From Casetext: Smarter Legal Research

MBIA Ins. Corp. v. Countrywide Home Loans, Inc.

Supreme Court, New York County, New York.
Dec 22, 2010
30 Misc. 3d 1201 (N.Y. Sup. Ct. 2010)


No. 602825/08.


MBIA INSURANCE CORPORATION, Plaintiff, v. COUNTRYWIDE HOME LOANS, INC., Countrywide Securities Corp., and Countrywide Financial Corp., Defendants.

Peter E. Calamari, Philippe Z. Selendy, Manisha M. Sheth, Nicholas F. Joseph and Eve S. Moskowitz of Quinn Emanuel Urquhart & Sulllivan, LLP for Plaintiff. Mark Holland, Christopher J. Garvey, Abigail K. Hemani, Larkin M. Morton, Paul F. Ware, Jr. and Sarah Heaton Concannon of Goodwin Procter LLP for Defendants.

Peter E. Calamari, Philippe Z. Selendy, Manisha M. Sheth, Nicholas F. Joseph and Eve S. Moskowitz of Quinn Emanuel Urquhart & Sulllivan, LLP for Plaintiff. Mark Holland, Christopher J. Garvey, Abigail K. Hemani, Larkin M. Morton, Paul F. Ware, Jr. and Sarah Heaton Concannon of Goodwin Procter LLP for Defendants.

Plaintiff MBIA Insurance Corporation (“MBIA”) moves in limine for a decision allowing MBIA to use statistical sampling to present evidence to prove its causes of action for fraud and breach of contract and to prove damages. Defendants Countrywide Home Loans, Inc., Countrywide Securities Corp. and Countrywide Financial Corp. (collectively, “Countrywide”) oppose.


The facts of this matter have been discussed extensively in previous decisions of this court. Thus, only details necessary to this motion are referenced herein.

Plaintiff filed the instant motion on April 30, 2010. After oral argument on June 16, 2010, the parties held the motion pending additional discussions. The parties filed supplemental memoranda of law of law in August 2010. An evidentiary hearing was held on September 27, 2010, at which Plaintiff presented its expert witness, statistician Charles D. Cowan, Ph.D., for direct and cross-examination. Dr. Cowan testified about his proposed method of sampling the fifteen residential mortgage-backed securitizations (“RMBS”) at issue in this matter. By request of the court, the parties submitted additional arguments by letter on October 13, 2010, the final submissions of this motion.


I.Timeliness of Plaintiff's Motion

A. New York Statute and Code Does Not Preclude the Instant Motion

Defendants first contest Plaintiff's motion on grounds of timeliness. Defendants claim that the motion is premature, that the timing of the instant motion is not contemplated by New York law and that New York courts have held that it is premature to rule on the admissibility of evidence prior to the determination of that evidence's relevance. Defendants' arguments are unavailing.

First, the governing rules of this court do not mandate an outside time limit for a movant to initiate a motion in limine. The Rules of the Commercial Division of the Supreme Court state that:

The parties shall make all motions in limine no later than ten days prior to the scheduled pre-trial conference date, and the motions shall be returnable on the date of the pre-trial conference, unless otherwise directed by the court.
22 NYCRR 202.70, Rule 27.

Plaintiff's motion was made rather well in advance of the pre-trial conference date, and this court has not mandated other times or limits for motions in limine. While the majority of motions in limine are made close to or during trial, neither New York statute nor code prevents a party from bringing a motion as their litigation strategy dictates. Neither does New York statute or code prevent the court from deciding that motion. B. New York Common Law Does Not Preclude the Instant Motion

Defendants' argument regarding the timing of pretrial conferences under 22 NYCRR 202.26 is not relevant to the issue at bar, and is disregarded.

Defendants cite second to New York cases Speed v. Avis Rent–A–Car, 172 A.D.2d 267 (1st Dep't 1991) and Grant v. Richard, 222 A.D.2d 1014 (4th Dep't 1995) for the proposition that the instant motion must be decided closer to or at time of trial. Defendants' Memorandum of Law in Opposition to Plaintiff's Motion In Limine Regarding Sampling (“Defendants' Opp. Memo.”), pp. 8–10. Both Speed and Grant address the admissibility of specific items of evidence. The court in Speed found that the relevance of the particular evidence at issue, notice of recall of the automobile central to the case, was best determined at trial. Speed, 172 A.D.2d at 268.Grant, relying on Speed, similarly found that deciding the question of admissibility of unspecified evidence and testimony “should await the trial, when the determination may be made in context.' “ Grant, 222 A.D.2d at 1014, quoting Speed, 172 A.D.2d at 268.

The court does not find these cases persuasive. The court sees no question, and Defendants do not argue, that proof regarding the securitizations and Plaintiff's contention of breach and/or fraud therein is relevant to this action. Plaintiff will present evidence to attempt to prove its contentions, and the trier of fact will hear that evidence and make its decision based thereon. The question is not of admissibility and relevance, but of the form that the relevant and admissible evidence will take at trial. New York common law does not prevent decision upon the current motion. C. Alleged Factual Issues Do Not Preclude the Instant Motion

Defendants further contend that legal and factual issues prevent decision upon Plaintiff's motion. Supplemental Memorandum of Law in opposition to Plaintiff's Motion In Limine Regarding Sampling (“Defendants' Supp. Opp. Memo.”), pp. 3–11.

Defendants first allege numerous “threshold disputed issues,” id. at. 4–6, which they contend must be resolved prior to the use of sampling. Defendants contend that should the court grant Plaintiff's motion, the court would then improperly “resolve myriad threshold legal questions not fully briefed by the parties or properly before the Court.” Id. at 4.

While the finder of fact may need to resolve the listed issues, Defendants provide no basis for its current contention that resolution must occur prior to decision upon the instant motion or that decision upon this motion would resolve the listed issues. Defendants do not argue how resolution, or non-resolution, of any of their purported issues will be affected by a statistically significant sampling of the securitizations at issue, nor do Defendants attempt to link the list of issues with their other arguments. Defendants simply state that if Plaintiff does not prevail upon all of the listed issues “MBIA's proposal will not work.” Id. at 6. Conversely, Defendants do not argue or show how a sample will be affected by resolution of the alleged issues.

Defendants' footnoted federal case, in which a magistrate judge denied a motion in limine to exclude a criminal record, is unpersuasive Id. at 3, n. 4. As stated above, this motion is not premature and enough is known about the shape of trial to allow this motion. Defendants' list of purported issues may be resolved upon summary judgment or at trial.

Second, Defendants argue that Plaintiff has failed to demonstrate how sampling will be useful at trial. Defendants claims that Plaintiff has not shown how it may use sampling to prove its claims for breach of contract and fraud or how it may use sampling to prove its alleged damages. Defendants contend that Plaintiff has failed to show explain what elements of its claims it will show through sampling or how sampling will prove those elements.

The use of sampling does not obviate Plaintiff's need to prove each element of its claims for breach of contract or fraud, and Plaintiff must prove entitlement to any damages. Should sampling be used, Plaintiff retains its obligation to demonstrate to the trier of fact that each element of each cause of action has been met. Plaintiff's possible use of sampling does not change Plaintiff's ultimate burden of proof, only how Plaintiff may present that proof.

Plaintiff has generally stated how it will use sampling of the securitizations at issue to demonstrate its claims for breach of contract and fraud as well as for damages. Plaintiff's Supplemental Reply Memorandum of Law in Further Support of Motion In Limine Regarding Sampling (“Plaintiff's Supp. Memo.”), pp. 4–8; Memorandum of Law in Support of Plaintiff's Motion in Limine Regarding Sampling (“Plaintiff's Memo.”), pp. 14–19. Plaintiff must prove its claims and actual damages. Defendants will have the ability to contest Plaintiff's proofs, and the trier of fact will decide the issues. Defendant has cited no authority to force Plaintiff to divulge its specific litigation strategy to prove its claims, nor any authority that Plaintiff should be required to do so in order to utilize sampling. Defendants' argument is unsupported and unavailing.

The court finds that Plaintiff's motion is timely and ready for decision on its merits.

II.Methodology of Plaintiff's Proposed Sampling Method

Defendants argue that Plaintiff's proposed sampling is improper for the case at bar and is methodologically infirm. A. Sampling is Appropriate for the Case at Bar

As a preliminary matter, Defendants first argue that Plaintiff has not shown that sampling is appropriate for this matter. Defendants' Supp. Opp. Memo., p. 12; citing People v. Wesley, 83 N.Y.2d 417, 436 (1994) (Kaye, Chief Judge, concurring). Defendants appear to misread Wesley.

Defendants cite to former Chief Judge Kaye's concurring opinion in Wesley. In relevant part, Chief Judge Kaye reiterates the primary opinion's holding that, in each inquiry as to the validity of scientific evidence, upon finding general acceptance of a scientific technique a foundational inquiry must then be completed to determine whether the generally accepted scientific techniques were actually used. Wesley, 83 N.Y.2d at 436. While Chief Judge Kaye stated that this must be done “in each case,” Chief Judge Kaye did not imply or state that scientific evidence was to be applied only in certain cases. Rather, she stated that the review was to ensure that when a generally accepted technique was used, the technique in question was correctly applied. Id. If the foundational inquiry revealed “infirmities in collection and analysis of the evidence not affecting its trustworthiness” then the weight of the evidence was affected—not its admissibility. Id. The evidence is then to be weighed by the trier of fact.

The admission of scientific evidence is not limited to certain cases. As mortgage-backed securities are a relatively new device, it is not surprising that Plaintiff did not cite a directly analogous New York case allowing the use of statistical evidence. It is also not surprising that Defendants' do not cite a case disallowing such evidence. Scientific evidence is to reviewed of its own accord. See, generally, Frye v. United States, 293 F. 1013 (D.C.Cir.1923); Wesley, supra, and it will be so reviewed here. B. Plaintiff's Methodology is Acceptable

Plaintiff proposes to use expert testimony based upon a statistically significant sample of the residential mortgage-backed securities.

New York uses the “general acceptance” test of the reliability and admissibility of expert testimony using scientific evidence. Frye, supra; see Nonnan v. City of New York, 32 AD3d 91, 102 n. 18 (1st Dep't 2006). Frye requires that expert testimony be based on scientific principal or procedure that has been “sufficiently established to have gained general acceptance in the particular field in which it belongs.” Frye, 293 F. at 1114.

The court must first determine whether the proffered scientific evidence at issue in the matter, here, statistical sampling, is novel. See People v. Wernick, 89 N.Y.2d 111, 115–16 (1996) citing Wesley, 83 N.Y.2d at 422. Should statistical sampling be determined novel, then the court must find whether the sampling is generally accepted in the particular field in which it belongs. Frye, 293 F. at 1114. The court must also determine whether the scientific evidence is reliable. Parker v. Mobil Oil Corp., 7 NY3d 434, 446–47 (2006); Wesley, 83 N.Y.2d at 422, 424. 1.Statistical Sampling Is Not Novel

Statistical sampling is a widely used method to present evidence from a large population of data. For example, the United States Court of Appeals for the Second Circuit has used statistical sampling to show gender-based salary disparity. Lavin–McEleney v. Marist College, 239 F.3d 476 (2d Cir2001) (allowing for both liability and damages); and numerous other courts have similarly accepted the use of statistical sampling, including our own Court of Appeals in 1856. See Muller v. Eno, 14 N.Y. 597, 598–99 (1856) (allowing sampling of “several” of fourteen bales of cotton to be used as evidence as to the state of all fourteen bales); see also Ratanasen v. State of Cal., Dept. of Health Services, 11 F.3d 1467 (9th Cir1993) (allowing random sampling to prove fraud, “provided the aggrieved party has an opportunity to rebut such evidence”); Evans v. Fenty, 714 F.Supp.2d 116 (D.D.C.2010) (allowing the use of statistically significant samples, over objection by expert witness); CBS Broadcasting v. Primetime 24 Joint Venture, 48 F.Supp.2d 1342 (S.D.Fla.1998) (approving the use of stratified sampling). Further, New York statute expressly recognizes the use of statistical sampling in certain administrative hearings. See, e.g., Mercy Hosp. of Watertown v. New York State Dept. of Social Servs., 79 N.Y.2d 197, 203–04 (1992).

The court finds that statistical sampling of large populations is not a novel concept. 2.Statistical Sampling Is Generally Accepted in the Scientific Community

It is undisputed that the use of statistical sampling is generally accepted in the scientific community. The court notes that while Defendants contest Plaintiff's specific methodology, Defendants contest neither the novelty nor the general acceptance in the scientific field of statistical sampling. As Plaintiff's expert, Dr. Cowan, provides, acceptance of statistical sampling is widespread. See Affidavit of Dr. Charles D. Cowan in Further Support of Plaintiff's Motion In Limine Regarding Sampling (“Cowan Aff.”), ¶ 36. Scientific literature and testing is replete with the use of statistical sampling; the validity of properly conducted sampling is not a question for debate. Sampling has also been used in examining pools of loans, id., pp. 14–14, Ex. 3, a fact that Defendants do not contest.

Statistical sampling is not novel and is generally accepted. The court thus focuses upon whether Plaintiff's proposed sampling methodology is reliable. Wesley, 83 N.Y.2d at 422. 3.Plaintiff's Methodology is Acceptable

Plaintiff has explained its proposed methodology in open court, in Plaintiff's Supplemental Memorandum of Law in Support of Motion In Limine Regarding Sampling and in Dr. Cowan's affidavit. Defendants put forward significant opposition to Plaintiff's methodology in its own memoranda of law as well as the Affidavit of Dr. Christopher M. James (“James Aff.) and the Affidavit of Roy Welsch, Ph.D. (“Welsch Aff.”).

In sum, Plaintiff has stated that for each of the fifteen securitizations here at issue it will take the following actions: First, for each securitization Plaintiff's method will select 400 loans. Plaintiff asserts that a sample of 400 loans per population will provide a confidence level of approximately 95%, with a 5% margin of error. Plaintiff's Supp. Memo., p. 17; Cowan Aff. ¶¶ 56–58. Plaintiff's method will stratify the samples of 400 loans selected into mutually exclusive subgroups. Plaintiff's Supp. Memo., p. 18–19; see Cowan Aff. ¶¶ 59–61. Plaintiff asserts that doing so will ensure that subgroups in each securitization will be properly represented in the sample. Plaintiff will stratify the loan samples for Plaintiff's fraud claim, insurance contract claim and repurchase contract through the use three variables: the borrower's credit score, combined loan to value ratio (“CLTV”) and Countrywide's documentation program). Each of these stratum will then be divided into further subgroups. Plaintiff's Supp. Memo., p. 20; Cowan Aff. ¶¶ 69, 71, 74–78. Plaintiff will stratify the loan samples for Plaintiff's servicing contract claim and implied covenant claim using delinquency status. Cowan Aff., ¶ 86.

Defendants respond with detailed criticism of Plaintiff's proposed methodology. Defendants assert, through Drs. Welsch and James, that Plaintiff's methodology flawed. Defendants state, inter alia, that Plaintiff has failed to establish that the proposed samples are of sufficient size to be reliable, that the method relies on the assumption that the outcome of the sampled loans will be binary (i.e., either “yes” or “no” answer with regard to breach), and that the representations and warranties at issue contain subjective elements and may not be easily reviewed for compliance. Defendants' Supp. Memo., p. 15–16, citing Welsch Aff. and James Aff. Defendants further assert that Plaintiff's proposed methodology to extrapolate results from the proposed sample to prove liability and damages is flawed. Id. at 16–21. Defendants' assertions, and the assertions of their experts, are not without merit.

The court finds that Plaintiff's proposed methodology of statistical sampling may be used at trial. Plaintiff has shown its methodology to be scientifically accepted, valid and reliable under the Frye standard. However, in so deciding, the court makes no finding that Plaintiff's proposed method is the only method by which Plaintiff (or Defendant) may present evidence or that Plaintiff's method is without flaw or unsusceptible to challenge. Defendants have raised significant valid challenges to Plaintiff's methodology; however the court finds that such challenges are premature to decide here. Defendants cited issues will be decided by the trier of fact as pertaining to the weight, rather than the acceptability, of the evidence. That decision will be made at trial.

The court notes that Defendants have argued that the instant motion “effectively attempt[s] to co-opt the Court as co-counsel for Plaintiff.” Defendants' Supp. Opp. Memo., p. 3. Defendants thus imply that should this decision not proceed as they would prefer, the court has then merely “assist[ed Plaintiff] in devising its sample” and has acted inappropriately in its rule as overseer, and has become advocate. Defendants further state that for its opposition to fail is this court's “blessing,” Defendants' Supp. Opp. Memo., pp. 2, 4, of Plaintiff's sampling motion. Defendants' hyperbole is more than troubling to the court, as the court would prefer that all parties focus upon the merits of the arguments before it, rather than impugning any part in the arguments to the court. The court welcomes questions, conferences and motions should Defendants have issues with what the court considers its own utmost impartiality.


This court has inherent power to regulate trial before it ( seeCPLR 4011 [“court may ... regulate the conduct of the trial in order to achieve a speedy and unprejudiced disposition of the matters at issue”] ). Deciding the current motion is completed within the court's inherent power. The court notes that the decision of the motion in limine has the possibility of saving the parties and the court from significant litigation time and may significantly streamline the action without compromising either party from proving its case. See Orange and Rockland Utilities, Inc. v. Assessor of Town of Haverstraw, 12 Misc.3d 1194(A), *6 n. 37 (N.Y. Sup.Ct., Richmond Co.2006). The court does not find any prejudice in deciding the motion before it and allowing the use of statistically significant samples of the securitizations at issue. No authority prevents the court from doing so, and the use of sampling is widespread as a valid method to prove cases with large amounts of underlying data.

As Plaintiff may present its case as it chooses, so may Defendants rebut Plaintiff's proffered arguments through Defendants own sampling chosen in a statistically valid manner. As is the nature of and proper in litigation, each party will then challenge the other's proofs. The ultimate trier of fact will then decide the issue. The court expects both Plaintiff and Defendants to be forthcoming to the other party in notifying and providing evidence of chosen loans in any sample undertaken to effectuate the fairness of these challenges.

Finally, while Plaintiff is permitted to present evidence as it so chooses, and the court will permit Plaintiff to present evidence of its claims through its chosen sampling methodology, the court does not necessarily endorse the Plaintiff's method as better or worse than any other method. Plaintiff must convince the trier of fact, whether it be a jury or this court, that it has proven each element of its separate claims. Plaintiff must then further show the trier of fact that the sample chosen is actually statistically significant and is applicable to each securitization as a whole. Proof of damages must then follow. Sampling itself is not proof, but merely a vehicle to present evidence.

Accordingly, it is

ORDERED that MBIA's motion in limine to be allowed to use and present evidence for its case through statistical sampling is granted.

Summaries of

MBIA Ins. Corp. v. Countrywide Home Loans, Inc.

Supreme Court, New York County, New York.
Dec 22, 2010
30 Misc. 3d 1201 (N.Y. Sup. Ct. 2010)
Case details for

MBIA Ins. Corp. v. Countrywide Home Loans, Inc.

Case Details


Court:Supreme Court, New York County, New York.

Date published: Dec 22, 2010


30 Misc. 3d 1201 (N.Y. Sup. Ct. 2010)
958 N.Y.S.2d 647
2010 N.Y. Slip Op. 52239

Citing Cases

In re RFC & RESCAP Liquidating Tr. Action

(Pls.' Reply at 26–27 n.24 [Doc. No. 3909].)And Plaintiffs claim that monoline insurers asserted…

Royal Park Invs. SA/NA v. HSBC Bank USA Nat'l Ass'n

They involve a monoline plaintiff, which is not limited to a repurchase remedy. See, e.g., Assured Guar. Mun.…