bertelson, the 30(b)(6) witness for DHI Mortgage Company, expressed similar astonishment, testifying that “[a]n 83 percent defect rate is enormous. . . I can’t imagine anybody would have [such a] defect rate.” Id. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 30 of 73 -23- material issues of fact—Goldman Sachs scrutinized the third-party vendors it utilized to conduct due diligence, and worked closely with the third-party vendors throughout the diligence process. III. GOLDMAN SACHS HAD STRONG INCENTIVES TO CONDUCT ROBUST DUE DILIGENCE. That Goldman Sachs had ample incentives to conduct thorough due diligence creates further factual issues as to the reasonableness of Goldman Sachs’ due diligence, which is defined as that required of a prudent man in the management of his own property. See 15 U.S.C. § 77k(c); In re WorldCom, 346 F. Supp. 2d at 664. First, Goldman Sachs retained first-loss positions in many of the securitizations it sold to the GSEs, meaning Goldman Sachs would be the first one to lose money if the loans did not perform well. See Stmt. ¶¶ 400-401. By the end of March 2007, Goldman Sachs estimated its residual positions across the mortgage whole loan credit desks at $751 million. Stmt. ¶ 401. On the basis of its due diligence, Goldman Sachs expected those loans to perform well enough to yield a return. Stmt. ¶ 402. Second, reputational interests provided incentives for Goldman Sachs to conduct thorough and robust due diligence. See, e.g., Stmt. ¶ 404. In a counterparty review of Goldman Sachs, Freddie Mac acknowledged that Goldman Sachs “[v]iews reputational risk as the most significant of all risks associated with this line of business.” Stmt. ¶ 405. IV. GOLDMAN SACHS MAINTAINED A REASONABLE PROCESS TO PRODUCE MATERIALLY ACCURATE OFFERING MATERIALS. FHFA’s claims that Goldman Sachs, including its Mortgage Finance Group, failed to conduct a sufficient review of the Offering Materials, see Pl. Br. at 10-11, 46-49, are at odds with the record and, at the very least, presents an issue of material fact that the jury must resolve. The Residential Mortgage Finance Group had “oversight responsibility for the offering materials at large” to ensure the documents were “materially accurate.” Stmt. ¶ 386. A “deal captain” was assigned within the Residential Mortgage Finance Group for each deal to ensure Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 31 of 73 -24- that the appropriate teams and product areas undertook incremental reviews of specific sections in the Offering Materials. Stmt. ¶ 387. The Residential Mortgage Finance Group worked with various teams both within and outside of Goldman Sachs. The collateral team reviewed the information provided about the collateral for the securities. Among other things, the team would “tie out” all the numbers related to the collateral information. Stmt. ¶ 388. The Residential Mortgage Finance Group would work with the structuring team to ensure that the structuring disclosures were accurate, and with external accountants to confirm the accuracy of the numerical information provided about the collateral. Stmt. ¶¶ 389, 390. This careful oversight by the Residential Mortgage Finance Group at a minimum raises additional disputed issues of fact regarding the reasonableness of Goldman Sachs’ due diligence efforts in ensuring the accuracy of the Offering Documents. In addition, the Residential Mortgage Finance Group would coordinate with various groups regarding descriptions in the Offering Materials of underwriting guidelines and the collateral underlying the securitization. Stmt. ¶ 391. The other parties involved in the transaction would also review sections of the prospectus supplement, and there were due diligence calls with the parties involved and legal counsel. Stmt. ¶ 391. The Mortgage Finance Group also reviewed the risk factors associated with a given transaction to, among other things, “highlight or make outside Counsel aware” of specific risks particular to a transaction. Stmt. ¶ 391. Both internal and external counsel participated in the preparation and review of offering materials. Stmt. ¶ 392. ARGUMENT Summary judgment is appropriate only where “there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a). “The party seeking summary judgment bears the burden of establishing that no genuine Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 32 of 73 -25- issue of material fact exists.” Vivenzio v. City of Syracuse, 611 F.3d 98, 106 (2d Cir. 2010) (citation omitted). Where, as here, the plaintiff seeks summary judgment on an affirmative defense, it must show a “complete failure of proof concerning an essential element of the [defense].” Brody v. Vill. of Port Chester, No. 00 Civ. 7481, 2007 WL 735022, at *5 n.12 (S.D.N.Y. Mar. 12, 2007) (quoting FDIC v. Giammettei, 34 F.3d 51, 54 (2d Cir. 1994)). In considering a motion for summary judgment, a district court may not “resolve disputed questions of fact but only [] determine whether, as to any material issue, a genuine factual dispute exists.” Ideal Steel Supply Corp. v. Anza, 652 F.3d 310, 326 (2d Cir. 2011). “In making this determination, the court must view all facts in the light most favorable to the non- moving party.” In re WorldCom, 346 F. Supp. 2d at 655 (citations omitted). In conducting its analysis, the court “may not make credibility determinations or weigh the evidence,” but instead “must draw all reasonable inferences in favor of the nonmoving party.” Ideal Steel Supply Corp., 652 F.3d at 326 (emphasis in original). “If there is any evidence in the record from any source from which a reasonable inference could be drawn in favor of the nonmoving party, summary judgment is improper.” Vivenzio, 611 F.3d at 106. FHFA does not come close to meeting its heavy burden of demonstrating a “complete failure of proof” with respect to Goldman Sachs’ due diligence and reasonable care defenses. FHFA’s 65-page brief and accompanying 239-page Rule 56.1 statement of purportedly “undisputed facts” make clear that there are innumerable important, disputed issues of fact for the jury to decide, and credibility determinations for the jury to make, concerning these defenses. On this record, the Court cannot properly determine on summary judgment that Goldman Sachs’ due diligence was unreasonable as a matter of law. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 33 of 73 -26- I. TO PROVE ITS REASONABLE CARE AND DUE DILIGENCE DEFENSES UNDER THE SECURITIES LAWS, GOLDMAN SACHS MUST SHOW IT EXERCISED REASONABLE CARE AND CONDUCTED A REASONABLE INVESTIGATION, RESPECTIVELY. A. The Reasonableness of the Care or Investigation Undertaken Pursuant to Sections 11 and 12 and the Blue Sky Laws Is a Question of Fact. The statutory defenses available to various Goldman Sachs entities under Sections 11 and 12 and the blue sky laws turn on whether the care or investigation undertaken was “reasonable.” “For decades, courts have recognized that what constitutes both a reasonable investigation and a reasonable belief in the accuracy of a registration statement hinges on a fact- intensive inquiry . . . .” In re WorldCom, Inc., 2005 WL 638268, at *8; see also Feinberg v. Katz, No. 01-cv-2739 (CSH), 2007 WL 4562930, *11 (S.D.N.Y. Dec. 21, 2007) (“Whether a party acted with objective reasonableness is a quintessential common law jury question.”) (quoting Kidder, Peabody & Co. v. IAG Int’l Acceptance Grp. N.V., 14 F. Supp. 2d 391, 398 (S.D.N.Y. 1998)); Quincy Co-op. Bank v. A.G. Edwards & Sons, Inc., 655 F. Supp. 78, 87 (D. Mass. 1986) (“Ordinarily, questions of reasonable care cannot be decided on a motion for summary judgment; they are, by their nature, fact-bound.”). The standard applicable to the due diligence defenses under Section 12 and the blue sky laws8 is different from the standard under Section 11. Section 12 of the Securities Act and the blue sky laws provide a “reasonable care defense.” In re Fuwei Films Sec. Litig., 634 F. Supp. 2d 419, 435 n.10 (S.D.N.Y. 2009) (citations omitted) (emphasis added); In re WorldCom, 346 F. Supp. 2d at 663. A seller is liable under Section 12 and blue sky laws only if it cannot show that it “did not know, and in the exercise of reasonable care could not have known, of” an 8 Blue sky laws should be interpreted in accordance with Section 12(a)(2). See Fed. Hous. Fin. Agency v. Bank of Am. Corp., No. 11 Civ. 6195 (DLC), 2012 WL 6592251, at *7 n.8 (S.D.N.Y. Dec. 18, 2012) (“[T]he D.C. and Virginia securities laws are generally interpreted in accordance with Section 12(a)(2).”) (citations omitted). Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 34 of 73 -27- untrue statement or omission of a material fact in the means used to offer or sell a security. Section 11 provides a distinct “reasonable investigation” standard under which a non-issuer must prove: (1) he conducted a reasonable investigation; (2) after that investigation and at the time the alleged misstatement in the registration statement became effective, he believed that the statements therein were true and there was no omission of material fact required to be stated in the registration statement or necessary to make the statements therein not misleading; and (3) he had reasonable ground for such a belief. The standard of reasonableness is that required of a prudent man in the management of his own property. 15 U.S.C. § 77k(c); In re WorldCom, 346 F. Supp. 2d at 664. Thus, due diligence is “[i]n effect . . . a negligence standard.” Id. at 662. FHFA attempts to conflate the Section 12 “reasonable care” requirement with the higher Section 11 “reasonable investigation” standard by arguing—in contravention of the plain language of the statute—that the analyses for the two standards typically are the same. Pl. Br. at 59. That is incorrect. As this Court has recognized, it is well established that the Section 12 reasonable care defense is “less demanding than the duty of due diligence imposed under Section 11.”9 In re WorldCom, 346 F. Supp. 2d at 663; see also SEC Release No. 8591, 2005 WL 1692642, at *79 (Aug. 3, 2005). None of the cases FHFA cites in support of its proposed reading negates this settled principle. Most glaringly, In re MetLife Demutualization Litig., 262 F.R.D. 217, 235 (E.D.N.Y. 2009), see Pl. Br. at 59, confirms that, while some courts “likened” 9 The SEC has stated that the standard of care under Section 12(a)(2) is less demanding than that prescribed by Section 11, or put another way, that Section 11 requires more diligent investigation than Section 12(a)(2). SEC Release No. 8591, 2005 WL 1692642, at *79 (Aug. 3, 2005). The SEC filed a brief to this effect in the John Nuveen & Co. cases in the Seventh Circuit. In his dissent from the Supreme Court’s denial of certiorari in the Nuveen case, Justice Powell, joined by Justice Rehnquist, noted that “Congress thus used different language for different situations. ‘Reasonable investigation’ is required for registered offerings under § 11, but nothing more than ‘mer[e] . . . ‘reasonable care’ is required by § 12(2). The difference in language is significant, because in the securities acts Congress has used its words with precision.” John Nuveen & Co., 450 U.S. 450 U.S. 1005, 1008-11 (1981) (internal citation omitted). Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 35 of 73 -28- the Section 11 reasonable investigation and Section 12 reasonable care standards, the “‘reasonable care’ standard is lower than the ‘reasonable investigation’ standard.” University Hill Foundation v. Goldman, Sachs & Co., 422 F. Supp. 879, 898, 900-01 (S.D.N.Y. 1976), did not, as FHFA implies, apply Section 11’s “reasonable investigation” analysis to Section 12’s “reasonable care” defense. Indeed, University Hill did not apply the due diligence defense; instead, it principally discussed whether the defendant’s representation that it had “made a reasonable investigation” of the security was a misrepresentation under Section 12.10 Notably, FHFA’s arguments for imposing a heightened standard under Section 11 on underwriter defendants who are affiliated with an issuer are inapplicable in the context of the “reasonable care” defenses. As discussed below, FHFA claims that when underwriters are affiliated with issuers, they are subject to “nearly absolute” liability that “approaches that of the issuer,” which is not afforded a due diligence defense under Section 11 at all. Pl. Br. at 45. This argument is inapposite with respect to Section 12 and the blue sky laws because those statutes make a reasonable care defense available to all “sellers” of a security including issuers.11 In re Fuwei Films Sec. Litig., 634 F. Supp. 2d at 435 n.10. In other words, FHFA’s claim that 10 Similarly, the court in Franklin Savings Bank of N.Y. v. Levy, 551 F.2d 521, 527 (2d Cir. 1977), recognized that “reasonable care” was the appropriate standard under Section 12, and never compared “reasonable care” to the Section 11 standard, or implied the two were the same. In re Software Toolworks Inc., 50 F.3d at 615, relied on authorities that recognized that the duties of due diligence under Sections 11 and 12 are not identical. See, e.g., Weinberger v. Jackson, No. C-89-2301-CAL, 1990 WL 260676, at *2 (N.D. Cal. Oct. 11, 1990) (quoting Sanders v. John Nuveen & Co., Inc., 619 F.2d 1222 (7th Cir. 1980)). 11 GSMSC, the issuer-depositor, cannot be liable under Section 12 or the blue sky laws, as liability “extend[s] only to the ‘immediate sellers’ of securities and ‘those who solicit purchasers to serve their own financial interests or those of the securities owner.’” In re Countrywide Fin. Corp. Mortg.-Backed Sec. Litig., 932 F. Supp. 2d 1095, 1118 (C.D. Cal. 2013). GSMSC, the issuer-depositor, “did not pass title directly to Fannie Mae or Freddie Mac, because the title passed to the [u]nderwriter, [Goldman, Sachs & Co.] before passing to the ultimate purchasers.” Id. Thus, GSMSC “cannot be liable under Section 12(a)(2).” Id. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 36 of 73 -29- “independence from an issuer” is a “core requirement,” Pl. Br. at 43, cannot apply to the reasonable care defenses because those defenses are available to issuers themselves. B. There Is No Heightened Standard for a Section 11 Due Diligence Defense For Affiliated Underwriters in an Asset-Backed Securities Transaction. FHFA’s arguments for applying “near absolute” liability to underwriters affiliated with issuers under Section 11 have no basis in the statutory text, are particularly inappropriate in the context of asset-backed securitizations, and at most raise genuine issues of fact as to what constitutes a “reasonable investigation” under the circumstances. For the 36 Securitizations in which Goldman Sachs affiliates served as both issuer and underwriter, FHFA argues that there is no Section 11 due diligence defense because underwriters affiliated with the issuer “will almost never be able to meet their burden of proof for a due diligence defense” under Section 11. Pl. Br. at 44. FHFA thus attempts to circumvent the separate, rigorous, fact-intensive analysis required to pierce the corporate veil—and similarly attempts to get around the factual reality of the high quality of Goldman Sachs’ due diligence—by inventing a “heightened standard” for Section 11 liability that neither the SEC, Congress, nor any court has recognized in the context of asset- backed securities.12 See FDIC v. Chase Mortgage Finance Corp., No. 12 Civ. 6166 (LLS), 2013 WL 5434633, at *9 (S.D.N.Y. Sept. 27, 2013) (stating that all underwriter defendants—including affiliated underwriters—were entitled a due diligence defense under 15 U.S.C. § 77k(b)); FHFA v. UBS Amer., Inc., 858 F. Supp. 2d 306, 329 (“Section 11 provides an affirmative defense of ‘due diligence’ that is available to defendants other than the issuer of the security.”). FHFA’s attempt to preclude Goldman Sachs from using its due diligence to defend itself is particularly 12 Because FHFA does not attempt to pierce the veil, no detailed discussion is undertaken here. But, as this Court knows, a party “asking the court to disregard the corporate structure faces ‘a difficult task’” that is “answered by the factfinder after trial.” NetJets Aviation, Inc. v. LHC Commc’ns, LLC, 537 F.3d 168, 176, 184 (2d Cir. 2008). Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 37 of 73 -30- unjustified and ironic given FHFA’s offensive reliance on that same due diligence to attempt to establish scienter for its fraud claims. See Am. Compl. ¶¶ 194-207, Case No. 11-civ-6198. Section 11 contains no requirement that the underwriter be unaffiliated with the depositor-issuer in order to raise a due diligence defense. As this Court stated in In re WorldCom: “In the Securities Acts Congress has used its words with precision,” so in fulfilling the Court’s role “to interpret the language of the statute enacted by Congress,” the Court must begin such interpretation “with the language employed by Congress . . . .” 346 F. Supp. 2d at 677-78 (internal citations omitted). Section 11(b) explicitly provides that the issuer cannot rely on a due diligence defense, but nowhere provides that entities affiliated with the issuer also cannot rely on that defense. See 15 U.S.C. § 77k(b) (emphasis added). Instead, directors and officers signing the registration statement—all of whom are “affiliates” of the issuer—do get the benefit of the due diligence defense. The statute and regulations promulgated by the SEC carefully define the “issuer” in an RMBS transaction, and that definition does not include underwriters or other affiliates of the sponsor.13 See 15 U.S.C. § 77b(a)(4 and 11); 17 C.F.R. § 230.191; 17 C.F.R. § 240.3b-19. Even where a plaintiff has alleged fraud, the Supreme Court has recognized that the plain text of the securities laws creates meaningful differences in liability regardless of whether corporate entities are affiliated. In Janus Capital Grp., Inc. v. First Derivative Traders, 131 S. Ct. 2296, 2304 (2011), the Supreme Court held that as long as “corporate formalities were 13 In promulgating rules regulating RMBS transactions within the existing framework of the securities acts, the SEC has recognized that affiliations between depositor, sponsor and underwriter are common. See SEC Release Nos. 33-8518, 34-50905 (Jan. 7, 2005) at 1550. Although Regulation AB requires disclosure of these affiliations, it does not bar them or change its liability scheme in light of them. See id. at 1550-1551. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 38 of 73 -31- observed,” the apportionment of liability among participants in a financial product should remain as explicitly dictated by Congress: For its part, [plaintiff] suggests that the “well-recognized and uniquely close relationship between a mutual fund and its investment adviser” should inform our decision. It suggests that an investment adviser should generally be understood to be the “maker” of statements by its client mutual fund, like a playwright whose lines are delivered by an actor. We decline this invitation to disregard the corporate form. Although [plaintiff] and its amici persuasively argue that investment advisers exercise significant influence over their client funds, it is undisputed that the corporate formalities were observed here. JCM and Janus Investment Fund remain legally separate entities, and Janus Investment Fund's board of trustees was more independent than the statute requires. Any reapportionment of liability in the securities industry in light of the close relationship between investment advisers and mutual funds is properly the responsibility of Congress and not the courts. Id. (citations removed) (emphasis added). As FHFA acknowledges, Pl. Br. at 44, the SEC has spelled out the “relevant circumstances” to “determin[e] whether or not the conduct of a person constitutes [due diligence],” and none of those circumstances suggest that an underwriter’s affiliation with an issuer creates “near absolute” liability. According to the SEC, relevant considerations include, inter alia: (i) the type of issuer; (ii) the type of security; (iii) when the person is an officer, “[t]he office held”; and (iv) when the person is an underwriter, “the type of underwriting arrangement, the role of the particular person as an underwriter and the availability of information with respect to the registrant.” 17 C.F.R. § 230.176(a)-(g). Nowhere does the SEC suggest that any of these facts would automatically elevate a person to the status and liability of an issuer. Rather, the question of whether the affiliations of any Goldman Sachs entity required a heightened degree of due diligence depends on an assessment of what is a reasonable investigation given the “type of issuer,” “type of security,” and “type of underwriting arrangement,” among other factors. Id. These are questions of fact for the jury. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 39 of 73 -32- Such a heightened standard would be particularly inappropriate in the context of asset-backed securitizations. The SEC has recognized that “[a]sset-backed securities and ABS issuers differ from corporate securities and operating companies.” SEC Release Nos. 33-8518, 34-50905 at 1508, 1510-11 (Jan. 7, 2005). In particular, In offering ABS, there is generally no business or management to describe. Instead, information about the transaction structure and the characteristics and quality of the asset pool and servicing is often what is most important to investors. . . . There are several distinguishing features between asset- backed securities and other fixed-income securities. For example, ABS investors are generally interested in the characteristics and quality of the underlying assets, the standards for their servicing, the timing and receipt of cash flows from those assets and the structure for distribution of those cash flows. As a general matter, there is essentially no business or management (and therefore no management’s discussion and analysis of financial performance and condition) of the issuing entity, which is designed to be a solely passive entity. GAAP financial information about the issuing entity generally does not provide useful information to investors. Information regarding characteristics and quality of the assets is important for investors in assessing how a pool will perform.” Id. Thus, unlike investors in corporate securities, investors in asset-backed securities do not rely on the credit or operations of the “passive” issuer, here the depositor, but instead look solely to the credit risk and cash-flow of the underlying assets. The financial statements or business plans of the depositor-issuer play no role in the investor’s decision to purchase RMBS; the relevant representations concern the assets, not the issuer. See Harold S. Bloomenthal & Samuel Wolff, Chapter 13A: Asset-Backed Securities, the Financial Crisis and the Dodd-Frank Act, Going Public and the Public Corporation § 13A:1 (2014). Further, the depositor-issuer plays a very limited role in enhancing the quality of the certificates post-securitization.14 See 17 14 See Gregory A. Markel & Gregory G. Ballard, Defending Claims Against Underwriters of Asset-Backed Securities: The Due Diligence Defense, Andrews Sec. Litig. and Reg. Rep. (2006) (“[M]atters such as finances, management and operations are not relevant in the ABS context, Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 40 of 73 -33- C.F.R. 229.1101(c)(2)(ii) (requiring that the entity issuing the asset-backed securities limit its activities to passive ownership or holding the pool of assets underlying the asset-backed securities and issuing the asset-backed securities). Indeed, the authorities on which FHFA relies confirm that, under the “[c]urrent legal regime under Section 11,” if “the underwriting function . . . is legally attributed to an affiliated entity separate from the arranger [sponsor] entity (or, where applicable, the intermediary [depositor] being considered the issuer) . . . the entity performing the underwriting function [] will be granted a due diligence defense under Section 11(b).” Merritt B. Fox, Due Diligence with Residential Mortgage Backed Securities 57 (Columbia Law & Econ. Working Paper No. 462, 2013) (“Fox Paper”) at 58.15 Here, the GSEs knew the parties in the transaction from the face of the Offering Materials as well as their affiliations. See, e.g., Stmt. ¶ 381. The differences between an asset-backed securitization and the typical corporate securitization require a different factual analysis of the “reasonable investigation” required of underwriters in each circumstance. See 17 C.F.R. § 230.176(a)-(g). As a result, Feit and the other cases on which FHFA relies are not applicable because they involve traditional corporate offerings in which underwriters reviewed representations about the activities and characteristics of an issuer which was an operating company whose performance was the source of financial return on the securities. Feit, for example, concerned a prospectus for an exchange offer issued by Leasco Data Processing Equipment Corporation (“Leasco”) in which Leasco failed to disclose the amount of liquid assets in the target company, even though Leasco’s primary motive for the offer was to acquire those where the issuer is a passive entity that has no ongoing business other than to hold financial assets that generate cash flows.”). 15 FHFA mischaracterizes Professor Fox’s thesis by citing his article for the proposition that where a defendant is heavily intertwined with the issuer, there should be strict liability. Pl. Br. at 45 n.29. What Professor Fox recommends is strict liability when the defendant is involved in all aspects of the securitization chain, including origination. Fox Paper at 56-57. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 41 of 73 -34- funds, and disclosure may have shown that the target company’s shares were “significantly undervalued.” Feit, 332 F. Supp. at 572-75. Here, the depositor (deemed the issuer under SEC regulations) purchases loans, sells them to the securitization trust, receives certificates and sells them to the underwriter. All the relevant representations are about the collateral backing the certificates and the terms of the certificates. In addition, the court in Feit also does not state that an underwriter must prove independence from the issuer to assert a due diligence defense. Rather, Feit explained that “[an underwriter] must make an investigation reasonably calculated to reveal all of those facts which would be of interest to a reasonably prudent investor. If he undertakes such an investigation, he will not be liable for material misrepresentations which his efforts did not uncover.” Feit, 332 F. Supp. at 581-82. Thus, the key for a due diligence defense under Section 11 is that the underwriter make an investigation reasonably calculated to reveal all facts of interest to a reasonably prudent investor—not a lack of affiliation between the underwriter and the issuer.16 Further, Feit held that, for inside directors of a corporate issuer, “liability will lie in practically all cases of misrepresentation” not due to any formal affiliation, but because of their “intimate knowledge of corporate affairs and of the particular transactions.” Id. at 578; Pl. Br. at 46. 16 The Court in Feit states that “a completely independent and duplicative investigation is not required, but, rather, that the defendants were expected to examine those documents which were readily available,” and not merely rely on “the oral word of management.” Feit, 332 F. Supp. at 576-77. This Court never held, as FHFA claims in its Brief, that “the underwriter in Feit was able to ‘barely’ meet its due diligence defense only because of ‘the underwriter’s independence from the issuer.’” Pl. Br. at 44 (citing In re Worldcom, 346 F. Supp. 2d at 675). Instead, this Court stated in Worldcom that “[g]iven the underwriter’s independence from the issuer” the Feit Court “affirmed that [] an underwriter is not ‘expected to possess the intimate knowledge of corporate affairs of inside directors,’” and “found that the Feit underwriters had ‘just barely’ satisfied th[e] standard [for an underwriter verifying the financials of a company] by completing a ‘thorough review of all available financial data . . .’” In re WorldCom, 346 F. Supp. 2d at 675 (emphasis added). Reliance on the oral word of management is vastly different from the due diligence Goldman Sachs conducted here. See generally supra Fact Secs. I-IV. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 42 of 73 -35- Thus, the cases FHFA cites in support of its invented standard merely underscore that the reasonableness of due diligence is a fundamentally fact-intensive inquiry that must be considered in light of the circumstances, even where it is clear that a defendant shares close knowledge with an issuer, e.g., an inside director. Here, as discussed above, the relevant representations concern assets rather than an issuer, there is no “business or management to describe,” and the issuer is “designed to be a solely passive entity.” SEC Release Nos. 33-8518, 34-50905 at 1508, 1510-11 (Jan. 7, 2005). These are among the many factual circumstances a jury should consider in determining what constitutes a “reasonable investigation” for the 36 Securitizations in which Goldman Sachs affiliates served as both issuer and underwriter. II. GOLDMAN SACHS’ DUE DILIGENCE PROVIDED A REASONABLE BASIS FOR ITS DISCLOSURES, WHICH GOLDMAN SACHS VERIFIED FOR MATERIAL ACCURACY. A. Whether Due Diligence at Time of Whole Loan Purchase Provided Reasonable Grounds for the Statements in the Offering Materials Is a Material Issue of Fact for the Jury. FHFA argues that the Court should find, on summary judgment, that Goldman Sachs’ due diligence was legally inadequate because Goldman Sachs primarily conducted due diligence before buying the pools containing loans in the SLGs. Pl. Br. at 8-10, 47-49. FHFA asks this Court to impose—in hindsight and contrary to widespread industry practice—a legal requirement that RMBS underwriters conduct a second round of duplicative due diligence in all cases and despite the lack of evidence that material misrepresentations or omissions not existing at the time of purchase would have existed as of the “cut off” date. See Pl. Br. at 47-49. Goldman Sachs’ acquisition-stage due diligence assessed exactly the attributes of the mortgage loans about which Goldman Sachs later made representations: whether the loans generally were originated in accordance with underwriting guidelines. There is no evidence that the GSEs, in their regular and thorough review of underwriters’ due diligence practices (including the Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 43 of 73 -36- practices of Goldman Sachs), asked whether any underwriter conducted this duplicative second round of diligence, or required it. FHFA also fails to identify any materially adverse changes in any loans in the SLGs between the time of acquisition and the cut-off date prior to securitization. At the very least, it is a question of fact as to whether reasonable care or investigation required not only a full loan-level due diligence review at the time of whole loan pool purchase, but also a second review at the time of securitization. See TSC Indus., Inc. v. Northway, Inc., 426 U.S. 438, 450 n.12 (1976) (Summary judgment is generally an inappropriate way to decide questions of reasonableness because “the jury’s unique competence in applying the ‘reasonable man’ standard is thought ordinarily to preclude summary judgment”). 1. FHFA Fails to Show That the Performance of Due Diligence at the Time of Loan Acquisition Was Unreasonable, or Inconsistent with Representations in the Offering Materials. Goldman Sachs typically conducted due diligence on pools of whole loans prior to purchase. Consistent with Goldman Sachs’ intention to hold those loans on its own books—as its own property and at its own risk—for a short period of time before securitization, this diligence assessed, among other things, the valuation of the underlying properties and compliance with underwriting guidelines. Conducting due diligence at the time of whole loan purchase was common in the industry, and wholly reasonable. See Escott v. BarChris Constr. Corp., 283 F. Supp. 643, 703 (S.D.N.Y. 1968) (with respect to Section 11, compliance with industry-wide norms should ordinarily satisfy the due diligence obligation); Fed. Hous. Fin. Agency v. UBS Am. Inc., et al., No. 11 Civ. 5201 (DLC) et al., 2013 WL 3284118, at *7 (S.D.N.Y. Nov. 15, 2013) (“[T]he Court expressed its view that while due diligence was subject to an objective standard, the size of the GSEs’ role in the industry made their procedures relevant in establishing what industry practice was.”); Hr’g Tr., 70:19-71:1, Dec. 17, 2012 (11 Civ. 6198 (DLC)) (recognizing a “very strong argument to be made . . . that [the] standard upon a Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 44 of 73 -37- reasonable inquiry . . . would be informed to some extent by the practices of major player[s] in the industry”); In re Conticommodity Services, MDL No. 644, 1988 WL 56172, at *1 (N.D. Ill. 1988). The GSEs, like most sophisticated RMBS investors, were well aware of that practice. See, e.g., Stmt. ¶ 437. It was similarly common for time to elapse between pre-purchase diligence and securitization. See Stmt. ¶ 438. Goldman Sachs’ Expert Charles Grice noted in his report that, based on his experience, “the information gathered during due diligence remains valid within the typical limited time lag between [pre-purchase due diligence and securitization].” Stmt. ¶ 438. Consistent with this widely-recognized practice, disclosures in the Offering Materials about compliance with underwriting guidelines generally specified that the description applied to when the loans were originated or acquired, not at the time of securitization. Stmt. ¶ 441. Investors were fully aware that the information in the disclosures was true as of the time of the loan’s origination and the time of whole loan purchase by Goldman Sachs. Likewise, the LTV ratios and owner-occupancy statistics included in the Prospectus Supplements were generally as of a statistical calculation or “cut-off” date specified in the prospectus supplement, not the closing date of the securitization. See Stmt. ¶ 442. The cut-off dates for the majority of Securitizations were at least twenty days prior to the securitization’s closing date and, in many cases, prior to the settlement date for certain underlying whole loan pools. See Stmt. ¶ 443. 2. The Time Between Whole Loan Purchase and Securitization Was Inconsequential for the Goldman Sachs Deals at Issue. FHFA speculates that a second round of due diligence would have revealed additional information that may have shed light on whether borrowers intended to occupy the subject properties or had undisclosed debts. For 55% of the pools at issue, the time between whole loan purchase and securitization was less than 90 days. Stmt. ¶ 437. Indeed, for some Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 45 of 73 -38- Securitizations, loan-level due diligence was conducted as close as nine days prior to securitization of the loan pools. Id. FHFA thus asks the Court to rule, purely based on its newly constructed legal theory, that whole loan diligence was inadequate no matter how close in time to securitization, without any evidence of material changes between purchase and the cut-off date. Moreover, to the extent loans were held on Goldman Sachs’ books for more than 90 days and continued to pay, they became “seasoned” as borrowers demonstrated their willingness and ability to pay. Contrary to FHFA’s suggestion, Pl. Br. at 46, 48, Goldman Sachs monitored loans in its books and culled those that experienced EPDs and delinquencies. Stmt. ¶¶ 444-448. Thus, the need to conduct any “further investigation” diminished over time. Id. FHFA attempts to manufacture an attack on Goldman Sachs’ conduct of due diligence at acquisition by claiming that “the percentages of loans in the final SLGs that had been diligenced were often substantially smaller than the percentages of loans reviewed in the acquisition pools that Goldman’s own policies required.” Pl. Br. at 16. This claim is inaccurate for several reasons. First, FHFA focuses on the percentages of due diligence conducted on the SLGs, not the whole loan purchase pools. As the GSEs were well aware, Goldman Sachs’ guidelines for sampling ranges apply to the pool level diligence, not diligence on the SLGs in particular. Second, to reach its conclusion that “fewer than 25% of loans in the SLGs had been subjected to credit diligence,” FHFA understates the percent of SLG loans subjected to diligence by dividing the number of loans in the SLGs that were subjected to due diligence for which the Grice Database had due diligence information by the number of loans in the SLGs, even though diligence material was not available for every loan pool. While FHFA argues that “files produced by Goldman’s expert, for example, show that Goldman conducted credit diligence on just 0.29% of the loans in the GSAMP 2005-HE5 SLG,” that securitization mostly included Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 46 of 73 -39- loans originated by SouthStar that were diligenced by Lydian Data Services, which could not produce documentation more than five years later. See Stmt. ¶ 115. Thus, the 0.29% statistic is not a meaningful number. In reality, 36.46% of loans in the GSEs’ supporting loan groups received credit and compliance due diligence: essentially the same percentage of diligence conducted on the pool as a whole.17 See Grice Database. Third, the “minimum” suggested guideline for credit diligence of Alt-A loan pools was 5%, not 25% as FHFA claims. See, e.g., Stmt. ¶ 451; see also Pl. Br. at 16. Thus, even the Alt-A pools FHFA cherry-picked in an attempt to show that diligence conducted on the SLGs was “far less than Goldman’s policies required,” had sample sizes consistent with Goldman Sachs’ policies. See, e.g., Stmt. ¶ 451. B. Goldman Sachs Verified the Material Accuracy of Representations in the Offering Materials. FHFA alleges that the due diligence performed by Goldman Sachs at the whole loan acquisition phase did not verify the accuracy of representations in the Offering Materials, and that there is no other evidence that Goldman Sachs properly verified the accuracy of those representations. Pl. Br. at 10-11. To the contrary, in addition to Goldman Sachs’ robust pre- purchase due diligence process generally, see supra Fact Secs. I-IV, Goldman Sachs had a detailed process in place to ensure that offering materials were accurate in all material respects, see supra Fact Sec. IV, which at least creates a material issue of fact improper for resolution on summary judgment.18 See Ideal Steel Supply Corp., 652 F.3d at 326. The due diligence process 17 By these standards, all except four of the Goldman Sachs sponsored deals were within Goldman Sachs’ recommended sampling guidelines for whole loan purchase pools. Even the four securitizations were within the sampling guidelines because, as noted above, those guidelines applied to diligence on the whole loan purchase pool, not the SLGs. 18 Michelle Gill’s statement that she “would not say there’s a tick and tie, you know, checking back to the underwriting guidelines” was a statement regarding who within Goldman Sachs compared the offering materials to the underwriting guidelines when reviewing the offering materials. See Pl. Br. at 10. Many people—both within and outside of Goldman Sachs— Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 47 of 73 -40- was designed to confirm that the loans Goldman Sachs purchased were originated in accordance with applicable guidelines and that the material information on the loan tape (e.g., income, appraisal value, credit score) was accurate, which provided a reasonable basis for Goldman Sachs to conclude that information in the Offering Materials about the loans and compliance with underwriting guidelines was accurate in all material respects. See generally Stmt. ¶ 395. The MCC, which was responsible for approving securitizations, discussed the due diligence results and disclosure requirements prior to securitization. See Stmt. ¶¶ 393, 395. The results of Goldman Sachs’ diligence, which were summarized in memos provided to the MCC, provided a reasonable basis for Goldman Sachs to conclude that its Offering Materials did not contain material misrepresentations. See Stmt. ¶ 395. Because the diligence samples were selected on the basis of adverse characteristics, higher-risk loans were overrepresented in credit and compliance due diligence samples; the portions of the pools randomly sampled or not sampled would have less risk than the diligence results based on an adverse sample would indicate. FHFA alleges that the MCC approved 30 of the 40 Securitizations with “no follow-up” because no document reflects follow-up. See Pl. Br. at n.11. At best, any lack of “follow-up” suggests that none was required because Goldman Sachs employees carefully prepared the information submitted to the MCC to address potential concerns and reduce the need for follow-up. Stmt. ¶ 398. Presenting a transaction to a management committee without addressing likely questions or expecting that the committee might reject the transaction as described would not have been a career-enhancing strategy. Any suggestion that the MCC was involved in the loan-level due diligence efforts compared the information in the loan files to the applicable underwriting guidelines during credit due diligence, and also reviewed for material accuracy the information on the loan tape—which was the basis for the loan-level information in the Offering Materials. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 48 of 73 -41- passive is contradicted by contemporaneous e-mails demonstrating that the MCC was highly active in making inquiries following their meetings. Stmt. ¶ 398. FHFA’s allegations that the MCC was “not presented with accurate information as to the number of conduit loans in the securitizations it approved,” see Pl. Br. at n.8, confuses loans originated to Conduit guidelines with loans originated by originators (such as SouthStar Funding) that were part of the Conduit program, but originated to their own guidelines. Contrary to FHFA’s assertions, there are no material differences between the descriptions provided to the MCC and those disclosed in the prospectus supplements. The MCC Memoranda and prospectus supplements for the securitizations cited by FHFA, see Pl. Br. at 15 n.8, list the same loan originators in roughly the same proportions for each deal. See FHFA Stmt. ¶¶ 115, 125, 126. For example, the prospectus supplement for GSAMP 2005-HE5 states that 75.5% of loans “were acquired by GSMC from SouthStar Funding, LLC,” and that “[t]he mortgage loans originated or acquired by SouthStar were generally in accordance with the underwriting criteria described in this section” entitled “Southstar Underwriting Guidelines.” Stmt. ¶ 115. This is consistent with the MCC memorandum, which states that “74.49 percent of loans were originated from SouthStar Funding” and “[l]oans originated by SouthStar were underwritten to their guidelines.” Stmt. ¶ 115. The MCC memorandum says that GSMC purchased 100% of the loans via the Goldman Sachs Mortgage Conduit, but never says that they were originated according to Goldman Sachs Mortgage Conduit Guidelines—to the contrary, the MCC Memorandum states that “[l]oans originated by SouthStar were underwritten to their guidelines, [and] [t]he remaining loans were originated according to Goldman Sachs Mortgage Company’s underwriting guidelines.” Stmt. ¶ 115. Simply because SouthStar was part of the Conduit program, does not mean that the loans originated from that seller were underwritten to Conduit Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 49 of 73 -42- guidelines, as is clearly indicated in both the prospectus supplement and corresponding MCC memorandum. Stmt. ¶ 115. Resolution of fact-intensive issues such as these must be left to the jury. See Ideal Steel Supply Corp., 652 F.3d at 326. III. GOLDMAN SACHS’ DUE DILIGENCE POLICIES WERE SOUND AND MATERIAL ISSUES OF FACT EXIST AS TO THEIR REASONABLENESS. A. The Reasonableness of Goldman Sachs’ Use of Sampling. FHFA’s argument that Goldman Sachs did not rely on statistically valid sampling in credit and compliance due diligence, Pl. Br. at 19-21, is another attempt to circumvent the high quality of Goldman Sachs’ contemporaneous diligence by imposing a standard that neither the SEC nor industry practice required. The GSEs not only approved similar sampling techniques by counterparties, they also used the same sampling techniques for their own securitizations. Thus, Goldman Sachs’ similar methodology in conducting mortgage-related due diligence was reasonable. See Hr’g Tr., 70:19-71:1, Dec. 17, 2012 (11 Civ. 6198 (DLC)). In this Action, FHFA relies on far smaller samples as its chief proof. FHFA mischaracterizes the record when it claims that Goldman Sachs employee William Moliski purportedly makes “clear” that “with sampling, [an originator] will be able to slip these loans [not subject to diligence] by.” Pl. Br. at 15. Mr. Moliski makes that statement while discussing the downside of sampling for a particular originator because Goldman Sachs recently had been rejecting a relatively large number of loans from that originator. This was far from a wholesale endorsement of reviewing 100% of loans in every case, as FHFA claims, and instead shows Goldman Sachs identifying, and refusing to purchase, potentially problematic loans. See Stmt. ¶ 206. FHFA suggests that, to support a due diligence defense, Goldman Sachs’ diligence samples had to be “statistically valid” or scientifically determined to permit Goldman Sachs to “extrapolate” its findings to the pool as a whole. Pl. Br. at 21-22. But the Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 50 of 73 -43- reasonableness of Goldman Sachs’ sampling practices is a classic question of fact. See SQP, Inc. v. Sirrom Sales, Inc., 130 F. Supp. 2d 364, 368 (N.D.N.Y. 2001) (“[r]eliability of [] sampling and testing methods . . . creates questions of fact”). This is particularly true given that neither the contemporaneous standards of Regulation AB—promulgated by the SEC in 2005 to regulate the asset-backed securities market—nor the post-financial crisis standards of Section 945 of the Dodd-Frank Act require that due diligence samples be statistically valid or that underwriters conform to any particular sampling protocol. See 15 U.S.C. § 77g. Indeed, the SEC declined to adopt suggestions that would have required samples to be “statistically valid” and explained that “sampling may be appropriate depending on the facts and circumstances.” SEC Release Nos. 33-9176, 34-63742, at 4235 (Jan. 25, 2011). The use of sampling—usually without any test of statistical validity—was widely accepted throughout the industry during the relevant time period, including by the GSEs, who used sampling in their bulk purchases and anti-predatory lending diligence of PLS purchases.19 See, e.g., Stmt. ¶¶ 696-703, 705-708. It is a question for the jury to decide whether Goldman Sachs’ application of sampling in conducting its mortgage-related due diligence was reasonable. See Sears, Roebuck and Co. v. Midcap, 893 A.2d 542, 554 (Del. 2006) (Under the “reasonably prudent” person standard, “custom or practice in a particular industry is probative of what conduct is reasonable under the circumstances.”); Ford Motor Co. v. Zahn, 265 F.2d 729, 732-33 (8th Cir. 1959) (whether defendant’s “sampling” protocol for inspecting products constituted “reasonable care” was “properly one for the jury to resolve”). 19 The GSEs were not only aware of Goldman Sachs’ sampling methodology, but also that those samples were not necessarily “statistically valid.” For example, in April 2007, Freddie Mac accepted Goldman Sachs’ characterization of its due diligence as “statistically relevant,” not “statistically significant.” See Stmt. ¶ 466. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 51 of 73 -44- B. The Reasonableness of Adverse Selection. Consistent with industry best practices, Goldman Sachs’ samples were selected, adversely and randomly, by experienced transaction managers applying the due diligence policies and procedures in place at Goldman Sachs.20 Stmt. ¶¶ 449-466. FHFA attempts to argue that, because Goldman Sachs did not use a strict formula for selecting samples, the sampling practices were unreasonable. Pl. Br. at 21-22, 61. FHFA provides no support for its argument that Goldman Sachs’ reliance on seasoned employees to select adverse and random samples was unreasonable as a matter of law. Pl. Br. at 19-21, 51-52. Goldman Sachs’ practice of selecting the majority of loans in its samples based on adverse characteristics resulted in overrepresentation of higher-risk loans in credit and compliance due diligence samples.21 See Fact Sec. I.B. Because Goldman Sachs focused its review on loans with a greater likelihood of defects, the unsampled portion of the pool would have had a far lower likelihood of defects. Stmt. ¶ 456. Furthermore, the sampling rates utilized by Goldman Sachs were consistent with others in the industry. Stmt. ¶¶ 455, 697. Whether selection of an adverse sample based on a review of loan characteristics by experienced employees, and selection of a random sample using a Microsoft Excel formula—rather than some scientific method FHFA fails to define, Pl. Br. at 20 FHFA makes the farfetched argument that, because Goldman Sachs explored automating its sample selection process in 2007, the practices it used to select samples before then were unreasonable as a matter of law. See Pl. Br. at 21. FHFA provides no justification for asserting that exploring new due diligence techniques supports a legal conclusion that those techniques were (or are) required. Goldman Sachs’ consideration of automating its processes merely illustrates the steps that Goldman Sachs took in continuing to elevate and improve its due diligence processes. See supra Facts Sec. I.C. Further, Goldman Sachs’ consideration of due diligence enhancements are inadmissible evidence of a potential “subsequent remedial measure” that should not be considered in deciding FHFA’s Motion. Fed. R. Evid. 407; Krouner v. Am. Heritage Fund, Inc., 899 F. Supp. 142, 147 (S.D.N.Y. 1995) (applying Fed. R. Evid. 407 in the securities context). 21 Conduit loan pools generally were all subjected to due diligence, although sampling of conduit pools could be approved on a case-by-case basis. See Stmt. ¶ 452. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 52 of 73 -45- 19-21, 51-52—constitutes reasonable care or investigation, is a material issue of fact for the jury to decide. See In re WorldCom, 2005 WL 638268, at *11. C. Whether Originators Limited Goldman Sachs’ Sample Sizes. FHFA implies that Goldman Sachs did not follow its own policies and instead caved to pressure from originators to lower its sample sizes. Pl. Br. at 16-19, 55-56. FHFA provides no basis for this conclusion, and Goldman Sachs has provided ample evidence to the contrary. See generally Grice Report. Goldman Sachs never agreed to limit its due diligence, nor would it have made any sense for Goldman Sachs to do so, given that it would hold the loans it bought on its books. Instead, the record shows that Goldman Sachs based sample sizes on an independent assessment of loan pool characteristics, such as internal pre-due diligence reports, geographic concentration reports, and loans previously dropped from pools offered by the same originator. Stmt. ¶¶ 453, 458. In the instances where bid stipulations for a given pool purported to limit the amount of diligence, Goldman Sachs generally reserved the right to increase, or “upsize,” the diligence sample, and always had the option not to purchase loans. Stmt. ¶¶ 462- 463. To support its claim, FHFA mischaracterizes the record. For example, FHFA cites an e-mail exchange with Countrywide about the due diligence sample size on an Alt-A pool. Pl. Br. at 17. FHFA misleadingly states that no “justification was provided” for a 10% credit and compliance sample for an Alt-A pool. Id. But no justification was required, given that the sample was consistent with Goldman Sachs’ policy, which was to sample 5% to 15% of the loans in Alt-A pools. See Stmt. ¶ 610. FHFA also fails to mention a subsequent e-mail confirming that “Countrywide has agreed to update DD to 15%.” Stmt. ¶ 632. FHFA also cites an e-mail where members of Goldman Sachs due diligence team discuss the appropriate level of review for the RFC 5Y MAY3007 FLX pool. Pl. Br. at 17-18. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 53 of 73 -46- FHFA alleges the Goldman Sachs’ Chief Underwriter Brian O’Brien thought a “minimum of 50% review” of the pool was appropriate, but Goldman Sachs’ trading desk refused to agree to such a level. Pl. Br. at 18. To the contrary, far from “refusing to agree” to a certain level of due diligence, the Goldman Sachs trader told the transaction manager to abide by the agreed upon sample “unless there is something systemic or something that necessitates increasing the sample.” Stmt. ¶ 224. Mr. O’Brien then asked members of the Goldman Sachs due diligence team (not any Goldman Sachs traders) what the correct sample size is—contrary to FHFA’s assertion, Pl. Br. at 18, Mr. O’Brien does not state that a “minimum of 50% review” of the pool was definitely appropriate, and no trader subsequently refuses to agree to a 50% sample size. See FHFA SUF ¶ 226. Goldman Sachs ultimately performed credit and compliance due diligence on 28% of the RFC 5Y MAY3007 FLX pool. Stmt. ¶ 227. The sample for this pool exceeded the range of 5 to 25% recommended by Goldman Sachs for Alt-A transactions.22 Even if two Goldman Sachs employees arguably thought Goldman Sachs should upsize the sample even more, that does not mean that the diligence conducted on that pool was unreasonable, let alone unreasonable as a matter of law. Interpretation of this e-mail, as well as its impact on the determination of the reasonableness of Goldman Sachs’ due diligence generally, “requires an exquisitely fact intensive inquiry into all of the circumstances surrounding the facts upon which [this claim] is premised.” See In re WorldCom, 2005 WL 638268, at *11; Abrams v. Reade, 419 22 Although FHFA relies on an e-mail in which a trader stated that he agreed to a 15% sample size rather than Goldman Sachs’ “typical” 20% sample size “to get the deal done,” Pl. Br. at 18, the 15% sample was within the sample size Goldman Sachs applied to Alt-A securitizations, and Goldman Sachs maintained the ability to upsize if necessary. Indeed, the e-mail FHFA cites states that the due diligence sample size was a minimum of 15%. See FHFA Ex. 170 (“*Min 15% DD Sample Size”). The sample sizes for the other pools FHFA lists in its Brief, see Pl. Br. at 18, were also well within the 5% to 25% sample size guidelines recommended by Goldman Sachs for Alt-A securitizations. Two of the pools listed (CW71 JAN302006 ARM and CW101 DEC142005 ARM) are not pools underlying the Securitizations. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 54 of 73 -47- F. Supp. 2d 476, 482 (S.D.N.Y. 2005) (“The inferences to be drawn from the adverse action taken against [plaintiff] and the . . . email are questions of fact and must be resolved at trial.”). D. Goldman Sachs’ Management of Third-Party Firms Was Effective. FHFA alleges that Goldman Sachs ignored concerns about the quality of work provided by third-party vendors and, thus, ignored its duty as underwriter to “look deeper and question more when confronted with red flags.” Pl. Br. at 52. FHFA again fails to show that this assertion is undisputed, and either ignores or mischaracterizes the overwhelming evidence to the contrary. This presents yet another clear issue of material fact inappropriate for resolution on summary judgment. See In re WorldCom, 346 F. Supp. at 679-80. The record shows that Goldman Sachs was actively involved in the diligence process, reviewing the results carefully, following up where it saw red flags, and working closely with third-party vendors to ensure that their processes met Goldman Sachs’ high standards. See supra Facts Sec. II. The documents FHFA cites as “undisputed evidence” of Goldman Sachs’ failure to manage third-party diligence vendors actually show the opposite, as the jury must decide. Pl. Br. at 25-28. FHFA cites an e-mail in which Goldman Sachs transaction manager Joseph Ozment says he is “having to really push [Clayton] just to get them to perform there [sic] jobs.” Pl. Br. at 25. This e-mail does not relate to Clayton’s due diligence reviews but rather its pace in reviewing cures of defects, and displays Goldman Sachs’ oversight of, and direction to, third- party due diligence vendors. Stmt. ¶ 321. Likewise, FHFA cites e-mails, Pl. Br. at 25-26, from transaction manager Lauren Carter, who reviews the work done by both Clayton and MDMC, and works with those diligence vendors to fix the problems identified. See, e.g., Stmt. ¶ 604. As FHFA admits, see Pl. Br. at 25-26, Goldman Sachs ceased doing business with some vendors when it found they were not providing quality due diligence in certain areas. See Stmt. ¶ 606. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 55 of 73 -48- FHFA also cites an e-mail where transaction manager Lauren Carter informed Linda Peterson that she had “4 loans in the Ameriquest scratch and dent where the borrowers are dead[;]” however, “Clayton review states loan is delinquent, but accept[s] as is and purchase.” Pl. Br. at 25. As an initial matter, the loans were unrelated to the Certificates; due to their delinquent status, they were all in a “scratch and dent” pool Goldman Sachs was considering for purchase. See Stmt. ¶ 577. The circumstances surrounding the four loans are unknown: the loans were well-seasoned and the borrowers may have died after origination, may have had co- borrowers, or may not have died at all. Goldman Sachs identified the issue, addressed it with Clayton, and resolved it. Stmt. ¶ 321. At Goldman Sachs’ request, Clayton modified its process to ensure that AppIntel flags were incorporated into its system and addressed. Stmt. ¶ 321. FHFA claims that a test of four due diligence vendors conducted by Goldman Sachs in September 2007 somehow indicates that the work conducted by those third-party due diligence vendors for the previous three years was inadequate, and that Goldman Sachs should have “review[ed] the vendors’ performance sooner.” See Pl. Br. at 27-28. FHFA’s claim is farfetched and inaccurate. This September 2007 test was part of Goldman Sachs’ continuing efforts to improve vendor quality, training and performance. Stmt. ¶¶ 328-334. Contrary to FHFA’s claim that Goldman Sachs failed to properly address red flags, see Pl. Br. at 53-54, Goldman Sachs conducted this review promptly after its purported need was identified in July 2007, and created a detailed plan for following up with each vendor after completing the loan test. Stmt. ¶¶ 328, 333. The test on which FHFA so heavily relies involved only 28 loans, all of which Goldman Sachs had excluded from a previous securitization. A “sample” of 28 loans would not constitute a “statistically valid” sample, as FHFA contends is essential to drawing reliable conclusions about a loan population. Stmt. ¶ 328. Furthermore, “[u]nderwriting is not Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 56 of 73 -49- formulaic or driven by a set of inflexible standards, instead, underwriters base individual lending decisions on a reasonable assessment of whether a loan complies with the applicable underwriting guidelines.” Id. Thus, there are legitimate reasons that the grades provided by the due diligence vendors during this test did not always match the grades Goldman Sachs applied. See id. FHFA fails to cite any consequence to this test because there was none—Goldman Sachs did not purchase any loan pools underlying the Securitizations after the September 2007 audit. See Stmt. ¶ 599. FHFA’s efforts to cast doubt on the competence of these vendors (which the GSEs also used) based on this 28 loan “sample” raise issues of fact for the jury. E. The Reasonableness of Goldman Sachs’ Due Diligence When It Acted Only as Underwriter. For the four Securitizations where Goldman Sachs was only the underwriter (not the issuer), FHFA claims it is entitled to a summary judgment finding that the sample Goldman Sachs used was unreasonable as a matter of law. Pl. Br. at 58; see also Stmt. ¶ 534. As stated by Goldman Sachs’ expert Charles Grice, “[t]he scope of due diligence conducted by an underwriter is typically more limited, commensurate with the underwriter’s limited role in the transactions, and is done primarily to satisfy the underwriter’s obligations under the securities law.” Stmt. ¶ 535. FHFA fails to show the absence of a material fact for the jury. Most notably, FHFA fails to address the fact that more than 1,800 loans underlying these four deals were subject to credit and compliance due diligence. Stmt. ¶ 537. These loans constitute 8.4% of the over 22,000 loans underlying the four deals collected by Goldman Sachs’ expert—more than four times larger than the 2% sample on which FHFA rests its case. The lower sample size conducted on deals where Goldman Sachs was only the underwriter does not establish as a matter of law that the diligence was insufficient; that determination is a question of material fact for the jury. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 57 of 73 -50- FHFA misrepresents the record when it argues that Goldman Sachs employees considered inadequate its due diligence procedures for securitizations for which Goldman Sachs was only the underwriter. Pl. Br. at 39-40. Jessica Feingold, a trader, expressed concern regarding the difference in sample sizes between Goldman Sachs-issued securitizations and securitizations where Goldman Sachs was only the underwriter—she never said there was an issue with non-random sampling and no trader said Goldman Sachs’ due diligence policies were “pathetic,” as FHFA claims. See Pl. Br. at 39-40. The e-mail in which a trader says “Its been pathetic,” never says that Goldman Sachs’ due diligence practices were pathetic, and FHFA cannot point to any testimony that the e-mail referred to the inadequacy of those practices. Pl. Br. at 39. Two e-mails from traders, even if interpreted as FHFA argues, do not establish that Goldman Sachs’ diligence should have been conducted differently for securitizations where Goldman Sachs was only the underwriter, especially in view of industry practice consistent with Goldman Sachs’ practices. See Stmt. ¶ 531. Although FHFA argues that certain loans in the ACCR 2005-4 and FHLT 2006-E securitizations were graded an event level “3” and thus not suitable for securitization, Pl. Br. at 40, Goldman Sachs identified compensating factors for loans to offset exceptions. Stmt. ¶ 497; see supra Argument Sec. III.F.1. Whether the practices Goldman Sachs used to conduct due diligence on loans where Goldman Sachs’ only involvement was as underwriter of a securitization were reasonable is a question for the jury. F. The Securitization of Loans Graded as “3s”. Ample evidence in the record demonstrates that underwriting involves reasonable judgment, see Assured Guaranty Municipal Corp. v. Flagstar Bank, FSB, 920 F. Supp. 2d 475, 505 (S.D.N.Y. 2013), and that determining whether a loan materially complies with guidelines, or includes appropriate compensating factors is a question of fact, see In re WorldCom, 2005 WL 638268, at *11. FHFA nevertheless argues that it is entitled to summary judgment on the Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 58 of 73 -51- reasonable care defense because a small number of loans graded by Clayton as “3s” were included in the Securitizations. Pl. Br. at 29, 40. This determination involves the resolution of numerous material issues of fact. See Vivenzio, 611 F.3d at 106. 1. Event Grade Three Does Not Mean the Loan Was Defective. As a threshold matter, FHFA erroneously equates an event grade “3” with a defective or “bad” loan that, if purchased and securitized, constitutes a de facto misrepresentation in the Offering Materials. See Pl. Br. at 29, 40. In doing so, FHFA ignores the meaning of an event grade “3.” As the GSEs recognized in their work with Clayton, an event grade “3” did not necessarily mean that the loans failed to comply with originator guidelines. See Stmt. ¶¶ 566, 568. An FHFA witness testified that he did not believe Freddie Mac had committed fraud by accepting loans that Clayton had initially graded as “3s,” and he did not believe the practice rendered Freddie Mac’s similarly-worded offering circulars false or misleading. Stmt. ¶ 567. A loan could be graded a “3” if it had an exception that was curable. For example, if a critical document such as a HUD-1 form was missing from the loan file, the loan would have been marked initially as an event grade “3”, but the grade could have been changed to a “1” or been waived in by Goldman Sachs if the originator supplied the HUD-1 form. Other information could be provided that further supported a borrower’s job history, income or assets, or confirmed the value of the mortgage property. See, e.g., Stmt. ¶ 569. The original event grade “3” also could have resulted from an error in the review, such as applying the incorrect loan underwriting guideline. Stmt. ¶ 558. In addition, loans classified as event grade “3s” included loans that fully complied with the originator’s underwriting guidelines, but did not comply with overlays that Goldman Sachs instructed the diligence provider to apply during its review. Id. For example, RFC 5Y MAY3007 FLX, the pool FHFA notes had 106 loans with a Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 59 of 73 -52- final grade of event grade “3” or “4” that were securitized in the SLG of GSR 2007-OA2, see Pl. Br. at 18 n.11, had a large number of loans graded “3” because the sample was checked against Goldman Sachs’ overlays, which “raised the bar” on the originator’s guidelines for making the loans. See Stmt. ¶ 573. Finally, loans graded as “3s” could have compensating factors that offset the exception identified by the diligence provider. See, e.g., Stmt. ¶ 574. The determination of whether compensating factors offset a material exception was left to Goldman Sachs, and Clayton and others frequently would mark loans as “3s” to escalate decisions on such loans to their client. Stmt. ¶ 570. FHFA argues that Goldman Sachs’ due diligence was legally insufficient because it “did not refer to any written policies or procedures” to apply compensating factors. Pl. Brief at 29; see also Stmt. ¶ 575. As Clayton testified, however, it requires discretion and judgment to determine whether a compensating factor offsets a given exception to the guidelines, and reasonable people can differ as to whether a compensating factor offsets such an exception. Stmt. ¶ 576. No rule can exist for every situation; the reasonableness of a given exception is, at the very least, a question of fact. See In re WorldCom, 2005 WL 638268, at *11. Even if FHFA could prove that loans with a final grade of “3” had material defects—a highly disputed issue of fact—it cannot get around the fact that the SLGs contained very few of those loans. Of the 199,556 ($47.9 billion) loans comprising the Grice Database, approximately 800 (or about $136 million) represented Credit Grade “3s” that supported certificates purchased by the GSEs, and only 213 (or about $41 million) represented Compliance Grade “3s” in the SLGs, with 38 loans (or about $7 million) overlapping. Stmt. ¶ 578. A review of those loans by Charles Grice revealed that loans with a final grade of “3” were included in the Securitizations for valid reasons, including after identification of compensating factors, Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 60 of 73 -53- additional documents or information, or pursuant to agreements to provide missing documents within a certain period. Stmt. ¶ 579. The GSEs used similar reasoning to re-grade loans.23 2. Goldman Sachs and Third-Parties Conducted Ample Due Diligence On All Loans in the Sample, Not Just Loans With Event Grade Threes. Although FHFA claims that it is “undisputed” that Goldman Sachs’ focus was to overturn vendors’ loan findings, Pl. Br. at 54, deposition testimony and contemporaneous e-mails show that Goldman Sachs reviewed all loans, not only those graded as a “3.” Stmt. ¶ 582; see Fact Secs. I, II. Goldman Sachs hired diligence vendors that performed multiple reviews of the diligenced loans, including those graded as event levels “1” and “2”. Stmt. ¶ 583; see also Fact Sec. II. FHFA’s farfetched theory contradicts common sense. Goldman Sachs would not have expended significant time, effort and expense reviewing due diligence results, adversely sampling loans with risk factors, or seeking missing documents and otherwise resolving defects if its goal was to securitize as many loans as possible. FHFA again mischaracterizes the record by citing e-mails purportedly showing that the “goal [of the Due Diligence Group] was to ‘clear as much as possible.’” Pl. Stmt. ¶ 312. Those e-mails show Goldman Sachs assuring originators that it would make reasonable efforts to clear loans, not that clearing potentially defective loans was the focus of the Due Diligence Group. Goldman Sachs worked closely with diligence vendors to ensure a careful review of potentially problematic loans, even if they otherwise complied with the applicable underwriting guidelines. For example, in November 2005, Goldman Sachs transaction manager Troy Grehalva instructed Clayton employee Joe Murray to “reject all loans that do not make 23 For example, Freddie Mac witness Ron Feigles testified that in a Freddie Mac deal called “Wells Fargo 2007-6,” 59% of the loans initially had material exceptions (i.e., initial grade “3s”) but were waived by Freddie Mac and re-graded to “2W.” The loan-level due diligence summary report for the deal identified various compensating factors as justification for the re-grades. Stmt. ¶ 581. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 61 of 73 -54- sense, regardless of meeting guidelines.” Stmt. ¶ 600. Grehalva also told Murray to “be conservative, [since Grehalva] can always overturn[.] If in questions [sic], just reject and I will review.” Stmt. ¶ 600. The third-party due diligence vendors testified to Goldman Sachs’ active participation in the due diligence process. Stmt. ¶ 591; see also Stmt. ¶¶ 622-626. Contrary to FHFA’s “accept all loans” theory, see Pl. Br. at 54, Goldman Sachs dropped nearly 10% of the loans from the pools underlying the Securitizations. See Stmt. ¶ 584. FHFA alleges, wrongly, that Goldman Sachs’ ability to waive loans into securitizations resulted in it purchasing the “worst of the worst” loans from a credit perspective. Pl. Br. at 29. Based on an e-mail in which Linda Peterson states that the Credit Committee generally reviews the “worst of the worst” loans, FHFA reckons that since the Credit Committee sometimes approved loans for purchase, it must have been securitizing the “worst” loans. Id. FHFA ignores that the function of the Credit Committee was to review the most difficult loan purchase decisions. Stmt. ¶ 563. Furthermore, the e-mail cited by FHFA demonstrates Goldman Sachs’ stringent due diligence process—a “pattern” of Goldman Sachs declining to purchase loans despite rebuttal information from originators. Pl. Br. at 29; Stmt. ¶ 317. FHFA also fails to show that any of the loans subject to review ended up in the SLGs. Stmt. ¶¶ 316-317. IV. GOLDMAN SACHS FOLLOWED ITS DUE DILIGENCE POLICIES. A. Goldman Sachs Accurately Disclosed the Values of Properties. As described above, see supra Argument Sec. I.B.1, Goldman Sachs conducted thorough due diligence to determine whether the appraisal value of a given loan was supported, by utilizing a variety of different tools including AVMs and BPOs. It relied in part on this information to make materially accurate disclosures in its Offering Materials. As an initial matter, FHFA relies on misleading and disputed statistics in support of its argument. See, e.g., Pl. Br. at 36 n.27. For example, FHFA claims that, “[f]or the loans in Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 62 of 73 -55- the SLGs at issue, Goldman did not obtain a BPO on 20.90% of the loans where no AVM value was received.” See, e.g., Pl. Br. at 36 n.27. These loans represent a tiny portion of the loans at issue—only 0.6% of the 60,267 total SLG loans in the Grice Database for valuation diligence. Similarly, although FHFA claims that “Goldman failed to obtain . . . a reconciliation on 22.39% of . . . loans in the SLGs” that a BPO indicated was out of tolerance, Pl. Br. at 36 n.27, that represents only 0.8% of the 60,267 total SLG loans in the Grice Database for valuation diligence. Again trying to conjure up undisputed “facts” from tiny grains of sand picked up along the beach, FHFA references only ten loans purchased by Goldman Sachs—nine of which were among almost 74,000 loans in the SLGs—that were discounted by less than 2% of the original purchase price, to support its sweeping assertion that Goldman Sachs routinely determined that the “actual values” of many of the mortgaged properties at issue were lower than reported by originators, but concealed these “true” values to negotiate lower prices for itself on the loans. See Pl. Br. at 34-36. These allegations are farfetched, unsupported by the record, and disputed.24 As an initial matter, this argument misconstrues the meaning of the “final value” Goldman Sachs derived for some properties. As the record demonstrates, value determinations for properties are estimates. See Stmt. ¶ 713. For each property, Goldman Sachs could consider several valuations, which frequently differed, sometimes materially. This could include the initial appraisal, AVM values, a third-party due diligence firm’s opinion of the appraisal, or a separate value from a BPO firm after review of the original appraisal. Goldman Sachs would 24 Although FHFA describes LTV as “critical in assessing a borrower’s ability to pay a mortgage and the adequacy of the collateral in the event of default . . . .” Pl. Br. at 2, the Truth in Lending Act factors for assessing a borrower’s ability to pay do not include LTV. 12 C.F.R. §1026.43. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 63 of 73 -56- review and reconcile all of the data to determine whether the appraisal was supported within a certain tolerance level—if it was not, the loan was typically dropped. 25 See Stmt. ¶ 485, 714. The documents FHFA cites in support of its implausible assertion that Goldman Sachs routinely renegotiated prices with originators based on its assessment of the “true value” of a property have no connection to the Securitizations or Offering Materials. See Pl. Br. at 34- 36, FHFA Stmt. ¶¶ 149-150. Loans with questionable valuations, when purchased by Goldman Sachs, typically were included in “scratch and dent” securitizations, which (as investors like the GSEs were well aware) consisted of “re-performing” loans or loans that did not meet guidelines, and which typically were securitized at a discount. See Pl. Exs. 151, 263, 264. None of the Securitizations were scratch and dent transactions. The only example FHFA could provide of Goldman Sachs’ supposed “scheme” to use internal final values to negotiate lower prices for the loans it bought, Pl. Br. at 34-36, is a pool where 10 loans26 were re-priced—by reducing the price less than 2%—after they were initially dropped post-valuation diligence, and 9 of those loans were subsequently included in the SLG. See Stmt. ¶ 717. Three of the loans FHFA identifies had “final values” within the 15% tolerance of the disclosed appraisal value; the remaining eight were within 6% of that tolerance range. Stmt. ¶ 152. Even if Goldman Sachs determined its own “value” of a given property, it cannot substitute that value for the information determined by an independent appraiser at the 25 Goldman Sachs included only 513 loans in the SLGs—0.3% of the 178,854 loans in the pools contributing to the SLGs on which Goldman Sachs performed valuation diligence—that it determined were out of tolerance for valuation due diligence reasons. Any number of reasons could have justified securitization of these loans, including that the seller provided additional information that the record does not reflect over five years later. See Stmt. ¶ 715. The inclusion in the SLGs of this small number of loans does not mean that, as a matter of law, Goldman Sachs’ due diligence was unreasonable. 26 FHFA argues that there were 11 loans that were re-priced, Pl. Br. at 36, but one loan was subsequently dropped by Goldman Sachs. Stmt. ¶ 152. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 64 of 73 -57- time of origination, as stated in the Offering Materials. Also, the Offering Materials explicitly warn investors that the LTV ratios will increase after origination. Stmt. ¶ 21. It is a material issue of fact whether this variance from the tolerance level and 2% discount on a miniscule number of loans was unreasonable. B. Goldman Sachs’ Representations Regarding Diligence of Loans Acquired Through Its Conduit Program Were Accurate In All Material Respects. FHFA again misses the mark when it argues that Goldman Sachs represented that substantially all loans bought through its conduit program were acquired generally in accordance with Goldman Sachs Mortgage Conduit Underwriting Guidelines, but that Goldman Sachs instead underwrote the loans to sellers’ guidelines and failed to do any comparison between the guidelines to ensure that the loans also met Goldman Sachs Mortgage Conduit Underwriting Guidelines. Pl. Br. at 31. FHFA fails to identify any evidence that the loans in the SLGs failed to comply with the Goldman Sachs Mortgage Conduit Underwriting Guidelines. As its “undisputed” evidence of failure to comply, FHFA mis-cites documents showing that Goldman Sachs analyzed or identified differences between its underwriting guidelines and those used by sellers. Pl. Br. at 55-56. FHFA quotes part of an e-mail in which Goldman Sachs employee Brian O’Brien states that “[w]e have largely abandoned the full guideline to guideline ‘gap analysis,’” see Pl. Br. at 55-56, but leaves out the passage in which Mr. O’Brien explains that: “Instead, we moved to observing actual differences at the file level on the current trade and considering those items as ‘stips’ for future trades.” Stmt. ¶ 724. Contrary to Plaintiff’s claims, Pl. Br. at 55-56, Goldman Sachs did not stop reviewing conduit guidelines; rather, it conducted its analysis on a loan-level basis instead of conducting a general analysis of the guidelines themselves. Moreover, Mr. O’Brien testified that he remembered “that we continued to do some element of gap analyses.” Stmt. ¶ 97. Linda Peterson and Mr. O’Brien both testified that any Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 65 of 73 -58- difficulties keeping up with an increase in the pace of necessary gap analyses were solved with additional staffing. Stmt. ¶¶ 102-105. Whether Goldman Sachs performed the “gap analyses” where necessary is an issue of material fact for the jury. Further, the prospectus supplements state that Goldman Sachs acquired loans that met certain underwriting criteria, not that originators underwrote all conduit loans to Goldman Sachs Mortgage Conduit Underwriting Guidelines. For example, the prospectus supplement for GSAA 2006-4 states: The mortgage loans were originated or acquired by . . . Goldman Sachs Mortgage Conduit . . . Pursuant to the residential mortgage loan conduit program, the sponsor purchases mortgage loans originated by the original mortgage loan sellers if the mortgage loans generally satisfy the sponsor’s underwriting guidelines. . . . Substantially all of the mortgage loans acquired by GSMC through its conduit program were acquired generally in accordance with the underwriting criteria described in this section. In certain instances, compensating factors . . . may warrant GSMC to make certain exceptions to these guidelines. In such instances GSMC would purchase a mortgage loan that did not completely conform to the guidelines set out below. Stmt. ¶ 19. The prospectus supplement goes on to describe the different programs and their “general” requirements. It is a question of fact for the jury whether any reasonable investor would understand this high level description of the loans Goldman Sachs generally acquired under the conduit program as a factual representation that every conduit loan was always originated to Goldman Sachs Mortgage Conduit Underwriting Guidelines. C. Information Obtained by Goldman Sachs From LownHome or Senderra Was the Same Information Known To the Public, Including the GSEs, and Goldman Sachs Did Not Conceal That Information To Gain Profits. FHFA proffers an absurd string of speculative inferences to suggest that Goldman Sachs had “unique and confidential access” from two affiliated originators about “unprecedented defaults and fraud,” which Goldman Sachs used to “short[] the market.” See Pl. Br. at 36-37. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 66 of 73 -59- The purportedly “inside” information about fraud in the subprime market that FHFA claims the affiliated originators provided to Goldman Sachs, id., was information already well known to the public.27 By late 2006 numerous public reports noted the high default rate of 2006 subprime loans and attributed these defaults, at least in part, to underwriting problems and low house price appreciation, and speculated that some might be attributable to fraud. For example, Fannie Mae employees received a Barclays Report on December 1, 2006, which attributed the poor performance of the 2006 Vintage to “[w]eaker loan characteristics and a deterioration in underwriting criteria in 2006, coupled with a significant increase in mortgage fraud.” Stmt. ¶ 647. An October 19, 2006 Wall Street Journal article reported “increasing concerns that lenders have been loosening their standards in an effort to boost loan volume,” and rising delinquencies. Stmt. ¶ 251. Freddie Mac’s former Chief Credit Risk Officer testified he had concerns about fraudulent owner occupancy representations and appraisal practices generally in 2006 and 2007, and he observed that “delinquency rates were rising” by the end of 2006 and is “sure others [could] observe the same event.” Stmt. ¶ 649. Volumes of additional evidence show that the GSEs were aware of the supposedly “secret information” FHFA alleges Goldman Sachs used to its sole benefit. Pl. Br. at 36-37; Stmt. ¶¶ 648, 652-653. FHFA’s claim that Goldman Sachs began shorting the market after receiving anecdotal information from Senderra has no support in the record. Pl. Br. at 36-37. The e-mails FHFA cites, see Pl. Br. at 36-37, were part of a large body of evidence concerning conditions in the housing and RMBS markets considered by Goldman Sachs, which is irrelevant to FHFA’s 27 FHFA insinuates that the use of “Internal Only” or “Do not distribute” at the top of e-mails is somehow evidence of wrongdoing. To the contrary, Goldman Sachs utilized that language in e- mails to ensure that, in line with its professional obligations to report materially accurate reporting information and in compliance with Goldman Sachs’ guidelines for written communications, documents were not improperly distributed outside the firm before the information they contained could be properly vetted. Stmt. ¶ 252. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 67 of 73 -60- claims. The December 14, 2006 meeting with Goldman Sachs’ chief financial officer, David Viniar, and the ensuing risk management decisions, are the subject of numerous contemporaneous e-mails which leave no doubt that the instructions given to the Mortgage Department were to reduce risk, not “short the market.” For example, a summary provided by one Goldman Sachs risk manager present at the meeting stated: “David Viniar just had an ad- hoc meeting on all things subprime . . . . [T]here was a lot of focus on reducing risk across the board . . . .” Stmt. ¶ 254. E-mails written by Dan Sparks, the head of the Mortgage Department, make clear that Mr. Sparks wanted to avoid being either too long or too short. For example: On January 23, 2007, Mr. Sparks specifically cautioned his traders: “I want to stay nimble, and not get too wedded to one way positions – getting massively short could be more painful [i.e., cause greater losses] than what we have experienced.” Stmt. ¶ 659 (emphasis added).) On February 22, 2007, Mr. Sparks sent an email to several traders in his Department, including Mr. Birnbaum, ordering them to cover large amounts of their short positions: “We need to buy back $1 billion single names and $2 billion of the stuff below – today. I know that sounds huge, but you can do it – spend bid/offer, whatever [it takes] to get it done. . . . I will not want us to trade property derivatives until we get much closer to home . . . .” Stmt. ¶ 659. Contemporaneous e-mails also reflect the significant uncertainty within Goldman Sachs as to the future direction of the mortgage market. In a March 3, 2007 email to himself, Mr. Sparks laid out points to go over at a meeting with his traders, stating: “Don’t add risk. Trade everything from short to flat. Get out of everything. Discuss liquidity of hedges.” Stmt. ¶ 659. Two days later, Tom Montag, a member of senior management, asked Mr. Sparks: “Do we think the business is net short, long or flat right now?” Stmt. ¶ 659. Mr. Sparks responded by providing his sense of the department’s position, but noting his uncertainty as to how the Mortgage Department’s diverse portfolio would perform under real-life market conditions: “We think the overall business is net short, but there is a lot of basis [risk].” Stmt. ¶ 659. Although one business within the Mortgage Department (the secondary trading desk) was net short Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 68 of 73 -61- subprime exposure in 2007—and FHFA quotes mainly from e-mails and self-evaluations by members of this desk—Mr. Sparks regarded this as “a protection hedge” that offset the Mortgage Department’s numerous long positions.28 Id. The profits of the Mortgage Department as a whole in 2007 reflected its hedged strategy, rather than the billions in profit made by hedge funds that truly shorted the market. 29 As Mr. Viniar testified before Congress: “In 2007, total net revenues from residential mortgage- 28 This view of the SPG Trading Desk’s position is consistent with that expressed in an August 2007 memorandum drafted by Goldman Sachs’ tax department, which was examining how to allocate certain gains and losses in the department’s book: As the sub prime mortgage market deteriorated in 2007, the ABS Business was short the market, which hedged the long exposure in the Residential Credit and CDO businesses. Senior management views the long and short mortgage exposure as a single business for purposes of assessing risk and determining overall risk limits. Specifically, Dan Sparks, the Head of Mortgage Department, confirmed that he took into account the long positions in managing the size of the short positions and viewed the short positions as protection against the risk of a “blow-up” in the mortgage book, particularly as the market went through significant turmoil this year. He further indicated that senior management was concerned about the size of the short position and considered reducing the short but he argued successfully that it should be viewed as disaster protection against the long positions including the warehouse lending positions. He said it would be difficult to isolate exactly how much of the short position was intended to offset each of the long positions but senior management was very aware of the offsetting nature of the positions and risk, particularly in a disaster in the mortgage business. Stmt. ¶ 659. 29 FHFA relies on the PSI Report to argue that Goldman Sachs “built and profited from a large net short position in mortgage related Securities” in December 2006 and early 2007. Stmt. ¶ 660. The PSI Report came to this erroneous conclusion by adding long and short positions on Mortgage Department “top sheets,” an informal tool prepared on a daily basis at the request of Dan Sparks to help him gauge, at a high level, the Mortgage Department’s overall positions in certain mortgage-related products. PSI Report at 419, n.1712. The “top sheets” did not include numerous substantial long positions (including mortgage loans and commercial mortgage-backed securities), while including the short hedges of those positions. Adding up long and short positions of different instruments also is a deeply flawed way to measure overall risk. Stmt. ¶ 660. Profits or losses in the face of steeply declining markets are a far better measure of net long or short positions, and by that measure Mr. Sparks’ sense that the Department was flat to slightly short was correct. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 69 of 73 -62- related products, both longs and shorts together, were less than $500 million, approximately 1 percent of Goldman Sachs’s overall net revenues.” By contrast, Paulson & Co., for example, which had a tiny fraction of Goldman Sachs’ capital—but actually did have the “big short”— reportedly made more than $15 billion from shorting residential mortgage-related products in 2007 alone. Stmt. ¶ 661. Other true “big short” funds also reportedly made many billions of dollars on their short positions in 2007. But even if Goldman Sachs had put on, and profited from, a supposed “big short,” FHFA cites no evidence showing any link between the “big short” and any information forwarded by Senderra and LownHome. Senderra and LownHome were not even among the top 500 originators of subprime loans in 2006, and hence had a market share less than the .016% share of the 500th originator. Goldman Sachs’ effort to reduce risk during a volatile time in the housing market did not mean it was betting that borrowers or bonds would fail, and it was wholly unrelated to either the reasonableness of Goldman Sachs’ due diligence, or the accuracy of the disclosures in the Offering Materials. These are quintessential issues of fact for the jury. D. Goldman Sachs Properly Applied Its Policies Concerning Originators on “Watch” and “Suspend” Lists. FHFA incorrectly alleges that Goldman Sachs “ignored” its findings regarding counterparties, Pl. Br. at 61, and continued to securitize loans from counterparties on its “watch” and “suspend” lists without the appropriate due diligence. Pl. Br. at 30-31. To support this allegation, FHFA wrongly assumes that if an originator was given a status of “watch” or “suspend” on the seller status report, Goldman Sachs should not have securitized loans from those originators. See Pl. Br. at 30-31. This assumption is misplaced for several reasons. First, as FHFA understands, a status of “suspend” meant Goldman Sachs should not be making further whole loan purchases. Stmt. ¶ 416. Neither “suspend” nor “watch” meant loans Goldman Sachs Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 70 of 73 -63- had previously purchased and was holding on its books, and which already had been subject to its rigorous due diligence process, could not be securitized. Id. Second, as FHFA acknowledges, Goldman Sachs could place an originator on its watch list for reasons that have nothing to do with the quality of the loans. See Fact Sec. I.A; Pl. Br. at 30. For example, Goldman Sachs could place originators on a watch list because they were slow to provide financial documents, or they had outstanding debts in connection with a prior, unrelated transaction. Goldman Sachs did not securitize loans from originators without performing reasonable due diligence on those loans first, and the examples FHFA provided demonstrate that practice. For example, although FHFA complains that loans originated by NovaStar were included in an SLG despite NovaStar being on Goldman Sachs’ “watch” list, Goldman Sachs performed diligence on 100% of loans in the NovaStar pools underlying the Securitizations— consistent with its policy. See Pl. Br. at 30-31; Stmt. ¶ 273.30 Whether Goldman Sachs conducted reasonable diligence on the originators from which it purchased loans, and whether it reasonably applied such findings to its decisions of which loans to include in the Securitizations, are questions for the jury. See In re WorldCom, 2005 WL 638268, at *11. FHFA’s assertion that Goldman Sachs tried to conceal that it was securitizing loans from “suspended” originators, see Pl. Br. at 31 n.21, blatantly mischaracterizes the record. FHFA cites an e-mail a Goldman Sachs employee wrote on January 26, 2007 that Fannie Mae was seeking a list of originators in Goldman Sachs deals to check whether they were approved for upcoming deals. Id; Stmt. ¶ 285. Contrary to FHFA’s assertions, nowhere does this document say Fannie Mae requested a list of originators for the GSAMP 2006-HE8 30 The other two examples FHFA provides regarding MLN and Weichert affect only 9 loans in the SLGs, at least 6 of which were subjected to credit and compliance due diligence. Stmt. ¶¶ 278-281. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 71 of 73 -64- securitization, which closed on December 27, 2006, a month prior to Fannie Mae’s request. Goldman Sachs could have had any number of reasons why it did not include Sebring or MLN on its list to Fannie Mae, including that Goldman Sachs already had suspended those originators and did not intend to continue purchasing loans from them. Sebring went out of business in December 2006, so it would not have made sense to include it on a list of originators for “upcoming deals.” Stmt. ¶ 285. V. SUMMARY JUDGMENT IS IMPROPER BECAUSE GOLDMAN SACHS HAS NOT BEEN AFFORDED A FULL OPPORTUNITY FOR DISCOVERY. Summary judgment is particularly unsuitable when a party has not been afforded sufficient discovery to gather facts essential to meeting its burden or necessary to establish an issue of material fact, see, e.g., Foresta v. Centerlight Capital Mgmt., LLC, 379 F. App’x 44, 46- 47 (2d Cir. 2010); Meloff v. N.Y. Life Ins. Co., 51 F.3d 372, 375 (2d Cir. 1995), and would constitute a denial of due process. As demonstrated in the Harsch Declaration, the limitations on discovery in this Action have foreclosed Goldman Sachs from obtaining discovery that would further show the reasonableness of Goldman Sachs’ due diligence by making direct comparisons to the GSEs’ due diligence processes and procedures on subprime and Alt-A loans of comparable credit quality, which they purchased directly from the same originators whose loans backed the PLS the GSEs purchased from Goldman Sachs. If, notwithstanding Goldman Sachs’ showing in opposition to the Motion, the Court considers that the evidence presented supports a grant of summary judgment, it should deny the Motion because of the limitations imposed on the scope of Goldman Sachs’ discovery and should permit Goldman Sachs to obtain that discovery.31 31 On the evening before the due date of this brief, FHFA made available more than 5,000 documents it previously had deemed privileged. Goldman Sachs has not been able to review all these documents, but a limited review demonstrates the GSEs’ understanding and acceptance of sampling and other due diligence practices that FHFA criticizes in its motion. Goldman Sachs reserve the right to supplement the record from this recent production. Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 72 of 73 -65- CONCLUSION For the foregoing reasons, there remain genuine issues of material fact as to whether Goldman Sachs exercised: (i) reasonable care sufficient for defenses under Section 12 or the blue sky laws; or (ii) a reasonable investigation sufficient for a due diligence defense under Section 11. Accordingly, FHFA’s Motion should be denied in its entirety. Dated: July 18, 2014 Respectfully submitted, /s/ Richard H. Klapper Richard H. Klapper Theodore Edelman Michael T. Tomaino, Jr. Tracy R. High SULLIVAN & CROMWELL LLP 125 Broad Street New York, New York 10004 Telephone: (212) 558-4000 Facsimile: (212) 558-3588 Counsel for Defendants Goldman, Sachs & Co., Goldman Sachs Mortgage Company, GS Mortgage Securities Corp., The Goldman Sachs Group, Inc., Goldman Sachs Real Estate Funding Corp., Howard S. Altarescu, Kevin Gasvoda, Michelle Gill, David J. Rosenblum, Jonathan S. Sobel, Daniel L. Sparks and Mark Weiss Case 1:11-cv-06198-DLC Document 817 Filed 08/01/14 Page 73 of 73