Opinion
Case No. 13-cv-3451 (SRN/HB)
08-14-2020
In Re: ResCap Liquidating Trust Litigation This document relates to: ResCap Liquidating Trust v. Primary Residential Mortgage, Inc., No. 16-cv-4070
SUSAN RICHARD NELSON, United States District Judge
FINDINGS OF FACT AND CONCLUSIONS OF LAW
TABLE OF CONTENTS
BACKGROUND ................................................................................................................ 1
FINDINGS OF FACT ....................................................................................................... 5
I. INTRODUCTION .............................................................................................. 5
A. The Parties ............................................................................................. 5
B. RFC's Role in the Mortgage Backed Securities Market .................... 6
C. Client Contracts ..................................................................................... 6
1. The March 2000 Client Contract ............................................... 7
2. The June 2001 Client Contract .................................................. 8
D. The AlterNet Guide and the Client Guide .......................................... 9
E. The Guides' Key Provisions ................................................................ 10
1. Sole Discretion ........................................................................... 10
2. Knowledge, Reliance, and Waiver ........................................... 11
3. Specific Representations and Warranties ............................... 13
4. Events of Default & Non-Exclusive, Cumulative Remedies .................................................................................... 14
5. Repurchase ................................................................................. 15
6. Indemnification .......................................................................... 15
F. RFC's Securitization Business ............................................................ 17
1. Teresa Farley, ResCap Trial Witness and ResCap's 30(b)(6) Witness on Securitization .......................................... 17
2. RFC Relied on the Truthfulness of Originator Representations and Warranties During Securitization ....... 18
3. The Key RFC Securitization Transaction Documents ......... 19
a. RFC Loan Sale Agreements .......................................... 19
b. RFC's Pooling and Servicing Agreements ................... 20
c. Loan Tape and Mortgage Loan Schedule .................... 22
d. RFC's Prospectus Supplements .................................... 24
4. RFC's Securitization Shelves .................................................. 26
5. The RMBS Trustees ................................................................. 26
6. The Monolines .......................................................................... 27
G. RFC's Bankruptcy and Pre-Petition Litigation ............................... 27
1. The Pre-Bankruptcy Claims and Litigation ........................... 28
2. The Original RMBS Settlement ............................................... 29
3. The Bankruptcy, and RMBS Trust and Monoline Claims ... 31
4. The 9019 Proceeding and the Mediation ................................. 34
5. The Settlements and Expert Assessment of the Settlements .......................................................................... 42
6. The Bankruptcy Court's Approval of the Settlements .......... 45
II. PRMI'S BREACHES OF ITS CLIENT CONTRACTS WITH RFC ........ 46
A. Mr. Butler's Reunderwriting Review of PRMI Loans and Global Loans ........................................................................................ 46
B. Ms. Keith, PRMI's Reunderwriting Expert ...................................... 48
C. Mr. Butler's Identifications of AlterNet Guide Breaches ................ 50
1. The Seven AlterNet Loans at Issue .......................................... 52
a. Loan 4375505 .................................................................. 52
b. Loan 4962418 .................................................................. 53
c. Loan 4117058 .................................................................. 54
d. Loan 4380515 .................................................................. 55
e. Loan 4115211 .................................................................. 55
f. Loan 4550405 ................................................................... 57
g. Loan 4413350 .................................................................. 59
D. Mr. Butler's Identification of Breaches of Four Assetwise Loans Made Under the Client Guide ............................................................ 62
1. Loan 5798390 ............................................................................. 62
2. Loan 8257547 ............................................................................. 63
3. Loan 10381337 ........................................................................... 65
4. Loan 6231616 ............................................................................. 67
E. Mr. Butler's Identification of Countrywide Loan Breaches: Loan 10226985 ...................................................................................... 69
III. PRMI'S BREACHES CONTRIBUTED TO RFC'S LIABILITY ............ 70
A. Bankruptcy Evidence Shows That PRMI's Breaches Contributed to RFC's Liability .......................................................... 71
1. Proofs of Claim in the Bankruptcy Asserted Origination-Level Breaches ..................................................... 71
2. The Bankruptcy Participants Measured RFC's Liability By Examining Origination Guideline Breaches ..................... 72
B. PRMI's Breaches Contributed to RFC's Risk of Liability on Specific Trust Representations ........................................................... 74
1. Mr. Butler Identified PRMI Breaches That Created Risk .... 74
2. PRMI's Experts Did Not Address the Issue of Risk of Liability ......................................................................... 78
C. The Court Finds ResCap's Evidence on the Proper Interpretations of the Trust Reps & Warranties More Persuasive ........................... 81
1. Client Contracts Gave Enforceable Loan-Level Reps ........... 81
2. The MLS Reps Were Not Merely Transcription Reps .......... 82
a. Ms. Farley's Testimony ................................................. 83
b. Mr. Butler's Testimony ................................................. 83
c. Judicial Interpretation of the MLS Rep ....................... 84
d. Actual Reliance by Stakeholders .................................. 84
e. MLS Rep Breach Remedy is a Loan-Level Remedy ... 85
f. No Persuasive PRMI Evidence Was Introduced To the Contrary ............................................................. 86
3. The No-Default Reps Were Not Merely Payment Default Reps .............................................................................. 87
4. The Loan Program Reps Were Enforceable For Single Loans .............................................................................. 93
5. The Credit Grade Reps Were Enforceable For Single Loans ............................................................................ 102
6. Trust Reps & Disclosures Were Passed Through To Monolines ................................................................................. 103
7. The Fraud Disclaimer Did Not Eliminate RFC's Risk of Liability ............................................................................... 105
8. The No-Fraud Rep .................................................................. 111
9. The Substantial Compliance Rep .......................................... 112
IV. RESCAP'S DAMAGES CALCULATION IS REASONABLE .............. 114
A. Dr. Snow Allocated to Breaching Losses ......................................... 114
1. The RMBS Trust Settlement Allocation ............................... 115
2. The Monoline Settlement Allocation ..................................... 118
3. The Total Allocation ................................................................ 120
B. Allocation to Breaching Losses is Reasonable and Reliable .......... 120
1. RFC's Settlement Was Allocated to Breaching Losses ........ 121
2. Similar Settlements Have Allocated to Breaching Losses ... 123
C. Allocating to Additional Factors is Unnecessary and Unreliable .. 124
1. Challenges to Certain Assumptions in Dr. Snow's Model .. 126
a. Legal Interpretations of Trust Reps ........................... 126
b. Causation and Materiality .......................................... 127
c. "Election of Remedies" Defense .................................. 128
d. Statute of Limitations .................................................. 129
i. Likelihood of RMBS Defendants' Success ...... 134
ii. Lower Settlement Value ................................... 138
2. Accounting For Only Select Unresolved Legal Issues & Defenses Would Have Introduced Bias and Speculation Into the Model ......................................................................... 139
3. Assigning Probabilities Would Be Speculative And Imprecise ................................................................................. 141
D. Dr. Snow Implemented His Methodology Reasonably .................. 144
1. The Treatment of the NDS Trust Breach Rate Is Reasonable ........................................................................... 144
2. The Definition of the At-Issue Population is Reasonable .... 146
3. The Monoline Allocation Approach is Reasonable .............. 147
4. The Point Estimate is an Appropriate Measure Of Damages ............................................................................. 149
5. Dr. McCrary's Prior Testimony Undermined His Credibility ......................................................................... 151
CONCLUSIONS OF LAW ........................................................................................... 154
I. OVERVIEW & ELEMENTS OF CLAIM FOR CONTRACTUAL INDEMNIFICATION .................................................................................. 154
II. A GOVERNING CONTRACT ................................................................... 156
III. PRMI COMMITTED EVENTS OF DEFAULT AND BREACHED REPS BY SELLING RFC DEFECTIVE LOANS .................................... 159
A. This Court Ruled That, as a Matter of Law, ResCap Had Sole Discretion, Under the Client Guide, to Determine Breaches ......... 159
B. The Assetwise Loans .......................................................................... 160
C. Countrywide Loan ............................................................................. 162
D. AlterNet Loans ................................................................................... 163
IV. PRMI'S BREACHES WERE A CONTRIBUTING CAUSE OF THE RISKS RFC SETTLED IN THE BANKRUPTCY SETTLEMENTS .... 167
A. Legal Standard ................................................................................... 167
B. The Evidence ...................................................................................... 168
1. The Evidence From the Bankruptcy Establishes the Requisite Causal Connection ................................................. 168
2. The Evidence Establishes that PRMI's Breaches Increased RFC's Risk of Liability to the RMBS Trusts and Monolines ......................................................................... 171
a. MLS Rep ....................................................................... 173
b. No-Default Rep ............................................................. 175
c. Loan Program and Credit Grade Reps ...................... 177
d. Monoline Reps .............................................................. 181
e. Fraud Disclaimer .......................................................... 182
f. Fraud and Substantial Compliance Reps ................... 183
V. RESCAP'S DAMAGES METHODOLOGY IS REASONABLY CERTAIN AND NON-SPECULATIVE ..................................................... 185
A. The Law .............................................................................................. 185
B. The Allocated Breaching Loss Approach ........................................ 189
1. Objective Standard Based on Losses and Relative Breach Rates ............................................................................ 190
2. Consistent With the Approach Used in Bankruptcy .......... 190
3. Consistent with Allocations in Other Court-Approved RMBS Settlements .................................................................. 195
4. PRMI's Objections to the Allocated Breaching Loss Approach Invite Unfairness, Speculation, and False Precision ......................................................................... 196
ORDER FOR JUDGMENT ......................................................................................... 202
SUSAN RICHARD NELSON, United States District Judge
The Court begins its analysis of this dispute by providing some important context and background on the residential mortgage securitization market. The majority of U.S. mortgages are financed through the securitization process. In Re: RFC & ResCap Liquidating Tr. Litig., 332 F. Supp. 3d 1101, 1117 (D. Minn. 2018) ("Common SJ Order") (citing Adam J. Levitin & Susan M. Wachter, Explaining the Housing Bubble, 100 GEO. L.J. 1177, 1182, 1187 (2012)). "Securitization" involves pooling large numbers of housing loans and then selling them to a trust. Id. (citing Baker v. CitiMortgage, Inc., No. 17-cv-2271 (SRN/KMM), 2017 WL 6886712, at *4 (D. Minn. Dec. 21, 2017)). A mortgage lender raises funds for new mortgages through this process, and the trust pays for the loans by issuing securities for which the loans serve as collateral. Id. (citing Baker, 2017 WL 6886712, at *4). "The right to receive trust income is parceled into certificates and sold to investors, called certificateholders." Id. (quoting BlackRock Fin. Mgmt. Inc. v. Segregated Account of Ambac Assur. Corp., 673 F.3d 169, 173 (2d Cir. 2012)).
Purchasers of the securities often require that they be insured by monoline insurers as a hedge against investment risk. Id. (citing In re Barclays Bank PLC Securities Litig, No. 09 Civ. 1989 (PAC), 2017 WL 4082305, at *4 (S.D.N.Y. Sept. 13, 2017), aff'd, 756 Fed. App'x 41 (2d Cir. 2018)).
In the early- to mid-2000s, a rise in home prices in the U.S. was "driven by increased demand, low interest rates, and easy credit access." Id. (quoting In re Barclays Bank, 2017 WL 4082305, at *4). While an initial mortgage refinancing boom from 2001 to 2003 led to increased earnings for mortgage originators and securitizers, when long-term interest rates began to rise, the mortgage industry sought other ways to maintain origination volumes. Id. (citing Levitin & Wachter, supra, at 1193-94). The solution required industry players "to find more product to move in order to maintain origination volumes and, hence, earnings." Id. (citing Levitin & Wachter, supra, at 1194).
A second mortgage boom ensued after 2003, but "[b]ecause the prime borrowing pool was exhausted, it was necessary to lower underwriting standards and look more to marginal borrowers to support origination volume levels." Id. (quoting Levitin & Wachter, supra, at 1194). During this time period, "[l]enders provided mortgage loans to many high-risk borrowers with questionable ability to repay, fueled in large part by the opportunity to package and sell those mortgages into the growing market for [residential] mortgage-backed securities ("[R]MBSs")." Id. (quoting In re Barclays Bank, 2017 WL 4082305, at *4).
In the mid-2000s, the "explosion in the market for [RMBS]" resulted in a securitization market frenzy. Id. (quoting Fed. Hous. Fin. Agency for Fed. Nat'l Mortgage Ass'n v. Nomura Holding Am., Inc., 873 F.3d 85, 96 (2d Cir. 2017)). It was not to last. Among other things, housing prices fell and
[l]ate 2006 and 2007 saw a dramatic rise in mortgage loan defaults, causing the value of the related securities, whose income depended on borrower payments, to deteriorate. Banks and other investors began to experience substantial losses; and many monoline insurers could not accommodate such loss, given its quick pace and dramatic size.Id. (quoting In re Barclays Bank, 2017 WL 4082305, at *4). The global economy experienced an unprecedented downturn in 2008 "that had a profoundly negative effect on the real estate and credit markets." Id. (quoting S.E.C. v. True North Fin. Corp., 909 F. Supp. 2d 1073, 1083 (D. Minn. 2012) (citations omitted)).
In December 2016, the ResCap Liquidating Trust ("ResCap") commenced this lawsuit against Primary Residential Mortgage, Inc. ("PRMI"), asserting claims of breach of contract and contractual indemnification. This lawsuit was part of a second wave of contract and indemnification lawsuits ("Wave Two") filed by ResCap in the District of Minnesota against numerous banks and mortgage lenders. See In re ResCap Liquidating Tr. Litig. ("PRMI SJ Order"), 428 F. Supp. 3d 53, 73 (D. Minn. 2019).
After ResCap began filing the first wave of lawsuits ("Wave One") in 2013, the judges of this District agreed, in January 2015, to consolidate the then-59 active cases before the undersigned Judge, Magistrate Judge Keyes, and Magistrate Judge Bowbeer. Id. A smaller group of non-consolidated cases, all of which settled prior to trial, were assigned to Judge Paul Magnuson, Magistrate Judge Keyes, and Magistrate Judge Bowbeer. Id.
Of the First-Wave cases in the Consolidated Action, the first and only case to proceed to trial was ResCap's case against Home Loan Center, Inc. ("HLC"), ResCap Liquidating Tr. v. Home Loan Center, Inc., 14-cv-1716 (SRN/HB). Id. at 74. The HLC jury trial took place from October 15 to November 7, 2018. Id. The jury returned a $28.7 million verdict in favor of ResCap, representing approximately 70% of the damages that Plaintiff's expert, Dr. Snow, testified was a conservative estimate of the damages to be allocated to HLC. Id. Following rulings on the award of attorneys' fees, pre-verdict pre-judgment interest, and post-verdict pre-judgment interest, the Court directed entry of judgment against HLC in the total amount of $68,484,502.06, inclusive of the jury's award of damages. Id. at n.12 (citing ResCap Liquidating Tr. v. Home Loan Ctr., Inc., 14-cv-1716, June 21, 2019 Order [Doc. No. 83].)
The instant matter was tried by way of a bench trial before the undersigned judge over thirteen days on February 10-14, 18-21, 2020, and March 3-4, 12-13, 2020. At trial, the parties introduced over 300 exhibits and testimony from 27 witnesses. Plaintiff introduced the testimony of thirteen witnesses, five of whom appeared live and six by deposition designation. Plaintiff also introduced, by stipulation, the prior testimony of Jeffrey Lipps, which the Court admitted. See Stip. re: Testimony of Jeffrey Lipps [Doc. No. 5413]; Testimony of Jeffrey Lipps [Doc. No. 5413-1]. PRMI introduced the testimony of seven live witnesses and thirteen witnesses by deposition designation. Based on the evidence presented at trial, and all of the files, records, and proceedings in this matter, the Court makes the following Findings of Fact and Conclusions of Law.
Plaintiff introduced the live testimony of the following witnesses at trial: Renee Bangerter, Teresa Farley, Steve Butler, Donald Hawthorne, and Karl Snow. Also, Plaintiff designated, and the Court admitted, the deposition testimony of the following witnesses: Yvonne Flitton, Lewis Kruger, Allen Pfeiffer, Frank Sillman, Andrew Swope, and Lori Zaloumis.
PRMI introduced the live testimony of the following witnesses at trial: Dave Zitting, Jim Crawford, Philip Burnaman, Kori Keith, Justin McCrary, Steven Schwarcz, and David Woll.
On March 12 and 13, 2020, two of these witnesses, Mr. Jim Crawford and Dr. Justin McCrary, testified live, via video link, from courthouses in Salt Lake City and New York City, respectively. In light of the two witnesses' concerns about the safety of traveling to the St. Paul courthouse during the COVID-19 pandemic, the Court ruled that they could appear by video. Counsel in St. Paul took their testimony via video link. The Court could see these witnesses on a large screen, hear them perfectly, and fully assess their credibility, as if they had appeared in person.
PRMI also designated, and the Court admitted, the deposition testimony of the following witnesses: Bradford Cornell, Timothy Devine, Tammy Hamzehpour, Robert Major, Sharon Maki, Kathlene Meadows, J.F. Morrow, Leigh Ann Richardson, Yvonne Flitton, Lewis Kruger, Frank Sillman, Andrew Swope, and Lori Zaloumis.
To the extent any finding of fact shall be determined to be a conclusion of law, it shall be so deemed, and vice versa.
FINDINGS OF FACT
I. INTRODUCTION
A. The Parties
1. This lawsuit stems from the 2012 bankruptcy (the "Bankruptcy") of the Minnesota company, Residential Funding Company, LLC ("RFC").
In re Residential Capital, LLC, No. 12-12020 (MG) (Bankr. S.D.N.Y.)
2. ResCap is a Delaware statutory trust that was created under the Bankruptcy Plan that was confirmed by United States Bankruptcy Judge Martin Glenn, United States Bankruptcy Court for the Southern District of New York, on December 11, 2013. See PTX-180 at 79-80, 106, 165 (Plan and Confirmation Order). Under the Plan, ResCap has been designated, on behalf of the Bankruptcy estates, to prosecute claims assigned to it by the former chapter 11 debtors in possession, including RFC. Id. at 160-61.
All exhibit pincites refer to the center page numbers stamped on each exhibit.
3. PRMI is a corporation organized under the laws of Nevada. PTX-009. Founded in 1998, PRMI is a mortgage lender that sold loans to RFC during the relevant time period. PTX-3; Zitting 1645:9-14.
All citations to trial testimony specify the testifying witness, followed by pincites to the relevant pages and lines of the trial transcript. All citations to testimony admitted by deposition designation specify the testifying witness, followed by "Dep.," followed by pincites to the relevant pages and lines of the designated testimony.
B. RFC's Role in the Mortgage Backed Securities Market
4. For a number of years prior to RFC's 2012 Bankruptcy filing, RFC participated in the RMBS market. Between 1990 and 2000, RFC was among the top national RMBS sponsors. Farley 101:19-102:2. As such, it functioned as a middleman in the RMBS market. In its role as a buyer, it purchased mortgages from banks and originating lenders, such as PRMI. See Common SJ Order, 332 F. Supp. 3d at 1101, 1118; Farley 123:22-25. Then, as a seller, RFC pooled the loans together into bundles of hundreds or thousands of home mortgage loans, and sold them into RMBS Trusts. Common SJ Order, 332 F. Supp. 3d at 1117-18; Farley 123:25-124:2. As noted, the RMBS Trusts paid for the loans by issuing securities to investors, for which the mortgage loans served as collateral. Common SJ Order, 332 F. Supp. 3d at 1117.
C. Client Contracts
5. In its role as a buyer of loans from originating lenders, or "Clients," RFC and the lenders, including PRMI, entered into agreements called "Client Contracts." Id. at 1118. The Client Contract formed the foundation of the business relationship between RFC and the originators. Bangerter 340:22-341:10. Among other things, the Client Contract set out the originator's obligations in connection with selling loans to RFC and incorporated by reference a much longer, more detailed document, sometimes called a "Seller Guide" or "Client Guide." PTX-003 at 1; PTX-010 at 1; Bangerter 344:19-345:23, 373:8-19.
6. It is undisputed that PRMI's sale of loans to RFC was governed by two Client Contracts: a March 30, 2000 Client Contract ("March 2000 Client Contract"), PTX-003; Zitting 1657:6-1658:18, followed by a June 25, 2001 Client Contract ("June 2001 Client Contract"), PTX-010; Zitting 1679:2-11. The March 2000 Client Contract incorporated by reference a different Guide—the AlterNet Guide. PTX-003 ¶ 1. The June 2001 Client Contract incorporated by reference the Client Guide. PTX-010 ¶ 1.
1. The March 2000 Client Contract
7. The parties entered into their first contract on March 30, 2000. PTX-003. RFC Marketing Assistant and then Sales Director Renee Bangerter (f/k/a Renee Popiolek) was involved on RFC's behalf in securing the parties' execution of this contract. Bangerter 338:8-22, 341:22-342:3, 343:15-344:4. The Court found Ms. Bangerter's testimony at trial to be highly credible. The March 2000 Client Contract was a standard-form contract that governed the relationship between RFC and PRMI. Id. at 341:2-10.
8. The March 2000 Client Contract approved PRMI to sell loans pursuant to the AlterNet Guide and incorporated all representations and warranties ("Reps") therein. PTX-003 ¶¶ 1, 3; Bangerter 344:19-345:23, 347:16-348:15. Among other terms, the Client Contract provided that it "may not be amended or modified orally, and no provision of this Contract may be waived or amended except in writing signed by the party against whom enforcement is sought." PTX-003 ¶ 2. Any "such [ ] written waiver or amendment must expressly reference this Contract." Id. (emphasis added). Further, in the event of PRMI's default under the Client Contract, RFC retained the right to exercise one or more of the remedies set forth in the AlterNet Guide. PTX-003 ¶ 4.
2. The June 2001 Client Contract
9. In June 2001, PRMI underwent a corporate restructuring in which the business shifted from a Utah corporation to a Nevada corporation. PTX-009. As a result, RFC and PRMI executed a letter agreement affirming that the new Nevada corporation would assume all obligations and liabilities of the prior Utah corporation ("Assumption Agreement"). Id.; Bangerter 370:25-371:13. The Assumption Agreement expressly referenced the March 2000 Client Contract and "the RFC Seller Guide incorporated therein by reference." PTX-009.
10. Five days after the Assumption Agreement, the parties executed an updated Client Contract, dated June 25, 2001. PTX-010. Ms. Bangerter oversaw the execution of this contract as well. Bangerter 372:18-24.
11. While similar in many ways to the March 2000 Client Contract, the June 2001 Client Contract expressly incorporated the Client Guide, rather than the AlterNet Guide. PTX-010 ¶ 1. By this point in time, RFC's subprime AlterNet program was available to all originators and RFC thus no longer needed a separate AlterNet Guide. Accordingly, the AlterNet Guide was merged into the Client Guide. Bangerter 373:12-374:3.
12. Similar to the March 2000 Client Contract, under the June 2001 Client Contract, PRMI made to RFC all the Reps set forth in the Client Guide. PTX-010 ¶ 4. Likewise, the June 2001 Client Contract stated it "may only be amended in writing signed by both parties." Id. ¶ 3; Bangerter 375:8-16. In addition, the June 2001 Client Contract allowed the parties to execute commitment letters, which could serve to modify specific terms of the sale of loans from PRMI to RFC. PTX-010 ¶ 2; Bangerter 375:1-7. In the event of PRMI's default, RFC had the right to exercise one or more of the remedies set forth in the Client Guide. PTX-010 ¶ 6. The June 2001 Client Contract also expressly provided that it "restates, amends, and supersedes any and all prior Contracts or agreements between the parties . . . ." Id. ¶ 9; Bangerter 375:18-376:8.
D. The AlterNet Guide and the Client Guide
13. The Court has already held that—barring the defenses of estoppel and waiver (discussed below)—all At-Issue Loans were governed by either the AlterNet Guide or the Client Guide. PRMI SJ Order, 428 F. Supp. 3d at 88 ("Setting aside, for a moment, those loans originated to Assetwise or Countrywide, there is no genuine dispute of material fact that the Guides governed all other loans, and ResCap's motion on that point is granted."); see also PTX-001; PTX-003; PTX-010; PTX-032; Bangerter 377:9-378:5. Pursuant to the March 2000 Client Contract, the AlterNet Guide governed all loans PRMI sold to RFC between March 30, 2000, and June 25, 2001. PRMI SJ Order, 428 F. Supp. 3d at 63-64; PTX-003; Zitting 1658:1-18. Pursuant to the June 2001 Client Contract, the Client Guide governed all loans PRMI sold to RFC on and after June 25, 2001. PRMI SJ Order, 428 F. Supp. 3d at 63-64; PTX-010; Zitting 1679:2-11.
As discussed in greater detail herein, the "At-Issue Loans" were loans that: (1) PRMI sold to RFC; (2) underlie the RMBS Trusts or Monoline claims; and (3) as of May 2013, had $500+ of realized losses or were 90+ days delinquent, in foreclosure, or real-estate owned.
14. At the end of 2005, PRMI closed its wholesale division and stopped selling subprime loans to RFC. Zitting 1727:23-1728:22. After that closure, PRMI continued to sell only a very small number of prime loans to RFC. Id. at 1729:1-6. The vast majority of loans sold by PRMI to RFC were sold in the early years of the relationship. See McCrary 2518:6-15 (noting that 69% of PRMI's At-Issue Loans were in earlier trusts).
15. Under the terms of the Client Contracts, PRMI acknowledged that it had "received and read" the applicable Guide, PTX-003 ¶ 1, and that it agreed to sell loans to RFC pursuant to the terms and conditions of the applicable Guide as amended from time to time, PTX-010 ¶ 1. The Client Guide was available to PRMI via a lender internet portal. Bangerter 382:2-13; Crawford 2452:23-2453:21. In addition, RFC had a practice of providing updates to the Guide directly to clients such as PRMI. See, e.g., PTX-361; accord Bangerter 382:2-384:25.
E. The Guides' Key Provisions
1. Sole Discretion
16. Section 141(B) and Section 113(B) of the Client Guide vests RFC with broad authority and sole discretion to make determinations of fact and decisions to act, stating:
(B) GMAC-RFC's Sole Discretion
Whenever any provision of this Client Guide contract requires []RFC to make a determination of fact or a decision to act, or to permit,
approve or deny another party's action such determination or decision shall be made in []RFC's sole discretion.See, e.g., PTX-032 § 141(B); PTX-1055 § 113(B); PRMI SJ Order, 428 F. Supp. 3d at 65
2. Knowledge, Reliance, and Waiver
17. The Guides' "knowledge" standard holds PRMI to a strict standard of both actual and constructive knowledge:
(A) "Knowledge" StandardSee, e.g., PTX-032 § 141(A); PTX-1055 § 113(A); PRMI SJ Order, 428 F. Supp. 3d at 64.
Whenever any representation, warranty, or other statement contained in this Client Guide is qualified by reference to a Client's "knowledge" or "to the best of" a party's "knowledge", such "knowledge" shall be deemed to include knowledge of facts or conditions of which Client, including (without limitation) any of its directors, officers, agents, or employees, either is actually aware or should have been aware under the circumstances with the exercise of reasonable care, due diligence, and competence in discharging its duties under this Client Guide and the Program Documents. All matters of public record shall be deemed to be known by the Client. Any representation or warranty that is inaccurate or incomplete in any material respect is presumed to be made with the knowledge of Client, unless Client demonstrates otherwise. "Due diligence" means that care which Client would exercise in obtaining and verifying information for a Loan in which Client would be entirely dependent on the Mortgaged Property or Mortgagor's credit as security to protect its investment.
18. In both Section 250 of the AlterNet Guide and Section A200 of the Client Guide, the originating lenders made substantively identical general Reps to RFC. PRMI acknowledged that RFC purchased loans in reliance on PRMI's Reps, and PRMI agreed to assume liability for any misrepresentations or breaches of the Guides, regardless of its or RFC's knowledge. See PTX-001 § 250; PTX-032 § A200; PRMI SJ Order, 428 F. Supp. 3d at 65.
PRMI disputes whether the 1997 version of the AlterNet Guide produced by ResCap and found at PTX-001 is the correct version of the AlterNet Guide applicable to the PRMI At-Issue Loans. The Court notes this disagreement, and addresses it in greater detail in its discussion of the loans in question.
19. Under both Guides, PRMI also agreed that no provisions of the AlterNet Guide or the Client Guide would be waived unless RFC did so in writing:
[AlterNet Seller] Representations and Warranties and CovenantsSee PTX-001 § 250; PTX-032 § A200; PTX-1055 § A200; PRMI SJ Order, 428 F. Supp. 3d at 65.
The [AlterNet Seller] Client acknowledges that [RFC] GMAC-RFC purchases Loans in reliance upon the accuracy and truth of the [AlterNet Seller's] Client's warranties and representations and upon the [AlterNet Seller's] Client's compliance with the agreements, requirements, terms and conditions set forth in the [AlterNet Seller] Client Contract and this [AlterNet] Client Guide.
All such [Reps] are absolute, and the [AlterNet Seller] Client is fully liable for any misrepresentation or breach of warranty regardless of whether it or [RFC] GMAC-RFC actually had, or reasonably could have been expected to obtain, knowledge of the facts giving rise to such misrepresentation or breach of warranty.
The [Reps] pertaining to each Loan purchased by [RFC] GMAC-RFC survive the Funding Date, any simultaneous or post-purchase sale of servicing with respect to the Loan and any termination of the [AlterNet Seller] Client Contract, and are not affected by any investigation or review made by, or on behalf of, [RFC] GMAC-RFC except when expressly waived in writing by [RFC] GMAC-RFC.
Consistent with this Court's summary judgment order, differences between the AlterNet Guide and Client Guide are denoted by placing varying AlterNet Guide text in brackets.
3. Specific Representations and Warranties
20. Under both the AlterNet Guide and Client Guide, PRMI affirmatively made specific Reps to RFC regarding "individual loans," including information about the loans' eligibility for sale to RFC and accuracy of information provided to RFC. PTX-001 § 251-1; PTX-032 § A202; PTX-1055 §§ A202, A200; PRMI SJ Order, 428 F. Supp. 3d at 66. Among other things, PRMI represented it had:
• Verified the accuracy of information used by borrowers to obtain the loans, PTX-001 § 251-1(A); PTX-032 § A202(A); PTX-1055 § A202(A); PRMI SJ Order, 428 F. Supp. 3d at 66;
• Ensured the proper completion and execution of loan forms, PTX-001 § 251-1(D); PTX-032 § A202(D); PTX-1055 § A202(D); PRMI SJ Order, 428 F. Supp. 3d at 66;
• Ensured that no default or other breach of loan terms existed in any loan, PTX-001 § 251-1(G); PTX-032 § A202(G); PTX-1055 § A202(G); PRMI SJ Order, 428 F. Supp. 3d at 66;
• Complied with applicable laws, PTX-001 §§ 251-1(D), 251-1(I); PTX-032 § A202(I); PTX-1055 § A202(I); PRMI SJ Order, 428 F. Supp. 3d at 66; and
• Confirmed the market value of the mortgaged property, PTX-001 § 251-1(T); PTX-032 § A202(T); PTX-1055 § A202(T); PRMI SJ Order, 428 F. Supp. 3d at 66
21. Section 251 of the AlterNet Guide and A202 of the Client Guide also required PRMI to represent and warrant:
• For each loan sold to RFC, "no circumstances exist involving the Loan Documents, the Mortgage Premises, the Borrower or the Borrower's credit standing that could: (i) cause private institutional investors to regard the Loan as an unacceptable investment, (ii) cause the Loan to become delinquent, or (iii) adversely affect the Value or marketability of the Mortgage Premises or the Loan[,]" PTX-001
§ 251-1(Q); PTX-032 § A202(Q); PTX-1055 § A202(Q); Zitting 1764:4-1765:5; and
• Each Loan was of "investment quality, has been prudently originated and has been underwritten in compliance with all requirements of the [] Guide[,]" PTX-001 § 251-1(T); PTX-032 § A202(T); PTX-1055 § A202(T); Zitting 1762:16-1763:24 ("Q. And PRMI had to make sure that the loan was prudently originated? A. It did.").
22. With respect to loans sold under the Client Guide, PRMI also warranted that it had not sold any "high risk" loans to RFC. PTX-1055 § A202(J)(1)(d); PRMI SJ Order, 428 F. Supp. 3d at 66.
4. Events of Default and Non-Exclusive, Cumulative Remedies
23. If originating lenders breached these Reps by committing an "Event of Default," both the AlterNet Guide and the Client Guide granted RFC wide-ranging recourse. Under AlterNet Guide Section 270 and Client Guide Section A209, "Non-Exclusive, Cumulative Remedies," the Guides broadly provided that "RFC may exercise any remedy outlined in this [] Guide," as well as "[a]ny other rights which it may have at law or in equity deemed appropriate to protect its interest." PTX-001 § 270(A); PTX-032 § A220(A); PTX-1055 § A209; PRMI SJ Order, 428 F. Supp. 3d at 66. RFC's exercise of one particular remedy did not prevent it from exercising "[o]ne or more other remedies in connection with the same Event of Default" and/or "[a]ny other rights which it may have at law or in equity." PTX-001 § A270(A); PTX-032 § A220(A); PTX-1055 § A209(A); PRMI SJ Order, 428 F. Supp. 3d at 66.
5. Repurchase
24. The Guides provided RFC with several remedies, including repurchase and indemnity. Under the repurchase provision of both the AlterNet Guide and Client Guide, if RFC determined that an Event of Default had occurred with respect to a particular loan, it could require the originating lender to repurchase the loan within 30 days of receiving notification from RFC. PTX-001 § 271(C); PTX-032 § A221(A); PTX-1055 § A210(A); PRMI SJ Order, 428 F. Supp. 3d at 66-67. Notably, "[]RFC is not required to demand repurchase within any particular period of time, and may elect not to require immediate repurchase. However, any delay in making this demand does not constitute a waiver by []RFC of any of its rights or remedies." PTX-001 § 271(C); PTX-032 § A221(A); PTX-1055 § A210(A); PRMI SJ Order, 428 F. Supp. 3d at 67.
25. Even if RFC determined that repurchase was not the appropriate remedy, under both the AlterNet Guide and Client Guide, the originator nonetheless remained obligated to pay RFC "all losses, costs and expenses incurred by []RFC and/or the Loan's Servicer as a result of an Event of Default," including reasonable attorneys' fees and related costs incurred in connection with any enforcement efforts. PTX-001 § 271(C); PTX-032 § A221(A); PTX-1055 § A210(A); PRMI SJ Order, 428 F. Supp. 3d at 67.
6. Indemnification
26. The Guides also provided RFC with wide-ranging indemnification in the event of an originating lender's default. The indemnification provision required the originating lender to indemnify RFC from "all losses, damages, penalties, fines, forfeitures, court costs and reasonable attorneys' fees, judgments, and any other costs, fees and expenses resulting from any Event of Default." PTX-001 § 274; PTX-032 § A223, PTX-1055 § A212; PRMI SJ Order, 428 F. Supp. 3d at 67.
27. Version of the Client Guide from December 1, 2005 forward provided examples of the type of Client conduct requiring indemnification:
This includes, without limitation, liabilities arising from (i) any act or failure to act, (ii) any breach of warranty, obligation or representation contained in the Client Contract, (iii) any claim, demand, defense or assertion against or involving []RFC based on or resulting from such breach, (iv) any breach of any representation, warranty or obligation made by []RFC in reliance upon any warranty, obligation or representation made by the Client contained in the Client Contract and (v) any untrue statement of a material fact, omission to state a material fact, or false or misleading information provided by the Client in information required under Regulation AB or any successor regulation.PTX-1056 § A212; PRMI SJ Order, 428 F. Supp. 3d at 67.
28. Versions of the Client Guide from July 1, 2002 forward contained additional language regarding PRMI's broad indemnification obligations to RFC:
In addition, Client shall indemnify []RFC against any and all losses, damages, penalties, fines, forfeitures, judgments, and any other costs, fees and expenses (including court costs and reasonable attorneys' fees) incurred by []RFC in connection with any litigation or governmental proceeding that alleges any violation of local, state or federal law by Client, or any of its agents, or any originator or broker in connection with the origination or servicing of a Loan. With regard to legal fees or other expenses incurred by or on behalf of []RFC in connection with any such litigation or governmental proceeding, Client shall reimburse []RFC for such fees and expenses. . . . Except for notices for reimbursement, []RFC is not required to give Client notice of any litigation or governmental proceeding that may trigger indemnification obligations. Client shall instruct its officers, directors and agents (including legal counsel) to cooperate with []RFC in connection with the defense of any litigation or governmental proceeding involving a Loan. []RFC has the right to control any litigation or governmental proceeding related to a Loan, including but
not limited to choosing defense counsel and making settlement decisions.PTX-032 § A223; PTX-1055 § A212; PRMI SJ Order, 428 F. Supp. 3d at 68.
F. RFC's Securitization Business
1. Teresa Farley, ResCap Trial Witness and ResCap's 30(b)(6) Witness on Securitization
29. Teresa Farley, former co-head of RFC's securitization group and a member of its leadership team, Farley 99:9-16, testified extensively about RFC's securitization business, including RFC's subjective understanding of certain representations, warranties, and disclaimers. Id. at 96:14-253:16, 282:17-328:24. Ms. Farley is an attorney who, prior to joining RFC, practiced law for several years at the Minneapolis law firm of Dorsey & Whitney, where she worked on RMBS transactions. Id. at 97:11-99:2.
30. During her tenure at RFC from 1989 to 2000, Ms. Farley negotiated RMBS-related contracts, including representations and warranties regarding mortgage loans. Farley 99:1-2, 101:6-7, 100:3-11. She also participated as a panelist in numerous industry conferences and seminars on issues involving securitization, and co-founded the Securitization Committee of the Mortgage Bankers Association. Id. at 103:7-107:8. After leaving RFC, she helped to start a smaller company, similar to RFC, that was soon acquired by CreditSuisse. Id. at 108:3-8. In her work with CreditSuisse, Farley was responsible for building an origination platform that could utilize securitization. Id. at 108:9-14. In 2004, she took a position at Merrill Lynch, building a warehouse lending platform in which she lent money to originators secured by and for the purpose of originating individual mortgage loans. Id. at 109:19-110:3. After leaving Merrill Lynch, Farley performed lending facility work for RFC/ResCap, oversaw compliance, and did some sales work Id. at 114:13-115:1, 118:2-11.
31. Following ResCap's Bankruptcy, Farley continued to work for ResCap, evaluating tax restructuring, and, in 2012, served as the senior director of the asset disposition group. Id. at 118:12-23. In her role with the asset disposition group, in approximately 2016 and 2017, Farley worked on liquidating assets, including RMBS securities issued by RFC in the 2000s. Id. at 121:2-21. She continues to perform work for ResCap at the present as a consultant. Id. at 327:22-328:24. The Court found Ms. Farley's trial testimony to be highly credible.
32. In addition to her trial testimony, Ms. Farley had previously testified as ResCap's corporate designee on the topic of securitization, in response to PRMI's Rule 30(b)(6) deposition notice. See In re ResCap Liquidating Tr. Litig. ("PRMI Daubert Order"), 432 F. Supp. 3d 902, 923 (D. Minn. 2020).
2. RFC Relied on the Truthfulness of Originator Representations and Warranties During Securitization
33. RFC relied on the representations it received from originators like PRMI in order to make truthful representations to the RMBS market. Farley 126:7-9. If RFC's securitization representations exposed it to liability, RFC "wanted the ability to put the loan back to the originator" under the originator's representations to RFC. Id. at 125:15-126:1. Moreover, if RFC could not rely on originator representations, as reflected in the Client Guide, RFC could not have securitized the loans, because "the rating agencies, the underwriters, [and] the investors couldn't have relied on [RFC's] rep[resentations]." Id. at 126:10-25.
34. Accordingly, whenever RFC made a new securitization representation, it attempted to ensure that it was "no broader than the representations [RFC] w[as] getting in the Client Guide from [RFC's] clients[,] [k]ind of back to back." Id. at 125:15-232:1-7. The representations made to RFC pursuant to the Client Guide were generally comparable to the representations RFC made in turn in its securitization transaction documents. Id. at 126:2-6; 231: 20-232:7.
35. RFC's securitization transactions were reflected in extensive and detailed transaction documents, which were variously reviewed by RFC's outside counsel, securities underwriters like Merrill Lynch and Bear Stearns and their counsel, a mortgage loan due diligence firm, credit rating agencies, a RMBS trustee and its counsel, accountants, and, depending on the transaction, monoline insurers and their counsel. Farley 127:6-128:19. The Court turns its attention to several of those key transaction documents.
3. The Key RFC Securitization Transaction Documents
a. RFC Loan Sale Agreements
36. For securitizations collateralized by first lien loans, RFC, as sponsor, used "Assignment and Assumption Agreements" to sell loans to a depositor entity. Farley 129:5-21; see, e.g., PTX-1150. The depositor was a special purpose entity established for the purpose of securitizing assets. Farley 130:19-131:5. For securitizations collateralized by second lien loans, RFC used either Assignment and Assumption Agreements or Home Equity Loan Purchase Agreements. Id. at 130:19-131:5; see, e.g., PTX-1156.
37. The Assignment and Assumption Agreement was a "fairly standard ... loan sale document" that was "[r]elatively short" and provided for the sale of the loans from RFC to the depositor. Farley 129:22-130:4, 145:5-17, 145:22-24. The Assignment and Assumption Agreements included both RFC's Reps to the depositor and the remedies for breach of those representation and warranties. Id. at 129:22-130:4, 145:25-146:2.
38. The mortgage loan representations in RFC's loan sale agreements concerned individual mortgage loans. Id. at 130:5-10, 147:8-14; see, e.g., PTX-1150 at 3 ("RFC represents and warrants to the [Depositor], with respect to each Mortgage Loan").
39. So too were the remedies, which required RFC to either cure, repurchase, or substitute individual mortgage loans. Farley 130:11-18, 147:15-22, 161:19-23; see, e.g., PTX-1150 at 8-9 ("Upon discovery by RFC . . . of a breach of the foregoing representations and warranties in respect of any Mortgage Loan . . . RFC shall . . . purchase such Mortgage Loan . . . ."); PTX-1156 at 21-22 ("Upon discovery by [RFC] . . . of a breach of any representation or warranty . . . with respect to any Home Equity Loan . . . [RFC] shall . . . repurchase such Home Equity Loan . . . .").
b. RFC's Pooling and Servicing Agreements
40. The depositor, in turn, sold the mortgage loans to RMBS trusts using a "Pooling and Servicing Agreement." Farley 133:8-134:1. The Pooling and Servicing Agreement was a sales document that also described the duties of the master servicer and trustee after the transaction closed, and set forth the obligations of the different classes of securities. Id. at 133:8-134:1, 149:19-150:9, 152:24-153:13, 153:24-154:3. The Pooling and Servicing Agreements were approximately 100 pages long—longer than the Assignment and Assumption Agreements because the latter were "pure sales contract[s]" and the former included obligations relating to cash flows, the master servicer, and the trustee. Id. at 134:11-25.
RFC's Assignment and Assumption Agreements, Home Loan Purchase Agreements, and Pooling and Servicing Agreements are collectively referred to as the "Trust Agreements," and RFC's representations and warranties contained therein are collectively referred to as the "Trust Reps."
41. Through the Pooling and Servicing Agreements, the RFC depositor entity assigned the Reps it received in the loan sale agreements to the RMBS trustee for the benefit of the certificateholders. Id. at 152:5-20. RFC's depositor entity also made additional loan-level representations to the RMBS trust in the Pooling and Servicing Agreements. Id. at 133:8-134:1, 150:10-151:1, 151:7-17, 154:4-6; see, e.g., DTX-1016.0115 at 36-38.
42. The Pooling and Servicing Agreements also contained the remedy for breaches of RFC's Reps. Farley 161:19-23. The Pooling and Servicing Agreements provided loan-level remedies that operated on individual mortgage loans. Id.; see, e.g., DTX-1016.0115 at 38 ("Upon discovery . . . of a breach of any of the [Reps] [the Depositor shall] purchase such Mortgage Loan . . . .").
43. The purpose of the loan sale agreements and the Pooling and Servicing Agreements was to provide meaningful representations to investors, which was reinforced by the repurchase remedy and the fact that RFC's Reps "were passed along to the [RMBS] trust." Farley 190:21-191:10. As such, RFC's Reps in the transaction agreements required a reasonable method of enforcement. Id. at 191:11-15.
c. Loan Tape and Mortgage Loan Schedule
44. As part of RFC's RMBS transactions, a loan tape was created and distributed to the transaction parties. Id. at 136:15-137:10. The loan tape set out 100-120 different attributes for each individual mortgage loan, including principal and interest, property address, credit scores, borrower income and debt, appraisers who were used, and the type of valuation used. Id. at 137:11-22. The vast majority of the information reflected in the loan tape came from the originators that sold RFC the loans. Id. at 137:23-138:4. RFC relied on the accuracy of the originators' information in compiling the loan tape. Id. at 138:5-7.
45. At the end of each RMBS transaction's lifecycle, a final loan tape was created, followed by the creation of a mortgage loan schedule ("MLS") from that final loan tape. Id. at 138:8-139:6. The MLS contained a subset of data fields from the loan tape and provided "the core fields one would need to disclose to really sell the loan [to a purchaser]." Id. at 139:7-25, 141:7-9, 141:18-22. The MLS had 20-22 data fields. Id. at 141:16-17. As RFC witness Theresa Farley explained, the MLS was the "heart and soul" of the RMBS transaction. Id. at 143:11-144:4.
46. The MLS was included with the Trust Agreements and "identified the loans that were being sold to the trust under that agreement." Id. at 139:7-25, 142:22-143:10. The MLS contained fields that identified the loans, their financial terms and origination dates, and fundamental credit risk data like loan-to-value ("LTV") ratio, owner occupancy status, and, for second lien loans, debt-to-income ("DTI") ratio. Id. at 139:7-25. The data fields included on the MLS were connected to specific risks associated with each loan and transaction. Id. at 167:13-18.
Loan-to-value ratio, or LTV, is a measure that compares the amount of a mortgage with the appraised value of the property. Consumer Financial Protection Bureau, What is loan-to-value ratio and how does it relate to my costs? (Nov. 15, 2019), https://www.consumerfinance.gov/ask-cfpb/what-is-a-loan-to-value-ratio-and-how-does-it-relate-to-my-costs-en-121/ (last visited August 11, 2020). Mortgage lenders use LTV, among several factors, to determine the risk of lending to the borrower. Id.
Debt-to-income ratio, or DTI, is a calculation of a borrower's monthly financial obligations divided by the borrower's gross monthly income. Consumer Financial Protection Bureau, What is a debt-to-income ratio? (Nov. 15, 2019), https://www.consumerfinance.gov/ask-cfpb/what-is-a-debt-to-income-ratio-why-is-the-43-debt-to-income-ratio-important-en-1791/ (last visited August 11, 2020). DTI is another means by which mortgage lenders evaluate the risk of lending to a borrower, by measuring a borrower's ability to manage monthly payments. Id.
47. Each loan's LTV ratio was "a key component in understanding the risk that is associated [with] a loan both in terms of speed and credit risk and then the loss." Id. at 140:24-141:6. The MLS included LTV information because it was a "critical component in a mortgage loan," and there was a direct correlation between the LTV ratio and credit risk and, if the loan fails to perform, "LTV is probably the biggest determinant in the loss you're going to take ... on that loan." Id. at 166:14-167:1.
48. A loan's DTI ratio "goes to the borrower's willingness and ability to pay." Id. at 140:24-141:6. The MLS for second lien ("RFMSII") transactions included DTI information because, for those loans, "the amount of discretionary income that a borrower has is going to be directly relevant to that person's ability to pay a home equity line or a second lien." Id. at 167:2-12.
49. Each MLS also included an occupancy code that reflected whether the property was owner occupied, because there was a strong correlation between owner occupancy and loss. Id. at 165:18-25.
d. RFC's Prospectus Supplements
50. RFC's RMBS trusts issued securities backed by the cash flows on the mortgage loans owned by the trusts. Id. at 135:1-7. RMBS investors purchased securities that had rights to the mortgage loan cash flows. Id. at 135:16-136:3.
51. As part of its RMBS transactions, RFC created "Prospectus Supplements," pursuant to SEC requirements, which disclosed and described the RMBS offering to investors. Id. at 135:8-15. Whereas the Trust Agreements made loan-level representations with loan-level remedies, the Prospectus Supplements made disclosures, and the remedy for inaccurate information in the Prospectus Supplement was understood to be a lawsuit under the Securities Acts, id. at 136:4-13, 161:24-162:5, such as the Securities Act of 1933 and the Securities Exchange Act of 1934, which provide for civil liability for false or misleading statements or omissions with respect to securities. See 15 U.S.C. §§ 77l; 78r (2018).
52. The Prospectus Supplements contained lengthy summaries describing "everything material about the mortgage loans that ... the securities rel[ied] on [for] their payments." Id. at 155:2-11. The Prospectus Supplements also contained disclosures and statistical tables reflecting the "characteristics of the individual loans in aggregate as they make up the pool of loans for which . . . the securities are relying on for their payment." Id. at 155:12-19, 156:6-13. The Prospectus Supplements included data collected from the loan tape, including weighted average percentages calculated based on individual loans, because "there was so much more [than the MLS data] that was important to understand when you were evaluating the risk of a transaction." Id. at 143:11-144:4.
53. In particular, Prospectus Supplements contained statistical tables disclosing information about loan documentation programs and credit grades, which were important disclosures for trusts with reduced documentation programs or borrowers with credit issues. Id. at 156:23-157:7, 159:16-161:6. These loan documentation programs and credit grades were defined in the Client Guide. Id. at 156:23-157:18, 193:9-12. Investors looking at documentation programs in the Prospectus Supplement "kn[e]w that they would be looking at . . . the Client Guide to understand what the program parameters were to be able to utilize reduced doc, no stated income, no income/no asset [loan programs]." Id. at 157:8-18. Credit grades were also categorized based on credit and underwriting characteristics. Id. at 193:13-22.
54. The statistical disclosures in the Prospectus Supplements were calculated based on the categorization of each individual loan. Id. at 158:3-20. If the data were wrong for any of the loans, it would change the percentages in the Prospectus Supplement. Id. at 158:21-159:3.
55. Investors reviewed these stratifications and disclosures, including those "based on the sponsor or originators' grading of mortgage loans," in "consider[ing] an investment" and modeling cash flow analysis. Burnaman 1563:8-22. The disclosures provided "comfort . . . on the disclosure items that [investors] look[ed] [at] to create the cash flow models." Id.
4. RFC's Securitization Shelves
56. There were five different RFC RMBS securitization shelves. Farley 131:6-133:1. Each of the shelves had different types of loans collateralizing the trusts: (a) the RFMSI shelf securitized "Jumbo A collateral," large loans to borrowers with higher credit quality, Butler 691:22-692:15; (b) the RFMSII shelf securitized closed-end second liens and home equity lines of credit, Farley 130:19-132:19, 167:2-12, Butler 692:23-693:2; (c) the RALI shelf securitized "Alt-A" loans for creditworthy borrowers but offered expanded, non-full documentation loan programs, Farley 131:6-132:19; Butler 692:16-22; (d) the RASC shelf securitized "subprime loans" for challenged borrowers with credit issues, Farley 159:16-21; Butler 693:13-21; and (e) the RAMP shelf securitized "miscellaneous" collateral that did not fit into the other shelves, Farley 131:6-132:19; Butler 694:2-12. The Trust Agreements remained consistent across deals for each shelf. Farley 133:5-7.
5. The RMBS Trustees
57. An RMBS trustee administered each of RFC's RMBS securitizations. See ResCap Liquidating Trust v. Home Loan Ctr., Inc., No. 14-cv-1716 (D. Minn.) ("HLC"), Lipps 1249:22-1250:4. The trustee was authorized to seek remedies against RFC to recover for investors' losses experienced by the securitization due to contractual breaches by RFC. Id. at 1374:7-16; see also Hawthorne 1009:19-1010:7. The RMBS Trustees for RFC's securitizations included: Deutsche Bank National Trust Company and Deutsche Bank Trust Company Americas, Bank of New York Mellon, U.S. Bank N.A., HSBC Bank USA, N.A., and Wells Fargo Bank, N.A. PTX-153 at 1-2; PTX-181 at 16, 30.
The Court cites to relevant testimony from a related 2018 jury trial before the undersigned judge, ResCap Liquidating Trust v. Home Loan Center, Inc., No. 14-cv-1716 (D. Minn.) ("HLC"). The HLC trial testimony of Mr. Lipps admitted pursuant to the Court's February 10, 2020 Order [Doc. No 5421] is attached as Appendix A [Doc. No. 5413-1] to the parties' February 7, 2020 Stipulation [Doc. No. 5413] and is cited herein as "Lipps HLC [citation]."
6. The Monolines
58. Certain of RFC's securitizations carried financial guaranty insurance furnished by monoline insurers. Lipps HLC 1217:23-1219:6. The monoline insurers for RFC's securitizations included MBIA Insurance Corporation ("MBIA"), Financial Guaranty Insurance Company ("FGIC"), Ambac Assurance Corporation ("Ambac"), and Syncora Guarantee Inc. ("Syncora") (collectively, the "Monolines"). Id. at 1350:3-8. Under the insurance policies, the Monolines generally guaranteed that investors would receive timely payments of principal and interest on their notes or certificates. Id. at 1218:4-23. If a defaulted loan caused a trust to be unable to timely pay its investors, the Monolines were to compensate the trust for the shortfall. Id.
Assured Guaranty Municipal Corporation (formerly known as Financial Security Assurance, Inc.) ("Assured") also filed a proof of claim and ultimately settled with RFC. Plaintiff does not seek indemnity for the Assured settlement.
G. RFC's Bankruptcy and Pre-Petition Litigation
59. Leading up to 2008, RFC's securitizations experienced substantial losses as loans began defaulting at a high rate. See Hamzehpour Dep. 125:12-126:11, 126:15-127:1. Following the financial crisis, RFC became subject to substantial litigation and threatened litigation from Monolines and RMBS investors, and RFC ultimately filed for bankruptcy. The relevant facts are summarized below.
1. The Pre-Bankruptcy Claims and Litigation
60. As described above, certain of RFC's securitizations were insured by the Monolines. In some circumstances, the Monolines were required to make payments to the RMBS Trusts on account of defective loans. By 2010, one of the Monolines, MBIA, had paid over a billion dollars. Lipps HLC 1223:15-25; see also id. at 1229:7-21, 1242:1-12; DTX-494.
61. MBIA sued RFC in 2008, alleging among other things, loan-level breaches of Reps. Lipps HLC 1217:13-22, 1219:7-20, 1224:13-14.
62. In late 2011, another Monoline, FGIC, filed the first of 12 lawsuits against RFC and its affiliates. Id. at 1247:19-1248:1, 1331:10-16. FGIC's lawsuits had not progressed as far as MBIA's lawsuit by the time RFC filed for bankruptcy. See id. at 1248:24-1249:12.
63. Three days before RFC filed for bankruptcy, a third monoline insurer, Assured Guaranty, sued RFC. Id. at 1248:17-23.
64. MBIA's, FGIC's, and Assured's lawsuits were pending at the time RFC filed for bankruptcy on May 14, 2012. Id. at 1248:2-19.
65. A group of investors in the RMBS Trusts also threatened RFC with litigation.
66. The investors (the "Gibbs Group") retained an RMBS litigator, Kathy Patrick, of the law firm Gibbs & Bruns LLP. Id. at 1250:10-20, 1251:4-13; Hawthorne 1097:15-23. In October 2011, Ms. Patrick sent a demand letter to RFC on behalf of the Gibbs Group, warning that her clients intended to direct the RMBS Trustees to commence litigation if RFC was not able to resolve their demands. Lipps HLC 1250:10-20. About one month later, attorney Talcott Franklin sent a similar letter on behalf of another group of RMBS Trust investors (the "Talcott Group," and with the Gibbs Group, the "Trust Investor Groups"). Id. at 1250:21-1251:3. Both letters alleged, among other things, claims for breaches of Reps. Id. at 1254:12-22.
2. The Original RMBS Settlement
67. RFC and its affiliates negotiated with the Trust Investor Group for a number of months prior to RFC's May 2012 bankruptcy. Id. at 1269:1-9.
68. RFC and certain of its affiliates entered into a settlement agreement with the Trust Investor Groups in May 2012 (the "Original RMBS Trust Settlement"). See id. at 1255:14-24. The Original RMBS Trust Settlement would have resolved all claims for up to 392 securitization trusts created between 2004 and 2007, for an allowed claim in the anticipated bankruptcy of RFC and its affiliates of $8.7 billion. Id. at 1255:14-21, 1257:19-1258:7. The $8.7 billion allowed claim reflected a settlement of the actual and projected losses of over $40 billion. Id. at 1258:8-13.
69. The Original RMBS Trust Settlement provided that the settlement would be allocated pro rata on the basis of the sum of the net losses that had been and were estimated to be experienced by each such RMBS Trust through the date of its termination. Hawthorne 1098:16-1099:17; PTX-179 ¶ 115. The Original RMBS Trust Settlement would not have resolved the claims of all RFC-sponsored RMBS Trusts, see Lipps HLC 1266:25-1267:10, 1278:8-12, nor would it have resolved the Monolines' claims, id. at 1258:18-24.
70. Tammy Hamzehpour, RFC's General Counsel, testified as to RFC's decision to enter into the Original RMBS Trust Settlement:
I think largely the uncertainty of the outcome of the litigation. They -- the [RFC] board members had a lot of . . . prior discussions about rep and warrant liability, . . . what it may or may not be, what our defenses were and . . . the unsettled state of the law.Hamzehpour Dep. 125:12-126:11.
We had the MBIA cases . . . that were coming in and . . . reports from loan level reviews were being done that reflected . . . higher defect rates than what we had originally assumed.
So I think there was a lot of . . . uncertainty around the outcome and so that was just the discussion of that.
And if you didn't settle these cases now or if you didn't settle the claims now, the reality of having a $44 billion proof of claim in the bankruptcy that would have to be litigated and dealt with at some point anyway. . . . [T]hat's largely I think . . . the discussion.
71. Ms. Hamzehpour also explained how RFC's internal analysis of its own exposure risk had evolved as the Monolines and Trust Investor Groups pressed their claims:
Q. [W]hat were the defect rates that you were seeing from the MBIA reunderwriting that you mentioned?Id. at 126:15-127:1.
A. You know, I don't remember exactly what they were. I just remember that they were coming out. . . . [M]y recollection is that they were shockingly higher than any of us believed. But nevertheless, even if they weren't right, there was still a great variation between what we thought they would be versus what the reports were showing.
3. The Bankruptcy, and RMBS Trust and Monoline Claims
72. On May 14, 2012, RFC and certain of its affiliates (collectively, the "Debtors"), filed chapter 11 petitions in the United States Bankruptcy Court for the Southern District of New York. Lipps HLC 1220:6-16, 1333:14-18. The Honorable Martin Glenn was the presiding judge. Id. at 1270:14-17.
73. The RMBS Trustees filed hundreds of proofs of claim against RFC and its affiliates for over 1,000 securitization trusts, including hundreds of additional trusts that were not part of the Original RMBS Trust Settlement. See id. at 1250:5-9, 1266:17-1267:6. The Monolines—MBIA, FGIC, Ambac, and Syncora—also filed proofs of claim, as did another monoline insurer, Assured. Id. at 1226:3-8, 1227:5-1228:10, 1242:17-1243:9, 1266:4-16, 1269:15-19, 1269:24-1270:9, 1369:19-21, 1675:19-1676:3; PTX-408.
74. The RMBS Trustees' proofs of claim against RFC were asserted in the Trustees' capacities as trustee for each of their respective RMBS Trusts. See, e.g., PTX-155 at 4; DTX-543 at 4; DTX-544 at 2. The RMBS Trustees asserted obligations that arose from certain governing agreements, including the Pooling and Servicing Agreements, Assignment and Assumption Agreements, and other transaction documents. See, e.g., PTX-155 at 4; DTX-543 at 4.
75. Under the governing transaction documents, RFC sold to a depositor mortgage loans "that were originated or otherwise acquired by the seller," and the depositor then "pool[ed] the mortgage loans and transfer[red] them into one or more RMBS Trusts." PTX-155 at 11-12; DTX-543 at 11. As set forth in the RMBS Trustees' proofs of claim, "[w]ith regard to the mortgage loans that it transfers to the depositor, the Seller [e.g., RFC] makes numerous representations, warranties and covenants (the 'Trust [Reps]')," and "[w]hen the depositor transfers the pools of mortgage loans to the RMBS Trust . . . it also assigns all of its rights in the mortgage loans and under the mortgage loan purchase agreements and other applicable agreements, including the Trust [Reps] . . . ." PTX-155 at 12; DTX-543 at 12-13. The RMBS Trustees did not identify specific Reps in the governing transaction documents that they alleged were breached. Rather, they identified categories of Reps:
Trust [Reps] vary based on the specific RMBS Trust Transaction Documents, but typically pertain to, among other things: (a) the standards and practices used in underwriting each mortgage loan; (b) the creditworthiness of the borrowers on the mortgage loans; (c) the percentage of a mortgage pool which has certain characteristics, such as owner-occupancy and documentation type; (d) the disclosure of information on the mortgage loan tapes; (e) the completeness of each mortgage loan file; (f) the origination of the loans in accordance with applicable federal and state laws; and/or (g) various characteristics of each specified mortgage loan such as loan-to-value ratio, debt-to-income ratio, lien position and whether the property mortgaged is owner-occupied.PTX-155 at 12-13; DTX-543 at 12.
76. The RMBS Trustees further asserted that:
Generally, if a defect with respect to the Trust [Reps] is discovered in any mortgage file that may adversely affect the value of the related mortgage loan, the interest of the RMBS Trustee (as legal owner or pledgee of the mortgage loans) or the Securities Holders, the Seller is required to cure such defect, repurchase the defective mortgage loan, or, subject to certain conditions, substitute a complying mortgage loan for the defective mortgage loan.PTX-155 at 13; DTX-543 at 13.
77. The RMBS Trustees asserted that "[b]reaches by one or more of the Debtors of such [Reps] . . . have given, and may in the future give, rise to claims (the '[Reps] Claims') by the RMBS Trusts . . . ." PTX-155 at 14-15; DTX-543 at 13. The RMBS Trustees referred to their potential mortgage repurchase claims as "Buyback Claims." PTX-155 at 6; DTX-543 at 13. As Mr. Hawthorne testified, the central claims being asserted by the RMBS Trustees and the Monolines were repurchase ("buyback") claims. Hawthorne 1009:19-1010:7.
The RMBS Trustees also asserted servicing-related claims. See PTX-155 at 18-19.
78. The RMBS Trustees alleged that the Seller (i.e., RFC) was responsible for "any and all [Reps] Claims relating to . . . the mortgage loans transferred to the RMBS Trusts . . . and any and all cure and substitution obligations and [repurchase claims] resulting from such breaches." See, e.g., PTX-155 at 20; DTX-543 at 17.
79. As set forth below, the RMBS Trustees measured their repurchase claims against RFC by assessing breaches of underwriting guidelines (i.e., origination defects). See Pfeiffer Dep. at 126:1-11, 146:19-147:4, 257:14-20.
80. Like the RMBS Trustees, the Monolines also asserted claims arising from loan-level defects in origination. See DTX-494 at 4 (asserting that RFC "made express [Reps] to MBIA in the transaction documents regarding loan-level characteristics and the characteristics of the loan pool as a whole, and agreed to repurchase all defective loans identified by MBIA[,]" and alleging flaws in RFC's "underwriting and origination processes.").
81. Jeffrey Lipps, who represented RFC in, among other matters, pre-bankruptcy litigation brought by MBIA, testified that the breaches alleged by MBIA "were [] a result of the characteristics of the loans[.]" Lipps HLC 1230:25-1231:2. Tammy Hamzehpour, testified that "reports from loan level reviews [in the MBIA litigation] were being done that reflected . . . higher defect rates than what we had originally assumed." Hamzehpour Dep. 125:20-25. She further testified that the defect rates coming from that MBIA litigation "were shockingly higher than any of us believed." Id. at 126:15-127:19.
82. FGIC similarly alleged that it had determined "that a shockingly high percentage of the Mortgage Loans . . . contained defects constituting breaches of loan-level [Reps]." PTX-146 at 9-10. Ambac made similar allegations, asserting that "the loans included in the RALI Transaction did not bear the attributes represented by RFC with respect to two key attributes," and that "the loans were of much poorer credit quality . . . than Ambac reasonably expected based on RFC's representations and disclosures." DTX-497 at 8. Syncora, in a late-filed proof of claim, asserted that "RFC is obligated to repurchase or substitute mortgage loans in the RALI 2006-QO4 mortgage pool for which there is a breach of an origination-related representation and warranty that materially adversely affects the value of the loan . . . ." DTX-569 at 17-18. Syncora further alleged that "RFC failed as Sponsor to repurchase loans that it knew were defective and materially breached [Reps] . . . ." Id. at 18.
4. The 9019 Proceeding and the Mediation
83. As described above, RFC and its affiliates entered into the Original RMBS Trust Settlement with the Trust Investor Groups prior to filing for bankruptcy.
84. RFC and its affiliates filed a motion in the Bankruptcy Court for approval of the Original RMBS Trust Settlement (the "9019 Motion" and the "9019 Proceeding"). Lipps HLC 1256:11-19, 1259:22-1260:6. RFC engaged an expert, Frank Sillman, to assess the reasonableness of the Original RMBS Trust Settlement. Id. at 1260:7-15. Mr. Sillman submitted a June 11, 2012 declaration. Sillman Dep. 38:1-13. As described below, Mr. Sillman subsequently submitted additional declarations after conducting a reunderwriting analysis.
85. A number of parties opposed the 9019 Motion, including the Official Committee of Unsecured Creditors (the "Committee"), the Monolines, and certain RMBS Trust investors. Lipps HLC 1259:6-16; Hawthorne 1281:19-1282:1.
The Committee was appointed by the United States Trustee to be the representative of all of RFC's creditors, including those who did not have RMBS-related claims. See 11 U.S.C. § 1102 (2018). The Committee had the statutory duty to, among other things, investigate the Debtors' liabilities. 11 U.S.C. § 1103(c)(2) (2018).
86. In its December 3, 2012 objection, the Committee asserted that the process that led to the Original RMBS Trust Settlement was flawed, that the $8.7 billion settlement amount was too high, and that the allocation based on net losses should not be approved. DTX-541 at 20, 34, 50; Lipps HLC 1259:17-21. The Committee referenced the "myriad [of] complex legal issues that may affect the outcome of the Rep claims," and discussed breaches, materiality, causation, election of remedies, and statute of limitations, among other issues and potential defenses. DTX-541 at 29-30, 47.
Pursuant to an order of the Bankruptcy Court, expert reports, objections, statements in support, and other pleadings and documents submitted in the 9019 Proceeding were to be served on interested parties by December 14, 2012, but not publicly filed until on or around February 1, 2013. See Fourth Revised Scheduling Order, at ¶¶ 2, 3, 10, No. 12-12020, Doc. No. 2528 (Bankr. S.D.N.Y. Dec. 27, 2012).
87. The Committee retained Bradford Cornell and J.F. Morrow as its experts to oppose the Original RMBS Trust Settlement. Cornell Dep. 22:3-18; Morrow Dep. 27:4-28:3. Professor Cornell testified that he had "no issue" with Mr. Sillman's calculation of over $40 billion in total losses, Cornell Dep. 135:22-136:13, and he also opined that there were $16.5 billion in losses attributable to material underwriting defects, id. at 139:17-140:5. Professor Cornell focused on material defects, "which counsel [for the Committee] ha[d] instructed [him] to use as a rough proxy for material [Rep] breaches[,] recognizing that this could lead [to] an overstatement, or conceivably an understatement of breach-related losses." Hawthorne 1233:5-16.
88. Professor Cornell opined as to a lower expected liability amount of $3.8 billion after a $12.7 billion discount for the causation defense. Cornell Dep. 140:2-21; DTX-541 at 37-42. Professor Cornell also looked at the impact on the claims of a reduction for a statute of limitations defense. Cornell Dep. 74:5-10; DTX-541 at 42-43. The amount of the reduction depended on whether he accounted for all tolling agreements, or just those tolling agreements executed by the RMBS Trustees (as opposed to those with investors). Cornell Dep. 72:23-73:13. In addition he also assessed the impact on the claims of a reduction for an election of remedies defense (i.e., the defense that the RMBS Trusts could not recover on loans that had been liquidated or charged-off). Id. at 74:15-76:13, 131:20-132:10; DTX-541 at 44-47.
89. Professor Cornell admitted in his deposition that he had no independent knowledge concerning the applicability of legal defenses, and that his statements on it were "just parroting what [the Committee's counsel] Mr. Bentley told me." Cornell Dep. 142:4-17; see also id. at 71:21-72:2, 125:7-23, 131:2-10.
90. Apparently based on the instructions of Committee counsel, Professor Cornell assumed that causation, statute of limitations, and election of remedies defenses would all be 100% effective. See id. at 131:20-133:14.
91. Mr. Morrow was asked by the Committee to calculate breach rates for Professor Cornell by "re-underwrit[ing] loans for their compliance with the applicable underwriting guidelines when purchased or made, if provided . . . ." Morrow Dep. 27:4-28:3. Mr. Morrow conducted a reunderwriting on a 1,500 loan sample from the RMBS Trusts subject to the Original RMBS Trust Settlement. Id. at 32:8-33:15. His reunderwriting was "narrowly tailored to identify material compliance with the ResCap guidelines." Id. at 49:21-50:7. And, by "narrowly tailored," Mr. Morrow meant that he and his team "didn't do a title search . . . [and] didn't do verification of employment" or other things one might do in a "broader" reunderwriting. Hawthorne 1310:2-23. As Mr. Hawthorne testified, if Mr. Morrow had done those things, that would tend to increase the breach rate. Id. at 1310:24-1311:10.
92. MBIA also objected to the Original RMBS Settlement. In its December 3, 2012 objection, MBIA asserted that the allocation methodology, based on net losses, failed to account for differences among the 392 RMBS Trusts that would have been subject to that settlement. DTX-539 at 21-25. Specifically, MBIA argued that the settlement did not account for differences among the trusts based on a potential statute of limitations defense, and based on purported differences in the Reps. Id.
93. As PRMI's expert David Woll acknowledged, MBIA was not a neutral party when it came to the Original RMBS Settlement. Woll 1902:22-25. MBIA had insured trusts primarily from 2006-2007, which were not subject to a potential statute of limitations defense. MBIA also insured trusts with Substantial Compliance Representations, which were Trust Reps made by RFC to certain RMBS Trusts warranting that the loans had been underwritten in substantial compliance with the Client Guide. Id. at 1939:22-24. Thus, MBIA stood to benefit from having more of the Original RMBS Trust Settlement allocated to later-dated trusts and to trusts with Substantial Compliance Reps. Id. at 1902:22-1903:10.
In this regard, MBIA's objection overstates the views of RFC's counsel to suit MBIA's arguments. Compare DTX-539 at 22 (MBIA mischaracterizing declaration of Jeffrey Lipps as stating that RFC would have made "strong arguments" based on the statute of limitations), with DTX-443 at 35-36 (describing statute of limitations as "one argument we likely would have considered making" in litigation, and discussing the arguments that both sides could make); see also Woll 1897:4-1902:8.
94. At least two investors in the RMBS Trusts also objected to the net loss allocation formula in the Original RMBS Trust Settlement. PTX-179 ¶ 115; Hawthorne 1099:19-1100:3.
95. As for the RMBS Trustees, they did not participate in the negotiation of the Original RMBS Trust Settlement and were not a party to that agreement. PTX-153 at 5. The Trust Investor Groups negotiated the settlement directly. Id. As noted above, however, the RMBS Trustees acted for the RMBS Trusts and represented the interests of the investors in those Trusts. Hawthorne 1100:14-1101:19; Lipps HLC 1249:22-1250:4. Thus, they undertook to analyze the Original RMBS Trust Settlement to determine whether to accept or reject it on behalf of the relevant RMBS Trusts. PTX-153 at 5.
96. The RMBS Trustees retained Duff & Phelps ("Duff") as their financial advisor to, among other things, "advise the RMBS Trustees regarding whether the RMBS Trust Settlement [was] a fair and reasonable settlement of the Settling Trusts' Repurchase Claims." Id.; Pfeiffer Dep. 34:18-20, 35:1-15. Duff advised the RMBS Trustees that the Original RMBS Trust Settlement fell within a reasonable range. Pfeiffer Dep. 316:20-317:23. Specifically, in their December 14, 2012 filing, the RMBS Trustees stated:
After careful consideration of relevant factors and analysis, including (a) the results of its review of a statistically significant number of loan files in the Settling Trusts provided by the Debtors, (b) the estimation of projected total collateral losses and underwriting breach rates in the Settling Trusts, (c) the estimation of likely agree rates with respect to the Settling Trusts (which take into account the litigation risk associated with the relative characteristics of the breach), and (d) consideration of causality factors (which take into account the litigation risk associated with a lack of causal relationship between the breach and loss), Duff advised the RMBS Trustees that the amount of the Allowed Claim is within a reasonable range to settle the Settling Trusts' Repurchase Claims, and that the Revised Claim Allocation Methodology would fairly allocate the Allowed Claim among the Settling Trusts.PTX-153 at 6.
97. As noted in the above quote from the RMBS Trustees, Duff conducted a reunderwriting analysis of a statistically significant number of loan files to determine underwriting breach rates. Id.; see also Pfeiffer Dep. 52:7-53:16. To do so, Duff sampled 6,500 mortgage loans. Pfeiffer Dep. 99:07-101:17. In assessing RFC's potential liability (and thus the reasonableness of the settlement), Duff was "focused on the reps between the debtor [RFC] and the trust . . . ." Id. at 146:19-147:4. To find "what [Reps] specifically applied to [a] loan," Duff looked to, among other things, the relevant Assignment and Assumption Agreement and Pooling and Servicing Agreement. Id. at 302:5-10. Duff was able to pull up those Reps through a database called Intex. Id. at 304:18-305:8. And, to assess RFC's potential breaches to the RMBS Trusts, Duff's reviewers conducted loan reviews "to determine whether the initial origination was done in conformity . . . with the guidelines." Id. at 126:1-11, 146:19-147:4 (Duff was "focused on the underwriting guidelines during [its] underwriting process"), 257:14-20 (agreeing that when Duff "was evaluating whether . . . one of the loans in its 6500 loan sample was in material breach, [Duff] was often referring to the client guide guidelines").
98. In response to the objections to the allocation methodology based on net losses, Duff also "evaluated the claim allocation methodology." PTX-179 at 48. As the RMBS Trustees stated in their December 14, 2012 filing, "[t]he RMBS Trustees, after consulting with Duff, asked the Debtors and the [Trust Investor Groups] to adjust the Claim Allocation Methodology." PTX-153 at 5. Thus, as of December 14, 2012, the Debtors and Trust Investor Groups had agreed with the RMBS Trustees' request to modify the allocation methodology. Id. "Based on [Duff's] suggestion, and after lengthy discussions with the Steering Committee Consenting Claimants, the Debtors, and other parties in interest, the claim allocation methodology in the Original RMBS Settlement Agreements was modified." PTX-179 ¶ 115; see Hawthorne 1098:21-1099:17.
99. Moreover, the RMBS Trustees stated that they "believe[d] that the Revised Claim Allocation Methodology addresses the objections filed to the Debtors' Claim Allocation Methodology because it sufficiently accounts for the differences among the characteristics of each Settling Trust," and that "the legal defenses to the Repurchase Claims asserted by certain of the objections do not alter the Duff analysis . . . ." PTX-153 at 6.
100. Like Duff, RFC's expert Mr. Sillman also conducted a reunderwriting to assess RFC's potential liability. Specifically, after the Committee and MBIA objected to Mr. Sillman's original analyses (see DTX-541 at 31-34; DTX-539 at 26-28), Mr. Sillman evaluated a sample of 1,500 loans to determine whether there were breaches of Reps, and to opine on the reasonableness of the settlement. Sillman Dep. 78:6-14; Lipps HLC 1264:16-24. Mr. Sillman considered the transaction documents and underwriting guidelines. Sillman Dep. 90:9-92:14. Mr. Sillman determined a settlement factor range based on what he had seen with other RMBS settlements, agency repurchase demands, and legal defenses. Id. at 148:8-149:20. Although Mr. Sillman's settlement factor of 41-47% took into account legal defenses generally, it did not differentiate between potentially time-barred trusts and trusts whose claims were not subject to any potential statute of limitations defense. Hawthorne 1106:4-1107:20; DTX-538 at 12.
101. Of the more than $40 billion in potential RMBS-related losses, Mr. Sillman determined that RFC's expected liability to the RMBS Trusts, after considering litigation defenses, was between approximately $7.8 and $10.2 billion dollars. DTX-538 at 3. Mr. Sillman did not opine as to RFC's exposure to the Monolines at that time.
102. The Bankruptcy Court did not ultimately approve the Original RMBS Trust Settlement. Lipps HLC 1256:20-21. Instead, after nearly six months of litigation over the Original RMBS Trust Settlement, RFC filed a motion to appoint a mediator to assist in resolving the dispute and other issues relating to RFC's bankruptcy case. Id. at 1270:18-24. The Bankruptcy Court appointed the Honorable James M. Peck of the United States Bankruptcy Court for the Southern District of New York as a mediator. Id. at 1271:3-1273:1; PTX-151. The mediation participants included RFC, the Committee, the Monolines, the RMBS Trustees, and the Trust Investor Groups. PTX-179 at 33, ¶ 77.
103. In connection with the mediation, in February 2013, Lewis Kruger was appointed as chief restructuring officer ("CRO") of RFC and the other Debtors. Lipps HLC 1274:18-1275:6. Mr. Kruger was an experienced commercial lawyer with a "long background in the insolvency and creditors' rights world." Kruger Dep. 35:14-36:21. Judge Glenn approved Mr. Kruger's appointment as CRO. Lipps HLC 1275:13-1276:20; PTX-158. Mr. Kruger was independent from RFC's existing management and pre-bankruptcy shareholders, and was brought in "to assist in the mediation." Lipps HLC 1276:24-1277:4. He had the power to make decisions on behalf of RFC in connection with the chapter 11 Plan. Kruger Dep. 26:11-13.
5. The Settlements and Expert Assessment of the Settlements
104. In May 2013, the parties agreed to the terms of a Plan Support Agreement (May 13), and subsequently to the terms of a Supplemental Term Sheet (May 23). See PTX-179 at 3, 34-35. The Supplemental Term Sheet set forth a proposed resolution of, among other things, the claims asserted by the RMBS Trustees, and MBIA and FGIC. Lipps HLC 1273:12-15, 1277:5-10.
105. The final terms of the RMBS Trust Settlement and the Monoline Settlements were ultimately set forth in the Chapter 11 Plan proposed by RFC and the Committee, and by way of a separate order for Syncora. The final Settlement amounts set forth in the Plan were: (1) $7,091,200,000 on behalf of the RMBS Trusts; (2) $1,450,000,000 on behalf of MBIA; (3) $415,000,000 on behalf of FGIC; (4) $22,800,000 on behalf of Ambac; and (5) $7,113,000 on behalf of Syncora. PTX-180 at 144-45, 150-51; see also Lipps HLC 1284:20-1285:1, 1295:1-14; PTX-179 at 35, 62. RFC, the Committee, the Monolines, the RMBS Trustees, and Trust Investor Groups all expressed their full support for the settlements. Lipps HLC 1281:21-1282:10.
106. RFC's expert Frank Sillman submitted a declaration opining on, among other things, the RMBS Trust Settlement. See Sillman Dep. 41:7-42:6. Mr. Sillman opined that "[t]he likely damages for breach of [Rep] claims after litigation defenses for all of the trusts in the RMBS settlement other than the wrapped bonds are from $7.4 billion to $8.7 billion." Id. at 48:9-18. Mr. Sillman's methodology was consistent with the methodology in his reply declaration in connection with the Original RMBS Trust Settlement. See DTX-538. He made high and low estimates of lifetime losses, multiplied those estimates by the defect rates, and then multiplied that number by a settlement factor that accounted generally for legal defenses. Sillman Dep. 53:24-55:18, 149:5-20.
107. As discussed, the RMBS Trustees' advisor Duff did an extensive evaluation of 6,500 mortgage loans. Mr. Pfeiffer testified that the reunderwriting alone cost between $5 to $10 million and took tens of thousands of employee hours. Pfeiffer Dep. 334:8-335:16. Among other things, Duff's task was to assess "the settlement as a whole and analyze the reasonableness of that, and then to, once we had the whole number and got comfortable that that was fair and reasonable . . . , to determine the best way to allocate, the fairest way to allocate that total sum to all the RMBS Trusts that were impacted by this claim." Id. at 35:1-15.
108. Consistent with Mr. Sillman's opinion, Duff determined that a range of $6.5 billion to $10.2 billion was "a range of what a reasonable settlement could look like" for the claims of the 392 RMBS Trusts that were subject to the Original RMBS Trust Settlement. Id. at 316:20-317:23. Duff also determined that the value of the claims of the additional RMBS Trusts that were not part of the Original RMBS Trust Settlement was approximately $976 million. Id. at 319:15-25, 320:4-21.
109. As described above, Duff also evaluated and assisted the parties in devising and implementing the breaching loss allocation methodology to allocate the RMBS Trust Settlement among the RMBS Trusts. Mr. Pfeiffer testified that Duff suggested using breaching losses rather than net losses, because Duff "felt it was fair to include the incident of breaches in the methodology to allocate the claim." Id. at 53:6-54:9. He further testified that Duff's approach "was the fairest way to [allocate] and reflected the information [Duff] had in front of [them]," and that "[i]f it were done another way, that may have been fair as well, just not as fair as the methodology we picked." Id. at 56:01-25. Duff did not allocate based on any legal defenses or purported differences in the strength of Reps. Hawthorne 1096:11-18.
110. As required by bankruptcy law, the settlements were conditioned upon bankruptcy court approval. See American Prairie Constr. Co v. Hoich, 594 F.3d 1015, 1025 (8th Cir. 2010). Thus, the settlements were not effective unless and until Judge Glenn approved them.
6. The Bankruptcy Court's Approval of the Settlements
111. The Bankruptcy Court held hearings on Plan confirmation. Lipps HLC 1285:8-19. On December 11, 2013, the Bankruptcy Court entered its order confirming the Plan, which also approved the RMBS Trust, MBIA, FGIC, and Ambac settlements. Id. at 1284:11-19; PTX-180 at 35-36. The Bankruptcy Court approved the Syncora settlement in a separate order. Lipps HLC 1295:1-14. The Bankruptcy Court also entered Findings of Fact in connection with the confirmation of the Plan. Lipps HLC at 1285:8-13; PTX-179.
112. The Bankruptcy Court's order approving the RMBS Trust Settlement included not just approval of the amount of the settlement, but also express approval of the breaching loss allocation of the RMBS Trust Settlement among the RMBS Trusts. See PTX-180 at 35; see also id. at 116 (defining the RMBS [Trust] Settlement as "the settlement that provides for the allowance, priority and allocation of the RMBS Trust Claims"). The Bankruptcy Court found that the RMBS Trust Settlement, including the allocation, was "fair and reasonable," id. at 34, and in the best interests of "the [i]nvestors in each RMBS Trust . . . each such RMBS Trust, the RMBS trustees, and all other parties in interest," id. at 24-25.
II. PRMI'S BREACHES OF ITS CLIENT CONTRACTS WITH RFC
A. Mr. Butler's Reunderwriting Review of PRMI Loans and Global Loans
113. ResCap retained Steven Butler as its reunderwriting expert. ResCap asked Mr. Butler to reunderwrite two sets of loans: (1) 157 loans that PRMI sold to RFC (the "PRMI Loans"), and (2) 410 loans (the "Global Loans") that 125 correspondent lenders sold and were contained in the Trusts subject to the Settlements. Butler 548:21-550:6.
The originating lenders of the mortgage loans sold to RFC are sometimes referred to as "correspondent lenders." Common SJ Order, 332 F. Supp. 3d at 1118 n.3.
114. The Court finds that Mr. Butler was well qualified to undertake this review. He has worked in residential mortgage lending for the past 45 years, id. at 548:14-17, and in the banking industry for 17 years, including as a lender, loan officer, underwriter, branch manager, and the president of four different banks, id. at 553:6-13. Subsequently, he worked for several accounting firms as an "industry expert in lending and banking," and then as a consultant "doing loan-related work." Id. at 553:14-554:1. For the past 15 years, Mr. Butler has been the president and owner of Butler Bank. Id. at 548:7-9. In that capacity, he has worked on numerous RMBS litigation matters since 2008. Id. at 548:10-23, 554:2-7.
115. Moreover, Mr. Butler has underwritten thousands of loans and reunderwritten tens of thousands of loans. Id. at 555:1-2, 556:23-25. Evaluating mortgage-related Reps has been a focus of the last 12 years of his RMBS litigation work. Id. at 555:3-12. And he has evaluated securitization Reps approximately 40 times in pre-litigation reviews, investigations for the Department of Justice, and litigation matters. Id. at 555:13-18. Mr. Butler is also a certified fraud examiner. Id. at 552:17-18.
A fraud examiner is certified to identify, prevent, investigate and testify about fraud. Butler 552:20-22.
116. Here, Mr. Butler conducted a reunderwriting review of the PRMI Loans and Global Loans. For both sets of loans, Mr. Butler directed that all loan files, applicable guidelines, and other documents relating to the loans be made available to him. Id. at 567:7-13. Mr. Butler supervised the work of his own subcontractors, as well as a vendor, Opus Capital Management. Id. at 566:21-567:6. He did the same for the Global Loans, in addition to supervising the work of two more vendors, Digital Risk, LLC and The Barrent Group. Id. at 570:1-22.
117. For both sets of loans, Mr. Butler devised a protocol to match each loan to the applicable guidelines. Id. at 567:14-21, 570:22. Under Mr. Butler's direction, the vendors used this protocol to match the appropriate guidelines to each loan. Id. at 567:22-568:6. The vendors then identified breaches by analyzing the contents of the loan files and other relevant documents and then comparing them to the applicable guidelines. Id. The vendors, in turn, prepared reports for Mr. Butler's review detailing their breach findings. Id. at 567:22-568:6, 570:22-571:7.
118. Mr. Butler reviewed all the breaches identified in each vendor's report and directed the vendor to incorporate any necessary changes. Id. at 568:7-9, 570:22-571:7. Mr. Butler then directed one of his highly experienced staff reunderwriters to review the revised report and provide comments. Id. at 568:9-14, 570:22-571:7. Beyond the time spent by the vendors and his own staff reunderwriters, Mr. Butler personally spent "close to two hours" on each loan reviewed. Id. at 569:16-22.
119. After Mr. Butler received the twice-reviewed report from his staff reunderwriters, he personally reviewed the report against the relevant guidelines, loan files, and other relevant documents, and then he determined which breaches were material. Id. at 568:15-24, 570:22-571:7.
120. In Mr. Butler's opinion, a material breach is one that "affects the [borrower's] ability to repay and the ultimate repayment of that loan." Id. at 573:2-4. Mr. Butler further testified that misrepresentations by the borrower of "income, debt, employment, or owner occupancy" were a material breach because they call into question "the veracity" of all information furnished by the borrower and casts doubt on the borrower's ability and willingness to repay. Id. at 573:15-17. In his years of experience, Mr. Butler stated he was unaware of any relevant guidelines that allowed for these types of borrower misrepresentations. Id. at 573:20-3, 576:1-3.
B. Ms. Keith, PRMI's Reunderwriting Expert
121. At trial, PRMI did not challenge Mr. Butler's process for reunderwriting either sets of loans, but contested certain findings of breach. To challenge these findings, PRMI's reunderwriting expert, Kori Keith, testified about her review of a subset of PRMI Loans.
122. Ms. Keith has worked in the residential mortgage industry for 22 years. Keith 2202:4-6. From 2007 to 2014, she worked at Steel Mountain Capital, a "buyer, holder, and seller of residential mortgage loans." Id. at 2203:4-8. From 1998 to 2007, Ms. Keith held various positions, including as a loan officer, senior loan officer, head of mortgage operations for two branches of US Bank, and owner of a mortgage broker company. Id. at 2207:25-2209:4. She is currently employed by Lakewood Mortgage Advisors, which provides due diligence, underwriting, litigation support, consulting, and expert services. Id. at 2201:13-18. For the past six years, she has conducted reunderwriting reviews that include assessing breaches of securitization Reps. Id. at 2211:5-13. She has previously served as an expert witness "with respect to residential mortgage loan underwriting and alleged breaches of Reps. Id. at 2211:14-17.
123. Although the Court finds that Ms. Keith was qualified to undertake this review, the weight and credibility accorded to her opinions will be discussed in the context of those opinions, infra.
124. In response to Mr. Butler's initial findings, Ms. Keith served rebuttal and surrebuttal responses for 77 of the PRMI Loans, but she did not review any of the Global Loans. Keith 2213:12-2214:4, 2214:18-23. Mr. Butler then reviewed Ms. Keith's rebuttal and surrebuttal responses for the subset of PRMI Loans. See Butler 569:3-8. Ultimately, Mr. Butler found that 101 of the 157 PRMI Loans (64%) materially breached PRMI's Reps to RFC. Id. at 571:19-25. Fifty of those 101 PRMI Loans contained "misrepresentations of either debt, income, occupancy or employment." Id. at 572:5-8.
Mr. Butler opined that underwriting guidelines "never allow for a misrepresentation," and that one cannot "underwrite misrepresented information" because "you don't really want to make a loan to somebody who doesn't tell the truth . . . ." Butler 573:18-574:2.
125. Ultimately, at trial, Ms. Keith disputed only 11 PRMI Loans that Mr. Butler found to be in breach of PRMI's Reps to RFC. Keith 2217:8-11 ("Q. So in total you are here today to talk about 11 loans as it relates to Mr. Butler's alleged breaches of PRMI [Reps]? A. Yes."). For five of these 11 loans, PRMI also contended that, even if the loans were in breach, Plaintiff is estopped from asserting those breaches. For one additional loan (the "Countrywide Loan," No. 10226985), PRMI raised only estoppel and waiver defenses. As to the remaining 89 PRMI Loans, Ms. Keith did not challenge Mr. Butler's findings of breaches because the Court had previously held, consistent with the Eighth Circuit's ruling in Residential Funding Co. v. Terrace Mortgage Co., 725 F.3d 910 (8th Cir. 2013), that ResCap had sole discretion under the express terms of the Client Guide to determine breaches. See, e.g., Keith 2215:6-2217:3, 2253:17-2254:10; PRMI SJ Order, 428 F. Supp. 3d at 90-93.
C. Mr. Butler's Identifications of AlterNet Guide Breaches
126. Both parties agree that ResCap's sole discretion does not apply to the AlterNet Loans (4115211, 4550405, 4413350, 4375505, 4962418, 4117058, and 4380515). Mr. Butler originally underwrote the AlterNet Loans to the Client Guide, but then re-reviewed them against the 1997 AlterNet Guide in response to Ms. Keith's objections. Butler 575:8-15, 576:9-25. Mr. Butler opined that the AlterNet Loans breach both Guides. Id. at 576:22-577:2.
127. For four of the AlterNet Loans (4375505, 4962418, 4117058, and 4380515), Ms. Keith did not dispute the underlying facts of the breaches but only Mr. Butler's use of the 1997 AlterNet Guide to reunderwrite loans originated and sold from 2000 to 2001. Id. at 577:5-578:3; Keith 2250:13-18, 2217:23-2218:6. For the remaining three AlterNet Loans (4115211, 4550405, 4413350), Ms. Keith disputed both Mr. Butler's breach findings and his use of the 1997 AlterNet Guide. Keith 2217:20-2218:19.
Ms. Keith disputed Mr. Butler's findings on Loan 4413350 based on an Assetwise approval certificate in the loan file, as discussed below.
128. At trial, Mr. Butler testified that he used the 1997 AlterNet Guide for two primary reasons: First, Mr. Butler and ResCap's counsel undertook extensive searches for later versions of the AlterNet Guide but could not locate any. Butler 576:12-21. Second, Mr. Butler determined that the Reps in the Client Guide and the AlterNet Guide "are very similar." Id. at 566:4-9. As noted above, each PRMI Loan that Mr. Butler determined breached the Client Guide also breached the 1997 AlterNet Guide.
129. In contrast, Ms. Keith testified that she was "unable to be convinced that" the 1997 AlterNet Guide was "the most recent guide to use." Keith 2219:12-19. Ms. Keith speculated that, "it [was] very unlikely that RFC went three years with no revisions to the AlterNet Seller Guide." Id. at 2220:1-3. Ms. Keith was not aware of a set of guidelines or underwriting matrices that had remained static over a three-year period. Id. at 2220:17-20.
130. On cross-examination, however, Ms. Keith conceded that she "do[es]n't know whether there was a revision [to the AlterNet Guide] or not," and she had not "seen any documents that actually refer to a later version of the AlterNet Guide." Id. at 2364:17-20, 2365:12-2366:1. She also admitted that "nobody here knows" if "RFC's AlterNet matrices remained static for three years," and she had "never seen an AlterNet Guide dated later than July 1997." Id. at 2364:21-2365:4.
131. Neither Ms. Keith nor PRMI offered a more recent version of the AlterNet Guide or even a document that references such a version. Having presided over this and related cases brought by ResCap for over half a decade, the Court finds the absence of a single document referencing a later version of the AlterNet Guide to be telling. The Court finds similarly noteworthy that PRMI failed to offer any fact testimony supporting Ms. Keith's opinion. In the absence of any documents or testimony suggesting the existence of a post-1997 version of the AlterNet Guide, the Court finds Ms. Keith's speculation about the potential existence of a post-1997 version of the AlterNet Guide not credible, and Mr. Butler's use of the 1997 AlterNet Guide to be reasonable.
1. The Seven AlterNet Loans at Issue
132. The Court now considers the evidence pertaining to each of the seven contested AlterNet Loans:
a. Loan 4375505
133. Loan 4375505 was originated on January 26, 2001, for $58,650. PTX-284 at 1. Mr. Butler testified that the loan violated the terms of the 1997 AlterNet Guide because: (1) the borrowers misrepresented their debt; (2) there was an excessive DTI ratio; (3) there were multiple unexplained credit inquiries; (4) there was an improper debt calculation; and (5) there was an ineligible flip transaction. Butler 585:14-23. Specifically, the borrower failed to disclose mortgages on seven other properties when applying for the loan. Id. at 586:6-15. A recalculation of the borrower's DTI ratio, accounting for those undisclosed mortgage obligations, yields a DTI ratio in excess of 300%, id., which made it highly unlikely that the borrower could afford the subject loan. Id. at 586:16-587:5.
134. Ms. Keith did not dispute these facts at trial. She only contested the applicability of the 1997 AlterNet Guide, criticizing Mr. Butler for not relying on a more contemporaneous set of guidelines for this loan. However, Ms. Keith did not nor could she opine that, under such a hypothetical guide, the identified breaches would be immaterial. The implication that the borrower's misrepresentation of debt here would somehow be permitted under any set of guidelines is in fact undermined by PRMI's own witnesses. Crawford 2451:1-14; Flitton Dep. 67:20-24, 69:24-70:5, 89:2-16. These witnesses all agreed that a prohibition on borrowers' misrepresentations in origination is contained in any set of industry guidelines. Id. Their testimony further aligns with that of Mr. Butler, who opined that "there are no guidelines" that would allow for such a misrepresentation by the borrower, resulting in a DTI ratio in excess of 300%. Butler 587:3-10. The Court therefore finds Ms. Keith's testimony not credible and assigns no weight to it.
b. Loan 4962418
135. Loan 4962418 was originated on May 30, 2001, for $59,500. PTX-292 at 18. Mr. Butler testified that the loan violated the 1997 AlterNet Guide because the borrower misrepresented their employment and because it was an ineligible flip transaction. Butler 587:11-18. The borrower represented that he owned a home improvement business, but (1) the borrower's bankruptcy filing within just five years of the loan did not list any self-employment income on the relevant income schedule for the relevant period; (2) the borrower's credit report did not reflect self-employment income; and (3) the Secretary of State records in Ohio, where the borrower resided, contained no record of the borrower's business. Id. at 587:18-588:6. Mr. Butler testified that the borrower's misrepresented employment "would violate the [1997] AlterNet Guide as well as any [other] set of guidelines." Id. at 588:12-13.
136. Again, Ms. Keith did not dispute these facts at trial, contesting only the applicability of the 1997 AlterNet Guide. For the reasons explained above, the Court finds Ms. Keith's testimony that another guide might have existed entirely speculative and assigns no weight to it. That the borrower's misrepresentation of employment here would be permitted under any set of guidelines could not be rebutted credibly by Ms. Keith.
c. Loan 4117058
137. Loan 4117058 was originated on November 22, 2000, for $49,700. PTX-283 at 1. Mr. Butler testified that the loan violated the 1997 AlterNet Guide because of PRMI's failure to verify the borrower's housing history and because of the borrowers' insufficient credit score. Butler 590:8-12. According to Mr. Butler, "the housing history is a very strong indicator of a borrower's ability and willingness to repay on a mortgage." Id. at 591:11-13. For the subject loan, the 1997 AlterNet Guide required a minimum credit score of 540 and 12 months of verified housing history. Id. at 590:12-20. The uncontested facts are that the borrower had a credit score of only 515 and there were only 11 months of verified mortgage payments in the loan file. Id.
138. Ms. Keith did not dispute these facts at trial, once again contesting only the applicability of the 1997 AlterNet Guide. For the reasons explained above, the Court assigns no weight to her speculation. Ms. Keith could not rebut PRMI's obligations to verify a borrower's housing history under any set of guidelines.
d. Loan 4380515
139. Loan 4380515 was originated on February 27, 2001, for $185,400. PTX-287 at 334. Mr. Butler determined that this loan was made for a non-arm's length transaction, which is prohibited by the 1997 AlterNet Guide. Butler 591:16-592:4. Such transactions call into question "the veracity of the information that's obtained" from the borrower. Id. at 592:25-593:1. Here, Mr. Butler opined that "the [borrower's] mother is the seller, the employer and the landlord. . . . [a]nd so the[] verifications of employment and rental history would be invalid because of the relationship between the parties." Id. at 592:10-17. Mr. Butler further testified that this type of conflict of interest is generally "not allowed by guidelines," but if "some guidelines" allow it, they require an "explanation of the relationship" and "proof" that the borrower did not receive any "favoritism" for the loan. Id. at 593:8-11.
140. Ms. Keith did not dispute these facts at trial, once again contesting only the applicability of the 1997 AlterNet Guide. For the reasons explained above, the Court assigns no weight to her speculation. Ms. Keith could not rebut that a loan for this type of non-arm's length transaction would not be approved under any set of guidelines.
e. Loan 4115211
141. Loan 4115211 was originated on November 1, 2000, for $102,000. PTX-282 at 1. PRMI represented to RFC that the loan was an owner-occupied, cash-out refinance, and that the subject borrower had a 526 FICO score and an 85% LTV ratio. Butler 610:20-611:5. However, the 1997 AlterNet Guide did not permit credit scores below 540. Id. at 611:6-10; PTX-282 at 63; PTX-001 at 210-11. The loan did not qualify for even the lowest AlterNet Guide credit grade, CM, which required a minimum credit score of 540 and a maximum LTV ratio of 65%. Butler 613:3-18; PTX-001 at 211. The borrower also did not qualify for an upgrade to a C credit grade, which required a maximum LTV ratio of 70%. Butler 613:19-614:3; PTX-001 at 210; see also Butler 614:4-11 (noting that "under either . . . credit grade, this loan did not qualify and breached the AlterNet Guide.").
142. Ms. Keith testified that the loan was properly approved with a B credit grade under the AlterNet Guide because, under the Client Guide, a "client may upgrade a loan if it fails to qualify for the higher grade based solely on the failure to meet a single credit characteristic [(here, the FICO score)] of that higher grade." Keith 2227:6-20. In support, Ms. Keith pointed to an internal RFC due diligence worksheet (PTX-282 at 60) that reflected a seller grade of B, and an Evaluwise grade of C. Id. at 2232:15-2233:10. Evaluwise, RFC's internal automated evaluation engine, selected this grade by pulling a different credit score (i.e., 546) for the borrower. Id. at 2232:15-2233:10, 2234:19-20. Based on the credit score pulled by Evaluwise, Ms. Keith testified that an RFC reviewer then upgraded the loan's credit grade from a "C" to a "B" and approved the loan for purchase as a "B" credit grade. See PTX 282 at 60; Butler 846:3-16; Keith 2227:17-23, 2232:15-2233:13. Ms. Keith also pointed to references to the borrower's 20 years of military employment and two years of owning the home on the worksheet as support for the B credit grade. Id. at 2234:6-16. Ms. Keith concluded that the loan was properly advanced one credit grade from a C to a B under the AlterNet Guide. Id. at 2239:2-10. Additionally, she concluded that the LTV ratio of the borrower met the 1997 AlterNet Guide criteria for "B" credit grades. See Keith 2227:17-23; PTX-1 at 209; PTX-282 at 60.
143. Ms. Keith, however, did not present any evidence that the due diligence worksheet was made available to PRMI. Nor could Ms. Keith dispute that, under the AlterNet Guide (PTX-001 at 33), PRMI acknowledged that RFC bought the loan in reliance upon PRMI's compliance with the AlterNet Guide and that it is liable, regardless of what may or may not have occurred at RFC's end, for any "breach of warranty regardless of whether it or RFC actually had, or reasonably could have been expected to obtain, knowledge of the facts giving rise to such . . . breach of warranty." Clearly, the loan was in breach of the AlterNet Guide when it left PRMI's hands.
144. The Court therefore finds that RFC's internal examination, after the fact, is irrelevant to a determination of breach. The borrower simply did not qualify for the loan.
f. Loan 4550405
145. Loan 4550405 was originated on March 19, 2001, for $105,600. PTX-291 at 17. Mr. Butler testified that the loan violated the 1997 AlterNet Guide because it was an impermissible flip transaction. Butler 619:20-620:14. He opined that the loan's appraised value was $132,000, there was a previous loan of $60,000 taken out on the property less than six months before, and these valuation changes were not addressed or explained on the appraisal. Id. at 621:16-623:13. Mr. Butler concluded that the loan was not prudently originated or of investment quality because "it's imprudent to originate a loan" without investigating the potential "churning or flipping [of] the property" indicated by the recent $60,000 loan. Id. at 621:16-20, 623:5-13.
146. Ms. Keith acknowledged that the AlterNet Guide addresses property churning. Keith 2241:17-2242:7; see also PTX-001 at 323. Nonetheless, she disagreed that there was a flip transaction because, in her opinion, the $60,000 loan did not evidence a prior sale, and the appraisal reflected that an underwriter looked at "the MLS, [which] shows previous sales" and noted "that there [were] no prior sales within one year of th[e] appraisal," Keith 2246:13-2247:2, 2247:14-18, 2249:5-15. She also contended that the appraised value could be supported by "market data" or "documented improvements to the property" under the AlterNet Guide, id. at 2242:25-2243:2, but pointed to no evidence of market data or improvements described in the appraisal. Cf. Butler 860:22-863:23 (opining that the "sales comparison approach" analysis in appraisal "doesn't analyze market data" and there were no improvements to the property "described in the appraisal.").
147. The 1997 AlterNet Guide requires that the loan be of "investment quality," "prudently originated," and "underwritten in compliance with all requirements of this AlterNet Guide." PTX-001 at 39, § 251-1(T). Unfortunately, Ms. Keith again speculated, without any notation or other evidence in the appraisal, that the $60,000 loan likely did not evidence a prior sale in an effort to retrospectively justify the loan. The issue for this Court to determine, however, is whether an investigation of this prior sale and the risk of a flip transaction actually occurred during underwriting. It is uncontested that the record contains no evidence of such an investigation.
148. Based on Mr. Butler's credible testimony, the Court finds that it was imprudent for PRMI to approve the loan without investigating the apparent fluctuation in appraised value in such a short span of time. See Butler 862:16-863:23 (stating that the "appraisal was incomplete because it didn't substantiate the increase in value of the property, which either is a flip [transaction] and erroneous or it was [an] improvement to the property and there's no evidence of that.").
g. Loan 4413350
149. At trial, Ms. Keith disputed Mr. Butler's breach analysis for one AlterNet Guide Loan (No. 4413350) that was also approved through the use of RFC's proprietary automated underwriting system, Assetwise.
150. Assetwise was a computer software system that allowed loan underwriters to speed up the underwriting process and provide a preliminary, often conditioned, credit decision on the loan based on the information input into it. Butler 581:22-582:6. It evaluated loan data to grade and slot loans intended for sale to RFC. Bangerter 358:19-23, 359:7-20. Assetwise offered originators a faster and easier way to upload information to RFC, which allowed for "more efficient and timely feedback" from RFC to the originator. Farley 204:25-205:4. Assetwise also improved the speed and accuracy of RFC's loan review by providing a uniform format for loan data transcription from originators to RFC. Id. at 203:8-204:11. RFC offered several ways for originators to access Assetwise, one of which was via the Internet. The Internet-accessed version of Assetwise was called Assetwise Direct. Bangerter 358:10-18.
Countrywide had its own proprietary underwriting software program, called "CLUES," and other originators used their own underwriting software programs. Butler 582:16-17.
151. After entering loan data into Assetwise, originators like PRMI received an Assetwise Findings Report providing information regarding the recommended loan program and credit grade, to allow for accurate pricing of the loan. Id. at 361:5-17; Richardson Dep. 33:2-5, 33:8-19. The Assetwise Findings Report also informed PRMI of mandatory next steps, including required credit documentation. Bangerter 361:5-17.
152. However, even when utilizing Assetwise technology, employees of loan originators such as PRMI were still responsible for reviewing the recommendations in the Assetwise Findings Report, checking the accuracy of the data, and ensuring that conditions for approval were satisfied. Butler 581:17-583:18; see also Zaloumis Dep. 143:17-144:7 ("There still has to be a human that underwrites a file."); Flitton Dep. 142:4-9 (noting "the file still had to be completely underwritten"). Moreover, Assetwise was not able to assist originators with certain aspects of loan underwriting, such as (1) properly calculating income, assets, and liabilities, Flitton Dep. 261:17-262:4; (2) evaluating fraud or misrepresentation, Bangerter 360:19-25, 431:14-432:3, 451:16-22; (3) assessing the adequacy of the loan collateral, id. at 358:24-359:6; (4) determining that all applicable laws and regulations had been complied with, PTX-001 § 251-1(D), PTX-032 § A202(I); (5) identifying whether the loan was non-arm's-length, Bangerter 360:19-25; (6) determining housing history, id. at 361:1-2; and (7) identifying "red flags" in the loan file. Butler 558:25-559:5.
153. Loan 4413350 was originated on March 14, 2001, for $69,700. PTX-290 at 3. Mr. Butler opined that the loan was approved by Assetwise under the AM credit grade, with a 561 credit score and an 85% LTV ratio. Butler 614:16-21; accord PTX-290 at 113- 16. Mr. Butler testified, however, that the Assetwise approval was invalid because the borrower's qualifying credit score of 548 was not used to generate the Assetwise approval, and it was the PRMI underwriter's responsibility to ensure Assetwise used the correct qualifying score. Butler 614:22-25, 616:10-19, 618:23-619:9. Furthermore, Mr. Butler concluded that the loan's actual credit grade under the AlterNet Guide was CM, which required a minimum FICO score of 540 and a maximum LTV ratio of 65%. Id. at 614:22-615:5, 615:9-24; accord PTX-001 at 208-11. Mr. Butler concluded that the loan violated the 1997 AlterNet Guide because the loan's 85% LTV ratio exceeded the requirements of the loan's credit grade.
154. In response, Ms. Keith testified that "the loan was properly underwritten to a maximum of 85 percent" LTV ratio. Keith 2255:4-12. Ms. Keith disagreed with Mr. Butler that the Assetwise approval was invalid, explaining that the credit grade is an "output field" of Assetwise that cannot be "force[d]" to obtain any specific credit score, id. at 2258:13-2261:18, and the Assetwise report reflected an 85% LTV ratio. Id. at 2259:3-11. In addition, Ms. Keith opined that an Assetwise approval cannot be invalidated based on an incorrect credit score because Assetwise itself pulls the qualifying credit score for the loan. Id. at 2259:17-2260:19.
155. Having considered the testimony and written evidence, the Court finds Mr. Butler's opinion on this loan to be more persuasive. To comply with its undisputed obligation to prudently underwrite loans, PRMI was not entitled simply to rely on an Assetwise approval in the face of contrary evidence in the file which would invalidate the approval. Here, the loan file contained a credit report showing that the qualifying credit score was 548, not 561, which rendered the loan ineligible for the AM credit grade under which it was approved. The Court finds that a prudent underwriter would have, at the very least, queried the discrepancy in credit scores before making the loan. Butler 619:1-17. It is uncontroverted that PRMI did not do so.
D. Mr. Butler's Identification of Breaches of Four Assetwise Loans Made Under the Client Guide
156. At trial, Ms. Keith disputed Mr. Butler's breaches for four loans (5798390, 8257547, 10381337, 6231616 ("Assetwise Loans")) solely on the basis of an Assetwise approval in the loan file. Both parties agree, however, that these loans were sold under the Client Guide, rather than the AlterNet Guide. As such, the Court has already held that ResCap had sole discretion to declare breaches with respect to these loans. PRMI SJ Order, 428 F. Supp. 3d at 92 ("RFC possessed the sole discretion under the Client Guide to make factual determinations including whether an Event of Default has occurred."). Moreover, irrespective of an Assetwise approval, the breaches on the loans identified by ResCap relate to Reps that PRMI concedes continue to apply to these loans, even under its theory of estoppel. See, e.g., Zitting 1762:16-1763:24, 1776:10-23 (conceding that PRMI was required to prudently originate Assetwise Loans under the AlterNet Guide and the Client Guide).
1. Loan 5798390
157. Loan 5798390 was originated on September 20, 2001, for $165,750. PTX-294 at 18. Mr. Butler determined that the loan violated the income documentation requirements of the Client Guide. Butler 594:3-595:9. The Assetwise Findings Report approved this loan on condition that PRMI provide the co-borrower's "[m]ost recent pay stub showing year to date earnings and W-2 form evidencing monthly salary of at least $1,733.00, or a Verification of Employment form, OR a Verification from VIE, Inc along with a verbal verification of employment." PTX-294 at 89; Butler 595:17-596:3.
158. Mr. Butler testified that even if the Assetwise approval was initially valid, it was conditional and PRMI failed to satisfy that condition—it failed to provide the most recent paystub, a complete verification of employment with year-to-date salary information, or a verbal verification of employment. Butler 596:3-18, 597:11-17 (noting "there were two pay stubs in the file, but the loan closed in October" and "the most recent pay stub we have is from July"); PTX-294 at 136-37, 141. While Ms. Keith testified that the verification of employment was in the file, she could not opine that it was complete because it did not contain year-to-date salary information. Keith 2297:1-4.
159. The documentary evidence is uncontested and demonstrates that PRMI approved the loan without complying with the Client Guide income documentation requirements and without satisfying the Assetwise Findings Report conditions. Accordingly, the Court finds that it was imprudent for PRMI to approve the loan—which ResCap could, in any event, declare a breach in its sole discretion. DTX-018 § A202(U).
2. Loan 8257547
160. Loan 8257547 was originated on November 21, 2002, for $90,750. PTX-304 at 1. Mr. Butler found that the loan did not meet the asset verification requirements of the Client Guide. Butler 597:20-598:12. The Assetwise Findings Report initially approved the loan on condition that PRMI provide $1,190.77 in verified assets. PTX-304 at 12. At trial, Mr. Butler testified that the verification of deposit ("VOD") returned by the bank raised "major questions" of concern. Butler 599:2-600:16. Specifically, the VOD listed two accounts but only one was verified, and the average monthly balance of $1,300 on the verified account was significantly less than the current balance of more than $3,600. Id. at 597:25-598:8, 599:15-600:16; PTX-304 at 146.
161. Section 471 of the applicable Client Guide, titled "Cash to Close," prescribes that, "[f]or all loan programs, if the verification of cash to close reveals a significant recent increase in the average balance of an existing account, recent large deposits or a newly-opened account with a significant balance, the [b]orrower must explain the increase and the Client must document this explanation." PTX-032 § 471. Mr. Butler testified that the condition on the Assetwise Findings Report could not be satisfied via the VOD in the file because there was no explanation in the file addressing either the failure to verify one account or the significant increase in account balance. Butler 600:17-601:5. While Ms. Keith disagreed that the failure to verify both accounts on the VOD resulted in a red flag, she speculated that the second account listed is "probably a joint account," which would need additional authorization from the non-borrower for the bank to provide information on this account. Keith 2269:21-2271:14; see also id. at 2272:8-13 (conceding that "we can't know . . . all these years later . . . what exactly transpired" but speculating that one account coming back verified is "usually because of lack of authorization from the co-owner of the account."). Ultimately, she could not dispute that the large increase in account balances reflected on the VOD should have been investigated.
162. The Court finds that ResCap had the sole discretion to declare that PRMI breached the Client Guide with respect to this loan. Regardless, the evidence demonstrates that PRMI failed to comply with the conditions for Assetwise approval and the asset verification requirements of the Client Guide. Therefore, even absent such sole discretion, in light of Ms. Keith's inability to dispute that the VOD discrepancies should have been investigated, the Court finds that the VOD could not be used to satisfy the condition on the Assetwise Findings Report.
3. Loan 10381337
163. Loan 10381337 was originated on December 23, 2005, for $448,000. PTX-280 at 7. Mr. Butler determined that pursuant to the Client Guide, the borrower was ineligible for the subject loan program, and that the loan did not comply with the verification requirements for income, employment, and cash reserves. Butler 602:23-603:4, 604:1-606:15. The Assetwise Findings Report approved the loan on condition that PRMI provide "the most recent 12 month employment history evidencing a monthly salary of at least $9,350.00 by providing the following: Most recent paystub(s) covering the last 30 days." PTX-280 at 96; Butler 604:10-22.
164. Mr. Butler testified that even if the Assetwise approval was initially valid, its condition was not met because the loan file did not show "12 months of employment where the borrower was earning at least $9,350." Butler 606:7-15. Specifically, the recent pay stub in the loan file contained "less than 6 months of income," id. at 605:15-16; PTX-280 at 171, and the verbal verification of employment from the borrower's previous employer did not reflect a salary. Butler 605:24-606:12; PTX-280 at 173. Furthermore, the specific version of the Client Guide that applied to this loan underscored that "clients who use the Assetwise electronic services are still bound by the [Reps] as set forth in the Client Guide" and "that the use of Assetwise does not relieve the clients of loan eligibility and underwriting requirements set forth in the Client Guide." Butler 603:8-25; PTX-1055 § A401.
165. Ms. Keith opined that the paystub in the file was sufficient to satisfy the Assetwise Findings Report condition for approval, even though she conceded that it reflects a $9,350 monthly salary for only six months, rather than a full year. Keith 2308:3-17. She opined that a paystub would "never cover a 12-month period" unless it was the last paystub of the year. Id. at 2308:1-22. As a result, she speculated that, from a practical standpoint, it would be "impossible" to close loans pursuant to this requirement "other than at the end of the year." Id. at 2308:18-22.
Despite various provisions in the Client Guide that made clear that loan originators such as PRMI remained responsible for ensuring compliance with Guide requirements even when using Assetwise, see, e.g., PTX-032 §§ A410, A420, Ms. Keith persisted in testifying that the Assetwise Loans did not need to comply with certain credit-related requirements of the Guide. Keith 2369:8-2372:8. The Court finds Ms. Keith's testimony in this regard lacking in credibility.
166. The Court finds that ResCap's sole discretion to declare breaches encompasses the discretion to determine that the Assetwise approval condition concerning employment history was not satisfied. PTX-1055 § 113(B); see also PRMI SJ Order, 428 F. Supp. 3d at 92. Even if ResCap did not have such discretion, the Court finds that the loan was not prudently originated—a Rep that PRMI itself concedes still applied to Assetwise loans. Zitting 1762:16-1763:24, 1776:10-23. Although Ms. Keith contended that the 12-month salary earnings requirement would only be satisfied "at the end of year" by a December paystub, the Court finds Mr. Butler's opinion in this instance to be more credible. Mr. Butler explained that a verbal verification of employment by the borrower could have also satisfied this requirement. Butler 595:1-9. The verification here, however, did not reflect the borrower's salary. The Court therefore finds that it was imprudent for PRMI to approve this loan when it did not meet the requirements of the Client Guide and did not satisfy the condition for approval on the Assetwise Findings Report demonstrating that the borrower had maintained a monthly salary of $9,350 for at least 12 months. PTX-1055 § A202(T).
4. Loan 6231616
167. Loan 6231616 was originated on October 2, 2001, for $104,890. PTX-295 at 37. Mr. Butler testified that the loan did not meet the verification requirements in the Client Guide for cash reserves and rental history. Butler 606:18-607:12, 609:24-610:19, 885:11-14. Regarding the verification requirement for the cash reserves, Mr. Butler explained, in his supplemental reunderwriting analysis, that PRMI failed to document two months of cash reserves for the borrower, which he contended violated the requirements under the Client Guide. Id. at 883:15-23. Specifically, he testified that the Client Guide required PRMI to verify that the borrower had both sufficient cash to pay closing costs and two months of additional cash reserves. Id. at 883:19-23. Although Mr. Butler agreed that PRMI documented sufficient cash to close, he maintained that PRMI failed to document two months of additional cash reserves. Id. at 883:24-884:2. On cross-examination, he admitted, however, that the Assetwise Findings Report required verification of only cash to close, and did not require verification of any cash reserves. Id. at 885:11-14; see also PTX-295 at 84-85.
168. As for the verification requirements for the rental history, Mr. Butler opined that the Client Guide required verification of 12 months of rental history. Butler 607:16-608:3; DTX-018 § 441. It is uncontested that PRMI verified only 11 months. Butler 609:19-23; compare PTX-295 at 73 (10/2/2001 Loan Application showing 12 months of residence at the borrower's current rental property), with PTX-295 at 118 (9/26/2001 Verification of Rent showing 11 months of verified rental history at the borrower's current rental property). Mr. Butler thus opined that the Assetwise approval was not valid because PRMI failed to obtain the required evidence of rental history under the Client Guide. Butler 606:20-609:8, 609:24-610:19.
169. While Ms. Keith conceded that Assetwise could not verify rental history, she testified that the one-year rental period reflected on the loan application was not an input into Assetwise and thus could not impact the approval on the Assetwise Findings Report. Keith 2286:12-15, 2288:4-11. She further claimed that while the Assetwise Findings Report could have conditioned the approval on verification of 12 months of housing history, it did not. Id. at 2282:25-2283:16, 2286:5-10. Ms. Keith thus disputed Mr. Butler's invalidation of the Assetwise approval.
170. Having considered the testimony and written evidence, the Court finds that ResCap had the sole discretion to declare PRMI's breach. Regardless, the uncontested evidence demonstrates that PRMI failed to obtain 12 months of rental history despite the fact that the Client Guide required that it do so and PRMI understood that Assetwise could not verify rental history. Butler 609:24-610:19; Keith 2286:12-15.
E. Mr. Butler's Identification of Breaches of the Countrywide Loan: Loan 10226985
171. During the course of the parties' relationship, RFC purchased a single pool of loans in late 2005 that PRMI originally underwrote to Countrywide guidelines ("Countrywide Pool"). The PRMI Loans reunderwritten by Mr. Butler contain one loan from the Countrywide Pool—Loan 10226985. Crawford 2430:17-25; Butler 588:14-19.
172. Loan 10226985 was originally reunderwritten by Mr. Butler to the Client Guide but, in response to Ms. Keith's rebuttal, he also reunderwrote the loan to the Countrywide Technical Manual. Butler 589:20-590:7. Mr. Butler identified four breaches of the Countrywide Technical Manual: (1) the borrower's misrepresentation of debt, (2) the borrower's excessive DTI ratio, (3) certain unexplained credit inquiries, and (4) an unreasonable stated income. Id. at 588:17-25. Specifically, Mr. Butler testified that the borrower failed to disclose another loan taken just one month before the subject loan. Id. at 589:1-8. Adding the monthly payment for the undisclosed loan, the borrower's DTI ratio exceeded that on the approval. Id. at 589:1-13. According to Mr. Butler, the borrower's misrepresentation of debt and DTI ratio "would disqualify the loan under any set of guidelines." Id. at 590:4-7; see also id. at 589:9-16 ("[the] excessive DTI ratio . . . exceeded the CLUES approval and would have exceeded any guideline").
173. At trial, Ms. Keith did not dispute the borrower's misrepresentation or the resulting DTI ratio, and agreed that no set of guidelines would allow fraud. Compare Butler 589:4-6 with Keith 2252:14-16 ("[T]here's no mortgage underwriting guide in the world that I've ever seen that says that fraud is okay."). PRMI's own witnesses also agreed that a prohibition on misrepresentation in origination is an industry standard Rep required by all investors, including Countrywide. See Crawford 2451:1-14; Flitton Dep. 67:20-24, 69:24-70:5, 89:2-16. The Court therefore finds that ResCap easily meets its burden to establish that the Countrywide loan was in breach.
III. PRMI'S BREACHES CONTRIBUTED TO RFC'S LIABILITY
174. The evidence at trial demonstrated that PRMI's origination-level breaches certainly contributed to RFC's liability in the Bankruptcy in several ways. First, the evidence from the Bankruptcy demonstrated that origination-level breaches contributed to RFC's risk of liability. In particular, proofs of claim filed in the Bankruptcy actually asserted origination-level breaches, and the parties to the Bankruptcy sought to assess RFC's risk of liability, in part, by looking to origination-level underwriting guidelines to evaluate the extent of such origination-level breaches. Second, the credible testimony of certain of ResCap's witnesses confirmed the extent to which PRMI's origination-level breaches contributed to RFC's risk of liability. Finally, in evaluating PRMI's argument that there was a gap between origination-level Reps and Trust and Monoline Reps, the Court finds ResCap's interpretations of its Trust and Monoline Reps to be more persuasive. At a minimum, there was certainly a risk that a judge or jury might interpret those Reps to assign liability to RFC.
A. Bankruptcy Evidence Shows That PRMI's Breaches Contributed to RFC's Liability
1. Proofs of Claim in the Bankruptcy Asserted Origination-Level Breaches
175. As described earlier, the Bankruptcy proofs of claim filed by the RMBS Trusts and the Monolines asserted origination defects against RFC. See, e.g., PTX-155 at 4, 11-13; DTX-543 at 4, 11-13. More specifically, the RMBS Trusts asserted breaches of "Trust Representations and Warranties" (i.e., those made by RFC to the RMBS Trusts and Monolines) arising from origination defects such as the creditworthiness of the borrowers, the standards and practices used in underwriting each mortgage loan, the characteristics of each mortgage loan, and information on the mortgage loan tapes. See PTX-155 at 4, 11-13; DTX-543 at 4, 11-13.
176. Similarly, it is uncontested that the Monolines asserted breaches of representations RFC made as to "loan-level characteristics," including alleged flaws in the "underwriting and origination processes," noting "a shockingly high percentage of the Mortgage Loans . . . contained defects constituting breaches of loan-level [Reps]." See DTX-494 at 4; PTX-146 at 9-10; DTX-497 at 8; DTX-569 at 17-18.
177. Witnesses from RFC's Bankruptcy testified that RFC faced claims asserting loan-level origination defects in both pre-bankruptcy litigation and the proofs of claim. For example, Mr. Lipps testified that the breaches alleged by MBIA "were [] a result of the characteristics of the loans." Lipps HLC 1230:25-1231:2. Likewise, Tammy Hamzehpour, RFC's General Counsel before the Bankruptcy, testified that the defect rates for the mortgage loans at issue in the MBIA litigation were "shockingly higher" than RFC had previously believed. Hamzehpour Dep. 126:15-127:19.
2. The Bankruptcy Participants Measured RFC's Liability By Examining Origination Guideline Breaches
178. Further establishing that PRMI's breaches contributed to RFC's liability is the fact that within RFC's Bankruptcy, the primary parties assessed loan-level origination breaches in order to gauge RFC's potential repurchase exposure. See Hawthorne 1233:5-16; Morrow Dep. 27:4-28:3; PTX-153 at 6; Pfeiffer Dep. 52:7-53:16, 99:07-101:17, 126:1-11, 146:19-147:4.
179. At trial, ResCap's expert, Donald Hawthorne, was the only expert to assess the work done by the reunderwriting experts in the Bankruptcy. Mr. Hawthorne testified that in order to measure RFC's potential repurchase exposure, the reunderwriters for the relevant parties in the Bankruptcy "determined breach rates by re-underwriting against the guidelines" used at the origination level. Hawthorne 1048:9-25, 1049:19-23 (explaining he reviewed the work of the Bankruptcy underwriting experts used by RFC, the Trusts, and the Unsecured Creditors Committee). In fact, Mr. Hawthorne explained that not only did the testimony from the Bankruptcy indicate that the Bankruptcy underwriters "relied on the guidelines for their re-underwriting," but it was also "apparent" from the breach rates the parties were identifying that "they were applying the guidelines to the loans [at issue] as a whole" regardless of the specific representations, warranties, and disclaimers provided by RFC to the Trust and Monolines. Id. at 1050:5-13.
180. Mr. Hawthorne's testimony is consistent with the underlying evidence from RFC's Bankruptcy. The Committee's reunderwriting expert, J.F. Morrow, was asked to "re-underwrite loans for their compliance with the applicable underwriting guidelines when purchased or made," such that his reunderwriting was "narrowly tailored to identify material compliance with the ResCap guidelines." Morrow Dep. 27:4-28:3. RFC's own expert, Frank Sillman, also "used a lack of substantial compliance with applicable guidelines as a proxy for breaches of contractual [Reps] . . . ." Sillman Dep. 144:19-145:3. And the RMBS Trustees' financial advisor, Duff, also analyzed RFC's potential liability and repurchase exposure by determining "whether the initial origination was done in conformity . . . with the guidelines." Pfeiffer Dep. 126:1-11. Importantly, the Bankruptcy Court's Confirmation Order and accompanying Findings of Fact relied on the foregoing experts' opinions. See PTX-179 at 44, 49; PTX-180 at 1-2 n.2.
181. Based on the foregoing, the Court finds that the claims asserted against RFC—which RFC settled in the Settlements—included claims asserting origination defects resulting from PRMI's underlying breaches. The parties to the Bankruptcy evidently believed the origination defect claims were significant, as both the RMBS Trusts and Monolines asserted widespread origination defects, and all parties assessed RFC's potential liability by examining breaches of the underlying guidelines. PRMI and other originators were responsible for compliance with those guidelines.
182. PRMI did not present any evidence to the contrary. While cross-examining Mr. Hawthorne, PRMI attempted to imply that the Bankruptcy experts were mistaken in assuming that guideline breaches were an appropriate proxy for breaches of RFC's Reps. See, e.g., Hawthorne 1241:3-6. But this implication is contradicted by the evidence. The RMBS Trustees' advisor, Duff, was specifically tasked with assessing RFC's Rep liability, but nonetheless utilized guideline breaches. Pfeiffer Dep. 126:6-11 (noting Duff utilized origination guidelines to determine whether "initial origination was done in conformity . . . with the guidelines"); Id. at 128:18-23 (noting Duff determined which origination guidelines matched which particular loan it was reviewing). Moreover, even though the Committee opposed the Original RMBS Trust Settlement, it had no objection to Mr. Sillman's reunderwriting to the guidelines, and in fact instructed its own expert, Dr. Cornell, to also reunderwrite to the guidelines. Hawthorne 1233:5-16.
183. It defies credibility to suggest that all of the experts and their attorneys in the Bankruptcy simply made a mistake, or engaged in the wrong analysis, and that the Bankruptcy Court erred in relying on those analyses in approving the Settlements. See Hawthorne 1241:7-15 (observing that "the notion that all the experts [in the Bankruptcy], and all of their attorneys and everyone involved" in an extended bankruptcy proceeding "somehow made a mistake or were looking at the wrong thing is [in his opinion] a little bit odd" (emphasis added)). That is particularly so in light of the lack of any evidence supporting PRMI's speculation.
B. PRMI's Breaches Contributed to RFC's Risk of Liability on Specific Trust Representations
1. Mr. Butler Identified PRMI Breaches That Created Risk
184. Beyond the origination-level reunderwriting noted above, Mr. Butler also determined, for each set of loans, whether PRMI's breaches of its Reps to RFC contributed to RFC's liability to the Trusts and Monolines. Butler 548:21-549:5 (noting assignment was to determine whether "any of the [reps and warranties] that were made by . . . [PRMI] to RFC were breached"), 549:9-18 (noting that he also reviewed whether PRMI's breaches to RFC did or "could be construed" to cause RFC to breach its representations "to the Trusts and the [Monolines]"). In particular, Mr. Butler addressed whether the facts underlying PRMI's breaches of its Reps to RFC also constituted breaches or could be construed to breach Reps RFC made to Trusts and Monolines. Id. at 548:21-549:5. Mr. Butler explained that the phrase "could be construed," in this context, means that a breach could be "considered to be a breach" by "third parties" including "the Court." Id. at 623:24-624:11, 655:12-15 (discussing his opinion "that RFC was at risk for having to repurchase loans that violated the representations that . . . RFC made to the trusts and the monolines."); 746:24-747:5 (similar).
185. For this review, Mr. Butler directed a vendor, Alix Partners, to provide him with the relevant documents that contained the Reps that RFC made to the Trusts and Monolines. Id. at 624:14-625:2. After reviewing RFC's Trust- and Monoline-level Reps, Mr. Butler identified which were related to loan quality and underwriting guidelines and examined which were breached, or could be construed to be breached as a result of PRMI's breaches of Reps to RFC. Id. at 624:18-625:8. Mr. Butler engaged in the same process for the PRMI Loans in Trusts that were insured by Monolines, using the Reps in the relevant insurance agreements. Id. at 694:20-695:8.
186. Although Mr. Butler's Trust-level review typically involved comparing PRMI's breaches to the text of the Rep RFC made to the Trust or Monoline, this was not the case for the MLS Rep. With slight, inconsequential variations in wording, RFC represented to every At-Issue Trust that the "information set forth on the [MLS] with respect to each Mortgage Loan is true and correct in all material respects." PTX-406 at 5; Butler 646:23-647:6 (same), 625:12-17 (noting an MLS consists of "all . . . of [the] information about each loan at a loan-by-loan level"). Mr. Butler thus needed to review PRMI's breach against the information in the MLS to determine if this Rep had been breached or could be construed to have been breached.
187. Again working with AlixPartners, Mr. Butler obtained the MLS for the at-issue Trusts from public databases and a third-party vendor called MBS Data. Butler 625:25-626:19. To the extent AlixPartners was able to locate the original MLS for the Trust, it provided that to Mr. Butler. Id. To the extent the original MLS was unavailable, AlixPartners created a proxy MLS from data obtained from MBS Data. Id. at 626:4-10. The Court has previously held, several times, that the MLS created by AlixPartners is a reliable proxy for the data in the original MLS. PRMI Daubert Order, 432 F. Supp. 3d at 925 (noting three prior orders holding the same).
188. Based on his review of the Trust-level Reps, Mr. Butler testified that 73 of the 101 breaching PRMI Loans breached or could be construed to breach RFC's Reps to the Trusts. Butler 627:3-7. He also opined that all 31 breaching PRMI Loans that were insured by Monolines breached or could be construed to breach RFC's Reps to the Monolines. Id. at 701:24-702:7. In total, Mr. Butler testified that 77 PRMI Loans (49%) breached PRMI's Reps to RFC, and breached or could be construed to breach RFC's Reps to Trusts and Monolines. Id. at 703:17-22.
189. Mr. Butler's determinations reflected his opinions regarding the risk posed by various Trust Reps. Accordingly if, for example, a Fraud Disclaimer was present in a given transaction, Mr. Butler testified that RFC still suffered from a risk of repurchase stemming from PRMI's breaches because "if there was a loan securitized with fraud [or misrepresentation], RFC would have . . . the risk of repurchase." Id. at 660:14-662:10 (explaining that RFC repurchased loans even where there was a Fraud Disclaimer), 663:4-664:11 ("RFC may have its fraud disclaimer, but at the same time, there are other reps that will probably be breached by the cause of the fraud."). Similarly, where the Trust Agreements contained a clause passing through an originator's "no fraud" representation, Mr. Butler testified that RFC still was at risk of repurchase. Id. at 937:15-21. Mr. Butler offered similar testimony identifying the risk of RFC liability stemming from each disputed Trust Rep. See, e.g., id. at 655:12-15 (MLS and No Default Reps), 685:6-15 (Credit Grade Reps).
190. Mr. Butler's opinions about the existence of risk of liability with respect to each Trust Rep—which the Court finds highly credible—were also corroborated by testimony from both Ms. Farley and Mr. Hawthorne. See, e.g., Farley 141:23-142:10 ("[E]very [Trust Rep] we had brought with it a chance of liability[.]"), 170:7-15 (noting that if RFC's representations on the MLS were inaccurate, RFC "would have to repurchase the loan or, if possible, cure it"), 200:24-201:8 (noting that Fraud Disclaimer was "not a silver bullet" and that RFC did not "necessarily [think it was] going to be effective"); Hawthorne 1030:22-1034:16 (discussing how a "no default" rep was, in many respects, equivalent to a "no fraud" or "misrepresentation" rep), 1041:9-1042:22 (same as to "MLS" rep).
2. PRMI's Experts Did Not Address the Issue of Risk of Liability
191. Although PRMI's experts opined about their preferred interpretation of the various Trust Reps, they never addressed whether RFC faced a reasonable risk that a judge or jury in litigation might arrive at a different interpretation—namely, whether PRMI's breaches created a risk of RFC liability to the Trusts and Monolines.
192. In fact, to the contrary, each of PRMI's experts explicitly denied that they were addressing any risk of liability. Ms. Keith, for example, disclaimed any opinion as to the risk of a court adopting any particular interpretation of the Trust Reps. Keith 2344:8-11 (testifying that she was "not offering an opinion as to RFC's potential legal liability at the time of the bankruptcy settlements for breaches of the trust representations"), 2352:13-18 (noting that "one of the things" Ms. Keith does not do and "[has] never done" in her work as an expert witness "is interpret judicial decisions" or "how they might apply in the future"), 2355:17-2356:11 (testifying that she did not have an opinion on "the risk to RFC" stemming from Trust reps, and asserting that she was not opining on RFC's liability). Thus, although Ms. Keith disagreed with many of Mr. Butler's securitization-level breach findings, her testimony did not address the key issue before this Court—whether PRMI's breaches contributed to a risk of RFC liability to the Trusts and Monolines.
For her Trust-level review, Ms. Keith assumed Mr. Butler's identified breaches of PRMI's Reps to RFC were accurate. Keith 2309:23-2310:4. She concluded, however, that of the 73 PRMI Loans that breached or could be construed to breach RFC's Trust Reps, Mr. Butler's Trust-level breaches should be cleared for 58 PRMI Loans. Id. at 2315:2-12 (clearing Trust breaches for 18 loans that breached only the Credit Grade Rep), 2318:12-24 (clearing Trust breaches for an additional three loans that breached only the Loan Program Rep), 2336:2-14 (clearing Trust breaches for an additional 27 loans that were securitized in a Trust with a Fraud Disclaimer), 2339:1-10 (clearing Trust breaches for an additional 10 loans that breached the No Default Rep or MLS Rep, and were securitized in Trusts without a No Fraud Rep). Ms. Keith also concluded that of the 31 loans that Mr. Butler found breached or could be construed to breach RFC's Reps to Monolines, all should be cleared. Id. at 2342:23-2343:1.
193. Much like Ms. Keith, Professor Schwarcz testified that RFC's risk of liability for claims premised on ResCap's interpretation of the Trust Reps was "beyond the scope of [his] report" and testimony, which did not "analyze how RFC should have reacted or not to particular repurchase demands" or address "potential liabilities." Schwarcz 2060:17-2062:18. Professor Schwarcz also explained that he was "not attempting to independently . . . interpret the [Trust Reps]," id. at 2106:18-20, or say "how a court would interpret" the Trust Reps. Id. at 2063:21-2064:3. Likewise, Professor Schwarcz also conceded he was not opining "as to how the industry would take into account various legal decisions" concerning Trust Reps. Id. at 2060:7-16.
194. Mr. Burnaman testified that he was "not qualified to offer" an opinion about "the legal interpretation of the representations," Burnaman 1604:11-14, and limited his opinions to purported industry custom and practice, divorced from any assessment of risk. Id. at 1544:12-1545:12.
195. PRMI's experts' failure to address the critical inquiry as to the issue of causation—the risk of liability faced by RFC—is perhaps best exemplified by their testimony about the Fraud Disclaimer. Despite relying heavily on the Fraud Disclaimer in closing arguments, see, e.g., PRMI Closing 2763, PRMI's experts expressly disavowed offering any opinion about the legal meaning or effect of a Fraud Disclaimer on RFC's risk of liability. Professor Schwarcz testified that he was "not giving an opinion as to risk of liability [with regards to the Fraud Disclaimer]." Schwarcz 2116:23-2118:3; id. at 2127:24-2129:7. Mr. Burnaman testified that he was not opining on the legal meaning of the Fraud Disclaimer, or its ability to mitigate RFC's risk. Burnaman 1622:4-1623:1. And Ms. Keith testified that her opinions regarding the Fraud Disclaimer "don't relate . . . to what [are] viable legal defenses in a court." Keith 2363:20-25.
Professor Schwarcz also testified that RFC "should not have liability for misrepresentations by borrowers in trusts with a fraud disclaimer," Schwarcz 2117:4-6, but explained his statement was "based upon [his] understanding of what these reps and warranties are interpreted to mean [in the industry]," and not a prediction of "how a court would come out." Id. at 2128:17-2129:1.
196. PRMI's experts' failure to address RFC's potential risk of liability is also confirmed by their documentary review. Unlike Mr. Butler, PRMI's experts did not review any repurchase demands made to RFC. See Schwarcz 2068:4-2071:5; Burnaman 1624:9-11. Nor did they review any entries in the RFC investor repurchase database, see Schwarcz 2071:6-10; Burnaman 1624:20-1625:15, despite acknowledging that repurchase demands are "helpful" in understanding investors' interpretation of the Trust Reps, Burnaman 1624:12-19.
To the limited extent Mr. Woll addressed these issues, he too did not deny that RFC faced a risk of liability on the relevant Trust Reps. See, e.g., Woll 1954:6-13.
197. Accordingly, based on the uncontroverted, credible testimony of Mr. Butler, Ms. Farley, and Mr. Hawthorne, the Court finds that PRMI's breaches of Reps to RFC created a material risk that RFC could be found liable based on its breached Reps to the Trusts and Monolines.
C. The Court Finds ResCap's Evidence as to the Proper Interpretations of the Trust Reps & Warranties More Persuasive
1. Client Contracts Gave Enforceable Loan-Level Representations
198. Although the Court addresses each of the disputed Reps in turn below, the Court first notes that ResCap provided largely undisputed evidence that the structure, function, and purpose of the Trust Agreements that contained the disputed Trust Reps was to provide meaningful, enforceable, loan-level remedies.
199. The Trust Agreements were designed to effectuate the sale of individual mortgage loans. Accordingly, they included both Trust Reps regarding those individual mortgage loans, expressed on a loan-by-loan basis, and remedies for breaches of the Trust Reps as to individual mortgage loans, in the form of repurchase and similar provisions that were expressed as and operated on a loan-by-loan basis. See, e.g., PTX-1150 at 3 ("RFC represents and warrants to the [Depositor], with respect to each Mortgage Loan . . . ."); PTX-1150 at 8-9 ("Upon discovery . . . of a breach . . . in respect of any Mortgage Loan . . . RFC shall . . . either cure such breach or . . . purchase such Mortgage Loan . . . ."); see Farley 129:22-130:18 (noting that the remedies in the Assignment and Assumption Agreements were "loan-by-loan remedies"), 133:8-134:1 (describing purpose of Pooling and Servicing Agreement), 145:25-147:22 (noting that the reps and warranties in the agreements at issue were "about each individual loan being sold under [the] document"), 150:10-151:17 (describing the reps and warranties in the Trust Agreements as "loan-level reps" and "individual reps and warranties relating to the mortgage loans being sold"), 154:4-6 (same), 161:19-23 (same), 190:21-191:15 (noting the importance of the loan-level reps and warranties). These representations, warranties, and remedies were necessary and appropriate because depositors and RMBS Trusts—which, again, were acquiring individual mortgage loans—needed to know "what they're buying." Id. at 167:19-168:6.
200. The Court finds Ms. Farley's testimony on the foregoing facts to be highly credible and well supported by the evidence. It has considered her testimony when evaluating the parties' divergent interpretations of the Trust Reps, discussed below.
2. The MLS Reps Were Not Merely Transcription Reps
201. The MLS Rep appears in all at-issue RFC trusts and generally provides that "[t]he information set forth on the [MLS] with respect to each Mortgage Loan is true and correct in all material respects." PTX-406 at 5; DTX-786 at 15; Schwarcz 2047:9-15. Whereas ResCap contends the MLS Rep warranted the substantive accuracy of information on the MLS, PRMI contends it only warranted the accuracy of RFC's transcription of information from the loan file to the MLS. The Court finds that the evidence at trial better supports ResCap's interpretation in several respects.
a. Ms. Farley's Testimony
202. First, the Court considers Ms. Farley's testimony on the meaning of the MLS Rep to be credible and persuasive. She testified that the MLS Rep "means what it says[,]" namely, "[t]hat the information on the [MLS] relating to each loan was true and correct." Farley 163:6-9; see also id. at 174:20-23. Throughout her 35 years in the RMBS industry, during which she drafted and negotiated some of the transaction documents at issue, Ms. Farley's understanding of the meaning of the MLS Representation has not changed. Id. at 206:23-207:10. Notably, Ms. Farley testified that during her 35 years in the RMBS industry, she has never heard anyone express the view that the MLS Rep was merely a transcription representation. Id. at 168:7-20. Ms. Farley supported her testimony with specific uncontradicted recollections that RFC consistently resisted adding fields to the MLS because RFC understood it would be exposing itself to liability for the substantive accuracy of the data to be provided in those fields, and recollections that RFC took steps to update and "substantiate [the objective accuracy of] any[] [data] that [it] thought had changed" between the date RFC acquired the loans and the date it sold the loans. Id. at 141:23-142:10, 168:21-171:17.
b. Mr. Butler's Testimony
203. Second, the Court also finds Mr. Butler's testimony on the MLS Rep to be both credible and persuasive. He explained, like Ms. Farley, that his understanding of the MLS Rep—based on his years of experience, the language of the Rep, and his knowledge of how it has been interpreted by other courts—was that it warranted substantive, not merely transcription, accuracy. Butler 647:17-648:13, 651:1-10. Moreover, his testimony was corroborated by other evidence he reviewed. As he explained, the testimony of Lisa Lundsten, a former Managing Director of Structured Finance at RFC, indicated that she understood it would be a breach of the MLS Rep if the MLS data was not "true and correct in all material respects," including, for example, "if a debt-to-income ratio were higher than what was represented," or if there was "fraud in the origination of the loan[.]" Id. at 648:17-649:11, 650:4-23. Additionally, Mr. Butler pointed to one documented instance in which a Monoline asserted an MLS Rep breach based on misrepresented income and an incorrect DTI ratio, and RFC repurchased the loan. Id. at 651:13-655:15. And he discussed an email from former RFC employee Pieter Van Zyl, identifying the MLS Rep as a representation that could be breached "in the event of fraud other than a specific fraud rep[.]" Id. at 655:18-656:6.
c. Judicial Interpretation of the MLS Rep
204. Third, the Court notes that both Mr. Hawthorne and PRMI's own experts acknowledged that plaintiffs in RMBS cases have repeatedly argued that the MLS Rep warrants substantive accuracy, and courts (including at least one that pre-dates the RFC settlements) have agreed. Hawthorne 1041:1-1043:4; Burnaman 1594:2-20, 1601:17-1602:11; Schwarcz 2048:10-2049:14, 2081:1-13; Keith 2350:12-24.
d. Actual Reliance by Stakeholders
205. Fourth, it is undisputed that investors, securities underwriters, Monolines, and credit rating agencies all actually relied on the substantive accuracy of information on the MLS. Farley 163:6-13 (noting the MLS Rep was "critical" because of the seller, investor, and rating agency reliance on the MLS), 172:9-174:19 (discussing rating agency, insurer, investor, and underwriter reliance on substantive accuracy of MLS); Butler 647:9-16 (noting the MLS is "information that people involved in the transaction, investors, [M]onolines, everybody, is relying on"); Burnaman 1585:24-1586:6 (noting purpose of MLS Rep was to provide comfort that the "information" in the MLS "was accurate" and "reflected what was in the underlying loan file"). As Ms. Farley testified, "when you have a sale contract . . . a purchaser . . . needs to be able to rely on the seller's rep as to what they're buying. Otherwise, they're not able to make an investment decision" or appropriately "price the asset." Farley 167:19-168:6.
e. The MLS Rep Breach Remedy is a Loan-Level Remedy
206. Fifth, the remedy set forth in the Trust Agreements for a breach of the MLS Rep is cure, repurchase, or substitution. The purpose of that remedy was "to keep the pool intact relative to its risk parameters at the time the pool was sold," id. at 164:2-14—namely, to replace the loan with another that had the substantive characteristics that had been represented. It would be illogical for the MLS Rep to merely warrant accurate transcription of loan characteristics because if that were true, the repurchase remedy for breaches of the Rep "wouldn't even work" and "just doesn't make sense," because the error could be corrected by changing a number on the MLS. Id. at 168:7-20.
f. No Persuasive PRMI Evidence Was Introduced to the Contrary
207. As against all the foregoing evidence, PRMI offered no fact testimony or documentary evidence in support of its position. Instead, it offered conclusory expert opinion, unsupported by corroborating evidence, which this Court deems unpersuasive. See Schwarcz 2045:16-2046:7 (opining the MLS Rep was a transcription representation); Burnaman 1585:14-1586:6 (opining that industry understanding of MLS Rep was designed to give "comfort that this information was accurate"); Keith 2346:12-20 (opining that MLS Rep involved "whether or not [the underlying data] had been transcribed for loan approval correctly, not whether or not the data itself was inherently correct").
208. While Mr. Burnaman opined that the MLS Rep was merely a transcription warranty, he acknowledged that investors relied on MLS data to model financial performance and cash flows, and as such would want the MLS data to be objectively accurate. Burnaman 1585:2-1586:6, 1600:10-1601:16. He was also effectively impeached with deposition testimony in which he admitted that a reasonable RMBS industry participant "would [] say without hesitation that if the objective information on the MLS schedule . . . was objectively untrue, [then] the MLS Rep was violated." Id. at 1598:10-1599:9 (emphasis added).
209. Likewise, Professor Schwarcz also admitted that securitization Reps were drafted using "plain language" and "avoided language that would require counterintuitive interpretations or that operated by implication rather than expressly." Schwarcz 2073:22-2075:23, 2078:3-15. Nonetheless, he argued that the MLS Rep is limited to transcription inaccuracies by implication, not the plain language of the Rep. Id. at 2045:10-2046:7. And while Ms. Keith attempted to support Mr. Burnaman's and Professor Schwarcz's testimony that the MLS Rep related to accurate transcription, see Keith 2323:8-18 (noting that people used to "hand-enter" MLS data), she nevertheless conceded that if certain fields in the MLS Rep were inaccurate, then there would be a breach of that representation. Id. at 2348:17-22. She was also confronted with her deposition testimony in which she stated that in her industry experience, the MLS Rep means that "all of that data is true and correct." Id. at 2349:15-2350:2.
210. Given the plain language and structure of the MLS Rep, and the testimony of Ms. Farley and the various experts introduced at trial, the Court finds ResCap's position that the MLS Rep warrants the substantive accuracy of the information in the MLS to be more persuasive.
3. The No-Default Reps Were Not Merely Payment Default Reps
211. The "No Default Rep" appears in every at-issue RFC Trust in one of three variations. See, e.g., PTX-1150 at 6; DTX-786 at 12; Schwarcz 2039:18-2040:3; Hawthorne 1034:17-22, 1038:6-12. One of the three versions states:
Other than with respect to a payment default, there is no material default, breach, violation or event of acceleration existing under the terms of any Mortgage Note or Mortgage and no event which, with notice and expiration of any grace or cure period, would constitute a material default, breach, violation or event of acceleration under the terms of any Mortgage Note or Mortgage, and no such material default, breach, violation or event of acceleration has been waived by RFC or by any other entity involved in originating or servicing a Mortgage Loan.PTX-107 at 5. The other two versions are nearly identical except that one replaces the opening phrase with the words, "Except as otherwise specifically set forth herein," see PTX-1150 at 6, and the other replaces the opening phrase with the words, "To the best of RFC's knowledge as of the Cut-off Date," see DTX-1001.0232 at 3.
212. ResCap contended that the No Default Rep in each of its variations warranted against any default as defined in the cited "Mortgage Note[s]," see ResCap Closing 2815:11-18, while PRMI argued that it warranted only against monetary defaults, see PRMI Closing 2758:1-5. The Court finds that ResCap's interpretation of the No Default Rep—in each of its variations—is more persuasive than PRMI's interpretation, for several reasons.
213. First, Ms. Farley testified credibly and persuasively that the No Default Rep means what it says—namely, that there was no default as defined under the mortgage note or the mortgage. Such a default occurs when the borrower provides materially false or inaccurate information in connection with the application for the loan, or fails to provide material information. See Farley 177:14-179:1. Throughout her 35 years in the RMBS industry, during which she drafted and negotiated some of the transaction documents at issue, Ms. Farley's understanding of the meaning of the No Default Rep never changed. Id. at 206:23-207:7. She also testified that she never heard the view expressed that the No Default Rep was somehow limited to only payment defaults. Id. at 179:18-21.
See, e.g., PTX-284 at 14 ("Borrower's Loan Application. Borrower shall be in default if, during the Loan application process, Borrower or any persons or entities acting at the direction of Borrower or with Borrower's knowledge or consent gave materially false, misleading, or inaccurate information or statements to Lender (or failed to provide Lender with material information) in connection with the Loan. Material representations include, but are not limited to, representations concerning Borrower's occupancy of the Property as Borrower's principal residence.").
214. Ms. Farley supported her testimony with a specific uncontradicted recollection of an incident in the early 1990s. A supervisor newly assigned to Ms. Farley's group "suggested that [RFC] try to limit [the No Default Rep] to just payment defaults," and assigned Ms. Farley "the [un]enviable task of calling up the rating agencies and asking them if they would mind limiting [the No Default Rep]" to just payment defaults. Id. at 179:22-180:13. Ms. Farley explained that the ratings agencies "minded and the rep was not limited." Id. at 180:10-13. Ms. Farley testified that the task was unenviable because "[i]t was obvious that this was a provision that was way beyond just payment, that [] misreps were important, and I knew and [the co-head of structured finance] knew and the lawyers['] external counsel . . . knew that there was no way the rating agencies were going to take a request like that seriously." Id. at 180:19-181:3.
215. Second, Mr. Butler testified to the same effect. Butler 635:4-638:9. His opinion is based on years of industry experience as corroborated by materials he reviewed relevant to this matter, including testimony from former RFC employee Ms. Lundsten, who testified that "if the borrower misrepresented information, that would breach the mortgage, and that in turn would breach the [No Default] representation[.]" Id. at 639:3-12, 640:9-13. Mr. Butler also supported his testimony by discussing a documented instance in which a Monoline asserted a No Default Rep breach based on misrepresented income, and RFC repurchased the loan. Id. at 651:13-655:15.
216. Third, the Court notes that as Mr. Hawthorne explained (and PRMI's experts acknowledged), courts hearing RMBS cases have interpreted the No Default Rep as warranting against borrower misrepresentations. Hawthorne 1031:23-1035:20; Burnaman 1594:2-20, 1614:15-20; Schwarcz 2105:22-2106:6. Moreover, at least one such court decision pre-dates the RFC settlements. Schwarcz 2081:1-13.
217. Fourth, ResCap's interpretation more accurately comports with the relevant commercial realities within the RMBS industry. As Ms. Farley explained, the No Default Rep was important because it related "to the integrity of the mortgage loan itself" because it went to "understanding the willingness [and] ability of the borrower to pay," which is "the heart of a [mortgage loan] transaction." Farley 179:2-9. Additionally, Ms. Farley testified that based on her dealings with Monoline insurers, the Monolines also understood the No Default Rep covered borrower misrepresentations. Id. at 181:22-182:7.
218. Fifth, the Court finds that ResCap's interpretation of the No Default Rep is the most logically sound, particularly given the version of the No Default Rep that explicitly excludes payment defaults. In this regard, the testimony of PRMI's expert, Professor Schwarcz, illustrated the weaknesses of PRMI's interpretation of the Rep when he conceded that "because of that payment carveout [] it's unclear [] what that no-default representation would mean," admitted that he is "not expressing a view" regarding No Default Reps with payment default carve-outs because he doesn't "think there would be a clear industry understanding of that," and stated that he "[is]n't sure how far [that version of the No Default Rep] would go beyond payment default." Schwarcz 2044:15-25, 2097:9-23, 2099:20-2100:11.
219. As against all the foregoing evidence, PRMI offered no fact testimony or documentary evidence in support of its position. Instead, it offered conclusory expert opinion that is unsupported by evidence, which this Court deems unpersuasive. See Burnaman 1591:13-1592:7 (opining that the No Default Rep was "meant to cover issues of [] payment default"); Schwarcz 2039:15-2040:3 (opining that the No Default Rep was "limited to payment defaults"); Keith 2325:8-25 (same).
220. With respect to Mr. Burnaman, the Court notes that contrary to PRMI's position, Mr. Burnaman admitted (1) that the No Default Rep covered other issues "that are not necessarily payment default issues but are nonetheless defaults under the underlying mortgage"; (2) that the No Default Rep "might . . . cover," as a "matter of [his] general understanding of the industry," criteria in the mortgage or note triggering a default; (3) that "things a borrower is required to do under the terms of a mortgage . . . might be covered" by the No Default Rep; and (4) that he is "[n]ot sure" whether the No Default Rep "covers the terms of the underlying mortgage note." Burnaman 1591:13-1592:7, 1608:10-1610:24.
221. Similarly, despite endorsing plain language readings of the relevant transaction documents, Professor Schwarcz acknowledged that "[t]here is no language explicitly limiting [the No Default Rep] to payment defaults." Schwarcz 2107:1-6.
222. Moreover, with respect to Ms. Keith, she testified that while "you have to read the actual language of each [No Default Rep] to try to interpret what types of defaults are covered . . . basically, the seller's representing that there's no default, breach, violation or event of acceleration existing under any mortgage note or mortgage." Keith 2325:22-2326:12.
223. In fact, PRMI's experts either did not review or did not recall reviewing the at-issue mortgages or notes in forming their opinions about the No Default Rep. Burnaman 1611:5-8; Schwarcz 2101:11-2103:6. Indeed, notwithstanding his criticism of ResCap for purportedly adopting a "contorted interpretation" of the No Default Rep, Professor Schwarcz testified that the default terms of the mortgage and note were "beyond the scope of [his] report." Schwarcz 2042:10-21, 2104:17-2105:14. Regardless, PRMI's experts acknowledged that under the standard industry form of mortgage and note, a misrepresentation by the borrower is an event of default. Burnaman 1612:5-8.
224. PRMI's experts also testified that the variation of the No Default Rep beginning with "To the best of RFC's knowledge as of the Cut-off Date" meant that RFC "eliminated [its] responsibility" if its knowledge of the default was not established. Keith 2328:15-2329:16; see also Schwarcz 2044:1-9. However, PRMI's experts subsequently disclaimed any opinion as to RFC's risk of liability under any of the Trust Reps. Keith 2344:8-11, 2352:13-18, 2355:17-2356:11; Schwarcz 2060:17-2062:18; Burnaman 1604:11-14. And as Mr. Butler testified, even this variation is "still a no-default rep" and "if there was a default, RFC was still at risk" because a plaintiff could establish that "RFC knew or should have known." Butler 823:9-824:5. Importantly, in attempting to rebut Mr. Butler, PRMI's underwriting expert did not assess whether RFC had knowledge of any borrower misrepresentations. Keith 2357:3-15.
225. Ultimately, PRMI's interpretation of the No Default Rep was that the Rep's purpose could not be to address borrower fraud or misrepresentation because such fraud and misrepresentation were addressed and allocated explicitly through a variety of different mechanisms. See Burnaman 1574:21-1575:23 (discussing the use of early payment default reps, specific fraud reserves for RMBS deals, and explicit fraud or misrepresentation reps and warranties). Still, even assuming that is true, the No Default Rep was, at the very least, simply another one of those mechanisms, one part of an "intertwined" and "overlapping" system of representations and warranties that accounted for fraud. Butler 642:2-10.
226. In view of the No Default Rep's express language and structure, and ResCap's witnesses' credible testimony, outlined above, the Court finds PRMI's experts' testimony unpersuasive. Moreover, it finds that ResCap's interpretation of the No Default Rep to be more persuasive than PRMI's interpretation based on the evidence adduced at trial.
4. The Loan Program Reps Were Enforceable For Single Loans
227. Beginning with the RALI shelf in 1999, and continuing several years thereafter for 244 at-issue securitizations on the RALI and RFMSI shelves, RFC made representations that typically took the following form: "Approximately [%] of the Mortgage Loans . . . were underwritten under a reduced loan documentation program, approximately [%] . . . were underwritten under a no-stated income program, [etc.] . . . ." DTX-1016.0115 at 37 (the "Loan Program Rep"); Farley 183:12-23; DTX-786 at 6. ResCap asserts that the Loan Program Reps were enforceable on a loan-by-loan basis. PRMI, by contrast, contends the Loan Program Reps warranted only the overall characteristics of the securitized loan pool and could not be enforced as to individual loans. See, e.g., Schwarcz 2007:12-24, 2150:23-2151:6; Burnaman 1564:19-1565:19, 1566:11- 1567:1; Keith 2316:16-2317:11. The Court finds that, based on the evidence adduced at trial, ResCap's interpretation of the Loan Program Rep is more persuasive.
228. Particularly persuasive to the Court was Ms. Farley's credible testimony on the subject. She personally negotiated the Loan Program Rep at issue, see Farley 183:7-184:2, and testified persuasively and credibly that it was intended, understood, and "meant to be read" as "a loan-level rep[resentation]," included in the loan-level representation section of the Trust Agreements, with "a loan-level remedy" that "relates to each individual mortgage loan, which in aggregate create a pool." Id. at 186:9-16, 186:22-187:2. She also explained that "the [Loan Program Rep] percentages . . . were calibrated so that if . . . a single loan fell out of a category, it would impact the[] percentages," indicating the Loan Program Rep statistics were compiled based on the categorization of individual mortgage loans. Id. at 186:9-16, 158:3-20 (testifying that the statistical disclosures in the Prospectus Supplement were calculated based on the categorization of each individual loan), 158:21-159:3 (noting "[t]he [loan program] percentages were brought out to . . . a tenth of a hundredth of [a] percent [] intentionally because even one violation alone that was mischaracterized . . . would affect these percentages"), 185:15-186:8 (testifying that the Loan Program Rep was designed to provide a loan-level remedy to inaccuracies in the Prospectus Supplement disclosures to "maintain the integrity of the pool"). During her 35 years in the industry, she had also never heard anyone express the view that the Loan Program Rep was only a "pool-wide" representation. Id. at 186:17-21.
229. Ms. Farley also supported her understanding of the Loan Program Rep with specific, uncontradicted testimony about the circumstances in which the Rep was drafted. As she explained, the RALI shelf—where the Loan Program Rep debuted—was a shelf she had designed to accommodate loans that had been issued under expanded loan programs. Id. at 131:6-133:4. Because expanded loan programs were the "key risk with respect to the RALI shelf," id. at 184:3-20, the disclosure of loan program type was a "critical disclosure" for participants and investors in transactions on that shelf, id. at 156:23-157:7, 131:6-132:19 (discussing unique nature of RALI shelf); Butler 692:16-22 (noting that because the RALI shelf contained "Alt-A" credit grade loans, which was a "step below A type credit," the "loan program rep would be of importance to investors"). This was accomplished in the ordinary course via disclosures in Prospectus Supplements. Farley 159:4-15, 184:21-25; see Butler 691:22-692:22.
230. However, ratings agencies or investors concluded that those Prospectus Supplement disclosures were insufficient on their own because "they were missing a loan-level remedy" and "wanted [] an immediate timely remedy where they could pu[t] a [loan] back and maintain the integrity of the pool." Farley 184:3-20, 185:15-21 (rating agencies wanted "a remedy in the event that a loan breached [its loan program] category such that the disclosure [in the prospectus supplement] was wrong. They wanted to add teeth to that disclosure."). Accordingly, RFC agreed to supplement the disclosure in the Prospectus Supplement with a loan-level representation and remedy in the Pooling and Servicing Agreement, the purpose and effect of which was to provide a "loan-level [repurchase] remedy" to "maintain the integrity of the pool," in the event a "mortgage loan . . . breached the loan doc[umentation] program rep[resentation]." Id. at 185:22-186:8. The Loan Program Rep was later extended to certain securitizations on the RFMS I shelf, which likewise involved loans originated to expanded loan programs. See Butler 691:22-692:15.
Although the Loan Program Rep did not explicitly identify the number of full documentation loans in each pool, RMBS industry participants understood that any loans not underwritten to reduced loan documentation programs "would be full doc[umentation]." Farley 192:8-12. These numbers are "borne out by the [prospectus supplement] disclosures." Id. at 192:1-12; see also Butler 672:5-16 (noting that where a Loan Program Rep stated, for example, 93% of the loans were underwritten to reduced documentation, the remaining 7% "would have to fall into the full documentation program"); PTX-099 at 9 (RALI 2007-QO2 Prospectus Supplement, disclosing reduced documentation numbers and full documentation numbers, the latter of which was not explicitly included in the corresponding PSA (PTX-097 at 57)). Similarly, although the relevant agreements did not explicitly cross-reference Client Guide definitions for the various disclosures, RMBS investors knew to review the Client Guide for that information. Farley 191:16-25.
231. Mr. Butler's testimony on the issue was also persuasive. He testified that the Loan Program Rep was a "loan-level representation" in the "loan-level rep section" of the Trust Agreements with a "loan-level . . . remedy." Id. at 674:14-676:10; see DTX-1016.0115 at 36-38. "RFC from a business perspective understood [the Loan Program Rep] [was] breached if a single loan fell outside the program parameters in which it had been placed[]." Butler 675:20-25 (relying on Ms. Farley's testimony). Indeed, the representation "would serve no purpose if it were not to trigger a repurchase cure remedy," because the Pooling and Servicing Agreements were not disclosure documents like the Prospectus Supplement. Id. at 676:5-10. "[A]ny one loan could breach the [Loan Program Rep]." Id.
232. Although Mr. Butler offered these opinions based on his own expertise, his opinions were corroborated by Ms. Farley and Ms. Lundsten, who testified that "if RFC slots a loan into a particular loan documentation program based on the underlying loan data it had received, and it turned out that the underlying data ended up being wrong[,] . . . [that] could make th[e] [Loan Program Rep] incorrect." Id. at 676:18-677:3 (discussing Lundsten's testimony in HLC trial). Moreover, Mr. Butler corroborated his testimony through an RFC spreadsheet assessing repurchase exposure for the purpose of settlement valuations reflecting that "the loan program rep would be considered to be an underwriting rep." Id. at 677:15-678:11, 731:9-13.
233. Mr. Hawthorne similarly testified that the Loan Program Rep was subject to a loan-level repurchase protocol. Hawthorne 1022:3-1024:5. As he explained, to determine whether there is a breach of the Loan Program Rep, a party must go to the loan file and determine which loan program applies, and then go to the applicable guidelines to determine whether the loan meets the underwriting criteria in those guidelines. Id. at 1023:15-1024:5. Accordingly, Mr. Hawthorne testified that, as of the "Settlement Period," it would "[have been] reasonable for RFC to not materially distinguish its risk of liability under a credit grade or loan program rep as opposed to an express guideline compliance rep[.]" Id. at 1027:2-1028:6.
The "Settlement Period" refers to the period of time from approximately May through December 2013, when the parties to the Bankruptcy were involved in settlement negotiations. Hawthorne 1007:7-9.
234. Faced with the foregoing evidence, PRMI adduced no fact evidence and relied entirely on unsupported expert opinion. The Court finds those opinions unpersuasive.
235. First, although PRMI's experts argued that the Loan Program Rep (and the Credit Grade Rep) can be enforced only on a pool-wide basis, they could not reconcile their position with the fact that the Reps are each explicitly subject to a loan-level remedy. The only expert that even attempted to do so, Professor Schwarcz, even admitted that "the repurchase protocol purports to apply to the [Loan Program Rep,]" Schwarcz 2137:24-2138:2, 2150:23-2151:6 (applying views about Credit Grade Rep to Loan Program Rep), and that "[t]he repurchase protocol by its terms is necessarily a loan level remedy." Id. at 2134:1-16. He also testified that "[the] repurchase protocol is the sole remedy" in the Trust Agreements he has reviewed, and he is not aware of any Trust Agreements where the repurchase protocol was not the sole remedy. Id. at 2133:15-25. Nonetheless, Professor Schwarcz dismissed the provision of a loan-level remedy for the Loan Program Rep as a "mistake[]" that "makes no sense," and was the product of "poor and inadequate drafting" that "would not be regarded as acceptable drafting in the industry." Id. at 2025:16-2026:4, 2031:19-2032:2, 2141:17-2142:18, 2149:24-2152:9.
236. Professor Schwarcz offered no evidence in support of this position, id. at 2163:15-2164:18, and conceded he has no experience with these unique kinds of representations. Id. at 2023:15-23, 2138:23-2144:5. More fundamentally, he conceded that securitization agreements are formal and extensively vetted documents, id. at 2153:23-25, drafted by "highly-sophisticated counsel," id. at 2155:6-9, and "closely scrutinized" by rating agencies, Monolines, and securities underwriters before closing. Id. at 2161:21-2162:14. He further testified that Securitization Reps are "highly scrutinized," id. at 2162:23-2163:3, and that RFC was a "highly-sophisticated sponsor" that knew how to make Trust Reps. Id. at 2154:1-24, 2158:18-2161:15 (discussing an exemplar set of transaction documents that had been circulated to at least five sets of counsel). Moreover, Professor Schwarcz disclaimed any opinion that RFC could have asserted that the drafting of the Loan Program Rep (or Credit Grade Rep) or Trust Agreements was a "mistake" as a defense in litigation. Id. at 2168:2-2169:3.
237. Importantly, the Court considers Ms. Farley's clear rejection of Professor Schwarcz's hypothesis to be particularly persuasive on this issue. Ms. Farley testified that the relevant language had been "reviewed by pretty much everyone involved with the shelf," including two sets of RFC outside counsel (Thacher Proffitt & Wood and Orrick Herrington & Sutcliff), "at least three rating agencies and their internal counsel," "underwriters' counsel," the "trustee and their counsel," GMAC and its counsel, and, if applicable, "[monoline] insurers and their counsel." Farley 189:2-25, 190:1-5, 190:10-20. Indeed, she explained that "[i]f any program was in danger of being overlawyered, [] RFC's programs [on this issue] were close." Id. at 189:8-10. Many of the lawyers involved were also "known experts in this securitization field." Id. at 189:21-22. In light of her testimony—and the absence of any evidentiary support for Professor Schwarcz's hypothesis—the Court declines to accept Professor Schwarcz's suggestion that the relevant parties and their counsel made multiple fundamental errors in hundreds of Trust Documents involving billions of dollars of RMBS over a period of many years.
Professor Schwarcz offered the same "poor drafting" theory as to every contract that contained a Credit Grade Rep. Schwarcz 2025:16-2026:14. He also testified that there are a host of other "pool-wide representations" that were "mistakenly placed" in the section of the Trust Agreements with a loan-level repurchase protocol. Id. at 2141:17-2142:18.
238. Second, the opinions offered by PRMI's witnesses on this issue would render the Loan Program Rep (and the Credit Grade Rep) effectively meaningless and unenforceable. As Ms. Farley explained, if the Loan Program Rep and Credit Grade Rep were each construed as a pool-wide representation with only a pool-wide remedy, enforcement would be "extraordinarily challenging," "extraordinarily complex," "time-consuming," and "expensive" because an investor "would have to re-underwrite every loan in the pool," then "figure out which loans would switch categories [in the event of a breach]," and "determine the proverbial straw that broke the camel's back" in order to identify "which loan [RFC would] repurchase." Id. at 187:3-189:1.
239. Even PRMI's experts largely agreed with that conclusion. Ms. Keith, for example, testified that it would be "kind of impossible" to identify breaching loans with a pool-wide analysis. Keith 2360:7-10; see also id. at 2358:3-5 ("So it's been my prior testimony that with the pool-wide rep, the only way you could know if that percentage is off is to analyze the pool."). Mr. Burnaman agreed, acknowledging that doing so would require "an analysis of every loan in the pool of loans to determine if the total number of loans within the relevant subcategories exceeds the pool-wide thresholds represented by RFC." Burnaman 1635:19-1636:2 (emphasis added), 1636:3-11 (testifying this would be "mind-numbingly complex," require "extensive discovery," and "be highly inefficient"). Even Professor Schwarcz had no answer when asked how the Loan Program (and Credit Grade Reps) could be remedied if they were construed as pool-wide representations with pool-wide remedies. He hypothesized that a court could "fashion a remedy" other than the loan-level repurchase remedy actually in the Trust Agreements, including breach of contract, event of default, indemnification, or even rescission of the entire transaction, but conceded this testimony was "speculation" and "not [an] attempt[] to give an . . . expert opinion." Schwarcz 2007:25-2008:12, 2026:20-2027:16, 2148:15-2149:4. Indeed, Professor Schwarcz disclaimed any opinion as to remedies for breach of the Trust Reps. Id. at 2146:5-2148:19, 2149:5-22.
Ms. Keith offered similar testimony that "rescind[ing] the entire transaction" was somehow more reasonable than using the loan-level repurchase remedies in the Trust Agreements to remedy breaches of the Loan Program and Credit Grade Reps. Keith 2359:14-2360:10. That testimony was neither credible nor persuasive in light of the other evidence discussed above.
240. Third, if the Loan Program Rep was merely a pool-wide representation with a pool-wide remedy, as PRMI's witnesses contend, it would have been unnecessary and duplicative because "[t]here already was a pool-level rep [in the prospectus supplement]." Farley 186:17-188:10.
241. Fourth, PRMI's experts contend that the Loan Program Rep (and Credit Grade Rep) cannot be interpreted as loan-level representations because of variance disclosures in the Prospectus Supplements. Burnaman 1568:17-1570:12. These disclosures provided that the sponsor may remove or add mortgage loans to the pool, but that the disclosures will be "substantially representative of the characteristics of the mortgage pool[.]" Butler 783:13-784:18. But these variance disclaimers do not change the fact that, as Ms. Farley testified, the Loan Program Rep was put in the Trust Agreements to give investors a loan-level remedy to enforce RFC's statistical disclosures. Farley 185:22-186:8. As Mr. Butler testified, a variance disclosure in the Prospectus Supplement "allows the depositor to make changes as needed . . . to balance the pool the way it was represented to investors." Butler 783:13-785:24. In other words, it allows the sponsor to take affirmative steps to align the quality of the mortgage pool with its representations. Mr. Butler also confirmed that this potential substitution "has nothing to do with the breach of any credit grade or loan doc program [representation]." Id. The Court agrees that a disclaimer in the Prospectus Supplement allowing RFC to knowingly substitute mortgage loans for the purpose of improving the quality of the loan pool does not mean the Loan Program Rep (or Credit Grade Rep) in the Trust Agreements could be understood to allow a sponsor to securitize loans with underwriting breaches, negatively impacting the quality of the loan pool—putting it at variance with its representations—without remedy.
242. Having carefully considered both the express language and structure of the Pooling and Servicing Agreements, and the Loan Program Rep's placement within it, as well as the testimony of the parties' witnesses, the Court finds persuasive ResCap's position that the Loan Program Rep can be breached by individual loans and provided a loan-level remedy.
5. The Credit Grade Reps Were Enforceable For Single Loans
243. In Assignment & Assumption Agreements for 72 at-issue Trusts on the RASC shelf, see DTX-786 at 3-4; Schwarcz 2021:12-20, RFC made a representation in the following form: "No more than [%] of the Mortgage Loans have been classified by RFC as Credit Grade [x], no more than [%] of any Mortgage Loans have been classified by RFC as Credit Grade [y] Mortgage Loans . . . in each case as described generally in the Prospectus Supplement." PTX-1150 at 7 (the "Credit Grade Rep"). As with the Loan Program Rep, ResCap contends the Credit Grade Rep provided a loan-level remedy. The Court again finds that the evidence at trial establishes that ResCap's interpretation is more persuasive.
244. The evidence the parties adduced as to the Credit Grade Rep is for the most part identical to the evidence they adduced as to the Loan Program Rep. Ms. Farley explained that the Credit Grade Rep was particularly germane to the RASC shelf, which she created, and which housed loans issued to borrowers with credit challenges. She offered testimony about its meaning and purpose, quite analogous to her testimony regarding the Loan Program Rep. See Farley 132:20-133:4, 159:16-21, 192:13-194:7, 196:11-16. Mr. Butler likewise mirrored his testimony about the Loan Program Rep. Butler 685:6-15, 685:19-686:18 (citing corroborative 30(b)(6) testimony), 687:13-690:8, 693:13-17, 769:5-13 (citing a corroborative repurchase demand, which RFC honored). The same was true for Mr. Hawthorne. Hawthorne 1013:4-1021:22. Similarly, the arguments PRMI's experts offered regarding the Credit Grade Rep were identical to the ones they offered regarding the Loan Program Rep. The Court accordingly makes the same findings as to the Credit Grade Rep as it did above for the Loan Program Rep, and concludes that ResCap's interpretation is more persuasive.
6. Trust Reps & Disclosures Were Passed Through To The Monolines
245. In its contracts with the Monolines, RFC represented that "[e]ach of the [Trust Reps were] true and correct," PTX-106 at 10, and warranted that statements in the Prospectuses and Prospectus Supplements were accurate and not misleading, PTX-1091 at 10 (together, the "Monoline Reps"). Mr. Butler testified that the Monoline Reps gave the Monolines the benefit of RFC's Trust Reps and warranted the truth of statements in the offering documents, including those warranting compliance with RFC's underwriting standards, procedures, and guidelines. See Butler 695:9-698:12; PTX-1091 at 10; PTX-1281 at 31 ("All of the mortgage loans were originated and underwritten in accordance with the Home Solution Program. . . . [RFC] reviewed the underwriting of each of the mortgage loans and determined that it was in conformity with the standards set forth . . . ."). Mr. Butler's opinion was based on his industry expertise, as well as corroborating evidence, including a repurchase demand from a Monoline, with which RFC complied, asserting breaches of "[Reps] [including the statement in the prospectus that] 'all mortgage loans were in substantial conformity with the standards set forth in [RFC's] AlterNet Program.' " Butler 698:16-701:20.
Offering documents are defined in the insurance agreements as relevant securities disclosures, including the Prospectus and Prospectus Supplement. PTX-106 at 6.
Contrary to Mr. Zitting's testimony, the Home Solution Program was included in the Client Guide as of January 1, 2003. Compare PTX-032 at 334, with Zitting 1677:12-16.
246. The Court finds Mr. Butler's testimony persuasive and well-founded. In so doing, the Court discounts Professor Schwarcz's unsupported opinion that the Monoline Reps were understood as only "10(b)-5" representations, "intended to give the monoline insurer comfort that it is not participating in [] federal securities law fraud." Schwarcz 2053:13-2054:24. Indeed, Professor Schwarcz conceded that the Prospectus Supplements' underwriting guideline representations applied to all loans in each Trust, and that they were subject to contractual remedies—a concession contradicting his testimony that they were mere 10(b)-5 representations. Id. at 2053:13-2054:5, 2171:2-2173:18; PTX-1091 at 28-30 (providing that a breach of the Monoline Representations constitutes an event of default, with remedies including acceleration of RFC's obligations); Id. at 22 (indemnifying Monoline for breaches).
While the language used for the Prospectus Rep mirrors federal securities law standards, that fact does not vitiate the force of the representation to the Monolines. Indeed, the plain language of the Monoline Insurance Contracts states that RFC represented to the Monolines that the "[o]ffering [d]ocument does not contain any untrue statement of a material fact and does not omit to state a material fact necessary to make the statements made therein, in light of the circumstances under which they were made, not misleading." PTX-1091 at 10. Moreover, RFC represented that the facts set forth within the offering documents—which included statements that "[a]ll the mortgage loans were originated and underwritten in accordance with the [applicable underwriting guidelines,]" PTX-1281 at 31—were materially true. See PTX-1091 at 10.
7. The Fraud Disclaimer Did Not Eliminate RFC's Risk of Liability
247. Beginning in the 1990s, certain of RFC's Trust Agreements stated that RFC "shall not be required to cure breaches, Repurchase Events or purchase or substitute for Mortgage Loans as provided above if the substance of such breach or Repurchase Event also constitutes fraud in the origination of the Mortgage Loan." PTX-023 at 9 (the "Fraud Disclaimer"); Farley 200:11-21. ResCap contends that this language was added to reduce RFC's exposure to fraud, but that it was not understood to cover misrepresentations or, in general, to be a "silver bullet" defense to liability. PRMI, to the contrary, contends that the Fraud Disclaimer extended to misrepresentations as well, and that it provided RFC with a complete defense. The Court finds RFC's position more persuasive based on the evidence presented at trial.
248. The Court finds particularly persuasive the testimony of Ms. Farley, who personally negotiated the Fraud Disclaimer. She testified that it was added to the Trust Agreements to create "an argument" against repurchasing loans with fraud, but that it "was not a silver bullet [against claims based on borrower fraud or misrepresentations]," and was not "necessarily going to be effective [at defending against those claims]." Farley 200:11-201:8. She further explained that RFC understood that "no [claimants] would be incented to want to prove [fraud]" when "all they would have to do" to remove a claim from the arguable ambit of the Disclaimer was to "claim [it as] a misrep[resentation]." Id. at 201:9-12. In fact, RFC itself "would not be incented to try to prove fraud on a loan" due to reputational risk. Id. at 201:19-22. And even if RFC attempted to use the Fraud Disclaimer as a shield, proving fraud is a "very high standard" that "required knowledge," and accordingly is "much harder to prove than misrep[resentation]." Id. at 201:3-8. Ms. Farley further testified that if the Fraud Disclaimer had effectively eliminated RFC's risk of liability for fraud, it would have dramatically changed RFC's risk profile, and the way RFC structured and priced its securitizations, in a manner that would have resulted in many meetings and business actions with the most senior personnel in the company. Id. at 202:7-21. Because that was not the purpose, or effect, of the Fraud Disclaimer, she noted that such meetings and business actions never occurred. Id.
249. Mr. Butler likewise testified that RMBS industry participants understood fraud and misrepresentation to mean different things, and therefore would have understood the Disclaimer as covering only the former. Butler 662:14-25. Thus, misrepresentation claims would remain viable even in the presence of a Fraud Disclaimer. Id. at 659:25- 660:6. Further, "RFC ha[d] risk" if there was fraud or a misrepresentation, despite the presence of a Fraud Disclaimer. Id. at 663:19-664:11, 937:15-21. "[If] fraud is discovered, [Monolines and investors are] going to put the loan back, and RFC is going to be at risk for repurchase of that loan." Id. at 665:11-19.
250. Mr. Butler offered those opinions based on his own expertise, including as a fraud examiner, but also based on extensive corroborating evidence. For example, Mr. Butler discussed Ms. Lundsten's testimony that the Fraud Disclaimer "d[id] not completely absolve [RFC] of any risk[,]" id. at 661:12-16, because " '[i]f an issue arises on a loan that appears to be related to fraud, there's a very good chance that one of the other customary reps [RFC has] given has been violated, so [RFC] still wind[s] up repurchasing these loans from deals and going back to clients[.]' " Id. at 663:14-18. He also discussed email evidence to RFC employees confirming that "RFC[] repurchase[d] loans with identified material misrepresentations despite the presence of a fraud disclaimer[,]" id. at 664:14-665:10, and ResCap's Rule 30(b)(6) deposition testimony demonstrating that RFC repurchased loans that had fraud or misrepresentations "despite the inclusion of a fraud disclaimer." Id. at 660:14-661:8. Finally, Mr. Butler explained at least one instance in which an RMBS Trustee demanded that RFC repurchase a loan based on borrower fraud, after which RFC repurchased the loan after confirming the fraud, even though the loan was subject to a Fraud Disclaimer. Id. at 668:10-669:6.
During cross examination, Mr. Butler was confronted with the testimony of RFC's in-house securitization expert, John Ruckdaschel, who appears to have used the word "misrepresentation" once when testifying about the scope of the Fraud Disclaimer during RFC's Bankruptcy. See Butler 798:1-12, 799:12-801:12. Mr. Butler asserted that Mr. Ruckdaschel was mistaken in using the word "misrepresentation" to describe the scope of the Fraud Disclaimer, see id. at 801:13-17, which the Court considers persuasive when considered alongside the (1) the plain language of the fraud disclaimer; (2) the legally significant difference between fraud and misrepresentation; and (3) the other corroborating evidence discussed above.
Mr. Butler was also confronted with correspondence showing that RFC had relied on the Fraud Disclaimer in rejecting repurchase demands based on purported borrower misrepresentations for several loans in the past. See id. at 922:1-925:15. However, while the correspondence illustrated that RFC initially denied repurchase, it did not show "the final disposition" of the dispute, and it was possible (absent any evidence showing otherwise) that RFC might have "ma[d]e a business decision and decide[d] to repurchase" the loan in the end. Id. at 926:9-22.
251. Mr. Hawthorne also opined that a reasonable defendant in RFC's position would not have placed much weight on the Fraud Disclaimer, and that plaintiffs seeking repurchase from RFC "had the much stronger hand." Hawthorne 1045:11-1048:8. Thus, whether a particular RMBS Trust was subject to a Fraud Disclaimer was "not a distinction that would make a lot of difference . . . ." Id. at 1045:17-18. Like Ms. Farley, he explained that fraud is difficult to prove due to the scienter requirement, and that invoking the Fraud Disclaimer would require RFC to contend that its own loans were fraudulent (and thus implicitly admit misrepresentation as well). Id. at 1029:14-20, 1045:11-1047:24.
252. As against the foregoing evidence, PRMI presented no fact testimony and relied only on unsupported and uncorroborated expert opinion. The Court finds that testimony unpersuasive.
253. First, PRMI's experts admitted that the text of the Fraud Disclaimer does not address misrepresentations and provided no support for their contention that it nonetheless disclaimed misrepresentations by implication. To the contrary, they either acknowledged that "legally [fraud and misrepresentation] are different," or disclaimed offering any opinion as to the difference between the two. Burnaman 1582:18-1583:5, 1620:15-24; Schwarcz 2122:16-2126:4; Keith 2332:8-13.
254. Second, none of PRMI's experts opined that the Fraud Disclaimer provided meaningful protection to RFC given the legal and practical difficulties of asserting it. Professor Schwarcz testified that it was "beyond the scope" of his opinion. Schwarcz 2119:5-2121:24. Mr. Burnaman testified that he was not offering an opinion as to whether RFC could have used the Fraud Disclaimer to defend against repurchase demands based on misrepresentations. Burnaman 1621:13-16. And Ms. Keith testified that her opinions regarding the Fraud Disclaimer "don't relate [] to what [are] viable legal defenses in a court." Keith 2363:8-2364:10.
255. Likewise, PRMI's experts had no opinion on the difficulty of proving fraud or its legal elements, see Burnaman 1620:25-1621:7; Schwarcz 2125:18-24, 2126:5-14, and no opinion as to whether someone would need to prove fraud for the Fraud Disclaimer to be effective, see Burnaman 1621:8-12; Schwarcz 2126:15-2127:23. Professor Schwarcz's testimony on these issues was particularly lacking in credibility, given he is a long-tenured member of the faculty at Duke Law School and doubtless would have been competent to address the issues directly.
256. Although Mr. Woll testified that a defendant in RFC's position, with the benefit of a Fraud Disclaimer, would have had "a strong argument that it cannot be held responsible for any repurchase obligation connected with an allegation of borrower fraud[,]" Woll 1874:15-20, the Court finds that opinion unpersuasive. Mr. Woll acknowledged that it presented "uncharted territory" and, therefore, legal risk, because there was no "decisional law" addressing a Fraud Disclaimer. Id. at 1878:13-17. And despite testifying that RFC could have asserted arguments in the alternative—i.e., that a borrower had committed fraud, and alternatively that there was no misrepresentation—Mr. Woll admitted that such a strategy would not have been available at trial. Id. at 1951:14-22, 1954:6-12 (noting that both "uncharted" and "charted" legal territory presents risk).
257. By contrast to Ms. Farley, who negotiated the Fraud Disclaimer and testified from extensive first-hand experience about its meaning, PRMI's experts offered little basis—corroborative or otherwise—for their opinions. Mr. Burnaman claims to have seen it in, at most, "one transaction from the '90s [from] Indymac." Burnaman 1580:6-14. Likewise, despite insisting that the Fraud Disclaimer was "very typical," Professor Schwarcz could not identify a single sponsor that used one, a single transaction that included one, or a single case addressing one. Schwarcz 2112:6-2115:16, 2121:25-2122:5, 2035:4-6.
258. For these reasons, the Court finds, first, that the Fraud Disclaimer did not offer RFC a defense to claims of misrepresentation in the origination of the securitized loans. And second, although the Disclaimer may have provided a technical defense to claims of borrower fraud, it was not certain. Third, it is unlikely that plaintiffs would have pursued the claim given the availability of a more easily proven claim of misrepresentation. Finally, it is unclear that RFC would have pursued the defense given the strategic considerations in doing so. In any event, it was not a "silver bullet" that eliminated all risk to RFC.
8. The No-Fraud Rep
259. Beginning in 2006, RFC made representations to certain Trusts that "[n]o fraud or misrepresentation ha[d] taken place in connection with the origination of any Mortgage Loan." PTX-1113 at 8 (a "Fraud Rep"). The Fraud Rep appears in 61 of the at-issue Trusts. See DTX-786 at 7-8 (noting 61 total trusts at issue with a No-Fraud Rep); Schwarcz 2033:9-20. PRMI contends that the Fraud Rep was the only way in which RFC warranted against fraud and misrepresentation in the securitized loans. It further contends that the existence of the Fraud Rep in only some securitizations creates an inference that RFC did not warrant against fraud and misrepresentation in other securitizations. ResCap disagrees. The Court finds that the evidence adduced at trial supports ResCap's position.
260. Ms. Farley persuasively testified that if there was fraud in a loan, it "could have been a breach of any number of representations . . . ." Farley at 199:14-202:6; see also id. at 175:3-23 (discussing the overlapping nature of various reps and warranties).
261. Mr. Butler and Mr. Hawthorne each testified to the same effect. See Butler 642:2-10 ("[E]ven if there's not a [F]raud [Rep] in a particular deal, but there's a misrepresentation, you're going to have a guideline [Rep] that's breached or a loan program [Rep] that's breached or an MLS [Rep] that's breached [or a] credit grade [Rep] that's breached. They are all, again, Ms. Farley said it, overlapping . . . [and] intertwined."), 650:13-23 (a borrower misrepresentation causing "the facts and figures about the loan [on the MLS to be] wrong" would breach the MLS Rep even where the trust did not have a "no-fraud" representation), 640:20-23 (noting "th[e] [No Default] representation [can] be breached [by borrower misrepresentations even] if the subject trust did not have a no-fraud representation"); Hawthorne 1030:18-1031:4 (noting that the No Default and MLS Reps "gave RFC a risk of liability on underlying misrepresentations").
262. PRMI's position at trial on this issue consisted exclusively of expert testimony unsupported by any evidence. See Burnaman 1573:7-22; Schwarcz 2037:20-2038:12; Keith 2326:13-2327:17. It was also premised on the unsubstantiated notion that the absence of an explicit Fraud Rep can be read as evidence that RFC did not intend to warrant against fraud. The Court perceives no basis on which to make that finding, particularly because it is undisputed that contracts of the type at issue here commonly included duplicative "belt-and-suspenders" contract terms, in which a given issue was addressed multiple times in different ways. See Hawthorne 1043:5-11.
9. The Substantial Compliance Rep
263. Beginning in 2003, certain Monoline-insured deals on the RFMSII shelf included a representation that "[a]ll of the Mortgage Loans have been underwritten in substantial compliance with the criteria set forth in the Program Guide" ("Substantial Compliance Rep"). The "Program Guide" is defined as the Client Guide. PTX-107 at 6; DTX-786 at 2 (noting 17 at-issue Trusts contain this Rep); Butler 628:20-23 (noting that "Program Guide" and "Client Guide" were synonymous), 692:23-693:12 (describing why a "Substantial Compliance" Rep was requested by Monolines). PRMI contends that the Substantial Compliance Rep was the only way in which RFC warranted against guideline breaches in the securitized loans, and that the existence of the Substantial Compliance Reps in only some securitizations creates an inference that RFC did not warrant against guideline breaches in other securitizations. See, e.g., Burnaman 1561:12-1562:9; Schwarcz 2190:17- 2191:1; Keith 2311:19-2313:3. ResCap disagrees. The Court finds that the evidence adduced at trial supported ResCap's position.
264. First, Ms. Farley testified to her contemporaneous understanding that there were many representations by which RFC could and did warrant compliance with its guidelines, and a Substantial Compliance Rep was not necessary to do so. As Ms. Farley explained, whether or not there was a Substantial Compliance Rep, "there were individual [Trust Reps] that overlapped and would have covered any material breach of the Client Guide and the underwriting standards therein." Farley 196:17-197:10. Indeed, Ms. Farley explained that "the vast majority, if not all of [RFC's Trust Reps] could be breached by a violation of . . . the Client Guide." Id. at 196:17-197:10, 197:17-198:18 (identifying hazard insurance, title insurance, and the MLS Rep, among others). Ms. Farley further testified that she had never heard the view expressed that the absence of a Substantial Compliance Rep immunized RFC from liability for guideline violations. Id. at 197:11-16 (noting that "[u]nderwriting violations would have been the heart of most . . . breaches").
265. Second, Mr. Butler testified to the same effect. See Butler 694:13-19 (noting that the Substantial Compliance Rep, Loan Program Rep, and Credit Grade Rep all overlapped because they "all related to underwriting in one way or another . . . [and were] all dependent on the underwriting of the loan and the criteria for underwriting the loans"), 726:3-20 (noting that Loan Program and Credit Grade Reps "were all guideline representations" because loan program and credit grade parameters were defined in the Client Guide).
266. Third, Mr. Hawthorne testified that the Substantial Compliance, Loan Program, and Credit Grade Reps were all representations about underwriting criteria, and that it would be reasonable for a party in RFC's position not to materially distinguish its risk between those Reps. Hawthorne 1011:16-1012:1, 1027:2-1028:6; see also DTX-737 at 112-14 (discussing representations contained in the parties' Governing Agreements, including Substantial Compliance, Loan Program, and Credit Grade Reps, that "either directly or indirectly represented that the loans had been originated based on, or conformed to, the relevant reunderwriting guidelines," and opining that "Plaintiffs, accordingly, would have been able to make strong arguments that these representations warranted that such loans conformed to the applicable Client Guide underwriting guidelines for their loan programs.").
IV. RESCAP'S DAMAGES CALCULATION IS REASONABLE
A. Dr. Snow Allocated to Breaching Losses
267. ResCap called Dr. Karl Snow to testify at trial regarding allocation and damages. Snow 1319 et seq.
268. Dr. Snow is an economic consultant who holds a Ph.D. in economics from the University of Chicago, where he focused on financial economics and econometrics. Id. at 1320:19-1321:21. He has held academic appointments in finance and economics at the University of North Carolina - Chapel Hill, Brigham Young University, and the Stockholm School of Economics, as well as adjunct professorships at the University of Maryland, Johns Hopkins, and American University. Id. at 1322:6-1323:5, 1324:9-16. He has also worked in the fields of financial economics and econometrics as an employee or consultant at UBS Investment Bank, Welch Consulting, Freddie Mac, and Bates White. Id. at 1323:8-1324:8, 1324:17-1326:22. In addition, he has published work in those fields in academic and practitioner journals and has given academic presentations in the United States and Europe. Id. at 1328:15-21. As a consultant, Dr. Snow has been involved in over 70 matters relating to mortgages or mortgage-backed securities, for which he has consulted or testified about issues including damages and statistics. Id. at 1327:22-1328:14. The Court accepted Dr. Snow as an expert in statistics, econometrics, and damages. Id. at 1328:22-1329:2.
269. Dr. Snow testified about a methodology he developed for the purpose of apportioning to PRMI an allocable share of the indemnifiable liability RFC incurred in its settlements with the RMBS Trusts and Monolines. Id. at 1319 et seq. As a general matter, Dr. Snow's methodology seeks to allocate RFC's relevant liabilities pro rata to dollars of breaching loan losses. Id. at 1331:11-14, 1338:5-17, 1339:1-7. PRMI did not present a competing damages model. McCrary 2596:11-16.
270. The Court summarizes Dr. Snow's methodology, for the RMBS Trust settlement and the Monoline settlements, below.
1. The RMBS Trust Settlement Allocation
271. To determine an RMBS Trust settlement allocation, Dr. Snow first considered the amount of settlement liability. Snow 1338:5-25. As noted, RFC settled the RMBS Trusts' claims in exchange for a gross allowed claim of $7.09 billion. Id. at 1339:8-17, 1343:7-1344:1. Dr. Snow made two deductions from that gross allowed claim for claims for which ResCap does not seek indemnification: (a) $269 million for claims
associated with NDS trusts; and (b) $73 million for servicing claims. Id. at 1343:7- 1345:14, 1396:1-24. At a high level, Dr. Snow based his NDS Trust deduction on a comparison of the losses in the NDS Trusts relative to the aggregate losses in all 539 trusts (including the NDS Trusts); the approach is discussed further below. Id. at 1344:6-1345:10. Dr. Snow's servicing claim deduction was based on bankruptcy schedules 1R and 4R. Id. at 1344:2-5. The net Allowed Claim, after these deductions, was $6.75 billion. Id. at 1343:7-1345:17.
272. Dr. Snow's methodology proceeded in three steps from there:
273. First, Dr. Snow identified the total breaching loan losses that were at issue in the RMBS Trust Settlement. Id. at 1345:15-17. To do so, Dr. Snow added up the dollar amount of losses on loans that were included in the RMBS Trust settlement and that, as of May 2013, had incurred at least $500 in actual losses, or had incurred expected losses, minus insurance payments. Id. at 1345:18-1347:7. Dr. Snow used May 2013 as the relevant date based on an instruction from counsel regarding the date on which the RMBS Trust settlement was agreed to in principle. Id. at 1346:24-1347:7. Dr. Snow chose $500 in actual losses as his cut-off based on an understanding that it would not have been economically feasible for the Trusts to have pursued claims for less than that amount. Id. at 1448:24-1449:1. Dr. Snow defined a loan with "expected losses" as one that was at least 90 days delinquent, in foreclosure, or had a status of "real estate owned"; and he calculated the expected losses on such loans by using a forecasting methodology that he developed, the reliability of which was not disputed at trial. Id. at 1346:11-14, 1347:15-1348:14. Under those assumptions, Dr. Snow determined that the total dollar amount of losses in the defined population (the "At-Issue Population") was $40.3 billion. Id. at 1348:15-1349:3.
274. Because Dr. Snow sought to allocate breaching losses, he next sought to determine the percentage of the $40.3 billion in loan losses that were associated with loans that had at least one material breach of a Trust Rep. Id. at 1349:4-6. To do so, he drew a random sample of 410 loans from the At-Issue Population, which Mr. Butler reunderwrote using the same methodology he used for the PRMI loan sample. Id. at 1349:7-1350:1; Butler 704:2-23. Mr. Butler found that 265 of the 410 Global Loans breached the Correspondents' Reps to RFC. Butler 704:9-17. For those 265 loans, Mr. Butler determined that 204 breached or could be construed to breach RFC's Reps to the Trusts. Id. at 705:10-14. Ms. Keith did not analyze any of the loans in the 410 Global Loan sample. Keith 2213:12-2214:4, 2214:18-23. Indeed, PRMI adduced no evidence at all concerning the Global Loans. Based on this information, Dr. Snow determined that approximately 59.7% of the relevant loan losses were associated with loans that breached a Trust Rep. Snow 1354:21-1356:2.
275. Second, Dr. Snow calculated a "settlement factor" for the RMBS Trust Settlements. This settlement factor, which consisted of the net allowed RMBS Trust Claim ($6.75 billion) divided by the total RMBS Trust Breaching Losses ($24 billion), equaled approximately 28%. Id. at 1356:1-1357:1. That is, RFC settled the RMBS Trusts' claims for 28 cents on the dollar for every dollar of breaching loss. Id. at 1357:9-12.
276. Third, Dr. Snow applied the foregoing settlement factor to the breaching losses on PRMI At-Issue Loans included in the RMBS Trust settlement. Id. at 1357:4-15. Dr. Snow calculated that there were $18.2 million in such breaching losses ($35.7 million in total trust losses in the 539 PRMI at-issue loans, multiplied by a 50.9% breach rate extrapolated from Mr. Butler's reunderwriting of a 150-loan random sample drawn from that population, excluding two loans in fully-insured trusts, all using the same methodologies described above). Id. at 1357:19-1359:13. Applying the 28% RMBS Trust settlement factor, Dr. Snow testified that PRMI's allocated share of the RMBS Trust Settlement was $5.1 million. Id. at 1359:14-19.
2. The Monoline Settlement Allocation
277. Dr. Snow allocated the relevant Monoline Settlements using the same general approach he used in allocating the RMBS Trust Settlement, albeit with some adjustments to account for the specific nature of the monoline claims. Snow 1361:1-1362:6.
278. As with the RMBS Trust allocation, Dr. Snow first identified the settlement liability amount for which he was calculating the allocation. This consisted of allowed claims to MBIA, FGIC, and Syncora on account of actual and expected insurance payments they had made to RMBS Trusts. Dr. Snow drew those relevant figures, which are not in dispute, from the relevant bankruptcy records. Id. at 1366:9-1367:10; PTX-180 at 150.
279. Next, Dr. Snow calculated the allowed claim for each pool by multiplying (a) the total allowed claim awarded to the monoline insurer of that pool, by (b) the percentage of insurance payments the monoline insurer made to that pool relative to that insurer's total payments to all pools. Id. at 1364:1-1367:16.
280. Dr. Snow then identified the total breaching loan losses in each monoline-insured pool that included a PRMI at-issue loan. Id. at 1367:20-24. To do so, Dr. Snow began by finding the total dollar amount of losses (actual and expected losses) on all at-issue loans in each pool, using the same methodology and definitions discussed above. Id. at 1367:25-1368:10. He multiplied that loan loss amount by a 66% global monoline breach rate, which he had extrapolated from Mr. Butler's reunderwriting of a 105-loan sample of monoline-insured at-issue loans (a subset of the global sample). Id. at 1368:1-1369:12.
Mr. Butler determined that 64 of the loans in that sample breached or could be construed to breach PRMI's representations to RFC, and that all 64 also breached RFC's representations to the Monolines. Butler 705:19-25.
281. Having identified the total breaching losses in each monoline-insured pool that included a PRMI at-issue loan, Dr. Snow then calculated a "settlement factor" for each such pool. Id. at 1361:4-14, 1371:20-1372:21. For each pool, this settlement factor consisted of the allowed claim to that relevant pool divided by the total breaching losses in that pool. Id.
282. Dr. Snow then determined PRMI's breaching losses for each pool by calculating PRMI's total losses in the pool and multiplying it by PRMI's monoline breach rate. Id. at 1373:2-11, 1373:24-1374:5. Dr. Snow determined the PRMI breach rate for this calculation based on Mr. Butler's reunderwriting of a 39-loan sample of monoline-insured PRMI at-issue loans (a subset of the 150-loan PRMI sample); that breach rate was approximately 67.3%. Id. at 1373:12-23.
283. Finally, Dr. Snow applied the settlement factor for each relevant monoline pool to the PRMI breaching losses in that pool and summed up the results across all relevant pools. Id. at 1372:16-1373:11, 1373:24-1375:12. Based on the foregoing analysis, Dr. Snow testified that PRMI's allocated share of the Monoline Settlements was approximately $401,000. Id. at 1375:13-1376:2.
3. The Total Allocation
284. Although PRMI's allocated share of the RMBS Trust settlement (approximately $5.1 million) plus its share of the Monoline settlements (approximately $401,000) totals approximately $5.5 million, Dr. Snow determined that this total may have been overstated by approximately $100,000 due to his use of statistical techniques referred to as ratio estimators. Id. at 1390:2-10, 1401:19-1403:21. Accordingly, Dr. Snow offered an adjusted PRMI allocation of approximately $5.4 million.
285. Because Dr. Snow's calculations were arrived at using statistical sampling, they are subject to a statistical confidence interval and a margin of error. Id. at 1408:11-1409:15. Dr. Snow testified that his $5.4 million calculation, referred to in statistics as a "point estimate," was subject to a 95% confidence interval and an approximate margin of error of +/- $1.35 million. Id. at 1409:16-1410:23, 1413:17-1415:8. Dr. Snow testified that the resulting range of damages was between $4.1 million and $6.8 million, with his $5.4 million estimate being the "most likely and conservative estimate' of PRMI's fair share of the RMBS Trust and Monoline settlements. Id. at 1415:15-16, 1417:17-22.
B. Allocation to Breaching Losses is Reasonable and Reliable
286. The Court finds that Dr. Snow's Allocated Breaching Loss Approach is reasonable and reliable because: (1) it is consistent with the approach used by the parties to actually allocate in the Bankruptcy, which was incorporated into the Plan and was approved by the Bankruptcy Court; and (2) it is consistent with the approach used in every other major RMBS settlement and approved by courts for each of those settlements.
1. RFC's Settlement Was Allocated to Breaching Losses
287. As described earlier, Duff prepared a methodology to allocate the RMBS Trust Settlement among the RMBS Trusts. That methodology was based on breaching losses—i.e., allocating pro-rata based on differences among the RMBS Trusts in losses and in the incidence of breaches of Reps. See Hawthorne 1096:11-15; Pfeiffer Dep. 52:7-53:16.
288. That allocation methodology was the subject of "lengthy discussions" with the Debtors, the Committee, and other parties in interest, and was incorporated by the Debtors and the Committee in their jointly-proposed Plan. PTX-153 at 5; PTX-179 ¶ 115; see also Hawthorne 1305:12-1307:21. Moreover, contrary to certain arguments made by PRMI's counsel, Duff specifically considered arguments made in connection with the Original RMBS Settlement that the allocation must account for differences among the various RMBS Trusts. DTX-539 at 21-25; PTX-179 ¶ 115; PTX-153 at 5; see also Hawthorne 1098:21-1099:17. Specifically, the RMBS Trustees concluded that the breaching loss allocation "sufficiently accounts for the differences among the characteristics of each Settling Trust," and that "the legal defenses to the Repurchase Claims asserted by certain of the objections do not alter the Duff analysis." PTX-153 at 6.
289. Certain investors objected to the Original RMBS Trust Settlement and the allocation proposed along with that settlement. PTX-179 ¶ 115; Hawthorne 1099:19-11:00-3. Investors in the RMBS Trusts received notice of the RMBS Trust Settlement, including the allocation, that was set forth in RFC's Plan. Hawthorne 1101:20-23. None of the investors or any other party objected to the breaching loss allocation in the Plan. See Hawthorne 1101:24-1102:5, 1307:22-25.
290. The Bankruptcy Court confirmed RFC's Plan, and in so doing expressly approved the RMBS Trust Settlement, including the allocation of that settlement. See PTX-180 at 34-35. The Bankruptcy Court found that the RMBS Trust Settlement, including the allocation, was "fair and reasonable," and in the best interests of "the Investors in each RMBS Trust, each such RMBS Trust, the RMBS trustees, and all other parties in interest." Id. at 24-25, 34.
291. Although the Court independently finds Dr. Snow's allocation to be reasonable, the allocation agreed to and approved in the Bankruptcy supports the reasonableness of Dr. Snow's similar methodology. Not only was the Bankruptcy allocation supported by both RFC and the Committee, it was also supported by the RMBS Trustees, who had fiduciary duties to the investors in the RMBS Trusts. Hawthorne 1100:14-21. Although PRMI's expert, Mr. Woll, speculated about the incentives of the RMBS Trustees, he admitted that: (1) the trustees had duties to all of the investors, Woll 1925:24-1927:1; (2) he was not offering the opinion that the trustees breached those duties, id. at 1926:4-1927:1; and (3) he had no basis to believe that the trust investors did not care what they received. Id. at 1928:23-1929:3. Nor is there any evidence that the RMBS Trustees breached their duties. Indeed, it is noteworthy that no investors objected to the allocation methodology.
2. Similar Settlements Have Allocated to Breaching Losses
292. Mr. Hawthorne testified that as part of his work he assessed other major RMBS trust settlements, including: the Countrywide settlement that predated the RFC Bankruptcy, the JPMorgan settlement that was reached in November 2013, the Citi settlement that was reached in 2014, and the WaMu settlement that was reached in 2016. Mr. Hawthorne testified that every one of these major RMBS trust settlements allocated among the trusts based on net losses. See Hawthorne 1102:6-1104:9.
293. Other than the allocation in RFC's Bankruptcy discussed above, no major RMBS trust settlement took breaches into account. Id. And no allocation has attempted to allocate based on the strengths or weaknesses of legal issues or defenses. Id. That is true notwithstanding the fact that there were some trusts in the Citi, JPMorgan, and WaMu settlements that were subject to a potential statute of limitations defense and potential differences among the trusts in the applicable Reps. Id. at 1104:4-23. Indeed, at the time of the 2016 WaMu settlement, the New York Court of Appeals had already issued ACE Securities Corporation Home Equity Loan Trust, Series 2006-SL2 v. DB Structured Products, Inc., 25 N.Y. 3d 581 (2015) ("ACE III") in RMBS defendants' favor on the statute of limitations issue, yet the court still approved the proposed WaMu settlement. Id.
294. Most recently, on January 3, 2020, Judge Richard Kyle Jr. in Ramsey County District Court approved an allocation of an RMBS settlement based on net losses. See In re MASTR Adjustable Rate Mortgs. Tr. 2006-OA2, No. 62-TR-CV-18-35, Order Granting Petition to Approve Settlement (Minn. Dist. Ct. Jan. 3, 2020) (Kyle, J).
295. As in RFC's Bankruptcy, the trustees in each of these cases had duties to the investors in the trusts. Hawthorne 1105:2-15. Thus, as Mr. Hawthorne testified, those trustees "can't benefit one trust at the expense of another." Id. And, in each case, a court approved the settlement, including the allocation. Id. at 1105:23-1106:3.
296. The Court finds that the allocations agreed to and approved in these other mass RMBS trust settlements support the reasonableness of Dr. Snow's methodology. No evidence was presented that any RMBS trust settlement allocated based on purported differences in the strength of trust representations, statute of limitations, or any other issue or defense raised by PRMI.
C. Allocating to Additional Factors is Unnecessary and Unreliable
297. PRMI contends that Dr. Snow's methodology improperly failed to account for certain purported differences among the RMBS Trusts in the strength of legal issues and defenses. McCrary 2509:25-2511:1. Specifically, PRMI's expert, Dr. McCrary, opined that Dr. Snow's methodology failed to account for certain differences among the RMBS Trusts in: (1) the strength of the applicable Trust Reps; and (2) the strength of a potential statute of limitations defense. Id. at 2517:7-12, 2518:16-21. As to the statute of limitations defense, he testified that PRMI's At-Issue Loans were concentrated in Trusts subject to this defense. Id. at 2517:22-2518:15.
298. The Court does not find that Dr. Snow's allocation was unreasonable for failing to account for certain additional factors that PRMI has identified.
299. In addition to the facts found above—i.e., that the Bankruptcy allocation did not take such additional factors into account, and that no other RMBS trust settlement has taken such additional factors into account—the Court finds that adding these additional factors to Dr. Snow's methodology, as advocated by PRMI, would be infeasible, potentially unfair, and would inject speculation and false precision into the determination of damages.
300. First, as described further below, there were numerous issues that could potentially vary on a trust-by-trust or loan-by-loan basis and that were uncertain as of the Settlement Period. The Court finds that Dr. Snow's methodology need not have accounted for more than a dozen of those different issues in order to be a reasonably certain and non-speculative allocation of damages.
301. Second, PRMI's assertion that Dr. Snow's allocation was unreasonable because it did not account for a specific subset of issues—cherry-picked by PRMI—also is incorrect. There was uncertainty as to numerous issues, not just the issues that PRMI has identified. And yet, Mr. Woll assessed only PRMI's self-selected issues—those issues that might favor PRMI's position. See e.g. Woll 1888:20-1891:20. Specifically, Mr. Woll focused almost exclusively on selected Trust Reps and the statute of limitations defense. Id. at 1801:11-1802:13. Because the outcome of litigation on these issues was substantially uncertain, and other issues potentially impacted RMBS claims, Dr. Snow reasonably did not account for these uncertain, isolated issues in making his allocation.
302. Third, the uncontroverted evidence demonstrated that attempting to quantify the probabilities of more than a dozen uncertain issues, or to draw potential correlations between them, would be utterly infeasible and so highly speculative that it would result in an unreliable expert opinion. PRMI's experts have not testified to the contrary. Nor did any of them testify as to what probabilities they would assign to the issues or if it was even feasible to do so. PRMI's expert, Dr. McCrary admitted that under such circumstances, he would allocate in the same manner that Dr. Snow did. McCrary 2711:8-24 ("I would have probably approached the problem the way Dr. Snow did, that is to say, if there is genuinely no ability to have any assessment of the relative strength of claims and defenses, then I think it's right that you can't actually take that into account.") As the evidence at trial illustrated, assigning probabilities to PRMI's narrow subset of issues and defenses that favored them would be unreliable and result in utter speculation.
1. Challenges to Certain Assumptions in Dr. Snow's Model
a. Legal Interpretations of Trust Reps
303. PRMI contends that Dr. Snow's allocation is not reasonable because it fails to account for purported differences among the Trusts with respect to certain specified Trust Reps. PRMI contends that an RMBS Trust with a Substantial Compliance Rep had more "settlement value" than an RMBS Trust with a Credit Grade or Loan Program Rep. PRMI also contends that an RMBS Trust with a No Fraud Rep had more "settlement value" than an RMBS Trust that only had an MLS Rep or one of three different types of No Default Reps. And, PRMI contends that an RMBS Trust with a Fraud Disclaimer had lower "settlement value" than a Trust without one.
304. As described earlier, however, RFC had a risk of liability on all of these Trust Reps. The legal meaning of many of the Trust Reps was contested as of the Settlement Period, and for others that were unique to RFC, there was no legal precedent one way or another. Thus, at best, there was uncertainty that could not be quantified, short of speculation.
b. Causation and Materiality
305. Mr. Hawthorne testified that causation and materiality were two defenses that RMBS defendants were raising as of the Settlement Period. Hawthorne 1010:13-18. PRMI, however, did not raise these issues as ones that should have been factored into an allocation.
306. As of the Settlement Period, there was a dispute between RMBS plaintiffs and defendants regarding causation in repurchase cases. Id. at 1053:22-25. Defendants argued that plaintiffs had to establish that the defect in the loan caused the loan to default, while plaintiffs argued that they only needed to establish that the defect "had a material and adverse effect on the noteholders' interest in that loan, which, in other words, made it materially more likely to default." Id. at 1054:1-10.
307. Also, as of the Settlement Period, although courts had issued decisions in plaintiffs' favor on causation, and Mr. Hawthorne testified that defenses based on causation were unlikely to be successful, RMBS defendants were nonetheless still asserting the causation defense, and the issue had not been decided by New York's highest court. Id. at 1054:20-1056:3, 1296:1-4. Moreover, if defendants were to succeed, it would have had "a very substantial impact" on plaintiffs' claims and been a "severe challenge" to those claims. Id. at 1056:4-1057:1. That impact would have varied among loans and trusts, and would depend on, among other things, the nature of the particular defect. Id. at 1056:23-1057:9.
308. Even if plaintiffs were not required to prove that a particular loan defect caused the loan to default, they nonetheless were required to prove that the defect materially increased the risk of default. Id. at 1057:17-1058:6. The issue of materiality would vary from loan to loan. Id. at 1058:7-1059:3. In other words, reunderwriters and ultimately factfinders would have different views about which breaches were material and which were not. Id. This was illustrated in the MARM case discussed by Mr. Hawthorne, Mastr Adjustable Rate Mortgages Trust 2006-OA2 v. UBS Real Estate Securities, Inc., No. 12-cv-7322 (PKC) (S.D.N.Y. 2012), in which Judge Castel (SDNY) reviewed a number of "bellwether loans" to establish guidance as to what he viewed as being material or immaterial breaches. Hawthorne at 1058:7-1059:3.
c. "Election of Remedies" Defense
309. Mr. Hawthorne also testified about the "election of remedies" defense—another defense that parties were litigating in repurchase cases. Id. at 1010:13-18.
310. As of the Settlement Period, RMBS defendants were arguing that an RMBS plaintiff could not assert repurchase claims on loans that had been charged-off, liquidated, or foreclosed. Id. at 1077:25-1078:22. One principle underlying this defense was that, because the RMBS trust decided to foreclose on a property and liquidate the loan, it had essentially elected that remedy and was thus precluded from asserting any other remedy under the repurchase protocol. Id.
311. Although Mr. Hawthorne did not view the election of remedies defense as particularly strong, "there were some judges who accepted it," including one court prior to the Settlement Period. Id. at 1078:23-1079:6. Moreover, RMBS defendants were making the argument during the Settlement Period, and the validity of the defense had not been conclusively decided by a higher court in New York. Id. at 1079:7-15. As discussed below, the Committee and RFC both discussed this defense as one of the primary issues in connection with the RMBS Trusts' claims.
312. Mr. Hawthorne also testified that the level of risk on an election of remedies defense would depend, in part, on the contractual language in the relevant securitization documents, and also could depend on the underlying state law governing the mortgage loan. Id. at 1079:16-1080:6. Thus, the application of the election of remedies defense could vary on a trust-by-trust or loan-by-loan basis. Id. at 1080:13-19.
d. Statute of Limitations
313. Mr. Hawthorne testified that the statute of limitations was another defense that RMBS defendants were asserting as of the Settlement Period. Id. at 1010:11-18. Tammy Hamzehpour, RFC's General Counsel, also testified that if the parties in Bankruptcy had litigated instead of settling, RFC would have raised this defense, and Lewis Kruger, the Debtors' CRO, was aware of the defense when he participated in Settlement negotiations. Hamzehpour Dep. 17:24-18:4, 97:5-15, 97:17-25, 98:2; Kruger Dep. 171:9-14.
314. Mr. Hawthorne testified that the "key question in statute of limitations is when does a breach occur for a repurchase claim, when does the claim accrue." Hawthorne 1059:15-21. As of the Settlement Period, RMBS defendants had been arguing that "the breach occurs and the claim accrues at the time that the rep and warranty is given, which is around the origination of the securitization." Id. at 1059:22-25. RMBS plaintiffs, on the other hand, had been asserting that the breach does not occur until the investor seeks cure or repurchase under the repurchase protocol, and the sponsor refuses the demand. Id. at 1060:1-9. According to the RMBS plaintiffs at the time, it was only at this point of refusal that the claim accrued and the statute of limitations began to run. Id.
315. As Mr. Hawthorne testified, as of the beginning of the Settlement Period in May 2013, "[t]here was not a lot of law and none of it was really authoritative in the sense of being governing authority in the State of New York." Id. at 1060:16-22. Mr. Hawthorne discussed five federal cases that were commonly cited—three of which gave some support to plaintiffs, and two of which gave some support to defendants. Id. at 1060:22-1061:1. The three cases that provided some support to plaintiffs were Lehman Bros. Holdings, Inc. v. National Bank of Arkansas, 875 F. Supp. 2d 911 (E.D. Ark. 2012), LaSalle Bank National Association v. Lehman Brothers Holdings, Inc., 237 F. Supp. 2d 618 (D. Md. 2002), and Federal Deposit Insurance Corporation v. Key Financial Services, No. Civ. 89-2366-DPW, 1999 WL 34866812 (D. Mass. Dec. 23, 1999, aff'd sub nom., Resolution Trust Corporation v. Key Financial Services, Inc., 280 F.3d 12 (1st Cir. 2002), while the two cases on which defendants relied were Lehman Bros. Holdings, Inc. v. Evergreen Moneysource Mortgage Company, 793 F. Supp. 2d 1189 (W.D. Wash. 2011), and Structured Mortgage Trust 1997-2 v. Daiwa Finance Corp., No. 02 Civ. 3232 (SHS), 2003 WL 548868 (S.D.N.Y. Feb. 25, 2003). Hawthorne at 1060:22-1061:15. Mr. Hawthorne testified that none of these cases definitely resolved this issue. Id. at 1060:22-1061:15, 1067:4-12.
316. PRMI's expert, Mr. Woll, distinguished or criticized the three cases supporting RMBS plaintiffs' statute of limitations arguments, while praising the two decisions favoring defendants' arguments. Woll:1840:19-1841:22. Mr. Woll noted that damages were at issue in Key Financial Services and LaSalle Bank, not the statute of limitations. Woll:1840:19-1841:22. However, Mr. Woll conceded that these two cases addressed the question of whether a breach occurs when a demand is made or when it is refused. Id. Although Mr. Woll acknowledged that the National Bank of Arkansas decision addressed the relevant statute of limitations accrual question, he stated that it failed to cite any relevant New York state or federal authority. Id. at 1842:14-21. By contrast, Mr. Woll testified that Daiwa, from the Southern District of New York, and Evergreen, from the Western District of Washington, were more on-point and relied on New York law. Id. at 1845:25-1846:4. Important to the issue of liability risk, the testimony of Mr. Woll and Mr. Hawthorne made clear, however, that at the beginning of the Settlement Period, there was not much authoritative New York law on the statute of limitations defense and what authority existed came down on both sides of the issue.
317. On May 14, 2013, the day after the Settlement Period began, two decisions were issued by New York trial courts, reaching different conclusions on the statute of limitations defense. Hawthorne: 1066:15-23. In ACE Securities Corporation Home Equity Loan Trust, Series 2006-SL2 v. DB Structured Products, Inc., 40 Misc. 3d 562 (N.Y. Sup. Ct. 2013) ("ACE I"), Justice Kornreich ruled in favor of the RMBS plaintiffs' position, and in Nomura Asset Acceptance Corporation Alternative Loan Trust, Series 2005-S4 v. Nomura Credit & Capital, Inc., 39 Misc. 3d 1226(A), (Sup. Ct. N.Y. Cty. May 10, 2013), Justice Sherwood ruled in favor of the RMBS defendants' position. Hawthorne at 1066:24-1067:6.
318. In ACE I, Justice Kornreich considered the intent of the parties and the structure of RMBS transactions. Id. at 1067:18-22. Mr. Hawthorne testified that the securitization structure was predicated "on the assumption that no one was doing due diligence [on] every one of the thousands of loans at the time the securitization was created." Id. at 1068:7-10. He stated that there simply was not money in the structure to do that. Id. at 1068:10-12. Instead, the agreed-upon process was that if the defect was discovered, a repurchase protocol would follow, and the repurchase protocol would be available for the life of the underlying mortgages. Id. at 1068:17-1069:10.
319. In Nomura, Justice Sherwood held that the breach occurred, if at all, when the representations were made, and that there was not an independent breach by a sponsor for refusing to repurchase a defective loan. Id. at 1141:24-1142:7.
320. Both the Nomura and ACE I decisions were available to the parties in RFC's Bankruptcy before they executed the Supplemental Term Sheet on May 23, 2013. Id. at 1072:6-17. The split between ACE I and Nomura had not been resolved by any higher court when RFC's Bankruptcy Plan was confirmed on December 11, 2013. Id. at 1072:11-17. Thus, as of RFC's Plan confirmation, RMBS litigators were "awaiting the outcome of the split between these two decisions." Id. at 1072:18-21. And, as of RFC's Plan confirmation, ACE I had been argued in front of the New York Appellate Division, and thus RMBS litigators were particularly focused on the outcome of the appeal of ACE I. Id. at 1072:22-1073:4.
321. After RFC's Plan was confirmed, the New York Appellate Division ruled on the appeal of ACE I in ACE Securities Corp. Home Equity Loan Trust, Series 2006-SL2 v. DB Structured Products, Inc., 112 A.D.3d 522 (N.Y. App. Div. 2013) ("ACE II"). The court ruled in favor of the defendant, reversing Justice Kornreich, and holding that the breach occurred, and the statute of limitations began to run, when the representation was made. Hawthorne at 1148:11-1149:3. In June 2015, the New York Court of Appeals ruled on the appeal of ACE II in ACE III, 25 N.Y.3d at 581, affirming the Appellate Division. Hawthorne at 1151:13-15. Importantly, however, these appellate rulings were not available to the parties during the Settlement Period.
322. Mr. Hawthorne testified that RMBS plaintiffs also argued that the statute of limitations should be equitably tolled because RFC, as master servicer, would have had actual or constructive knowledge of the loan defects and had an obligation to disclose them. Id. at 1076:17-1077:15. As Mr. Hawthorne testified, the RMBS investors had asserted the equitable tolling argument in RFC's Bankruptcy. Id.; see also PTX-145 at 20-21. Mr. Hawthorne's testimony demonstrates that this injected further uncertainty into the statute of limitations issue.
323. PRMI's expert, David Woll, offered two primary opinions as to the statute of limitations. First, Mr. Woll opined that as of the Settlement Period "it was substantially more likely that the defendant's position on the statute of limitations in these repurchase cases would ultimately prevail." Woll 1808:11-19. Second, Mr. Woll testified that "a reasonable defendant . . . would ascribe a lower settlement value to potentially time-barred claims subject to the [statute of limitations] defense than it would to claims not subject to the defense." Id. at 1808:6-10. In particular, he testified that the defense applied to 339 of the 506 Trusts covered by the Bankruptcy Settlements. Id. at 1864:2-12.
i. Likelihood of RMBS Defendants' Success
324. Mr. Woll testified that three basic principles established before the Settlement Period made it more likely that RMBS defendants would succeed on the statute of limitations issue. Id. at 1816:4-8. First, Mr. Woll testified that under New York law, contract claims accrue at the time of breach, regardless of when the injury occurs or is discovered, id. at 1816:16-1818:12, and that in his view, as of the Settlement Period, New York courts favored clear accrual rules, id. at 1818:13-20. Second, Mr. Woll testified that prior to the Settlement Period, New York courts had held that any breaches of "reps and warranties made in a contract as to facts as they existed at the time of contracting or the closing date" occur when they are made. Id. at 1819:8-1821:19. Third, Mr. Woll testified that prior to the Settlement Period, the New York Court of Appeals' decision in Hahn Automotive Warehouse, Inc. v. American Zurich Insurance Company, 916 N.Y.S.2d 678 (N.Y. App. Div. 2011), stood "for the proposition . . . that in New York if there's for instance, a pre-suit demand requirement in a contract, that New York will look to when the plaintiff possessed the right to make the demand as opposed to when the plaintiff makes the demand for purposes of starting the statute of limitations clock." Id. at 1823:4-20. According to Mr. Woll, RMBS defendants cited to Hahn to refute arguments being made by RMBS plaintiffs that the breach accrued upon the defendant's refusal to repurchase the loan under the repurchase protocol. Id. at 1825:21-1826:5.
325. Mr. Woll's testimony demonstrated that indeed there was substantial uncertainty regarding the statute of limitations defense during the Settlement Period. As both Messrs. Woll and Hawthorne acknowledged, there was legal authority during the Settlement Period supporting the RMBS defendants' position that the claim accrues when the representation is made. But, Mr. Woll did not dispute that there was also contrary legal authority supporting the RMBS plaintiffs' position, and that indeed there was a direct split of authority in the New York trial courts. Moreover, neither Mr. Woll nor Mr. Hawthorne could state with certainty which side would prevail. Although Mr. Woll asserted that the RMBS defendants were "substantially" more likely to prevail, Justice Kornreich certainly disagreed, and as of the Settlement Period, her opinion was only one of two New York state court decisions directly addressing the statute of limitations in this context.
326. Mr. Woll purported to offer an objective assessment of the cases and relevant legal principles as of the Settlement Period and then retroactively predicted the outcome. However, as this Court has held in connection with a similar expert proffered in a related case, "[a] retroactive examination of the likelihood of predicting a court decision, in most cases, would be too speculative to justify its own probative force." In re RFC and Rescap Liquid. Tr. Litig. ("Common Daubert Order"), Nos. 13-cv-3451 (SRN/HB), 14-cv-1716 (SRN/HB), 2018 WL 4489685, at *27 (D. Minn. Sept. 19, 2018). That is true here as well.
327. Although Mr. Woll testified as to particular principles of New York statute of limitations law, it was the application of those principles in the RMBS context that was uncertain. Indeed, as Mr. Hawthorne testified, two New York trial court judges reached opposite conclusions regarding the application of those principles. Hawthorne 1074:13- 1075:23. Moreover, during the Bankruptcy case, RFC acknowledged that the case law on "accrual" in the context of RMBS litigation was "largely undeveloped." DTX-536 at 54.
328. Moreover, in assessing Mr. Woll's objectivity, the Court also notes Mr. Woll's prior role as an advocate in the ACE cases, Woll 1967:7-1968:13, as well as in Wave One of this Consolidated Action, on behalf of defendants that ultimately settled with ResCap. Id. at 1958:15-1960:14. Granted, Plaintiff's expert, Mr. Hawthorne, had previously advised plaintiff's counsel in ACE II on statute of limitations arguments, and had also previously advanced statute of limitations arguments on behalf of another RMBS plaintiff, Syncora. Hawthorne 1131:3-1133:6. However, unlike Mr. Hawthorne's advising work in ACE II and his role as an advocate in an earlier RMBS action, Mr. Woll was the lead litigator in the very case for which RMBS litigants during the Settlement Period were awaiting the outcome. Further, Mr. Woll served as counsel for defendants in Wave One of this Consolidated Action on the very same claims asserted against PRMI here. In that capacity, Mr. Woll also argued the statute of limitations issue under New York law, advancing some of the same arguments he now offers as expert opinions. Id. at 1960:18-1963:24. Mr. Woll, in his role as advocate in Wave One of this Consolidated Action, also attacked Dr. Snow's methodologies based on the statute of limitations. Id. at 1965:15-23. Although Mr. Hawthorne's experience in the RMBS area was plaintiff-oriented, unlike Mr. Woll, he did not represent parties in this Consolidated Action. Mr. Woll's direct personal involvement on the statute of limitations issue therefore raises concern about his objectivity as an expert.
329. Mr. Woll also either dismissed summarily or failed to discuss key New York cases and principles that RMBS plaintiffs relied upon. For example, although he focused on Hahn, Mr. Woll did not discuss Continental Casualty Co. v. Stronghold Insurance Co., Ltd., 77 F.3d 16 (2d Cir. 1996), a case relied on by RMBS plaintiffs and Justice Kornreich. See ACE I, 40 Misc.3d at 568. Nor did Mr. Woll address the condition precedent argument that RMBS plaintiffs asserted—an argument that the New York Court of Appeals in ACE III referred to as the plaintiff's "strongest argument." See ACE III, 25 N.Y.3d at 630. Reviewing events in hindsight ignores the uncertainty that existed at the time.
330. Mr. Woll also focused on evidence from RFC's Bankruptcy that supported his opinions, while ignoring evidence that contradicted them. For example, in both his expert report and his testimony, Mr. Woll relied on excerpts of the Bankruptcy declaration of Mr. Lipps regarding equitable tolling. Woll 1885:7-1891:11. Mr. Woll specifically relied on the discussion by Mr. Lipps of the arguments RFC could have made against equitable tolling. Id. He did not address, however, the surrounding portions of Mr. Lipps' declaration, which discussed the arguments that RMBS plaintiffs could have made in opposition to equitable tolling, and the risks and uncertainties those arguments presented for RFC, other than simply noting that equitable tolling was only available in rare and exceptional circumstances. Id.
331. Mr. Woll also relied on briefs filed by MBIA in the Bankruptcy. Id. at 1895:1-3. Mr. Woll acknowledged that because MBIA was an RMBS plaintiff that had asserted claims against RFC, it was not a neutral party. Id. at 1902:22-1903:7. Understandably, MBIA advanced arguments favorable to the trusts that MBIA was insuring. Id. He further acknowledged that a plaintiff's view may be relevant on the issue of how a reasonable defendant would have valued the settlement. Id. at 1904:6-14. Nonetheless, Mr. Woll also testified that he did not consider the positions of the RMBS Trustees to be relevant because they were "plaintiff-side" trusts, and because they had their own motivations as trustees. Id. at 1904:15-25.
ii. Lower Settlement Value
332. As noted, Mr. Woll testified that even if RMBS defendants were not "substantially" more likely to prevail on the issue of the statute of limitations, "a reasonable defendant . . . would ascribe a lower settlement value to potentially time-barred claims subject to the [statute of limitations] defense than it would to claims not subject to the defense." Id. at 1808:6-10.
333. Mr. Woll acknowledged that his opinion assumed that the claims were otherwise "equivalent." Id. at 1931:25-1932:7. He understood that his opinion was rendered in a vacuum—it presumed that there were no other defenses or issues that could differentiate a claim on a loan-by-loan or trust-by-trust basis. Id. at 1932:25-1933:14. Mr. Woll also acknowledged, however, that the statute of limitations issue was by no means the only issue at play regarding repurchase claims as of the Settlement Period. Id. at 1933:10-18. Indeed, Mr. Hawthorne testified at length regarding numerous legal issues and defenses that could have potentially impacted claims, varying across loans or trusts.
334. Many of these legal issues and defenses remained unresolved as of the Settlement Period, and as such, they presented uncertainties. As discussed below, the undisputed testimony at trial demonstrated that somehow trying to assign particular percentages to the different probabilities as to the strengths and weaknesses of these defenses would be entirely speculative. No witness or expert on either side attempted to do so.
2. Accounting For Only Select Unresolved Legal Issues and Defenses Would Have Introduced Bias and Speculation Into the Model
335. PRMI contends that it "has never argued Plaintiff must account for 'every possible strength or weakness of the settled claims.' " (PRMI Reply Mem. in Support of Summ. J. ("PRMI SJ Reply") [Doc. No. 5327] at 5.) Rather, PRMI has asserted that ResCap must take into account issues that purportedly are "few in number, clear in significance, and easy to account for." Id.; see also PRMI Closing 2774:16-24 (arguing that allocation should have accounted for "essential issues").
336. The evidence at trial does not support PRMI's position.
337. First, Mr. Woll considered the statute of limitations defense and the defense view of the Trust Reps in a vacuum, without also considering the effect of multiple other unresolved legal issues and defenses on the parties' assessment of the claims. Litigation, however, is not conducted in a vacuum, and a settlement must consider all potential risks of liability, as illustrated by Mr. Hawthorne's testimony and the evidence from RFC's Bankruptcy. Mr. Hawthorne was the only expert for either party to address other important, unresolved issues that were being litigated in RMBS cases as of the Settlement Period and that were being contested in RFC's Bankruptcy case. None of PRMI's experts did so. Mr. Hawthorne's testimony demonstrated the significant uncertainty surrounding those myriad issues.
338. Second, Mr. Woll testified that the statute of limitations was "different . . . from other defenses that typically are raised in litigation" because it had "the potential for eliminating liability," as opposed to defenses that may just reduce liability. Woll 1813:2-14. As just noted, however, Mr. Woll did not assess the statute of limitations defense in context with all the other defenses at issue in RMBS litigation as of the Settlement Period. See id. at 1932:25-1933:18. Thus, he could not offer any opinion about the comparative impact of those defenses based on his limited analysis.
339. Moreover, as Mr. Hawthorne testified, if RMBS defendants had prevailed on the applicable causation standard, which was being litigated during the Settlement Period, the impact on repurchase claims would have been very substantial and presented a severe challenge for RMBS plaintiffs' claims. Hawthorne 1010:13-18, 1056:4-1057:1. This was further illustrated by the opinion of Dr. Cornell, an expert retained by the Committee in challenging the Original RMBS Settlement. Dr. Cornell found that if successful, the causation defense would reduce the RMBS Trusts' claims from $16.5 billion (which reflects breaching losses) to just $3.8 billion, wiping out the vast majority of the claims. Cornell Dep. 140:2-21; DTX-541 at 37-42. Like the statute of limitations defense, the applicable causation standard was a question of law that could be decided early on, without discovery.
340. Similarly, the election of remedies defense, discussed previously, if successful, would have eliminated all claims on liquidated loans, that were a substantial portion of the RMBS Trusts' claims. Hawthorne 1077:25-1078:22.
341. The emphasis that PRMI and Mr. Woll placed on the statute of limitations, as opposed to other defenses, also does not align with how the Bankruptcy parties actually assessed and argued these issues at the time. For example, in its objection to the Original RMBS Trust Settlement, just prior to the Settlement Period, the Committee discussed three "major defenses" (causation, statute of limitations, and election of remedies) and their potential impact on the claims of the RMBS Trusts. See DTX-541 at 35-41. The Committee did not place any extra emphasis on the statute of limitations defense, and concluded that "there are myriad complex legal issues that may affect the outcome of the Rep claims . . . ." Id. at 47. In support of Plan confirmation toward the end of the Settlement Period, RFC and the Committee submitted a brief discussing the RMBS Trust Settlement. DTX-578 at 86-94. That brief discussed the same major three defenses and did not place any particular weight on the statute of limitations defense over the causation and election of remedies defenses. Id.
342. Thus, the Court finds that it is neither fair nor reasonable to consider issues in isolation or to consider only certain issues (i.e. statute of limitations and differences in certain Trust Reps) as being potentially applicable to allocation, while ignoring others. Such an approach ignores not only the way reasonable litigants assess cases, but it would improperly favor some defendant-originators over others in the allocation of damages.
3. Assigning Probabilities Would Be Speculative and Imprecise
343. Mr. Hawthorne credibly testified that "assigning probabilities—precise quantitative probabilities to different litigation factors like these inevitably requires sort of false precision," Hawthorne 1081:16-18, and that "trying to put specific numbers to a claim is, first of all, always going to be contestable and, second of all, gets increasingly unrealistic if you try to deal with a number of intersecting probabilities and numerically derive some kind of quantitative assessment of the likelihood of success as a whole," id. at 1081:24-1082:5. This testimony was uncontradicted.
344. Mr. Woll did not offer any opinion on how much the statute of limitations would purportedly impact "settlement value" or offer any sort of percentage discount that should be taken based on this defense. Nor did Mr. Woll testify that it would be feasible to assign percentage discounts to individual issues or defenses with any sort of certainty. Although Mr. Woll testified that "differential weight" could be given to a claim depending upon whether the statute of limitations defense applied, Woll 1894:12-15, he made clear that he was not testifying that "a reasonable defendant could assign, with some precision, percentages to the probabilities on the outcome of the statute of limitations," id. at 1894:18-21. Mr. Woll was questioned by PRMI's counsel about different hypothetical percentages, but he never assigned any probability to the statute of limitations defense, and instead used broad, qualitative terms such as "stronger" or "less settlement value." Id. at 1827:12-22, 1927:15-20. Neither Mr. Woll nor any other witness explained how these probabilities could be calculated and applied to an allocation, or whether that was even feasible.
345. Although Mr. Woll selectively relied on the declarations of RFC's pre-bankruptcy litigation counsel, Mr. Lipps, for certain points, id. at 1885:7-1891:11, he ignored that Mr. Lipps also stated in his declaration that "[i]n order to conduct Litigation Risk Analysis on an RMBS case, an attorney would have to assign highly speculative percentages to the resolution of all of these issues," and "[t]he end product is little better than guesswork and therefore provides no meaningful guidance." DTX-537 at 10. This Court has not admitted Mr. Lipps' testimony for the truth of the matter asserted. Tr. at 2652:23-2653:19. Rather, as PRMI itself has argued throughout this case, the statements of Mr. Lipps are relevant as information available to the parties during the Bankruptcy. Mr. Lipps' declaration was filed in the Bankruptcy Court prior to the Settlement Period. See DTX-537.
To the extent that PRMI continues to object to the Court's ruling with respect to the admissibility of this exhibit for a limited purpose, or any other evidentiary rulings at trial, the Court will not revisit those rulings here.
346. As Mr. Hawthorne's testimony illustrated, to account for even a few of the issues that PRMI contends should have been part of the allocation, requires the assignment of several speculative probabilities to a variety of interrelated issues. See Hawthorne 1082:12-1084:21, 1087:3-1091:20.
347. Finally, and of particular note, PRMI's expert, Dr. McCrary, admitted that where estimates of probabilities are guesswork, he "would have probably approached the problem the way that Dr. Snow did." McCrary 2711:8-2711:24. Thus, even PRMI's own expert admits that absent assigning actual probabilities with some precision (which no PRMI witness contends was feasible), Dr. Snow's allocation is reasonable.
348. Based on the foregoing, the Court finds that Dr. Snow's allocation approach is reasonable. As set forth above, Dr. Snow's approach is based on objective factors of breaches and losses. Moreover, it is consistent with the methodology used by the parties in RFC's Bankruptcy, which was incorporated in the Plan and approved by the Bankruptcy Court. It is also consistent with the approaches used in every other major RMBS trust settlement.
D. Dr. Snow Implemented His Methodology Reasonably
349. PRMI offered trial testimony from Dr. Justin McCrary, who criticized various technical decisions Dr. Snow made in implementing his methodology and calculating damages. McCrary 2486 et seq. The Court finds those criticisms unavailing.
1. The Treatment of the NDS Trust Breach Rate is Reasonable
350. As discussed above, Dr. Snow deducted $269 million from the RMBS Trust Allowed Claim to account for settlement proceeds attributable to NDS Trusts for which ResCap does not seek indemnity. Snow 1343:9-23, 1396:1-21. Dr. Snow calculated that deduction pro rata based on loan losses, in a manner equivalent to assuming that the breach rate of loans in the NDS Trusts was equivalent to that outside the NDS Trusts. Id. at 1396:6-15. Dr. McCrary criticized that assumption as unfounded. Id. at 1396:1-24, 1398:5-10; McCrary 2624:11-25.
351. The Court finds Dr. McCrary's criticism unavailing for several reasons. First, in contrast to Dr. McCrary, who offered no evidence that the breach rate in the NDS population was different than the breach rate outside the NDS population, Dr. Snow conducted an empirical analysis suggesting that the breach rates were unlikely to differ materially. Snow 1398:11-19; McCrary 2628:3-9. Specifically, Dr. Snow examined breach rate estimates that Duff prepared as part of the Bankruptcy, based on reunderwriting that was conducted in the Bankruptcy. Snow 1398:21-1399:11. The Duff analysis accounted for loan vintage and product type. Id. at 1398:21-24. As applied here, Dr. Snow determined that the NDS Trust breach rate implied by the Duff data was the same as the non-NDS Trust breach rate implied by that data. Id. at 1399:12-16. In this regard, Dr. McCrary admitted that the two factors Duff examined—vintage and product type—were the only two factors Dr. McCrary himself had sought to consider when he was developing his criticisms of Dr. Snow's monoline methodology. McCrary 2623:10-2624:10, 2626:18-24, 2627:20-24.
352. Second, Dr. Snow testified, without contradiction from Dr. McCrary, that it was not possible for him to obtain a reunderwriting-based breach rate for the NDS Trusts, because loan-level data for those Trusts was not available. Snow 1397:11-18; McCrary 2626:7-13. Dr. McCrary did not explain how Dr. Snow could have addressed this issue other than through the approaches that Dr. Snow implemented.
353. Third, Dr. Snow conducted quantitative scenario analyses to determine what PRMI's damage allocation would have been under different NDS Trust breach rates. Snow 1400:2-15. Even assuming the NDS Trust breach rate was 100%—the most extreme possible assumption in PRMI's favor, which was not a "reasonable or real world possibility"—the total impact on damages was to decrease PRMI's allocation by only $125,000. Id. at 1400:16-1401:4. In the absence of testimony or other evidence suggesting that the NDS Trust breach rate was meaningfully different than the 66% rate Dr. Snow assumed, id. at 1399:23-25, the Court concludes that Dr. Snow's treatment of this issue was proper and that his results are reliable.
2. The Definition of the At-Issue Population is Reasonable
354. As discussed above, Dr. Snow's allocation methodology was based on samples drawn from a specified "At-Issue" population, which Dr. Snow defined as loans included in the Bankruptcy Settlements that had actual losses of at least $500, or expected losses, as of May 2013. Snow 1344:9-18, 1346:24-1347:7. Dr. McCrary criticized this approach, arguing that Dr. Snow should have considered every loan settled in the Bankruptcy. Id. at 1390:20:1391:3; McCrary 2543:7-18, 2544:1-22, 2549:9-18. The Court rejects this criticism for several reasons.
355. First, Dr. McCrary's criticism itself assumed that the population of interest for purposes of Dr. Snow's analysis was the set of all loans settled in the Bankruptcy. McCrary 2549:9-18. Dr. McCrary offered no support for that assumption and could not do so because the question turns not on an issue of statistics, but rather on factual and legal issues concerning the Bankruptcy parties' settlement negotiations. As this Court previously held in rejecting Dr. McCrary's criticism in the context of a Wave One Daubert motion, "Dr. Snow's decision to sample from the At-Issue Loans makes good sense given that the purpose of his study is to allocate the Bankruptcy claims among Defendants, and those claims are premised on losses to loans sold by RFC. Thus, conceptually, those damages would necessarily have flowed from the loans that actually experienced economic losses, i.e. the At-Issue Loans." Common Daubert Order, 2018 WL 4489685, at *5.
356. Second, it is undisputed that "the losses on the [A]t-[I]ssue [L]oans are 98 percent of the losses on all the [loans settled in the Bankruptcy.]" Snow 1392:3-5. Indeed, of the approximately 1.5 million non-At-Issue loans that Dr. McCrary contended should have been included in Dr. Snow's analysis, approximately 1.2 or 1.3 million were paid in full as of May 2013; as Dr. Snow testified "[t]here was no loss associated with those loans, nor would there be a future loss associated with . . . those loans." Id. at 1391:10-15. These numbers support the reasonableness of Dr. Snow's method.
357. Third, as with his other criticisms, Dr. McCrary presented no analysis suggesting that PRMI's allocation would have meaningfully changed—much less decreased—if Dr. Snow had included performing loans in his calculations. By contrast, Dr. Snow presented scenario analyses indicating that including performing loans would not have had a meaningful impact. Id. at 1392:21-1393:7. Specifically, (a) assuming that performing PRMI and Global Loans had the same breach rates as the respective categories of non-performing loans, PRMI's damages allocation would increase by $24,000; (b) implementing an extreme assumption in PRMI's favor that PRMI's performing loans had a 0% breach rate and the performing Global Loans had the same breach rate as the non-performing Global Loans, PRMI's allocation would decrease by approximately $135,000; and (c) even implementing the most extreme possible assumption in PRMI's favor, i.e., that PRMI's performing loans had a 0% breach rate and the performing Global Loans had a 100% breach rate, the impact on PRMI's damages would only be approximately $207,000. Id. at 1393:13-1394:21. These figures further support the reliability of Dr. Snow's methodology.
3. The Monoline Allocation Approach is Reasonable
358. As discussed above, Dr. Snow's monoline allocation methodology relied on blended monoline breach rates that included loans insured by different monoline insurers. Snow 1493:20-1494:4. Dr. McCrary criticized that approach on the basis that breach rates may have varied across monoline insurers. McCrary 2572:12-23. The Court rejects that criticism for several reasons.
359. First, Dr. Snow's methodology did not ignore potential differences between the breach rates in different monoline settlements. As Dr. Snow explained, his blended monoline breach rates in fact incorporated "in a sense a weighted average of the individual monoline breach rates." Snow 1371:5-7.
360. Second, the cost of drawing and reunderwriting monoline-by-monoline samples, in the manner Dr. McCrary advocated—or indeed of drawing samples for each monoline pool—would have far outstripped the approximate $400,000 monoline allocation that ResCap seeks in this action. Dr. Snow testified that such an analysis would have added approximately $11 million in expense, relying on undisputed evidence, in the form of a declaration from counsel, that it would cost approximately $8,000 to underwrite a single loan in this action. Id. at 1378:4-1379:18. On cross-examination, Dr. McCrary conceded that he had not been aware of that cost when offering the criticisms in his rebuttal report. McCrary 2677:13-16. At his deposition, Dr. McCrary appeared to have never reviewed the relevant attorney declaration until it was brought up by opposing counsel, even though it was attached to Dr. Snow's rebuttal report. Id. at 2677:17-2679:16.
361. Third, as with his other criticisms, Dr. McCrary performed no quantitative analysis to demonstrate that his criticisms, if addressed, would have reduced PRMI's monoline allocation. Snow 1379:19-1380:4. Such analysis, which Dr. Snow did perform, confirmed the reliability of Dr. Snow's methodology. Id. at 1380:11-13. Specifically, Dr. Snow developed a monoline-by-monoline approach, based on data already available to him from existing samples, to provide a sense of whether his use of a blended breach rate had adversely impacted PRMI. Id. at 1380:14-1383:22. Far from benefitting PRMI, the monoline-by-monoline approach generated a number somewhat higher than the blended breach rate approach: approximately $441,000 in the former versus approximately $401,000 in the latter. Id. at 1383:23-1385:2. Further, Dr. Snow also performed a non-extrapolated monoline calculation that (1) assumed all 46,000 non-PRMI loans relevant to the monoline allocation were in breach; and (2) assumed all PRMI loans relevant to the monoline allocation were not in breach, except for 22 loans Mr. Butler actually deemed to be in breach based on a reunderwriting. Id. at 1385:3-1388:25. This non-extrapolated calculation, which reflected the smallest possible monoline allocation to PRMI based on unreasonable and extreme assumptions, would generate an award of approximately $286,000. Id. at 1388:20-1389:5. Dr. McCrary offered no meaningful criticism of this latter approach during his testimony, thereby confirming Dr. Snow's testimony that the maximum possible impact of Dr. McCrary's monoline criticisms was $115,000—i.e., the difference between Dr. Snow's baseline $401,000 allocation and his $286,000 non-extrapolated calculation. Id. at 1528:2-1530:7. In light of all the foregoing, the Court concludes that Dr. Snow's baseline monoline methodology was reasonable and reliable.
4. The Point Estimate is an Appropriate Measure of Damages
362. As discussed above, Dr. Snow's allocation methodology generates a point estimate of $5.4 million, with a confidence interval of 95% and a margin of error of approximately $1.35 million. Id. at 1409:16-1410:23, 1413:17-1415:8. In his testimony, Dr. McCrary opined that Dr. Snow "ought to have considered" using the lower bound of the confidence interval, rather than the point estimate, as the measure of damages. McCrary 2672:16-2673:4. The Court finds no evidence suggesting that Dr. Snow erred in rejecting that approach.
363. First, Dr. McCrary never opined that Dr. Snow should actually have used the lower bound of the confidence interval, and indeed expressly disclaimed such an opinion. Id. at 2672:21-2673:4 ("A. It's certainly my view that in light of the imprecision associated with Dr. Snow's methodology, that he ought to have considered that. As to whether or not he should have, I think it's right that that's closer to asking me an affirmative question and I wasn't asked that. Q. Okay. And you've not offered that opinion? A. What I've offered instead is that I don't think that's a bad idea and, in fact, it's certainly one that ought to have been considered."). In fact, as to the margin of error for the RMBS Trust Settlement, Dr. McCrary conceded that Dr. Snow's calculation was "in the right ballpark." Id. at 2667:21-23.
364. Second, Dr. McCrary provided no meaningful support for his view. He conceded that he has "not cited any textbook or any peer-reviewed journal article stating that statistics always requires using the lower bound of a confidence interval in a circumstance like this one," that he was "not aware of examples in which agencies or entities other than [the Center for Medicare and Services ("CMS"), addressed in his testimony] use the lower bound," and that, in his entire career, he has "never once advocated for the use of the lower bound of a confidence interval that [he] had developed." Id. at 2674:21-2675:19, 2676:7-12. His opinion in this respect was consistent with that of Dr. Snow, who testified that there is not a principle of statistics or economics pursuant to which "it would be appropriate for [ResCap] to request or the Court to [a]ward . . . [the] lower bound[] number." Snow 1415:9-13. Indeed, the only purported support Dr. McCrary provided for his view about the lower bound relates to a purported practice of the CMS—but the CMS policy manual upon which Dr. McCrary relied does not require use of the lower bound in all instances, and is instead "a bit open-ended" and "has language that might be taken to support exceptions," as Dr. McCrary conceded. McCrary 2673:13-2674:20.
365. Third, it is undisputed that awarding the lower bound as a measure of damages would be "overwhelmingly likely" to understate PRMI's actual allocation, and indeed would be expected to understate the allocation in 97.5% of outcomes. Snow 1416:8-17.
5. Dr. McCrary's Prior Testimony Undermined His Credibility
366. Although Dr. McCrary advanced some other critiques of Dr. Snow's methodology, the Court finds those and the ones discussed above unpersuasive and, in some instances, not credible, due to the contradictory testimony Dr. McCrary offered in prior proceedings.
367. First, to the extent Dr. McCrary suggested here that Dr. Snow's method for allocating the RMBS Trust Settlement was unworkable, that testimony contradicted his testimony in a previous jury trial in the Consolidated Action. Specifically, at trial in ResCap's action against Home Loan Center (No. 14-cv-1716 (SRN/HB)), Dr. McCrary testified on behalf of the defense that his critiques of Dr. Snow's Trust methodology were minor in nature:
• "[I]f we analogize Dr. Snow's [RMBS Trust Settlement] methodology to a car, you would say that the car has some problems, like maybe the windows don't go down or something like that and that it's in need of a tune-up . . . ." McCrary 2713:1-10.
• "Dr. Snow's trust settlement analysis has some problems attached to it, but [I] don't think those problems are something that would render the trust settlement analysis unusable . . . ." Id. at 2713:11-15.
• "[T]he trust settlement analysis, setting aside the strength of claims and defenses, is one that is workable." Id. at 2713:15-17.
368. At trial here, Dr. McCrary made no effort to distinguish the Trust methodology that Dr. Snow used here from the one that Dr. Snow used in the Home Loan Center action. Nor did Dr. McCrary attempt to distinguish his testimony there from the issues relevant here. The Court therefore finds that Dr. McCrary's testimony in Home Loan Center provides a further basis upon which to find that Dr. Snow's Trust methodology here was reasonable and reliable.
369. Second, as to the NDS Trust and At-Issue Population issues discussed earlier, Dr. McCrary testified during the Home Loan Center trial that the maximum possible impact of those issues was "minor": "[I]n connection with the trust settlement damages estimate . . . there should probably be some adjustment. It's minor for the fact of the population is actually the wrong one . . . ." Id. at 2616:19-2617:15, 2618:8-17 (noting that these issues "do[]n't render the underlying approach fundamentally unreliable"). Because the extent of those issues in Home Loan Center (where Dr. McCrary admitted they had a maximal impact on damages of $350,000) was larger than the maximal impact here (which Dr. McCrary conceded was $330,000), Dr. McCrary's prior testimony again provides support for a finding that Dr. Snow's methodology here was reasonable and reliable. Indeed, after having testified that a larger $350,000 impact in Home Loan Center was minor, Dr. McCrary undermined his credibility here by insisting that it was "not for [him] to say" whether a $330,000 impact here would be small. See id. at 2615:15-2619:4.
370. Third, Dr. McCrary testified here that Dr. Snow's monoline methodologies violated five "pillars" of statistics. Id. at 2613:17-19. However, in Wave One, Dr. McCrary opined that there were a total of three pillars. Id. at 2605:16-2606:25, 2608:2-6, 2609:4-8. Subsequently, in a December 21, 2018 expert report he issued in a case titled Lopez v. Liberty Mutual, Dr. McCrary opined that "[s]tatistical analysis rests on four pillars." Id. at 2609:9-2611:12 (emphasis added). Dr. McCrary also conceded that he has developed the concept of "pillars" himself in order to better communicate information about statistics. See id. at 2600:16-2601:1.
371. Fourth, although Dr. McCrary criticized Dr. Snow's failure to allocate based on the strength and weaknesses of litigation claims and defenses—and indeed testified that Dr. Snow's failure to do so violated a fundamental "pillar" of statistics—Dr. McCrary's credibility in offering that opinion was undercut based on his prior testimony. Specifically, notwithstanding Dr. McCrary's extensive work for approximately 20 defendants in Wave One, where he was given the assignment of assessing the reliability of the same methodology at issue here, Dr. McCrary never once raised the "strength of claim or defense" issue as a problem until this case. Id. at 2688:6-15, 2689:1-11. Indeed, Dr. McCrary lodged his criticism here without even knowing whether it would have been possible for Dr. Snow to assess the likelihood of success on the statute of limitations, id. at 2698:12-16, and he conceded that "if there is genuinely no ability to have any assessment of the relative strength of claims and defenses, then . . . it's right that you can't actually take that into account," in which case he would have instead implemented Dr. Snow's approach, id. at 2711:8-22.
CONCLUSIONS OF LAW
I. OVERVIEW AND ELEMENTS OF CLAIM FOR CONTRACTUAL INDEMNIFICATION
372. ResCap has asserted a claim for contractual indemnification against PRMI.
373. To prevail on its contractual indemnification claim, ResCap was required to establish the following elements by a preponderance of the evidence: (1) a Governing Contract: the existence of a contract (i.e., the Client Contract and the Guides); (2) Breach and Causation: that the losses, damages, or liabilities that RFC incurred fall within the scope of that contract, that PRMI was in breach of that contract, and that PRMI's breaches were a contributing cause of the liability RFC settled in its Bankruptcy proceeding; (3) Reasonableness of the Bankruptcy Settlements: that the Bankruptcy Settlements for which ResCap sought indemnification were reasonable, and entered into in good faith; and (4) Reasonably Certain and Non-Speculative Damages: the amount of reasonably certain and non-speculative damages ResCap is entitled to recover in indemnification from PRMI. See Common SJ Order, 332 F. Supp. 3d at 1129-30, 1191; see also Rice Lake Contracting Corp. v. Rust Env. & Infrastructure, Inc., 616 N.W.2d 288, 291 (Minn. Ct. App. 2000) (citing Altermatt v. Arlan's Dep't Stores, 169 N.W.2d 231, 232 (1969)).
374. As to the first element—the existence of a governing contract—the Court has already ruled as a matter of law that "[b]arring the defenses of waiver and estoppel, . . . the Guides apply to all At-Issue loans." PRMI SJ Order, 428 F. Supp. 3d at 90. PRMI does not dispute this finding generally. In addition, and as discussed in more detail below, the Court has rejected PRMI's argument as speculative, and without a basis in the record, that the seven AlterNet Loans at issue were subject to a later version of the AlterNet Guide, and not the 1997 version used by Mr. Butler to determine PRMI's breaches. Accordingly, Plaintiff has satisfied the first element of its contractual indemnification claim—the existence of a governing contract.
As will be discussed in more detail below, the Court finds that it need not rule on whether PRMI has established that ResCap should be estopped from enforcing certain provisions of the Client Guide as to the five Assetwise Loans at issue in this case, because even if ResCap was so estopped, the evidence shows that PRMI was in breach in any event. With respect to each of the Assetwise Loans at issue, one or more of the following facts were established at trial: (1) the loans were Client Guide loans for which ResCap had sole discretion to determine breaches; (2) one or more of the conditions set forth in the initial Assetwise Approval were not met; (3) there was a breach of the Client Guide or AlterNet Guide that PRMI concedes was applicable to all loans, including Assetwise loans; or (4) there was a borrower misrepresentation—which even PRMI's witnesses conceded was not permissible under any Guide.
375. As to the second element—breach and causation—that the losses, damages, or liabilities incurred by RFC fell within the scope of the Guides' indemnification provisions, ResCap has met its burden of proof that PRMI was in breach and that PRMI's breaches were a contributing cause of the liability RFC settled in its Bankruptcy. Pursuant to the Guides, PRMI agreed to indemnify RFC for "losses and liabilities 'resulting from,' 'arising from,' or that were 'a result of' Events of Default or breaches of Reps" by PRMI. Common SJ Order, 332 F. Supp. 3d at 1158-60, 1164; PTX-001 § 274; PTX-1056 § A212; PTX-032 § A223; PTX-1055 § A212; see also PRMI SJ Order, 428 F. Supp. 3d at 67. As will be discussed below, ResCap established at trial the requisite Events of Default (i.e., PRMI's breaches), and that those breaches were a contributing cause of the liability that RFC settled in its Bankruptcy.
376. As to the third element, the reasonableness of the Bankruptcy Settlements, the Court has already ruled as a matter of law that ResCap has met its burden of proving that the Bankruptcy Settlements at issue were reasonable and entered into in good faith. See PRMI SJ Order, 428 F. Supp. 3d at 87. Thus, the third element is satisfied.
377. As to the fourth element—the amount of damages ResCap is entitled to recover from PRMI by way of indemnification—as will be discussed below, ResCap proved the existence of reasonable and non-speculative damages at trial.
II. A GOVERNING CONTRACT
378. As noted, it is undisputed that PRMI's sale of loans to RFC was governed by two Client Contracts: (1) the March 2000 Client Contract, PTX-003; Zitting 1657:6-1658:18; and (2) the June 2001 Client Contract. PTX-010; Zitting 1679:2-11. In turn, the Client Contracts, respectively, incorporated by reference two Guides: the March 2000 Client Contract incorporated the AlterNet Guide, PTX-003 ¶ 1; Bangerter 343:3-345:23, and the June 2001 Client Contract incorporated the Client Guide. PTX-010 ¶ 1.
379. The Court has already ruled as a matter of law that "the Guides apply to all At-Issue loans," absent the defenses of estoppel and waiver. PRMI SJ Order, 428 F. Supp. 3d at 90. Because PRMI contends that its defenses of estoppel and waiver apply to only five Assetwise Loans and one Countrywide Loan, the Court concludes that as to all other loans, the provisions of the Guides were incorporated into the governing contracts.
380. Regarding the seven loans sold under the AlterNet Guide (4115211, 4550405, 4413350, 4375505, 4962418, 4117058, and 4380515), Plaintiff has established that the March 2000 Client Contract incorporated the AlterNet Guide. See PTX-003 at 1; Bangerter 343:3-345:23. Plaintiff has also established that the 1997 AlterNet Guide was the version of the AlterNet Guide incorporated into the March 2000 Client Contract. See Bangerter 348:16-349:1. The AlterNet Guide, as supplemented or amended, "sets forth the terms and conditions under which" PRMI sold AlterNet mortgage loans to RFC. PTX-001 at 16.
381. Moreover, as noted earlier, Mr. Butler explained that (1) he and Plaintiff's counsel undertook extensive searches for any later versions of the AlterNet Guide and did not locate any, and (2) the Reps in the Client Guide and AlterNet Guide "are very similar." Butler 576:12-21, 566:4-9.
382. PRMI has argued that "it believe[s]" that the 1997 AlterNet Guide was "not the AlterNet Seller Guide that was [in force] in the year 2000." PRMI Objection 349:13-16. But PRMI offers no evidence to support its assertion, and thus its argument fails. See generally Sokol & Assocs., Inc. v. Techsonic Indus., Inc., 495 F.3d 605, 610 (8th Cir. 2007) ("[T]he well-established rule in [Minnesota] [is] that a party asserting the parol modification of a written contract has the burden of proving the modification by clear and convincing evidence. The burden is not met by a mere preponderance of the evidence.") (quoting Merickel v. Erickson Stores Corp., 95 N.W.2d 303, 305 (Minn. 1959)).
383. PRMI relies exclusively on the speculation of its expert, Ms. Keith. As stated earlier, the Court finds this speculation not credible. See Bennett v. Comm'r of Human Servs., No. A09-442, 2010 WL 10406, at *2 (Minn. Ct. App. Jan. 5, 2010) (finding that contentions based on mere speculation are not persuasive). Ms. Keith admitted that she did not see any document that actually referred to a later version of the AlterNet Guide. Keith 2364:17-20, 2365:12-2366:1. PRMI did not offer any evidence of a subsequent version of the AlterNet Guide between the 1997 AlterNet Guide and the year 2000. Accordingly, the Court concludes that Plaintiff has met its burden of establishing that the 1997 AlterNet Guide was incorporated into the March 2000 Client Contract, and PRMI has failed to demonstrate otherwise. Therefore, it was reasonable for Mr. Butler to utilize the 1997 AlterNet Guide to determine breaches of the AlterNet Loans.
384. Even if RFC were estopped from enforcing certain provisions of the Guides as to the five Assetwise Loans, as will be discussed in the next section, they are still in breach.
385. Accordingly, the Court concludes that ResCap has established the existence of the applicable governing contracts.
III. PRMI COMMITTED EVENTS OF DEFAULT AND BREACHED REPS BY SELLING RFC DEFECTIVE LOANS
386. ResCap has established that PRMI sold RFC loans that materially breached PRMI's Reps to RFC, resulting in Events of Default under the Client Contracts and Guides, and triggering PRMI's indemnity obligations.
387. Under the Guides, it was an Event of Default if, among other things, PRMI: (1) failed to comply with any requirement, term, or condition set forth in the Guides; (2) breached any agreement outlined or incorporated by reference in the Client Contracts; or (3) breached or made any false or misleading Reps to RFC, or failed to provide RFC with information that was true, complete, and accurate. PTX-001 at 51; PTX-032 at 57.
A. This Court Ruled That, as a Matter of Law, ResCap Had Sole Discretion, Under the Client Guide, to Determine Breaches
388. As set forth above, ResCap's expert, Mr. Butler, reunderwrote 157 PRMI Loans and found that 101 materially breached PRMI's Reps to RFC. Butler 548:21-550:6, 571:19-25.
389. PRMI did not challenge Mr. Butler's breach findings as to 89 PRMI Loans that were sold to RFC pursuant to the June 2001 Client Contract and Client Guide. Nor could it, because this Court has already held that ResCap had "sole discretion under the Client Guide to make factual determinations including whether an Event of Default has occurred." PRMI SJ Order, 428 F. Supp. 3d at 92. That discretion also "necessarily encompasse[d]" the determination that an Event of Default or breach was material. Id.; see also Residential Funding Co. v. First Mortg. Corp., No. 13-cv-3490 (SRN/HB), 2018 WL 6727065, at *14 (D. Minn. Dec. 21, 2018). Moreover, PRMI did not assert any waiver or estoppel defense as to those 89 PRMI Loans. See, e.g., Keith 2215:6-2217:3, 2253:17-2254:10. Accordingly, ResCap has established that those 89 PRMI Loans materially breached PRMI's Reps to RFC.
390. In addition, the parties did not dispute that four of the Assetwise Loans at issue (5798390, 8257547, 10381337, 6231616) were sold under the Client Guide, rather than the AlterNet Guide. Therefore, irrespective of PRMI's defense of waiver and estoppel as to these four Assetwise Loans, RFC had sole discretion to declare breaches as to these loans under the Client Guide. PRMI SJ Order, 428 F. Supp. 3d at 92.
B. The Assetwise Loans
391. Even setting aside RFC's sole discretion to declare breaches with respect to the four Assetwise Loans sold under the Client Guide (5798390, 8257547, 10381337, 6231616), the breaches on these loans relate to Reps that PRMI concedes continue to apply to these loans, even under its theory of estoppel. See, e.g., Zitting 1762:16-1763:24, 1776:10-23 (conceding that PRMI was required to prudently originate Assetwise Loans under the AlterNet Guide and the Client Guide).
392. Moreover, the uncontested evidence at trial showed that the Assetwise Findings Report approved Loans 5798390, 8257547, and 10381337 on the condition that they comply with certain requirements—requirements that PRMI failed to meet with respect to complete income documentation, see Butler 596:3-18, 597:11-17; PTX-294 at 141; PTX-294 at 136-37, asset verification, see Butler 597:20-598:12, 599:2-600:16; PTX-304 at 12, 146, and employment history, see Butler 605:15-606:15. With respect to Loan 6231616, the evidence at trial showed that PRMI failed to obtain 12 months of rental history, despite the fact that the Client Guide required it to do so, and PRMI knew that Assetwise could not verify rental history. Butler 609:24-610:19; Keith 2286:12-15.
393. Thus, RFC had sole discretion to determine breaches of the four Assetwise Loans, and the loans independently failed to satisfy the conditions or capabilities of Assetwise. Even assuming, for argument sake, that RFC was estopped from enforcing certain provisions of the Client Guide, the Court concludes that Loans 5798390, 8257547, 10381337, and 6231616 were nonetheless in material breach.
394. One of the contested loans, Loan 4413350, was an AlterNet Guide loan that was also approved by Assetwise. Again, as an AlterNet Loan, the determination of breaches was not subject to RFC's sole discretion. However, Mr. Butler testified that the loan's Assetwise approval was invalid because the borrower's qualifying credit score of 548 was not used to generate the Assetwise approval, and it was the PRMI underwriter's responsibility to ensure Assetwise used the correct qualifying score. Butler 614:22-25, 616:10-19, 618:23-619:9. Furthermore, Mr. Butler concluded that the loan's actual credit grade under the AlterNet Guide was CM, which required a minimum FICO score of 540 and a maximum LTV ratio of 65%. Id. at 614:22-615:5, 615:9-24; accord PTX-001 at 208-211. The Court concludes that PRMI was not entitled to simply rely on Assetwise approval in the face of contrary evidence in the file that would invalidate that approval. A prudent underwriter would have questioned the discrepancy in credit scores before making the loan, which PRMI failed to do. Accordingly, the Court concludes that Loan 4413350 was in material breach of PRMI's Reps to RFC.
395. As to all of the Assetwise Loans, the Court need not rule on whether ResCap should be estopped from enforcing certain provisions of the Client Guide as to the five Assetwise Loans at issue, or whether ResCap waived its right to enforce certain provisions of the Client Guide.
396. Regarding estoppel, even if ResCap was so estopped, as PRMI argues, the evidence at trial showed that PRMI was in breach in any event with respect to each of the Assetwise Loans. With respect to each of the Assetwise Loans at issue, one or more of the following facts were established at trial: (1) the loans were Client Guide loans for which ResCap had sole discretion to determine breaches; (2) one or more of the conditions set forth in the initial Assetwise Approval were not met; (3) there was a breach of the Client Guide or AlterNet Guide that PRMI concedes was applicable to all loans, including Assetwise Loans; or (4) there was a borrower misrepresentation—which even PRMI's witnesses conceded was not permissible under any Guide.
397. As to waiver, it is undisputed that both the Client Guide and the AlterNet Guide required that any waiver be in writing. See PTX-001 § 250; PTX-032 § A200; PTX-1055 § A200; PRMI SJ Order, 428 F. Supp. 3d at 65. No evidence of any written waiver was introduced at trial.
C. Countrywide Loan
398. PRMI originally underwrote Loan 10226985 to Countrywide's guidelines, as it had acquired the loan in a Countrywide Pool. Crawford 2430:17-25. Mr. Butler identified a borrower misrepresentation with respect to this loan, as well as an excessive DTI ratio, unexplained credit inquiries, and an unreasonable stated income. Butler 588:17- 25. Ms. Keith did not dispute the borrower's misrepresentation or the resulting DTI ratio, and agreed that no set of guidelines would allow fraud. Keith 2252:14-16. PRMI's witnesses agreed that prohibiting misrepresentation in origination is a standard industry Rep required by all investors, including Countrywide. Crawford 2451:1-14; Flitton Dep. 67:20-24, 69:24-70:5, 89:2-16. Accordingly, based on the borrower's misrepresentations, the Court concludes that Loan 10226985 was in material breach.
399. As to the Countrywide Loan, the Court need not rule on whether ResCap should be estopped from enforcing certain provisions of the Client Guide, or whether ResCap waived its right to enforce certain provisions of the Client Guide. The evidence showed that the borrower's misrepresentations were impermissible under any set of guidelines, and there was no evidence of written waiver.
D. AlterNet Loans
400. The AlterNet Loans that Mr. Butler found to have materially breached PRMI's Reps to RFC were subject to the March 2000 Client Contract and the AlterNet Guide. PTX-001; PTX-003. Accordingly, they were not subject to ResCap's sole discretion to determine breaches. However, as set forth below, the Court concludes that ResCap has established that the AlterNet Loans materially breached PRMI's applicable Reps based on borrower misrepresentations or due to PRMI's failure to meet the AlterNet Guide's requirements in other respects. Mr. Butler determined that each of PRMI's breaches "affects the ability to repay and the ultimate repayment of that loan." Butler: 573: 2-4; see also Hawthorne 1057:17-1058:6.
See MASTR Adjustable Rate Mortgs. Tr. 2006-OA2 v. UBS Real Estate Sec. Inc., No. 12-cv-7322 (PKC), 2015 WL 764665, at *15 (S.D.N.Y. Jan. 9, 2015) (stating that to prove material and adverse effect, "[plaintiffs] may rely upon proof that as to a specific loan, there is a material or significant increase in the risk of loss"); Homeward Residential, Inc. v. Sand Canyon Corp., 298 F.R.D. 116, 131 (S.D.N.Y. 2014) (finding plaintiff can prove material and adverse effect by showing "increased credit risk"); Wells Fargo Bank, N.A. v. Bank of Am., N.A., No. 10-cv-9584 (JPO), 2013 WL 1285289, at *10 (S.D.N.Y. Mar. 28, 2013) (finding that to prove material and adverse effect, "the breach need only cause harm, though not necessarily default"), vacated on other grounds, 627 Fed. App'x 27 (2d Cir. 2015); Assured Guar. Mun. Corp. v. Flagstar Bank, FSB, 892 F. Supp. 2d 596, 603 (S.D.N.Y. 2012) (holding a breach causes a material and adverse effect if it significantly increases a loan's risk of loss); Syncora Guarantee Inc. v. EMC Mortg. Corp., 874 F. Supp. 2d 328, 335 (S.D.N.Y. 2012) (finding Syncora could establish a material and adverse effect by showing a breach increased its risk of loss on a certificate insurance policy, regardless of whether the breaches caused any loans to default).
401. For four of the AlterNet Loans (4375505, 4962418, 4117058, and 4380515), PRMI's sole contention was that Mr. Butler improperly utilized the 1997 AlterNet Guide to determine breaches of PRMI's Reps to RFC. As stated previously, the Court rejects PRMI's contention and concludes that Mr. Butler appropriately used the 1997 AlterNet Guide.
402. As noted, for four of the AlterNet Loans (4375505, 4962418, 4117058, and 4380515), Ms. Keith did not dispute Mr. Butler's findings of breach. Loan 4375505 was in breach for several reasons, including borrower misrepresentations. Butler 585:6-15. PRMI's witnesses agreed that a borrower's misrepresentation of debt would not be permitted under any guidelines. Crawford 2451:1-14; Flitton Dep. 67:20-24, 69:24-70:5, 89:2-16. Loan 4962418 likewise contained a borrower misrepresentation of employment—another violation under any set of guidelines. Butler 587:11-18, 588:12-13. Loan 4117058 was in breach for PRMI's failure to verify the borrower's housing history and because of the borrower's insufficient credit score. Id. at 590:8-12. Loan 4380515 involved a non-arm's length transaction, calling into question the veracity of the information obtained from the borrower, including the verification of employment and rental history. Id. at 591:16-592:4, 592:25-593:1. Again, Ms. Keith did not dispute these findings of breach. The Court concludes that ResCap has established that these four loans materially breached PRMI's Reps to RFC contained in the 1997 AlterNet Guide and the March 2000 Client Contract as follows:
• Loan 4375505 materially breached Sections 251-1(A), 251-1(B), 251-1(G), 251-1(T), 254, and 260 and Chapter 5 of the AlterNet Guide (PTX-001), and thus Sections 1 and 3 of the March 2000 Client Contract (PTX-003 at 1-2). See Butler 585:14-23, 586:6-587:10.
• Loan 4962418 materially breached Sections 251-1(A), 251-1(B), 251-1(G), 251-1(T), 254, and 260 of the AlterNet Guide (PTX-001), and thus Sections 1 and 3 of the March 2000 Client Contract (PTX-003 at 1-2). See Butler 587:11-588:6, 588:12-13.
• Loan 4117058 materially breached Sections 251-1(A), 251-1(B), 251-1(T), 254, 260, and 335(f) of the AlterNet Guide (PTX-001), and thus Sections 1 and 3 of the March 2000 Client Contract (PTX-003 at 1-2). See Butler 590:8-20, 591:11-13.
• Loan 4380515 materially breached Sections 251-1(A), 251-1(B), 251-1(T), 254, 260, and 312 of the AlterNet Guide (PTX-001), and thus Sections 1 and 3 of the March 2000 Client Contract (PTX-003 at 1-2). See Butler 591:16-592:4, 592:10-593:11.
403. The Court further concludes that ResCap has established that AlterNet Loans 4115211 and 4550405 also materially breached PRMI's Reps to RFC contained in the 1997 AlterNet Guide and the March 2000 Client Contract. For these loans, Ms. Keith argued that they were not subject to the 1997 AlterNet Guide and also disputed Mr. Butler's findings of breach.
Because Loan 4413350 was an AlterNet Loan that was also approved by Assetwise, the Court addressed it previously.
404. For the reasons set forth above, the Court concludes that Loans 4115211 and 4550405 were subject to the 1997 AlterNet Guide, and the evidence at trial established that PRMI materially breached the 1997 AlterNet Guide's requirements. Loan 4115211 breached the credit score requirement. Butler 610:20-611:10. And, as the Court previously found, because the borrower simply did not qualify for the loan, RFC's after-the-fact internal examination of this loan is irrelevant to the determination of breach. See id. at 614:4-11. Loan 4550405 was an impermissible flip transaction that was not prudently originated or of investment quality. Id. at 619:20-620:14, 621:16-20, 623:5-13. As noted, the Court has rejected Ms. Keith's speculation to the contrary. See Keith 2242:25-2243:17. Accordingly, the Court concludes that ResCap has undisputedly established that these two loans materially breached PRMI's Reps to RFC contained in the 1997 AlterNet Guide and the March 2000 Client Contract as follows:
• Loan 4115211 materially breached Sections 251-1(A), 251-1(B), 251-1(T), 254, and 260 and Chapter 5 of the AlterNet Guide (PTX-001), and thus Sections 1 and 3 of the March 2000 Client Contract (PTX-003 at 1-2). See Butler 613:3-614:11. Based on the Court's prior ruling that precluded reliance- or knowledge-based defenses, the Court finds that the due
diligence worksheet for loan 4115211, discussed earlier, is irrelevant to the above breaches of Reps. See PRMI SJ Order, 428 F. Supp. 3d at 120-21.
• Loan 4550405 materially breached Sections 251-1(A), 251-1(B), 251-1(T), 254, 260, and 473(F) of the AlterNet Guide (PTX-001), and thus Sections 1 and 3 of the March 2000 Client Contract (PTX-003 at 1-2). See Butler 619:20-620:14, 621:16-623:13, 862:16-863:23.
IV. PRMI'S BREACHES WERE A CONTRIBUTING CAUSE OF THE RISKS RFC SETTLED IN THE BANKRUPTCY SETTLEMENTS
A. Legal Standard
405. As this Court has already held, "ResCap and PRMI have, through the applicable Guides, contracted for a 'contributing cause' standard of causation and the Court is bound to enforce the contract's plain language." PRMI SJ Order, 428 F. Supp. 3d at 103 (citing Common SJ Order, 332 F. Supp. 3d at 1165); see also First Mortg., 2018 WL 6727065, at *10; Residential Funding Co. v. Universal Am. Mortg. Co., No. 13-cv-3519 (PAM/HB), 2018 WL 4955237, at *2 (D. Minn. Oct. 12, 2018). The Client Guide's relevant sections regarding indemnity—A212 and A202(II)—use the phrases "resulting from," "arising from," and "as a result of," which, under Minnesota law, are synonymous with contributing cause, not proximate cause. Common SJ Order, 332 F. Supp. 3d at 1164. Accordingly, "ResCap need only show that PRMI's breaches of any [Reps] were a contributing cause of the liabilities and losses settled in the Bankruptcy Settlements." PRMI SJ Order, 428 F. Supp. 3d at 103.
406. The "fairly liberal" causation standard that PRMI agreed to in the Client Guides "does not require ResCap to prove that each [PRMI] breach of its [Reps] caused RFC to breach a particular trust-level [Rep], which in turn caused the Trusts or Monolines to assert a claim against RFC relating back to that specific [PRMI] breach." Jan. 24, 2019 Pretrial Order [Doc. No. 215] at 7 n.2, First Mortg, No. 13-cv-3490 (SRN/HB). "To require such evidence would not only run afoul of a 'contributing cause' standard, but would inappropriately turn this trial into the very loan-by-loan repurchase trial obviated by the bankruptcy settlements." Id. Rather, it is sufficient that ResCap prove that PRMI's breaches increased RFC's risk of liability on the RMBS Trust and Monoline claims that were settled in the Bankruptcy. PRMI SJ Order, 428 F. Supp. 3d at 103. As set forth herein, ResCap has met that standard.
B. The Evidence
1. The Evidence From the Bankruptcy Establishes the Requisite Causal Connection
407. The evidence from the Bankruptcy establishes that the Settlements resolved claims resulting from, arising from, and incurred as a result of PRMI's breaches. The RMBS Trusts and Monolines asserted claims against RFC for origination defects (i.e. that the loans failed to comply with underwriting guidelines). See PTX-155 at 12-15; DTX-543 at 12-13; DTX-494 at 4; PTX-146 at 9-10; DTX-497 at 8; DTX-569 at 17-18; Pfeiffer Dep. at 126:1-11, 146:19-147:4, 257:14-20. Those claims were resolved in the Bankruptcy Settlements, which settled "RMBS R+W Claims" that "ar[ose] from any obligations or liability in respect of the origination or sale of mortgage loans to the RMBS Trusts." PTX-180 at 116.
408. PRMI was responsible for the origination defects in the loans it sold to RFC. See DTX-018 §§ A201(M), A202(A), A202(B), A202(G), A206, 401 ("The Client [PRMI] represents and warrants that all Loans sold to GMAC-RFC meet the eligibility and underwriting guidelines as set forth in this Guide," and PRMI "is responsible for the credit and property underwriting"). As the Court has found, the evidence at trial showed that numerous PRMI loans were defective and contrary to the Reps PRMI made to RFC. PRMI also acknowledged RFC's reliance on PRMI's false Reps in securitizing the loans and making its own Reps. See Common SJ Order, 332 F. Supp. 3d at 1181-82; see also DTX-018 at 38, § A200 ("The Client [PRMI] acknowledges that [RFC] purchases Loans in reliance upon the accuracy and truth of the Client's warranties and representations and upon the Client's compliance with the agreements, requirements, terms and conditions set forth in the Client Contract and this Client Guide.").
409. Moreover, all of the relevant parties in the Bankruptcy assessed RFC's potential repurchase liability for breaches of Trust Reps by examining origination defects (i.e., breaches of the underwriting guidelines). See Hawthorne 1048:9-25, 1049:19-23. PRMI did not present any evidence to the contrary. Rather, PRMI, through cross examination and argument, invited speculation that perhaps those parties were mistaken in looking to underwriting guidelines breaches. See id. at 1237:18-1241:15; PRMI Closing 2752:12-22. That speculation, however, is not supported by the evidence. See, e.g., Tommassello v. Stine, No. 8-cv-1190 (PJS/LIB), 2011 WL 383001, at *16 (D. Minn. Feb. 3, 2011) ("But this is speculation, [it is] not evidence."). The Court agrees with Mr. Hawthorne that it defies common sense to believe that the counsel and experts for RFC, the Committee, and the RMBS Trusts were all incorrect in concluding that the breaches of the underwriting guidelines were a proxy for breaches of the Trust Reps. Hawthorne 1241:7-15. To the contrary, the evidence at trial—including the expert testimony of Messrs. Hawthorne and Butler—demonstrates that it was reasonable for the parties in the Bankruptcy to equate underwriting guideline breaches with RMBS Trust breaches.
For example, PRMI's counsel argued that RFC's expert Mr. Sillman "made an assumption based on the few trusts he had time to look at that all trusts had a compliance rep." PRMI Closing 2752:16-22. But Mr. Sillman testified that his assumption was made after looking at not a "few" trusts, but rather "dozens and dozens of [trusts], maybe more," and that a guidelines representation appeared in "all of the ones that [he] reviewed." Sillman Dep. 145:5-17. That is pertinent given that, according to PRMI's witness, only 17 trusts had a Substantial Compliance Rep. Schwarcz 2018:1-8. Thus, Mr. Hawthorne's inference that Mr. Sillman treated other Trust Reps as warranting compliance with guidelines, Hawthorne 1239:6-14, is reasonable. It is also consistent with (1) the testimony of ResCap's witnesses that a Substantial Compliance Rep is not the only way to warrant compliance with guidelines, and (2) the fact that every reunderwriter in the Bankruptcy utilized guidelines breaches as a proxy for Trust Rep breaches. PRMI's speculation that Mr. Sillman simply made a mistaken assumption is unsupported by the record.
410. The evidence thus is clear that (i) the claims against RFC were centered on the quality of the mortgage loans, (ii) PRMI was contractually responsible for the quality of the loans it sold, and (iii) PRMI's breaches contributed to the liabilities that were resolved by RFC's Bankruptcy Settlements.
411. PRMI nonetheless has argued that ResCap failed to establish causation because the RMBS Trusts and Monolines did not expressly assert breaches of certain specific Trust Reps (e.g., the Credit Grade and Loan Program Reps) in their proofs of claim. That argument is factually erroneous and, in any event, tries to impose a heightened causation standard that is inconsistent with the Guides.
412. PRMI has not disputed that the Monolines asserted claims for breaches of Trust Reps based on underlying origination defects for which PRMI is responsible. See DTX-494 at 4; Lipps HLC 1230:25-1231:2; PTX-146 at 9-10; DTX-497 at 8; DTX-569 at 17-18. As to the RMBS Trusts, they asserted origination defects in the form of categories of Reps regarding the mortgage loans. See PTX-155 at 12-13, DTX-543 at 12. Those categories included Reps regarding the credit characteristics of the loans, underwriting criteria, prudent origination, and encompassed the very so-called "pool-wide" Reps that PRMI argues were never asserted. Id. The RMBS Trusts were not required to specify each particular Rep for each particular loan for which they were asserting claims, loan-by-loan and representation-by-representation. See In re DJK Residential LLC, 416 B.R. 100, 106 (Bankr. S.D.N.Y. 2009) (applying general federal pleading standards to proofs of claim). It was sufficient that the RMBS Trusts asserted multiple origination defects for which RFC was potentially liable, that PRMI in fact breached numerous Reps, and that RFC settled liability for origination breaches. Matching particular loans to particular Trust Reps was not necessary and would have improperly turned the Client Guide's liberal causation standard into a proxy for a loan-by-loan proximate cause analysis.
2. The Evidence Establishes that PRMI's Breaches Increased RFC's Risk of Liability to the RMBS Trusts and Monolines
413. In addition to the evidence from RFC's Bankruptcy, the undisputed testimony establishes that RFC faced a risk of liability as a result of PRMI's breaches. ResCap presented credible fact and expert testimony that through the Trust Reps, RFC incurred a risk of liability for PRMI's origination breaches. See, e.g., Butler 655:12-15, 660:14-662:10, 685:6-15, 937:15-21; Farley 141:23-142:10, 170:7-15, 200:24-201:8.
414. Although PRMI presented testimony from multiple experts regarding the interpretation of the Trust Reps, those experts did not refute the fact that the Trust Reps presented a risk of liability for RFC based on PRMI's guideline breaches. Rather, those experts either disclaimed to be offering an opinion regarding legal risk, see Keith 2344:8-11, 2352:13-18, 2355:17-2356:11; Schwarcz 2060:17-2062:18, 2060:7-16; Burnaman 1604:11-14, or they acknowledged that there was risk even if they did not think the risk was as significant as ResCap's witnesses, see Woll 1954:6-13. The relevant question for causation is whether PRMI's breaches of its Reps to RFC contributed to RFC's risk of liability, and none of PRMI's witnesses could refute that PRMI's breaches created at least some risk. Accordingly, PRMI's experts do not rebut ResCap's evidence on the issue of causation.
415. Because PRMI's witnesses cannot reasonably refute that PRMI's breaches caused at least some risk, the Court need not determine which party offers the better interpretation of the Trust Reps. Nevertheless, the Court concludes that ResCap's interpretations are more persuasive than PRMI's because they are supported by (1) the language and structure of the Trust Reps and transaction documents, and (2) the testimony of ResCap's witnesses regarding the scope of the Trust Reps, which the Court finds to be more persuasive than the testimony offered by PRMI's witnesses.
a. MLS Rep
416. As described above, ResCap and PRMI dispute the proper interpretation of the MLS Rep. ResCap contends that the MLS Rep warranted the substantive accuracy of information on the MLS, and PRMI contends it only warranted the accuracy of RFC's transcriptions. Compare Farley 163:6-9, 168:7-20; Butler 647:17-648:13, 651:1-10, with Schwarcz 2045:16-2046:7; Keith 2346:12-20. The Court concludes that the evidence supports Plaintiff's position.
417. First, the language and structure of the MLS Rep supports ResCap's position. The MLS Rep warrants that the "information set forth on the [MLS] is true and correct in all material respects." PTX-406 at 5. It does not limit the representation to mere transcription or contain any other qualifier. Rather, it represents the truthfulness of the information in all respects. Further, the MLS warrants the accuracy of the information "with respect to each Mortgage Loan." PTX-406 at 5. The MLS Rep is not warranting pool-level accuracy, but rather the substantive truthfulness of individual characteristics on a loan-by-loan basis. Moreover, the remedy for a breach of the MLS Rep is cure, repurchase, or substitution. The Court agrees with Ms. Farley's testimony that the repurchase remedy would not make sense if the MLS Rep merely warranted accurate "transcription" because the remedy for such an error would be merely fixing the transcription. Farley 168:7-20; see Maven Techs., LLC v. Vasile, 147 A.D.3d 1377, 1378 (N.Y. App. Div. 2017) ("[A] contract must be read as a whole to give effect and meaning to every term . . . [and] should be interpreted in a way [that] reconciles all [of] its provisions, if possible.") (citations and quotations omitted).
418. Second, the trial testimony supports ResCap's position. Ms. Farley established that it was RFC's understanding that the MLS Rep means that "the information on the [MLS] relating to each loan was true and correct," and that investors, underwriters, monoline insurers, and credit rating agencies relied on the substantive accuracy of the MLS. Farley 163:6-13; 172:9-174:19. Mr. Butler testified based on his industry experience and review of relevant evidence that the MLS Rep warranted substantive accuracy. Butler 647:17-648:13, 651:1-10. Mr. Hawthorne testified that RMBS plaintiffs as of the Settlement Period also viewed the MLS Rep as warranting substantive accuracy, and that those RMBS plaintiffs had cited case law to support that view. Hawthorne 1041:1-1043:4.
419. Third, ResCap's position is supported by persuasive authority both before and after the Settlement Period. See Deutsche Bank Nat'l Tr. Co. v. Morgan Stanley Mortg. Capital Holdings LLC, 289 F. Supp. 3d 484, 509-10 (S.D.N.Y. 2018) ("Like the majority of courts to have considered this exact issue, this Court agrees with [plaintiff's] interpretation."); U.S. Bank, Nat'l Ass'n v. UBS Real Estate Sec. Inc., 205 F. Supp. 3d 386, 429 (S.D.N.Y. 2016) ("MLS Warranty imposes a form of strict or absolute liability for a materially untrue or incorrect statement on the MLS."); Bank of N.Y. Mellon v. WMC Mortg., LLC, No. 654464/2012, 2013 WL 6153207, at *3 (N.Y. Sup. Ct. Nov. 21, 2013) ("The truth of the specific facts in the loan tape is the very thing the originator . . . warrants in the [MLS Rep]."), aff'd, 136 A.D.3d 1, 8 (N.Y. App. Div. 2015) (MLS Rep guaranteed the "veracity of information"), aff'd, 65 N.E.3d 1275 (N.Y. 2016); MBIA Ins. Corp. v. Countrywide Home Loans, Inc., No. 602825/08, 2013 WL 1845588, at *27-29 (N.Y. Sup. Ct. Apr. 29, 2013) (granting summary judgment to Plaintiff regarding MLS Rep breaches based on loan-level inaccuracies for 1,414 loans). In so holding, those Courts have rejected the argument, advanced by defendants there and PRMI here, that MLS Reps merely guarantee the accuracy of a sponsor's transcription of the information.
b. No-Default Rep
420. As described above, ResCap and PRMI dispute the proper interpretation of the No Default Rep. ResCap contends that the No Default Rep warranted against any default as defined in the relevant mortgage note or mortgage, and PRMI contends it only warranted against payment defaults. Compare ResCap Closing 2815:11-18, with PRMI Closing 2758:1-5. The Court concludes that the evidence supports ResCap's position.
421. First, the language and structure of the No Default Rep supports ResCap's position. The No Default Rep warrants that the "there is no material default, breach, violation or event of acceleration existing under any Mortgage Note or Mortgage and no event which, with notice and expiration of any grace or cure period, would constitute a material default, breach, violation or event of acceleration under the terms of any Mortgage Note or Mortgage." PTX-107 at 5. It does not limit the representation to only payment defaults. Rather, it represents there is no default, without limitation, under the terms of the underlying mortgage note or mortgage. The mortgage note and mortgage state that the borrower is in default if he or she provides false, misleading, or inaccurate information to the lender or makes a material omission. See, e.g., PTX-284 at 14. Indeed, one variation of the No Default Rep specifically excludes payment defaults as a breach of the No Default Rep. PTX-107 at 5.
422. Second, the trial testimony supports ResCap's position. Ms. Farley established that it was RFC's understanding that the No Default Rep meant there was no default under the mortgage note or mortgage, and that RMBS industry participants—such as monoline insurers and credit rating agencies—understood that the No Default Rep covered borrower misrepresentations. Farley 177:14-179:1, 179:18-21, 181:22-182:7, 206:23-207:7. Mr. Butler testified based on his industry experience and review of relevant evidence that the No Default Rep warranted that there was no default under the mortgage note or mortgage, including borrower misrepresentations. Butler 635:4-638:9. Mr. Hawthorne testified that RMBS plaintiffs as of the Settlement Period also viewed the No Default Rep as warranting against borrower misrepresentations, and that those RMBS plaintiffs had case law to support that view. Hawthorne 1031:23-1035:20.
423. Third, ResCap's position is supported by persuasive authority both before and after the Settlement Period. See Countrywide Home Loans, Inc., 2013 WL 1845588, at *23-26 (granting summary judgment as to existence of No Default Rep breaches based on borrower misrepresentations in 610 loans); Tr. for Certificate Holders of Merrill Lynch Mortg. Passthrough Certificates Series 1999-C1 v. Love Funding Corp., No. 04 Civ. 9890 (SAS), 2005 WL 2582177, at *6-7 (S.D.N.Y. Oct. 11, 2005) (holding defendant strictly liable under No Default Rep due to borrower fraud); see also MBIA Ins. Corp. v. Credit Suisse Sec. (USA) LLC, 165 A.D.3d 108, 115 (N.Y. App. Div. 2018) (holding scope of the No Default Rep was a jury issue, as to which the RMBS sponsor therefore faced risk of loss at trial). As those holdings demonstrate, the No Default Rep warrants the absence of any borrower default—not just a payment default as argued by defendants there, and PRMI here.
c. Loan Program and Credit Grade Reps
424. As described above, ResCap and PRMI dispute the proper interpretation of the Loan Program and Credit Grade Reps. ResCap contends that these representations warranted compliance with underwriting criteria and were enforceable on a loan-level basis, while PRMI contends they only warranted the aggregate characteristics of the securitized loan pool and could not be enforced as to individual loans. Compare Farley 186:9-16, 186:22-187:2; Butler 674:14-676:10, with Schwarcz 2007:12-24, 2150:23-2151:6; Burnaman 1564:19-1565:19; 1566:11-1567:1; Keith 2316:16-2317:11. The Court concludes that the evidence supports ResCap's position.
425. First, the language and structure of the Loan Program and Credit Grade Reps and the Trust Agreements supports ResCap's position that the Reps are enforceable loan-level representations warranting compliance with underwriting criteria. The Loan Program and Credit Grade Reps were in the loan-level representation section of the Trust Agreements. Farley 130:5-10, 147:8-14; see also, e.g., DTX-1016.0115 at 36-38 ("The Company hereby represents and warrants to the Trustee for the benefit of Certificateholders"); PTX-1150 at 3-9 ("RFC represents and warrants to the [Depositor], with respect to each Mortgage Loan"). The remedy for breach of the Loan Program or Credit Grade Reps was a loan-level remedy that operated on individual mortgage loans and required RFC to either cure, repurchase, or substitute individual mortgage loans. Farley 130:11-18, 147:15-22, 154:4-6, 161:19-23, see also, e.g., DTX-1016.0115 at 38 ("Upon discovery . . . of a breach of any of the [Reps] [the Depositor shall] purchase such Mortgage Loan"); PTX-1150 at 8-9 ("Upon discovery by RFC . . . of a breach of the foregoing [Reps] in respect of any Mortgage Loan . . . RFC shall . . . purchase such Mortgage Loan").
426. The loan-level repurchase remedy applied to the Loan Program and Credit Grade Reps without limitation, and was the sole remedy for breach of the representations. See Farley 185:15-187:2; Butler 674:14-25, 675:19-676:10; Hawthorne 1022:3-1023:23; Schwarcz 2133:15-25. Further, the Loan Program and Credit Grade Reps warranted that the loans were "underwritten under . . . loan documentation program[s]" and "classified . . . as Credit Grade[s]," respectively. DTX-1016.0115 at 37; PTX-1150 at 7. The loan documentation programs and Credit Grades were defined in the applicable underwriting guidelines. Hawthorne 1022:3-1024:5. The representations, therefore, warranted that the loans met the loan documentation program and Credit Grade criteria provided in the relevant guidelines, and were subject to a loan-level repurchase remedy if that criteria was breached.
427. The Court agrees with Ms. Farley's (and PRMI's experts') testimony that, if the Loan Program and Credit Grade Reps were construed as pool-wide representations with only pool-wide remedies, enforcement of the Loan Program and Credit Grade Reps would be "extraordinarily challenging," "highly inefficient," and "kind of impossible," effectively rendering the representations and repurchase remedy meaningless. Farley 187:3-189:1; Keith 2360:7-10; see Maven Techs., 147 A.D.3d at 1378 ("It is well settled that a contract must be read as a whole to give effect and meaning to every term" (quotation omitted)).
428. Second, the trial testimony supports Plaintiff's position. Ms. Farley established that it was RFC's understanding that the Loan Program and Credit Grade Reps were loan-level representations with loan-level remedies, and that investors and rating agencies relied on the Loan Program and Credit Grade Reps to "add teeth" to the disclosures in the Prospectus Supplements relating to loan documentation programs and credit grades, which were "key risk[s]" with respect to the RALI, RFMSI, and RASC securitization shelves. Farley 184:3-20, 185:15-21. Ms. Farley also established that the Loan Program and Credit Grade Rep statistics were calculated based on the categorization of individual loans, with the resulting percentages finely calibrated to the tenth or hundredth percentile, such that an individual loan, wrongly categorized, would impact the percentages. Id. at 158:21-159:3, 186:9-16, 186:22-187:2. This supports her testimony that the Loan Program and Credit Grade Reps related to the categorization of individual loans. Mr. Butler testified, based on his industry experience and review of relevant evidence, that the Loan Program and Credit Grade Reps were loan-level representations warranting compliance with the underwriting parameters, had a loan-level remedy, could be breached as to a single loan, and "would serve no purpose if . . . not to trigger a repurchase [] remedy." Butler 674:14-676:10, 676:5-10. Mr. Hawthorne testified that, as of the Settlement Period, a reasonable sponsor in RFC's position would have viewed the Loan Program and Credit Grade Reps as comparable to express guideline representations in terms of assessing risk of liability. Hawthorne 1027:2-1028:6.
429. Third, PRMI's failure to address the legal risk posed by the Loan Program and Credit Grade Reps supports Plaintiff's position that these representations did, in fact, create risk to RFC. PRMI's experts conceded that their interpretation of the Loan Program and Credit Grade Reps as pool-wide representations rendered them effectively unenforceable and meaningless, and so contended that the inclusion of the Loan Program and Credit Grade Reps in the loan-level representation section of the Trust Agreements (with a loan-level repurchase remedy) was a "mistake[]" and "poor drafting." Schwarcz 2007:12-24, 2025:16-2026:4; 2141:17-2142:18, 2150:23-2152:20; see also Burnaman 1564:19-1565:19, 1566:11-1567:1; Keith 2316:16-2318:2. But Professor Schwarcz disclaimed any opinion that RFC could have asserted a drafting mistake as a viable defense in litigation. Schwarcz 2168:2-2169:3. Nor could he have credibly asserted such an opinion, as mistake is not a defense where the interests of a third party good faith purchaser, such as a certificateholder, would be impacted. Restatement (Second) of Contracts § 155 (1981) ("Where a writing that evidences or embodies an agreement in whole or in part fails to express the agreement because of a mistake of both parties as to the contents or effect of the writing, the court may at the request of a party reform the writing to express the agreement, except to the extent that rights of third parties such as good faith purchasers for value will be unfairly affected.") (emphasis added); see also Abramson v. Nelson, 116 N.W.2d 405, 409 (Minn. 1962). Further, parties may not assert unilateral mistakes as basis for reforming a contract. Nichols v. Shelard Nat. Bank, 294 N.W.2d 730, 734 (Minn. 1980) (finding it would be unfair to allow a party to benefit from its own mistaken understanding because it would "destroy" the counter-party's "right to rely on [] written assent to the agreement").
d. Monoline Reps
430. As described above, ResCap and PRMI dispute the proper interpretation of the Monoline Reps. Plaintiff contends that the Monoline Reps warranted to the Monolines that the Trust Reps and statements in the Prospectuses and Prospectus Supplements were accurate and not misleading. Butler 695:9-698:12, 698:16-701:20. PRMI does not dispute that the Monoline Reps passed through the Trust Reps, but does contend that the Monoline Rep warranting the accuracy of the Prospectus and Prospectus Supplement statements was only a "10(b)-5" representation and cannot be enforced on a loan-by-loan basis. Schwarcz 2053:13-2054:24, 2171:2-2173:18. The Court concludes that the evidence supports Plaintiff's position.
431. First, the language and structure of the Monoline Reps support ResCap's position. The Monoline Reps warranted that the Prospectuses and Prospectus Supplements "d[id] not contain any untrue statement of a material fact." PTX-1091 at 10. The Prospectus Supplements, in turn, warranted that the loans complied with underwriting guidelines. PTX-1281 at 31. The Monoline Reps were not limited to 10(b)-5 claims. Rather, they represented that the Trust Reps and statements in the Prospectuses and Prospectus Supplements were accurate and not misleading. Indeed, the monoline insurance agreements contained remedies for breach of the Monoline Reps. PTX-1091 at 22, 28-30.
432. Second, the trial testimony supports Plaintiff's position. Mr. Butler testified based on his industry experience and review of relevant evidence that the Monoline Reps passed through the Trust Reps and also warranted the truth of statements in the offering documents, including those warranting compliance with RFC's underwriting standards. See Butler 695:9-698:12.
e. Fraud Disclaimer
433. At trial, ResCap and PRMI also disputed the proper interpretation of the Fraud Disclaimer. Plaintiff contended that the Fraud Disclaimer was not a "silver bullet" defense and did not cover misrepresentation, Farley 200:11-201:22, while PRMI maintained that it applied to both fraud and misrepresentation. Burnaman 1582:18-1583:5, 1620:15-24; Schwarcz 2122:16-2126:4; Keith 2332:8-13. The Court concludes that the evidence supports Plaintiff's position.
434. First, the language of the Fraud Disclaimer supports ResCap's position. The Fraud Disclaimer provided that RFC "shall not be required to [repurchase] Mortgage Loans . . . if the substance of [the] breach . . . also constitutes fraud in the origination of [the] Mortgage Loan." PTX-023 at 9 (emphasis added). The Fraud Disclaimer did not purport to exculpate RFC from breaches based on misrepresentations.
435. Second, the trial testimony supports Plaintiff's position. Ms. Farley established that it was RFC's understanding that the Fraud Disclaimer "was not a silver bullet" against claims premised on borrower fraud or misrepresentations, did not cover misrepresentations, and did not eliminate RFC's risk of liability for fraud. Farley 200:11-201:22, 202:7-21. Mr. Butler testified based on his industry experience and review of relevant evidence that the Fraud Disclaimer did not cover misrepresentations and would not have eliminated RFC's "risk for repurchase" if borrower fraud was identified. Butler 662:14-25, 665:11-19. Mr. Hawthorne testified that a reasonable defendant in RFC's position would not have placed much weight on the Fraud Disclaimer. Hawthorne 1047:18-23, 1048:7-8.
436. Third, ResCap's position is supported by case law. Fraud requires clear and convincing evidence of intent. Crigger v. Fahnestock & Co., 443 F.3d 230, 234 (2d Cir. 2006). A misrepresentation is not fraud without proof of intent, and therefore not ostensibly covered by the Fraud Disclaimer.
437. Fourth, PRMI's failure to address the legal efficacy of the Fraud Disclaimer supports Plaintiff's position that the Fraud Disclaimer did not eliminate RFC's risk of liability for borrower fraud and misrepresentation. See Schwarcz 2116:23-2118:3, 2127:24-2129:7; Burnaman 1620:11-1622:13. As described above, PRMI's experts disclaimed any opinion that the Fraud Disclaimer was a viable defense in litigation or eliminated RFC's risk of liability for claims premised on borrower fraud or misrepresentation. Burnaman 1621:13-16; Keith: 2363:20-25; Schwarcz 2128:17-2129:1. RFC could not have asserted that it securitized fraudulent loans without significant reputational risk, and moreover, could not have attempted to prove fraud as a defense without also establishing the existence of misrepresentations, thereby increasing its potential liability. Farley 200:24-201:22; Butler 663:19-664:11, 665:11-19, 937:15-21; Hawthorne 1029:14-20, 1045:11-1047:24. PRMI's argument that RFC could have admitted to securitizing fraudulent loans as a defense is not credible or persuasive.
f. Fraud and Substantial Compliance Reps
438. ResCap and PRMI also disputed the import of the Fraud and Substantial Compliance Reps. PRMI contended that the Fraud Rep was the only way RFC could warrant against fraud and misrepresentation, and its existence in some securitizations implied that other securitizations did not contain a warranty relating to fraud or misrepresentation. See Burnaman 1573:7-22, Schwarcz 2037:20-2038:12; Keith 2326:13-2327:17. Similarly, PRMI argued that the Substantial Compliance Rep was the only way RFC could warrant against guideline breaches, and its existence in some securitizations implied that other securitizations did not contain a warranty relating to compliance with underwriting guidelines. See Burnaman 1561:12-1562:9; Schwarcz 2190:17-2191:1; Keith 2311:19-2313:3. Plaintiff maintained that other representations could warrant against fraud, misrepresentation, or guideline breaches, and that a warranty against those underwriting defects did not require the specific formulation of the Fraud and Substantial Compliance Reps. See Farley 196:17-197:10, 199:14-202:6; Butler 642:2-7, 640:20-23, 650:13-23, 694:13-19; Hawthorne 1030:18-1031:4. The Court concludes that the evidence supports Plaintiff's position.
439. First, the language and interpretation of the various other Trust Reps supports ResCap's position. The MLS and No Default Reps were broader than PRMI's experts contend—for example, both of those representations covered borrower misrepresentations. PTX-406 at 5; DTX-786 at 15; PTX-1150 at 6; DTX-786 at 12; Butler 635:4-638:9, 647:17-648:13, 651:1-10; Farley 177:14-179:1; Hawthorne 1031:23-1035:20. Thus, the assertion by PRMI's experts that the Fraud Rep is the only way to warrant against fraud or misrepresentation is not persuasive. Similarly, the Loan Program and Credit Grade Reps were enforceable loan-level representations with loan-level remedies that required compliance with underwriting criteria. See Farley 185:22-186:8, 186:9-16, 186:22-187:2; Butler 674:14-676:10; Hawthorne 1022:3-1024:5. Therefore, PRMI's assertion that the Substantial Compliance Rep was the only way to warrant against guideline breaches is also not persuasive. There was nothing about the Trust Agreements lacking the specific Fraud and Substantial Compliance Reps that limited the application of other Trust Reps to underwriting defects arising from fraud, misrepresentation, or guideline breaches.
440. Second, the trial testimony supported Plaintiff's position. Ms. Farley established that it was not RFC's understanding that the only way to warrant against fraud and misrepresentation or guideline breaches was with the Fraud and Substantial Compliance Reps, respectively. Farley 196:17-197:10, 199:14-202:6. Ms. Farley also established that fraud, misrepresentation, and guideline breaches could have breached any number of Trust Reps. Id. Mr. Butler and Mr. Hawthorne testified to the same effect. See Butler 640:20-23, 642:2-10, 650:13-23; Hawthorne 1030:18-1031:4.
V. RESCAP'S DAMAGES METHODOLOGY IS REASONABLY CERTAIN AND NON-SPECULATIVE
A. The Law
441. This is a claim for contractual indemnity. In an action on a contract, the plaintiff bears the burden of proving damages to a reasonable degree of certainty. Everyday Learning Corp. v. Larson, 242 F.3d 815, 819 (8th Cir. 2001) (citing N. Cent. Co. v. Phelps Aero, Inc., 139 N.W.2d 258, 263 (Minn. 1965)); see also Common SJ Order, 332 F. Supp. 3d at 1191. "[D]amages which are speculative, remote, or conjectural are not recoverable." Brown v. Diversified Distribution Sys., LLC, 801 F.3d 901, 910 (8th Cir. 2015) (quoting Leoni v. Bemis Co., 255 N.W.2d 824, 826 (Minn. 1977)). However, "[t]he law does not require mathematical precision." Poppler v. Wright Hennepin Coop. Elec. Ass'n, 834 N.W.2d 527, 546 (Minn. Ct. App. 2013), aff'd, 845 N.W.2d 168 (Minn. 2014); see also Common SJ Order, 332 F. Supp. 3d at 1203 ("To prove damages, a plaintiff must offer a reasonably certain, though not necessarily mathematically precise, basis to demonstrate who owes what for its claims."). "Once the fact of loss has been shown, the difficulty of proving its amount will not preclude recovery so long as there is proof of a reasonable basis upon which to approximate the amount." Children's Broad. Corp. v. Walt Disney Co., 245 F.3d 1008, 1016 (8th Cir. 2001) (quoting Leoni, 255 N.W.2d at 826); see also Poppler, 834 N.W.2d at 546 (same); Common SJ Order, 332 F. Supp. 3d at 1191.
442. The foregoing principles of Minnesota law apply where, as here, damages must be allocated among multiple defendants. As with damages generally, a plaintiff "need not prove allocation with precision." UnitedHealth Grp. Inc. v. Exec. Risk Specialty Ins., 870 F.3d 856, 863 (8th Cir. 2017); see also PRMI SJ Order, 428 F. Supp. 3d at 135, 143, 147. Rather, an allocation need only be reasonable and nonspeculative. UnitedHealth Grp. Inc., 870 F.3d at 863; see also RSUI Indemn. Co. v. New Horizon Kids Quest, Inc., 933 F.3d 960, 966 (8th Cir. 2019) (directing the district court, on remand, "to allocate 'as best it can' [an] unallocated jury award between covered and uncovered claims," without indicating which party bears burden of allocation); Krause v. Merickel, 344 N.W.2d 398, 403 (Minn. 1984) ("[W]e have affirmed settlement allocations if they appear reasonable in light of the total award to the plaintiffs . . . and not patently arbitrary.") (quotation and citation omitted); PRMI SJ Order, 428 F. Supp. 3d at 135, 143.
443. PRMI has relied almost exclusively on UnitedHealth throughout this litigation to challenge ResCap's damages methodology. UnitedHealth, however, does not alter long-standing Minnesota law on the allocation of damages, or on the calculation of damages generally. As this Court previously held in the Consolidated Action:
UnitedHealth does not impose as rigid a standard for assessing contract damages or mandate that a specific formula be applied to allocate them among multiple parties, as Defendants argue. Rather, UnitedHealth stands for the proposition that an "insured . . . must present a non-speculative basis to allocate a settlement between covered and non-covered claims," but "need not prove allocation with precision." 870 F.3d at 863. Thus, UnitedHealth's holding regarding an insured's responsibility to sort covered claims from non-covered claims in a multi-party settlement is entirely consistent with the long standing requirements of Minnesota law of contract damages. To prove damages, a plaintiff must offer a reasonably certain, though not necessarily mathematically precise, basis to demonstrate who owes what for its claims. See [Eklund v. Vincent Brass & Aluminum Co., 351 N.W.2d 371, 379 (Minn. Ct. App. 1984)]; Poppler, 834 N.W.2d at 546-47; [Hydra-Mac, Inc. v. Onan Corp., 450 N.W.2d 913, 921 (Minn. 1990)]; Leoni, 255 N.W.2d at 826.Common SJ Order, 332 F. Supp. 3d at 1203
444. Moreover, the circumstances of this case are very different from those the Eighth Circuit considered in UnitedHealth. See Common SJ Order, 332 F. Supp. 3d at 1203-04.
445. First, UnitedHealth was an insurance coverage case, involving an innocent insurer, in which the court had to decide whether the settled claims were covered by the applicable policy. See UnitedHealth Grp. Inc. v. Columbia Cas. Co., 47 F. Supp. 3d 863, 865 (D. Minn. 2014). Here, by contrast, ResCap seeks contractual indemnification from PRMI for PRMI's wrongful conduct in breaching its contractual obligations to RFC. The allocation question before the Court is how to apportion liability among PRMI and other breaching originators who also sold defective loans to RFC.
446. Second, although UnitedHealth involved insurance coverage, it did not involve the allocation of a settlement among multiple insurers. In cases requiring the allocation of damages among multiple insurers, the Minnesota Supreme Court has recognized "that damages are by nature fact-dependent and that trial courts must be given the flexibility to apportion them in a manner befitting each case." N. States Power Co. v. Fid. & Cas. Co. of N.Y., 523 N.W.2d 657, 663 (Minn. 1994). In Northern States, the Minnesota Supreme Court approved of a "pro rata by time on the risk" allocation approach among multiple insurers. Id.
447. Third, UnitedHealth predominately involved unrelated ERISA and antitrust claims from two separate cases in different jurisdictions. See UnitedHealth, 47 F. Supp. 3d at 868, 878 (noting that the two settled litigations were each a "separate case . . . before a different judge in a different district," and each involved separate defendants). By contrast, the Settlements at issue here involved related claims asserted in a single action—RFC's Bankruptcy—and principally involved breaches of Reps by PRMI and the sale of defective loans by PRMI and other originators.
448. Fourth, in UnitedHealth, the parties to the litigation did not allocate the settlement, and thus the court in the subsequent coverage litigation would have been required to do so from scratch. Id. at 867. Here, however, as discussed further below, the parties to the Bankruptcy did allocate the Settlements among RMBS Trusts and Monolines, and the Bankruptcy Court approved that allocation.
449. Fifth, unlike the insured in UnitedHealth, id. at 885-88, ResCap presented substantial evidence and expert opinions in support of its proposed allocation, including the testimony of multiple witnesses and experts as to (i) the Reps PRMI made to RFC, and that RFC made to the RMBS Trusts and Monolines, (ii) the breaches of those Reps, and PRMI's relative breach rate as compared to those of other originators, and (iii) the strength of the claims and defenses. Moreover, ResCap provided the Court with its Allocated Breaching Loss Approach, which provided a fair, practical, reasonable, and non-speculative way to allocate damages among more than 80 different originators.
B. The Allocated Breaching Loss Approach
450. ResCap's Allocated Breaching Loss Approach satisfies the requirements of Minnesota law for the determination of damages, and for the allocation of settlements among multiple defendants. As set forth above, ResCap's Allocated Breaching Loss Approach is reasonable and non-speculative because: (1) it allocates the Bankruptcy Settlements based on objective, measurable standards—each defendant's breach rate and the losses associated with those breaches, as adjusted to take into account the discount that RFC was able to negotiate in its Bankruptcy, (2) it is consistent with the approach used by the parties to actually allocate in the Bankruptcy, which approach was incorporated into the Plan and was approved by the Bankruptcy Court; and (3) it is consistent with the approach used in every other major RMBS settlement and approved by courts for each of those settlements. Altering Dr. Snow's methodology, as advocated by PRMI, would be infeasible, potentially unfair, and would inject speculation and false precision into the determination of damages.
As such, the "relative value" of each defendant's claims and defenses was effectively, and practically, subsumed in the allocation. See UnitedHealth, 870 F.3d at 865 (stating that allocation in that case required "either contemporaneous evidence of valuation or expert testimony on relative value to provide a reasonable foundation for a jury's decision.").
1. Objective Standard Based on Losses and Relative Breach Rates
451. The Allocated Breaching Loss Approach is a reliable methodology upon which the Court can determine PRMI's liability. See Common SJ Order, 332 F. Supp. 3d at 1204. As described earlier, the model allocates liability among PRMI and more than eighty other originators pro rata based on (1) the amount of breaching loans each sold to RFC, (2) each originator's relative breach rate, as compared to the breach rate for a global sample of loans, and (3) the discount that RFC achieved in its Settlements. Each of these criteria is concrete and verifiable. Given these features, the Court finds that the Allocated Breaching Loss Approach provides a reliable, non-speculative basis for calculating damages in this case.
2. Consistent With the Approach Used in Bankruptcy
452. ResCap's Allocated Breaching Loss Approach is consistent with the breaching loss allocation approach used by the parties in RFC's Bankruptcy to allocate among the RMBS Trusts. As noted, this approach was incorporated into the Plan proposed by RFC and the Committee, and was approved by the Bankruptcy Court as being fair and reasonable to each of the RMBS Trusts and their investors. See PTX-153 at 5; PTX-179 ¶¶ 115, 178; PTX-180 at 34-36, 145-51. As with the Allocated Breaching Loss Approach, the Bankruptcy allocation apportioned liability based upon relative breaching losses, without considering additional factors such as purported differences in the availability of certain defenses, such as the statute of limitations or other issues raised by PRMI in this case. See Hawthorne 1102:6-1104:23.
453. The Court concludes that the Allocated Breaching Loss Approach is reasonable and non-speculative even if there had been no allocation of the Settlements in the Bankruptcy. The fact that the Allocated Breaching Loss Approach is consistent with the actual Bankruptcy allocation, however, further supports the Court's conclusion. That is particularly so considering the involvement of numerous parties and professionals in formulating the methodology in the Bankruptcy, the lack of any objections by the investors with an economic stake in the allocation, and the approval by the Bankruptcy Court.
454. The Court finds it significant that the Committee, which represented the interests of all unsecured creditors in RFC's Bankruptcy, supported the breaching loss allocation. Indeed, the Committee had objected to the Original RMBS Trust Settlement including, specifically, the proposed allocation because, at that time, the allocation was based solely on net losses. See PTX 179 ¶¶ 101-03, 115; Hawthorne 1098:16-1099:17. The Committee argued that such an allocation was deficient because it purportedly failed to account for certain differences between the RMBS Trusts. Id. The Committee, however, did not object to the revised breaching loss allocation incorporated in the final RMBS Trust Settlement, and indeed jointly proposed the Plan (including the RMBS Trust Settlement and allocation) for approval by the Bankruptcy Court. See Lipps HLC 1281:21-1282:10; PTX-180 at 82, 140-51.
455. Similarly, certain RMBS Trust investors that had objected to the net-loss allocation proposed as part of the Original RMBS Trust Settlement withdrew those objections in light of the breaching loss allocation in the final RMBS Trust Settlement. See Hawthorne 1101:20-1102:5, 1307:22-25. In formulating his opinion, Mr. Woll failed to consider these facts. Woll 1906:12-1907:24.
456. Notably, most reported decisions that consider allocations of settlements involve situations where the parties to the settlement did not themselves allocate, or where evidence of the parties' allocation had not been presented to the court. See, e.g., UnitedHealth, 870 F.3d at 862-63 ("[B]ecause there was no allocation of damage monies . . . that would establish the amount Bor-Son paid toward the settlement of that loss of rental claim, and since Bor-Son, in this case, furnished no evidence to establish that fact, Bor-Son failed to meet its burden of proving its claim for reimbursement" (quoting Bor-Son Bldg. Corp. v. Emp'rs Commercial Union Ins. Co. of Am., 323 N.W.2d 58, 64 (Minn. 1982))); UnitedHealth Grp. Inc. v. Columbia Cas. Co., 941 F. Supp. 2d 1029, 1033-34 (D. Minn. 2013) ("Bor-Son thus implies that a contemporaneous allocation is not the only method by which an insured can establish that it incurred a covered loss."); UnitedHealth, 47 F. Supp. 3d at 875 n.11 (noting that the insured had suggested the possibility of introducing evidence "as to how the Settlement proceeds were actually distributed to members of the settlement class" as evidence in support of an allocation, but failed to do so). Here, however, there was extensive evidence presented at trial regarding the breaching loss allocation that was done in the Bankruptcy, including credible testimony from Allen Pfeiffer, and the Plan itself. Pfeiffer Dep. 35:1-15, 52:7-54:9, 56:01-25; PTX-180 at 82, 144-51.
457. PRMI has argued that the allocation in RFC's Bankruptcy is irrelevant because it was a "plaintiff-side" allocation among the RMBS Trusts. See, e.g., Woll 1904:15-25. The Court disagrees.
458. First, PRMI's argument is erroneous as a factual matter. The Bankruptcy allocation was part of the RMBS Trust Settlement set forth in the Plan proposed by RFC and the Committee as joint Plan proponents. PTX-180 at 82, 140-51; PTX-153 at 5; PTX-179 ¶ 115; see also Hawthorne 1305:12-1306:14 (confirming that "the debtors and the committee propose[d] the allocation to be approved and confirmed by Judge Glenn"); Woll 1923:5-1924:20. Thus, the Bankruptcy allocation was not a "plaintiff-side" allocation—it was an allocation that was advocated by all sides and approved by the Bankruptcy Court.
459. Second, PRMI's argument is based on a misreading of UnitedHealth. UnitedHealth states that "[t]he allocation inquiry examines how a reasonable party in UnitedHealth's position would have valued the covered and non-covered claims." 870 F.3d at 863. That does not mean, however, that how other parties in the litigation may have valued the claims is irrelevant. Indeed, the next sentence in UnitedHealth states that the Court should consider what the "parties" knew at the time of the settlement. Id. UnitedHealth also makes clear that the Court may consider a wide panoply of evidence including "testimony from attorneys involved in the underlying lawsuits, evidence from those lawsuits, expert testimony evaluating the lawsuits, a review of the underlying transcripts, or other admissible evidence." Id. Certainly, proposed allocations of the settlement are relevant. Evidence of the Bankruptcy allocation is highly probative, because it was the RMBS Trust investors who were receiving the distributions and who, as a result, had the most at stake in ensuring the allocation was fair and reasonable.
460. PRMI also has argued that the Bankruptcy allocation is not relevant because it was not contemporaneous with the Settlements. PRMI is incorrect. UnitedHealth holds that "[e]vents and circumstances happening after settlement are relevant only insofar as they inform how a reasonable party would have valued and allocated the claims at the time of settlement." 870 F.3d at 864. Here, the evidence shows that the breaching loss approach used in the Bankruptcy was developed prior to the parties' execution of a Plan Support Agreement in May 2013. PTX-153 at 5; PTX-179 ¶ 115; see also Hawthorne 1305:12-1307:21. That the parties may have adjusted allocation amounts after that date does not undo the fact that the principles underlying the allocation had already been developed. Moreover, the Bankruptcy Settlements were not entered into on a single date, but rather involved a series of settlement agreements with different entities over a period of a few months, and which were approved by the Bankruptcy Court in December 2013. See PTX-179 ¶¶ 71-83; see also Am. Prairie Constr. Co. v. Hoich, 594 F.3d 1015, 1024 (8th Cir. 2010) ("It is a recognized principle of bankruptcy law that a bankruptcy court is required to approve any compromise or settlement proposed in the course of a Chapter 11 reorganization before such compromise or settlement can be deemed effective.").
3. Consistent with Allocations in Other Court-Approved RMBS Settlements
461. As ResCap's expert Mr. Hawthorne testified, every other major RMBS trust settlement both before and after RFC's settlements has allocated damages among trusts based on losses or net losses. Hawthorne 1102:6-1104.9. No settlement has allocated based on purported differences among trusts in the strength of a statute of limitations defense, the applicable Trust Reps, or any other factors. Id. at 1103:2-5.
462. Judge Kyle in Ramsey County District Court recently approved an allocation based on net losses in an RMBS litigation over the objection of certain investors. See In re MASTR Adjustable Rate Mortgs. Tr. 2006-OA2, No. 62-TR-CV-18-35, slip op. at 53 (Minn. Dist. Ct. Jan. 3, 2020). Judge Kyle explained that the allocation based on net losses "is straight-forward and objective and is not susceptible to competing interpretations regarding complex legal or factual issues . . . ." Id. Judge Kyle concluded that the allocation methodology was reasonable and cited to several courts that had "approved the use of estimated lifetime losses as an allocation methodology in multi-trust RMBS settlements." Id. (citing In re Bank of N.Y. Mellon, No. 651786/2011, 2014 WL 1057187 (N.Y. Sup. Ct. Jan. 31, 2014), aff'd as modified sub nom., In re Bank of N.Y. Mellon, 127 A.D.3d 120 (N.Y. App. Div. 2015); In re U.S. Bank Nat'l Ass'n, 27 N.Y.S.3d 797 (N.Y. Sup. Ct. 2015); [U.S. Bank Nat'l Ass'n v. Fed. Home Loan of Boston] ("JPM Article 77"), No. 652382/2014, 2016 WL 9110399 (N.Y. Sup. Ct. Aug. 12, 2016)).
4. PRMI's Objections to the Allocated Breaching Loss Approach Invite Unfairness, Speculation, and False Precision
463. PRMI asserts that ResCap's Allocated Breaching Loss Approach is unreasonable because it does not consider purported differences among the RMBS Trusts in (1) the strength of the applicable Trust Reps and (2) the strength of a potential statute of limitations defense. See McCrary 2509:25-2511:1, 2685:7-19, 2686:20-2687:5. PRMI further asserts that Dr. Snow's decision to sample only from the At-Issue Loans was flawed. See id. at 2550:22-25, 2551:1-3.
464. As discussed earlier, PRMI's assertions, and the evidence offered in support of those assertions, do not alter the Court's conclusion that ResCap's Allocated Breaching Loss Approach is reasonable, fair, and non-speculative, and thus satisfies the requirements of Minnesota law.
465. As the Court previously held, "Plaintiff need not factor every single difference in trust representations and the strength of certain defenses into its allocation in order to" satisfy the requirements of Minnesota law for damage and allocation. PRMI SJ Order, 428 F. Supp. 3d at 143 (citing Common SJ Order, 332 F. Supp. at 1203-04; Univ. Am. Mortg. Co., LLC, 2018 WL 4955237, at *4; First Mortg., 2018 WL 6727065, at *7-9). The alleged differences among the RMBS Trusts that PRMI proffered at trial are too speculative to alter the Court's view.
466. The testimony of ResCap's witnesses—including Mr. Hawthorne, Ms. Farley, and Mr. Butler—established that it was reasonable for the Allocated Breaching Loss Approach to not adjust for these purported differences. PRMI's experts could not refute that the Trust Reps at issue posed risk to RFC for origination breaches, and the evidence showed that in many cases there was no material difference in the risks posed by one Rep as opposed to another.
467. Moreover, PRMI's proposed interpretations of the Trust Reps at issue are inconsistent with the language and structure of the documents, and PRMI's experts could not reconcile or explain the inconsistencies. See, e.g., Schwarcz 2025:16-2026:4, 2031:19-2032:2, 2141:17-2142:18, 2149:24-2152:9, 2163:15-2164:18.
468. The allocation in the Bankruptcy did not differentiate based on the purported differences in the strength of the Trust Reps. Hawthorne 1096:7-18. Moreover, as Mr. Hawthorne testified, other major RMBS trust settlements also involved trusts with varying Reps, but the allocations of those settlements did not account for those variances. See id. at 1102:22-1103:5, 1104:18-23.
469. Finally, the Court agrees with Mr. Hawthorne that any attempt to statistically account for any difference in the level of risk posed to certain Trust Reps would be highly speculative and an exercise in false precision. Hawthorne 1081:16-18, 1081:24-1082:5. In fact, a hypothetical question posed to Mr. Hawthorne at trial that tried to account for just a few of the issues illustrated that assigning numerical values to the full set of potential differences among trusts would not be reliable. Id. at 1087:8-1091:20. No PRMI witness disputed Mr. Hawthorne's testimony, nor did any witness for PRMI offer any possible statistical method to account for such differences.
470. Notably, throughout the case, PRMI never proposed an alternative allocation model. Nor did PRMI's witnesses offer any testimony that it would be feasible to make a fair, reasonable, or non-speculative allocation based on the issues that PRMI identified. Although PRMI did not have the burden to prove damages, the fact that PRMI's witnesses could not articulate an alternative damages model further illustrates the reasonableness of ResCap's approach. On summary judgment, this Court invited PRMI to "offer non-speculative evidence relevant to the effect, if any, of the strength and weaknesses of the trust representations and certain defenses on ResCap's damages allocation." PRMI SJ Order, 428 F. Supp. 3d at 144. Although PRMI did offer testimony regarding the Trust Reps, it failed to offer any explanation as to how that testimony could be translated into a workable damages allocation that could satisfy the requirements of Minnesota law.
471. As noted, PRMI also asserted that ResCap's allocation failed to account for purported differences in the strength of the statute of limitations defense against the RMBS Trusts. For many of the same reasons, this argument does not alter the Court's conclusions.
472. PRMI's focus solely on the statute of limitations defense ignored that there were other issues and defenses being litigated in RMBS cases as of the Settlement Period. As discussed earlier, there were multiple defenses that a party in RFC's position could have asserted as of the Settlement Period, and in fact RFC did assert a number of them. Mr. Hawthorne's testimony established that many of these issues and defenses could vary among loans or RMBS Trusts. PRMI, however, never addressed these other issues and defenses, but rather focused solely on the defense that favored them—the statute of limitations defense. PRMI's cherry-picking of one defense that may disproportionately benefit its position does not render ResCap's Allocated Breaching Loss Approach unreasonable. To the contrary, it illustrates the objective benefits of the Allocated Breaching Loss Approach, which relies on basic objective criteria—breaches and losses. As Judge Kyle recently held in upholding an allocation based on net losses, attempting to consider multiple additional criteria is "susceptible to competing interpretations regarding complex legal or factual issues." In re MASTR, No. 62-TR-CV-18-35, slip op. at 53. This Court agrees, and thus believes that ResCap's allocation based on breaching losses is reasonable.
473. Moreover, as with the differences in the Trust Reps, neither the allocation in RFC's Bankruptcy nor in any other major RMBS settlements accounted for differences in statute of limitations or other legal defenses. Hawthorne 1102:6-1104.9.
474. Finally, as Mr. Hawthorne testified, trying to assign numerical values to legal defenses is speculative and leads to false precision. Id. at 1081:16-1082:5.
475. The Court also rejects PRMI's contention that Dr. Snow's allocation as to the Ambac Settlement was flawed. See PRMI Closing 2791:4-19. PRMI argued that (i) Ambac only asserted an "origination based proof of claim" as to one RMBS Trust and only servicing claims as to others, (ii) PRMI's loans were not in that RMBS Trust, and (iii) Ambac never filed an amended proof of claim asserting origination claims as to the other trusts. Id.
476. Ambac did not need to file an amended proof of claim to negotiate with RFC over additional origination claims it could assert during the confidential mediation. See Hawthorne 1194:9-15. The settlement and the Bankruptcy Court's order approving it reflected that Ambac was asserting not less than $223,301,022 in unliquidated origination claims beyond those asserted in its proofs of claim. PTX-171 at 5. Ambac's claims were asserted as to all of its "Policies," not just on one transaction. Id. at 2. Absent the settlement, Ambac could have sought to amend its claim and argue that any amendments related back to its timely claim or timely claims filed by the RMBS Trustees. Hawthorne 1094:23-1096:2; see In re Haugen Constr. Servs., Inc., 876 F.2d 681, 682 (8th Cir. 1989) (noting "[g]reat liberality in permitting amendments of claims in bankruptcy"); In re Texaco, Inc., 218 B.R. 1, 7 (Bankr. S.D.N.Y. 1998) (same); see also McHale v. Anthony, 839 N.Y.S.2d 33 (N.Y. App. Div. 2007) (noting insurer's subrogated claim related back to insured's original lawsuit for statute of limitations purposes).
477. As to whether the sample on which Dr. Snow based his damages model was properly chosen and defined, the Court concludes "that a proper 'universe' was examined and a representative sample was chosen." Lutheran Mut. Life Ins. Co. v. United States, 816 F.2d 376, 378 (8th Cir. 1987). As the Court has previously explained, Dr. Snow's decision to sample from the At-Issue Loans was sensible, "given that the purpose of his study [was] to allocate the bankruptcy claims among [d]efendants, and those claims are premised on losses to loans sold by RFC. Thus, conceptually, those damages would necessarily have flowed from the loans that actually experienced economic losses, i.e. the At-Issue Loans." PRMI Daubert Order, 432 F. Supp. 3d at 917 (quoting Common Daubert Order, 2018 WL 4489685, at *5). Dr. Snow's methodology also properly excluded performing loans, as he presented scenario analyses indicating that including performing loans would not have had a meaningful impact on damages. Snow 1392:21-1393:7. And, Dr. Snow's methodology also properly excluded NDS Trust loans, as his analysis found nothing to suggest that the breach rates would have materially differed, it was not possible for him to obtain a reunderwriting breach rate for the NDS Trusts, and, in a quantitative scenario analysis, even assuming the highest NDS Trust breach rate, he found the total impact on damages was minimal. Id. at 1397:11-18, 1398:11-19, 1400:16-1401:4.
478. As to PRMI's criticisms of Dr. Snow's calculation of damages, the Court finds that Dr. Snow's calculation of damages was reasonable and non-speculative. The calculation satisfies Minnesota law, which does not require absolute precision, but "reasonable exactness" in ascertaining a non-speculative amount of damages. Faust v. Parrott, 270 N.W. 2d 117, 120 (Minn. 1978).
479. As discussed earlier, Dr. Snow's allocation methodology generates a point estimate of approximately $5.4 million, with a confidence interval and a margin of error. The Court finds that the margins of error associated with Dr. Snow's calculated breach rates are reasonable. See, e.g., Snow 1407:12-14 (10.8% for the PRMI-specific trust breach rate), 1408:5-7 (7.2% for the global trust breach rate), 1477:23-1478:1 (19.1% for the PRMI-specific monoline breach rate), 1478:24-1479:4 (12.9% for the global monoline breach rate); cf. Fed. Hous. Fin. Agency v. JPMorgan Chase & Co., 2012 WL 6000885, at *9 (S.D.N.Y. Dec. 3, 2012) (accepting potential margins of error between 10% and 19.2%). Given that the cost to narrow the margins of error "would have dwarfed Plaintiff's total damages claim . . . . Dr. Snow was entitled to balance the benefit of any additional precision against its cost." PRMI Daubert Order, 432 F. Supp. 3d at 919 (citations and quotations omitted). Finally, the point estimate of $5.4 million, not the lower bound, is the appropriate measure of damages. See United States v. Bell, 602 F.3d 1074, 1085 (9th Cir. 2010) (remanding for recalculation where district court accounted for statistical uncertainty by subtracting the confidence interval where "[t]here was no evidentiary basis for preferring values at the lower bound of the margin of error").
480. Based on the foregoing, Plaintiff has established reasonable, non-speculative damages for its indemnification claim.
ORDER
ResCap has successfully proven the elements of its contractual indemnification claim. Accordingly, based on all the files, records, and proceedings herein, IT IS HEREBY ORDERED that:
1. Judgment on claim for contractual indemnification be entered in favor of Plaintiff, and against Defendant Primary Residential Mortgage, Inc. in the matter 16-cv-4070;
2. Defendant shall pay Plaintiff $5.4 million in damages; and
3. Within 10 days, the parties shall meet and confer regarding a briefing schedule on Plaintiff's motion for an award of attorney's fees, costs, and pre-judgment interest, and shall inform the Court of the schedule forthwith.
LET JUDGMENT BE ENTERED ACCORDINGLY.
Dated: August 14, 2020
s/Susan Richard Nelson
SUSAN RICHARD NELSON
United States District Judge