FTC Capital GMBH v. Credit Suisse Group AG

UNITED STATES DISTRICT COURT

SOUTHERN DISTRICT OF NEW YORK

---------------------------------------X

IN RE LIBOR-BASED FINANCIAL INSTRUMENTS

ANTITRUST LITIGATION MEMORANDUM AND ORDER

---------------------------------------X 11 MDL 2262 (NRB)

THIS DOCUMENT RELATES TO: ALL CASES

---------------------------------------X

NAOMI REICE BUCHWALD

UNITED STATES DISTRICT JUDGE

Pursuant to the July 5, 2022 scheduling order, the parties

are currently engaged in discovery relating to two upstream issues:

(1) the alleged existence of a 16-bank conspiracy to persistently

suppress LIBOR, and (2) the effect of the Second Circuit’s decision

in United States v. Connolly, 24 F. 4th 821 (2d Cir. 2022).1 See

ECF No. 3425. To date, defendants have collectively produced over

1 In Connolly, the Second Circuit reversed the defendants’ convictions for

conspiracy to commit wire fraud and bank fraud, concluding that the Government

failed to prove that the trader-influenced LIBOR submissions were false on the

Government’s theory that “there was only one true interest rate that [each bank]

could submit.” 24 F.4th at 836, 843. The Second Circuit explained: “The

precise hypothetical question to which the LIBOR submitters were responding was

at what interest rate ‘could’ [the bank] borrow a typical amount of cash if it

were to seek interbank offers and were to accept. If the rate submitted is one

that the bank could request, be offered, and accept, the submission,

irrespective of its motivation, would not be false.” Id. at 835 (emphasis

added).

On April 4, 2022, the Court instructed the parties to focus their initial

discovery efforts on the effect of Connolly as well as the existence of a 16-

bank conspiracy to persistently suppress LIBOR, given the potential impact of

the anticipated summary judgment motions related to those issues. See ECF No.

3386 at 1. Once discovery and motion practice related to the two upstream

issues and OTC class certification are complete, the parties will, if necessary,

turn to whether plaintiffs suffered injuries as a result of the alleged LIBOR

suppression, whether plaintiffs had notice of and relied on an allegedly

inaccurate LIBOR, and any other remaining issues (the “downstream issues”).

See ECF No. 3425 at 7.

3.4 million documents, spanning 15.9 million pages and 88,000 audio

files, from over 500 custodians, generally from a period of August

2007 to May 2010. See ECF Nos. 3557 at 2; 3559 at 1. For the most

part, these documents were previously produced to various government

regulators, including the Department of Justice, Securities and

Exchange Commission, Commodity Futures Trading Commission, and New

York Department of Financial Services, in connection with their

investigations into parallel allegations of LIBOR manipulation at

issue here.2 See ECF Nos. 3549 at 1, 3559 at 1, 5.

On October 6, 2022, the Court received two applications to

compel the production of additional documents. See ECF Nos. 3547,

3549. In the first application, the Direct Action Plaintiffs’

(the “DAPs”) seek an order compelling all defendants to produce

documents from an additional 17-month time period of June 2010

through October 2011 that relate to the two upstream issues. See

ECF No. 3547. In the second application, all plaintiffs seek an

order compelling 11 bank defendants to produce additional

documents relating to the two upstream issues that are responsive

2 The Exchange-Based, OTC, and Lender plaintiffs previously acknowledged that

the regulatory investigations were “based on the same conduct underlying

plaintiffs’ civil claims” and “directly relate to whether the Panel Banks

engaged in unlawful manipulation of LIBOR and the quantum of that manipulation.”

ECF No. 1415 at 2, 4. The DAPs have also described the regulatory productions

as “undeniably relevant to [their] claims because they go to such core issues

as who was involved in LIBOR manipulation, over what time period, by what means,

to what extent, and the like.” ECF No. 1410 at 1. Defendants state that the

government regulators provided input into the search terms that were used for

their regulatory productions, which “were constructed to identify documents

related to the LIBOR-setting process and the suppression of LIBOR.” ECF No.

3559 at 5.

to more than 100 additional search terms and/or are from more than

40 additional custodians.3 See ECF No. 3549. On December 15,

2022, the Court requested additional information necessary to

conduct a Rule 26(b)(1) analysis of the second application, see

ECF No. 3603, which was provided by the parties on January 12,

2023 and January 19, 2023, see ECF Nos. 3624, 3626, 3632.4

For the reasons stated below, the first application for

documents from June 2010 through October 2011 is denied in its

entirety, and the second application for additional search terms

and custodians is granted in part and denied in part.

BACKGROUND

The Court assumes familiarity with the factual allegations in

this MDL, and only provides the background necessary to resolve

the instant applications.

A. DAPs’ Application to Compel Documents from June 2010 through

October 2011

In the first application of October 6, 2022, the DAPs

acknowledge that the Court previously denied certain DAPs’ requests

3 The 11 bank defendants from whom plaintiffs seek relief on this application

are: Bank of America, Citibank, Deutsche Bank AG, HSBC, JPMorgan, Norinchukin,

Portigon AG, Royal Bank of Canada, Royal Bank of Scotland, Société Générale,

and UBS AG. See ECF Nos. 3549, 3626. Plaintiffs’ application does not seek any

relief with respect to the other five bank defendants—Barclays, Lloyds, Credit

Suisse, Rabobank, and MUFG—all of whom negotiated agreements with plaintiffs on

this application. See ECF Nos. 3559 at 1 n.1, 3567, 3571, 3597, 3648.

4 As plaintiffs have sufficiently described the motions they wish to bring and

the reasons therefore in their applications of October 6, 2022 and their

supplemental submissions of January 12, 2023 and defendants have responded,

formal motions would be superfluous. See ECF Nos. 3547, 3549, 3624, 3626, 3632.

for leave to amend their complaints to extend the end of the

alleged suppression period from May 2010 to October 2011. See ECF

3457 at 1 (citing In re LIBOR-Based Fin. Instruments Antitrust

Litig., No. 11 MDL 2262 (NRB), 2019 WL 1331830, at *18 (S.D.N.Y.

Mar. 25, 2019) (“LIBOR VIII”)). Nevertheless, the DAPs argue that

an order compelling all defendants to collect and produce all non-

privileged, responsive documents relating to the two upstream

issues from all of their custodians for the additional 17-month

period of June 2010 to October 2011 is warranted because they have

discovered new evidence showing defendants suppressed LIBOR after

May 2010. See id.

The DAPs specifically cite to: (i) two reports from the

Federal Bureau of Investigation (“FBI”) used as exhibits in the

Connolly trial which reference recorded conversations from June

28, 2011 of a Deutsche Bank trader discussing banks’ reluctance to

“move LIBOR submissions up to the correct levels” and from November

23, 2010 of a Deutsche Bank trader stating that there is a “strong

incentive” among all USD LIBOR setters to keep LIBOR low, see ECF

Nos. 3547-1, 3547-2; (ii) an August 2011 research note stating

that banks continue to submit lower LIBOR rates than what they are

paying in the market, see ECF No. 3547-3; and (iii) an economic

diagram seemingly created by the DAPs that is based on a figure

previously used in another plaintiff’s complaint to allegedly

“show[] the magnitude of suppression when compared to an

alternative benchmark—the Eurodollar Deposit Rate published by the

Federal Reserve,” see ECF No. 3547 at 4.5 Based on this “new

evidence,” the DAPs assert that documents from the additional 17-

month period would show the existence of a conspiracy to suppress

LIBOR through 2011, establish that defendants knowingly submitted

false answers to the LIBOR question, and be relevant to the

parties’ expert economists’ models by clarifying whether June 2010

through October 2011 was a “clean period” that can be used as a

comparison to the alleged suppression period. See id. at 2.

Defendants filed a response to the DAP’s application on

October 13, 2022. See ECF No. 3557. In their response, defendants

argue that the “new evidence” identified by the DAPs is “woefully

insufficient” to justify the unduly burdensome expansion of the

discovery period by an additional 17 months, which defendants

estimate would result in the production of approximately 1.7

million documents spanning 8 million pages. Id. at 1, 2.6

B. Plaintiffs’ Application to Compel Additional Search Terms and

Custodians

In the second application of October 6, 2022, all plaintiffs

seek an order compelling 11 bank defendants to collect and produce

5 The Court notes that one of the experts proffered by the Lender Plaintiffs on

their class certification motion utilized this so-called new evidence in his

expert report. See In re LIBOR-Based Fin. Instruments Antitrust Litig., 299 F.

Supp. 3d 430, 557 (S.D.N.Y. 2018) (“LIBOR VII”).

6 Without the Court’s permission and contrary to the parties’ purported agreement,

the DAPs filed a reply in support of their application on October 18, 2022, see

ECF No. 3561, which defendants asked the Court to disregard, see ECF No. 3564.

Even if the Court considered the DAPs’ reply, it would not change the result.

all non-privileged, responsive documents relating to the two

upstream issues that hit on additional search terms and/or are

from additional custodians. See ECF No. 3549. Plaintiffs argue

that their requested additional search terms target common gaps in

defendants’ productions across all cases, which they group into

six categories:

1) “Defendants’ suppression of LIBOR, in order to show that

they collusively set their LIBOR contributions at rates

that were not truthful answers to the LIBOR question;”

2) “Defendants’ efforts through interbank brokers to

facilitate, monitor, and police their coordinated

persistent suppression of LIBOR;”

3) “Thomson Reuters’ efforts to carry out so-called

‘tolerance checks’ that Defendants used to monitor and

enforce their coordinated persistent suppression of

LIBOR;”

4) “the roles of the British Bankers Association (‘BBA’)

and related committees, including, but not limited to,

the Foreign Exchange and Money Markets Committee

(‘FXMMC’), in the Defendants’ coordinated persistent

suppression of LIBOR;”

5) “the role of the Bank of England (‘BOE’) in the

Defendants’ coordinated persistent suppression of

LIBOR;” and

6) “communications with the Federal Reserve Bank of New

York (‘NY Fed’) regarding Defendants’ persistent

suppression of LIBOR.”

Id. at 1-2. Plaintiffs claim that, through their thorough review

of defendants’ productions, they have “identified numerous

relevant terms that the Defendants did not search” in each of these

categories. Id. at 6.7

With respect to custodians, in their original submission,

plaintiffs identified categories of custodians whose files were

not searched, including actual LIBOR submitters, bank employees

who communicated with employees of other banks about anticipated

LIBOR submissions, and high-level executives who allegedly issued

directives to suppress LIBOR. Id. at 7. However, plaintiffs

failed to identify those custodians by name and title. Id.

Defendants filed a response to plaintiffs’ application on

October 13, 2022. See ECF No. 3559. In their response, defendants

argued that documents responsive to plaintiffs’ requests have

already been captured by the broad search terms used by defendants,

and that plaintiffs have not identified any specific gaps in their

productions beyond conclusory allegations of insufficiency. See

id. at 1-3. Specifically, defendants highlight that all 62

7 Plaintiffs argue that, to the extent defendants express concerns about burden,

they have offered defendants multiple strategies to reduce costs, including

allowing them to de-duplicate documents hitting on the new search terms that

have already been produced, to produce all documents that hit on the proposed

search terms while maintaining the ability to claw back documents without

waiving any privilege, and to narrow the proposed search terms by using hit-

count reports and responsiveness samples. See ECF No. 3549 at 3. At the

outset, the Court notes that de-duplication would be expected in the normal

course and rejects the suggestion that defendants should be compelled to produce

documents before conducting a review for privilege and confidential supervisory

information (“CSI”) between banks and their regulators. See infra Section

(B)(1)(ii). Such an order would unfairly advantage plaintiffs who would very

likely see privileged communications, could expose defendants to significant

regulatory consequences if CSI is disclosed, and could result in even more

litigation expenditures arising from subsequent disputes over claw back

requests.

exhibits attached to plaintiffs’ application do “not include a

single relevant, document that does not contain the term ‘LIBOR,’”

which was used as a standalone search term by almost all

defendants.8 Id. at 8; see also ECF No. 3624 at 4. In addition,

defendants claim that the production of additional documents would

be unduly burdensome and disproportionate to the needs of the case,

id. at 4-12, particularly with respect to communications with the

Federal Reserve and Bank of England given those communications

“also raise potential bank examination privilege,” id. at 8 n.10.9

C. The Court’s Request for Additional Information on Plaintiffs’

Application for Search Terms and Custodians

After reviewing the applications and defendants’ responses,

on December 15, 2022, the Court requested additional information

from the parties necessary to conduct a Rule 26(b)(1) analysis of

plaintiffs’ application for additional search terms and

custodians. See ECF No. 3603. Among other requests, the Court

directed the parties to provide: the search terms used by each

defendant; the search terms plaintiffs seek to compel; the

custodians plaintiffs seek to compel and defendants’ position with

8 UBS is the only defendant from whom plaintiffs seek relief that did not use

“LIBOR” as a standalone search term. See ECF No. 3624 at 4. Instead, UBS

produced documents from 57 custodians which were responsive to 222 search terms,

many of which include the term “LIBOR.” See ECF No. 3632 at 4.

9 Plaintiffs also filed a reply in support of the second application on October

18, 2022, see ECF No. 3562, which defendants similarly asked the Court to

disregard, see ECF No. 3564. Likewise, here, plaintiffs’ reply does not alter

the Court’s decision as all additional information needed to conduct a 26(b)(1)

analysis was gleaned from the parties’ supplemental submissions of January 12,

2023 and January 19, 2023. See ECF Nos. 3624, 3626, 3632.

respect to each custodian request; a list of significant documents

not produced by defendants along with the requested search terms

that would have resulted in their production; descriptions of each

defendant’s respective ability to run search terms across their

original collections;10 and the number of documents and pages

produced by each defendant. See id. at 3-4.

On January 12, 2023 and January 19, 2023, the parties

responded to the Court’s December 15, 2022 requests. See ECF Nos.

3624, 3626, 3632. In their submissions, plaintiffs requested more

than 100 additional search terms, see ECF No. 3624, and more than

40 additional custodians, see ECF No. 3626. The Court recognizes

the very substantial effort that went into the responses to the

Court’s December 15, 2022 requests.

LEGAL STANDARD

“A district court has wide latitude to determine the scope of

discovery.” In re Agent Orange Prod. Liab. Litig., 517 F.3d 76,

103 (2d Cir. 2008); accord EM Ltd. v. Republic of Argentina, 695

F.3d 201, 207 (2d Cir. 2012), aff’d sub nom. Republic of Argentina

v. NML Cap., Ltd., 573 U.S. 134 (2014). Discovery is generally

limited to “any nonprivileged matter that is relevant to any

party’s claim or defense and proportional to the needs of the case,

10 In their response, defendants noted that ordering them to produce additional

search terms would result in many defendants needing to re-collect documents

and essentially re-do their document reviews as the “original workspaces used

for review and production of documents no longer exist, as a result of changing

e-discovery processes and technology.” See ECF No. 3559 at 12.

considering the importance of the issues at stake in the action,

the amount in controversy, the parties’ relative access to relevant

information, the parties’ resources, the importance of discovery

in resolving the issues, and whether the burden or expense of the

proposed discovery outweighs its likely benefit.” Fed. R. Civ. P.

26(b)(1). “The burden of demonstrating relevance is on the party

seeking discovery.” Trilegiant Corp. v. Sitel Corp., 275 F.R.D.

428, 431 (S.D.N.Y. 2011) (internal citations omitted). “Once

relevance has been shown, it is up to the responding party to

justify curtailing discovery.” Id.

Relevance “has been construed broadly to encompass any

matter that bears on, or that reasonably could lead to other matter

that could bear on, any issue that is or may be in the case.”

Oppenheimer Fund, Inc. v. Sanders, 437 U.S. 340, 351 (1978).

However, discovery requests must be “properly tailored,” Martinez

v. Robinson, No. 99 Civ. 11911 (DAB) (JCF), 2002 WL 424680, at *2

(S.D.N.Y. Mar. 19, 2002), as “[p]roportionality focuses on the

marginal utility of the discovery sought,” Vaigasi v. Solow Mgmt.

Corp., No. 11 Civ. 5088 (RMB) (HBP), 2016 WL 616386, at *14

(S.D.N.Y. Feb. 16, 2016). “Proportionality and relevance are

‘conjoined’ concepts; the greater the relevance of the information

in issue, the less likely its discovery will be found to be

disproportionate.” Id. (internal citation omitted).

With these legal standards in mind, the Court turns to the

specific discovery requests.

DISCUSSION

A. DAPs’ Application to Compel Documents from June 2010 through

October 2011

The Court denies the DAPs’ application, finding that a 17-

month extension of the discovery period is wholly disproportionate

to the needs of the case. See Fed. R. Civ. P. 26(b)(1). Indeed,

granting the DAPs’ request would extend the discovery period well

beyond the relevant time period asserted in the DAPs’ complaints.

See ECF No. 3547 at 1; see also Kingsway Fin. Servs., Inc. v.

Pricewaterhouse-Coopers LLP, No. 03 Civ. 5560 (RMB) (HBP), 2008 WL

4452134, at *5 (S.D.N.Y. Oct. 2, 2008) (denying discovery request

for documents outside of relevant time period in complaint).

“To justify discovery requests for documents generated

outside the time period alleged in the [the DAPs’] Complaint[s],

the [DAPs] must demonstrate the relevance of the potential

discovery to the claims it has asserted for the [August 2007 to

May 2010] time period.” United States ex rel. Bilotta v. Novartis

Pharms. Corp., No. 11 Civ. 0071 (PGG), 2015 WL 13649823, at *4

(S.D.N.Y. July 29, 2015). Here, the DAPs argue that they have

identified “new evidence” showing that defendants suppressed LIBOR

through October 2011 and drawing into question whether June 2010

through October 2011 can be used as a “clean period” in the expert

economists’ models. See ECF No. 3547 at 1. The Court is not

persuaded.

As an initial matter, the three recently-discovered documents

cited by the DAPs are of limited relevance as they constitute mere

speculation about banks’ LIBOR submission practices by bank

employees who were not directly involved in the LIBOR submission

process. See ECF Nos. 3547-1, 3547-2, 3547-3. Even if the DAPs

presented more persuasive evidence of LIBOR suppression after May

2010, however, it is still not clear how that evidence would affect

the merits of the DAPs’ claims, which are limited to August 2007

through May 2010. See In re Weatherford Intern. Sec. Litig., No.

11 Civ. 1646 (LAK) (JCF), 2013 WL 5788687, at *3 (S.D.N.Y. Oct.

28, 2013) (“[P]ointing to one example of potentially discoverable

material does not transform an overbroad request into one that is

properly tailored.”) (internal quotations and citation omitted).

Without clear relevance to the DAPs’ causes of action, the

DAPs’ request is therefore only arguably relevant to establishing

a “clean period” in furtherance of their expert economists’ models,

which they argue will “show that LIBOR was lower than it would

have been but for Defendants’ conduct.” See ECF No. 3547 at 2.

Simply put, the DAPs desire to establish a “clean period” does not

justify requiring defendants to review presumably millions of

documents from a 17-month period from hundreds of custodians.

While “additional years’ worth of information [may] increase[] the

likelihood that relevant trends and fluctuations will become

clear,” Novartis Pharms. Corp., 2015 WL 13649823, at *4, “discovery

is not boundless, and a court may place limits on discovery demands

that are unreasonably cumulative or duplicative, or in cases where

the burden or expense of the proposed discovery outweighs its

likely benefit.” Kingsway Fin. Servs., Inc., 2008 WL 4452134, at

*4 (internal quotations and citations omitted). Accordingly, the

DAPs’ application is denied in its entirety.

B. Plaintiffs’ Application to Compel Additional Search Terms and

Custodians

The Court next turns to plaintiffs’ application to compel the

production of documents that are responsive to more than 100

additional search terms and/or are from more than 40 additional

custodians. See ECF No. 3549.

1. Search Term Requests

As noted, in their application of October 6, 2022, plaintiffs

bucketed their search term requests into the following six

categories:

1) “Defendants’ suppression of LIBOR, in order to show that

they collusively set their LIBOR contributions at rates

that were not truthful answers to the LIBOR question;”

2) “Defendants’ efforts through interbank brokers to

facilitate, monitor, and police their coordinated

persistent suppression of LIBOR;”

3) “Thomson Reuters’ efforts to carry out so-called

‘tolerance checks’ that Defendants used to monitor and

enforce their coordinated persistent suppression of

LIBOR;”

4) “the roles of the British Bankers Association (‘BBA’)

and related committees, including, but not limited to,

the Foreign Exchange and Money Markets Committee

(‘FXMMC’), in the Defendants’ coordinated persistent

suppression of LIBOR;”

5) “the role of the Bank of England (‘BOE’) in the

Defendants’ coordinated persistent suppression of

LIBOR;” and

6) “communications with the Federal Reserve Bank of New

York (‘NY Fed’) regarding Defendants’ persistent

suppression of LIBOR.”

Id. at 1-2. To begin, the Court rejects plaintiffs’ request for

additional documents related to categories three (Thomson

Reuters), five (Bank of England), and six (Federal Reserve Bank of

New York). Then, the Court considers plaintiffs’ specific search

term requests related to categories one (LIBOR suppression), two

(use of interbank brokers), and four (use of the BBA and related

committees) as well as plaintiffs’ custodian requests.

i. Category Three (Thomson Reuters)

The Court finds that the requests related to Thomson Reuters

are neither relevant nor proportional to the needs of the case.

See Fed. R. Civ. P. 26(b)(1). Despite having already received

productions from non-party Thomson Reuters and four bank

defendants that ran either “Thomson Reuters” or “Reuter*” as a

search term, see ECF Nos. 1537, 3624 at 48, plaintiffs cite no

hard evidence supporting their suggestion that Thomson Reuters’

“efforts to carry out so-called ‘tolerance checks’” were

nefarious, see ECF No. 3549 at 2; ECF No. 3626-2 at 22. Indeed,

plaintiffs’ support for their allegation is limited to: (i) a

transcript of a call between John Ewan from the BBA and Miles

Storey from Barclays, in which Storey states that he received a

call from Thomson Reuters to confirm that he is “happy” with his

LIBOR submission, see ECF No. 3549-33 at 6; and (ii) bare-bones

call logs produced by Thomson Reuters which merely provide for

each call the date, the name of the bank, the revision to the

bank’s LIBOR rate, and the number of queries made per bank, see

ECF No. 3549-34.11 The call logs do not contain the names of the

call participants, identify who initiated each call, or

memorialize any conversation that occurred on the call. See id.

Contrary to plaintiffs’ arguments, none of these documents

show that “Defendants required Thomson Reuters to perform so-

called ‘tolerance’ checks on each Defendant’s submissions before

publication” or that the purpose of Thomson Reuters’ calls with

the defendant banks was to “improve alignment with the other Panel

Banks.” See ECF No. 3549 at 10. Rather, it is far more likely

that, to the extent that Thomson Reuters did contact LIBOR

11 Plaintiffs cite an additional document in support of their theory that does

not reference Thomson Reuters. See ECF No. 3549 at 10 (citing ECF No. 3549-

31). The Court rejects plaintiffs’ suggestion that it should make the logical

leap of assuming that, when a Rabobank submitter states in this document that

the BBA would call if his submission was out of line with other banks, he was

actually “referring to Thomson Reuters acting as agent of the BBA.” ECF No.

3549 at 10 n.22.

submitters, it did so in order to confirm that they did not make

any mistakes when submitting their LIBOR rates. Indeed,

plaintiffs’ further argument, supported only by a news article

from 2012, that “Thomson Reuters alerted the BBA on a regular basis

about LIBOR manipulation” undermines its claim that Thomson

Reuters acted on behalf of defendants to forward the alleged

conspiracy to manipulate LIBOR. See id. (citing ECF No. 3549-35).

In fact, the cited news article states that the reason Thomson

Reuters would occasionally contact LIBOR submitters was to

“check[] [] implausible rates”—not to encourage banks to modify

their submissions to align with others. ECF No. 3549-35 at 5.

Accordingly, the Court denies plaintiffs’ application to obtain

additional documents related to category three.

ii. Categories Five (Bank of England) and Six (Federal

Reserve Bank of New York)

The Court next finds that categories five and six, targeting

communications with the Bank of England and Federal Reserve Bank

of New York, raise specific issues of privilege, which make their

proposed review unduly burdensome and disproportionate to the

needs of the case. See Fed. R. Civ. P. 26(b)(1). Indeed, “agency

opinions and recommendations and banks’ responses thereto” are

protected by a “qualified bank examination privilege,” which

“arises out of the practical need for openness and honesty between

bank examiners and the banks they regulate, and is intended to

protect the integrity of the regulatory process by privileging

such communications.” Sharkey v. J.P. Morgan Chase & Co., No. 10

Civ. 3824 (RWS), 2013 WL 2254553, at *1 (S.D.N.Y. May 22, 2013);

see also Wultz v. Bank of China Ltd., 61 F. Supp. 3d 272, 282

(S.D.N.Y. 2013) (collecting cases); 12 C.F.R. 261.2(b) (defining

“confidential supervisory information”). Because this privilege

“belongs to the regulatory authority and not to the banks that it

regulates,” Bank of China v. St. Paul Mercury Ins. Co., No. 03

Civ. 9797 (RWS), 2004 WL 2624673, at *4 (S.D.N.Y. Nov. 18, 2004),

as modified on reconsideration sub nom., 2005 WL 580502 (S.D.N.Y.

Mar. 10, 2005), the regulator “must be allowed the opportunity to

assert the privilege and the opportunity to defend its assertion,”

Wultz, 61 F. Supp. 3d at 282.

While a court may ultimately “override the privilege if the

requesting party demonstrates ‘good cause,’” id., the Court finds

that the asserted need for these documents does not outweigh the

burden and delays that will necessarily arise from the multi-

layered review of documents that fall into these categories—

particularly in light of the 3.4 million documents that have

already been produced by defendants. The Court also recognizes

that, apart from these established privilege issues, any effort to

take third-party discovery would presumably face additional

hurdles. As such, plaintiffs’ requests related to categories five

and six are also denied.

iii. Categories One (Suppression of LIBOR), Two (Use of

Interbank Brokers), and Four (Use of BBA and Related

Committees)

Defendants do not dispute the general relevance of

plaintiffs’ remaining requests for documents related to

defendants’ alleged suppression of LIBOR, alleged use of interbank

brokers to further their conspiracy, and alleged use of the BBA

and related committees as forums to further their conspiracy. See

ECF No. 3559 at 1, 8. Nor does the Court.12 However, in the

context of a request for additional discovery, plaintiffs must

point to “specific information that is relevant to their claims

and would be found solely in the unproduced documents.” Alaska

Elec. Pension Fund v. Bank of Am. Corp., No. 14 Civ. 7126 (JMF),

2016 WL 6779901, at *3 (S.D.N.Y. Nov. 16, 2016) (citing Fort Worth

Emps.’ Ret. Fund v. J.P. Morgan Chase & Co., 297 F.R.D. 99, 104

(S.D.N.Y. 2013) (granting additional discovery because plaintiffs

“provided sufficient justification for expanding search terms” by

“highlight[ing] both general categories and concrete examples of

documents . . . that are relevant but would not be discovered” by

the current search”) (emphasis in original)). Only then can the

Court consider the “marginal utility” of each of plaintiffs’ search

term requests when weighing proportionality in light of

12 To the extent that plaintiffs seek documents related to the alleged

suppression of LIBOR in order to show that banks “collusively set their LIBOR

contributions at rates that were not truthful answers to the LIBOR question,”

ECF No. 3549 at 1 (emphasis added), the Court notes that that the concept of

truthfulness must be understood in light of Connolly, 24 F.4th at 837.

defendants’ already voluminous productions. Vaigasi, 2016 WL

616386, at *14.

Here, however, plaintiffs have failed to establish the unique

relevance of most of their search term requests, despite being

given multiple opportunities to do so. As noted, after plaintiffs

submitted their initial application along with 62 exhibits that

did not “include a single relevant document that does not contain

the term ‘LIBOR,’” ECF No. 3559 at 8, the Court directed plaintiffs

to provide a list of significant documents not produced by

defendants and the additional search terms that would have resulted

in their production, ECF No. 3603 at 3. While plaintiffs did

provide a chart of documents, that chart’s usefulness in evaluating

the relative importance of many of plaintiffs’ search term requests

is limited.

As an initial matter, many of the documents identified by

plaintiffs: (i) appear to have been sent and received by

individuals that defendants did not collect documents from, making

their exclusion a custodians issue, rather than a search terms

issue, see e.g., ECF No. 3626-2 at 2, 5, 6; (ii) were produced by

at least some defendants, making their proposed production by other

defendants cumulative, see e.g., id. at 2, 23, 24, 26-33; or

(iii) may not actually exist as their identification is based

solely on second-hand references to those communications, meaning

those communications may have occurred in-person, by phone, or not

at all, see e.g., id. at 8-21. For the limited number of documents

that do not fall into those categories, plaintiffs have failed to

specify precisely which of their search term requests would have

led to the production of each of the documents they identified.

Id. at 1-37. As such, the Court continues to struggle to decipher

which requested search terms would have resulted in the identified

documents’ production. It is still simply not clear to the Court

that all of the identified documents would not have been produced

through the use of a few, broad search terms like “LIBOR,”

“interbank,” and “BBA,” and that more than 100 additional search

terms are truly needed.13

Accordingly, the Court denies all of plaintiffs’ search term

requests, which it deems unreasonably cumulative and duplicative

and therefore lacking in “marginal utility,” except for the search

terms listed in Appendix 1 to this memorandum and order. See In

re Commodity Exch., Inc., Gold Futures & Options Trading Litig.,

14 MD 2548 (VEC), 2021 WL 2481835, at *2 (S.D.N.Y. June 17, 2021);

Treppel v. Biovail Corp., 249 F.R.D. 111, 117 (S.D.N.Y. 2008).

With respect to the search term requests granted by the Court, the

13 In addition, only one of the documents plaintiffs identified as related to

defendants’ efforts to collusively suppress LIBOR even arguably involves

communications between multiple bank defendants. See ECF No. 3626-2 at 1-12;

see also ECF No. 3559 at 6 (“It is telling that Plaintiffs have not identified

in their Motion a single document containing any evidence—direct or

circumstantial—of a panel-wide conspiracy to suppress LIBOR.”). In that single

document, the Bloomberg chat of an RBS employee, who was not a custodian, was

“shared” with Credit Suisse and Deutsche Bank employees, according to

plaintiffs. ECF No. 3626-2 at 2. As such, it is not clear to the Court that,

even there, a direct communication between bank defendants occurred.

Court exercises its discretion and concludes that their relevance

outweighs their burden to a degree that warrants them being used

by the defendants listed in Appendix 1.14 The Court understands

that all of the granted search term requests have already been

used or substantially used by some defendants, and those

defendants, which have been omitted from Appendix 1, are not

directed to re-run those search terms. Nevertheless, the Court

concludes that the granted search term requests are so highly

relevant to plaintiffs’ allegations that the defendants that have

not used those terms should do so now.

2. Custodian Requests

The Court next turns to plaintiffs’ application for documents

from more than 40 additional custodians from nine defendants. See

ECF No. 3626-3.15 In evaluating each custodian request, the Court

14 The Court understands that this order will require certain defendants that

ran search terms but no longer have access to their original document

collections to re-collect documents from their original custodians. See ECF

Nos. 3559 at 12, 3632 at 2-3. The Court has considered the additional burden

imposed by this order on those defendants, but ultimately does not believe that

the defendants that failed to preserve their original collections should get a

pass from participating in this stage of discovery. The Court further

understands that two defendants, Bank of America and Citibank, conducted

document-by-document reviews of certain custodians’ documents without the

application of any search terms. See ECF No. 3632 at 5-6. Bank of America and

Citibank thus need not apply any additional search terms to those custodians’

documents. See id. at 6. With respect to Portigon, the Court understands that

search terms were applied before any documents were collected from custodians.

See id. at 6. Therefore, Portigon should re-collect and review documents

hitting on the additional search terms from its custodians. To the extent

feasible, defendants are permitted to de-duplicate documents that have already

been reviewed against documents that hit on the additional search terms.

15 Plaintiffs do not seek any additional custodians from Portigon and Société

Générale. See ECF No. 3626-3 at 1 n.1.

has considered whether plaintiffs have met their burden in showing

that the requested custodian is “likely to have non-cumulative

relevant documents.” Assured Guar. Mun. Corp. v. UBS Real Est.

Sec. Inc., No. 12 Civ. 1579 (HB) (JCF), 2013 WL 1195545, at *3

(S.D.N.Y. Mar. 25, 2013) (denying custodian requests as

cumulative); see also Mortg. Resol. Servicing, LLC v. JPMorgan

Chase Bank, N.A., No. 15 Civ. 293 (LTS) (JCF), 2017 WL 2305398, at

*2 (S.D.N.Y. May 18, 2017) (“[A] party seeking to compel another

party to search the files of additional custodians bears the burden

of establishing the relevance of the documents it seeks from those

custodians.”). As with plaintiffs’ search term requests, the Court

is then “obligat[ed] to balance its utility against its cost.”

Coventry Cap. US LLC v. EEA Life Settlements, Inc., No. 17 Civ.

7417 (VM), 2021 WL 961750, at *2 (S.D.N.Y. Mar. 15, 2021).

The Court grants plaintiffs’ requests for custodians that:

(i) were directly involved in the LIBOR submission process (Yasmine

Nahouli and Ignacio Rodrigo, NatWest/Royal Bank of Scotland);

(ii) were directly involved in alleged efforts to manipulate LIBOR

(Andrew Rivas, Deutsche Bank; Kevin Liddy, NatWest/Royal Bank of

Scotland; Reto Stadelmann, UBS); (iii) had alleged unique and

substantial communications about accusations of LIBOR suppression,

including with other panel banks or at BBA meetings (Michael

Kirkwood, Citibank; Dyfrig John, HSBC; Harry Samuel, Royal Bank of

Canada; Johnny Cameron and Brian Crowe, NatWest/Royal Bank of

Scotland); or (iv) were involved in an internal audit related to

LIBOR (Chris Riley, Norinchukin). For all of these custodians,

plaintiffs have demonstrated that “the discovery sought is of

sufficient importance to justify the burden and cost that discovery

will impose on the responding party.” Blackrock Allocation Target

Shares: Series S Portfolio v. Bank of New York Mellon, No. 14 Civ.

9372 (GBD) (HBP), 2018 WL 2215510, at *12 (S.D.N.Y. May 15, 2018).

However, the Court denies plaintiffs’ requests for

custodians: (i) whose stated relevance is based on their

involvement in internal investigations that did not concern

allegations of LIBOR suppression (William Broeksmit and Jonathan

Kahlberg, Deutsche Bank); (ii) who were not involved in the LIBOR

setting process but nevertheless received or communicated

information about anticipated LIBOR rates (Kevin McBunch, Thomas

McGee, Brian Wichnern, and John Widmeier, Bank of America; Robbie

Anderson, Olivier Pouzoulet, and Tim Townsend, NatWest/Royal Bank

of Scotland); (iii) who merely expressed or received an opinion

about LIBOR (Peter Kretzmer, James Tedrake, and Joseph Randazzo,

Bank of America; Darren Gebler and Mark Zarb, JPMorgan; Michael

Andrew, Jeff Kulkarni, and John Retter, NatWest/Royal Bank of

Scotland); (iv) who were so far removed from the LIBOR submission

process that they are unlikely to possess unique information, such

as junior team members, high-level executives, employees from

unrelated departments, and a member of one panel bank’s press

communications team (Joanna Bushby, Bank of America; Jeffrey

French, Citibank; David Folkerts-Landau, Deutsche Bank; Mark

Garvin and Achilles Macris, JPMorgan; Clark Foran, NatWest/Royal

Bank of Canada); (v) whose stated relevance raises specific issues

of privilege (Stuart Gulliver, HSBC); or (vi) whose documents are

likely to be highly cumulative of documents from other custodians

(Heinz Srocke, James Forese, and Francisco Fernandez De Ybarra,

Citibank; Knut Pohlen, Deutsche Bank; Gary Lomas, Norinchukin;

Fred Goodwin and John Hourican, NatWest/Royal Bank of Scotland).

None of these proposed custodians were directly involved in the

LIBOR submission process, and plaintiffs’ reasoning for their

inclusion does not sufficiently show that a search of their

documents “would provide unique, relevant and noncumulative

evidence.” Coventry Cap. US LLC v. EEA Life Settlements Inc., No.

17 Civ. 7417 (VM) (SLC), 2020 WL 7383940, at *7 (S.D.N.Y. Dec. 16,

2020). Accordingly, “requiring a search of the files of these

custodians cannot be justified given the minimal marginal value of

the information sought.” Assured Guar. Mun. Corp., 2013 WL

1195545, at *5; see also Mortg. Resol. Servicing, LLC, 2017 WL

2305398, at *3 (“Even if the plaintiffs had cleared the hurdle of

showing that the proposed additional custodians had some unique

relevant information, discovery of their files would not be

warranted because the cost and burden would be disproportionate.”).

The Court also denies plaintiffs’ remaining custodian

requests (Paul Glands and Colin Harrison, JPMorgan) as moot, given

defendants have since agreed to produce the requested custodians’

responsive documents. See ECF No. 3632-1 at 11. To the extent

plaintiffs seek a broader production for these custodians than

what has currently been produced by defendants, that request is

denied as disproportionate to the needs of the case.

Lastly, in the absence of a showing that any documents were

improperly withheld, the Court declines to order defendants to

produce their entire regulatory productions, from which the Court

understands defendants have already produced responsive documents.

See ECF Nos. 3626 at 2, 3626-3 at 1 n.1. The Court does not find

that the simple fact that a custodian’s files were made available

to regulators makes every document in that custodian’s files

relevant.

Accordingly, plaintiffs’ application to compel additional

search terms and custodians is granted in part and denied in part.

CONCLUSION

For the foregoing reasons, the DAPs’ application for an order

compelling the production of documents from June 2010 through

October 2011 is denied in its entirety, and plaintiffs’ application

for an order compelling the production of documents responsive to

more than 100 additional search terms and/or from more than 40

additional custodians is granted in part and denied in part.

Complete lists of the search term and custodian requests that are

granted as to certain defendants are provided in the appendices.

The Clerk of Court is respectfully instructed to terminate the

motions pending at ECF Nos. 3547 and 3549.

SO ORDERED.

Dated: New York, New York

April 10, 2023 ,

hk dL chal

NAOMI REICE BUCHWALD

UNITED STATES DISTRICT JUDGE

26

APPENDICES

A. Appendix 1: Granted Search Term Requests

Search Term Defendant(s) Ordered to Run

Search Term16

BBA* UBS

British bankers association Citibank

HSBC

Portigon

UBS

Collu* Bank of America

Deutsche Bank

Portigon

Royal Bank of Canada

NatWest/Royal Bank of Scotland

Société Générale

Err /5 (lo* or hi* or side) Bank of America

Portigon

Royal Bank of Canada

Société Générale

UBS

Fix* Bank of America

Portigon

UBS

(Interbank or inter*bank) and Deutsche Bank

(loan* or lend* or borrow* or HSBC

bid* or offer*) JPMorgan

UBS

Middle /5 (keep* or line* or Bank of America

remain* or stay* or within) Deutsche Bank

Norinchukin

Portigon

Royal Bank of Canada

NatWest/Royal Bank of Scotland

Société Générale

UBS

16 Bank of America and Citibank need not run any additional search terms across

the documents of its custodians whose responsive documents were produced without

the application of search terms. See supra footnote 14.

Outlier* /7 (libor* or submi* Bank of America

or set* or fix* or rate* or Royal Bank of Canada

level* or pack* or mov*) Portigon

Pack /7 (stay* or remain* or Bank of America

keep* or part* or in* or HSBC

middle* or within or mov* or Norinchukin

stand* or out*)

Portigon

Royal Bank of Canada

NatWest/Royal Bank of Scotland

Perceiv* /5 (strong or weak Bank of America

or fix* or level* or libor* or Portigon

market* or rate* or set* or Royal Bank of Canada

submi*) Société Générale

UBS

Perception /5 (strong or weak Bank of America

or press or market or public HSBC

or client* or customer* or JPMorgan

“out there”) Portigon

Royal Bank of Canada

Société Générale

UBS

Reputation Bank of America

Deutsche Bank

Norinchukin

Portigon

Royal Bank of Canada

NatWest/Royal Bank of Scotland

Société Générale

UBS

“Too high” and (libor* or Bank of America

submi* or set* or fix* or Deutsche Bank

rate* or level* or mov* or Portigon

err*)

Royal Bank of Canada

“Too low” and (libor* or Bank of America

submi* or set* or fix* or Deutsche Bank

rate* or level* or mov* or Portigon

err*)

Royal Bank of Canada

B. Appendix 2: Granted Custodian Requests

Custodian Defendant Ordered to Produce

Documents from Custodian

Michael Kirkwood Citibank

Andrew Rivas Deutsche Bank

Dyfrig John HSBC

Chris Riley Norinchukin

Harry Samuel Royal Bank of Canada

Yasmine Nahouli NatWest/Royal Bank of Scotland

Ignacio Rodrigo NatWest/Royal Bank of Scotland

Johnny Cameron NatWest/Royal Bank of Scotland

Brian Crowe NatWest/Royal Bank of Scotland

Kevin Liddy NatWest/Royal Bank of Scotland

Reto Stadelmann UBS