OPINION & ORDER

DENISE COTE, District Judge.

This Opinion addresses the fairness of an almost $2 billion settlement (the "Settlement") reached in antitrust class action litigation arising from the purchase and sale of credit default swaps ("CDS").1 Plaintiffs bring this antitrust action individually and on behalf of all persons who, during the period from January 1, 2008 through September 25, 2015 (the "Class Period"), bought CDS from, or sold CDS to, certain banks in the United States (the "Class"). The defendants are those banks (the "Dealer Defendants"),2 as well as the International Swaps and Derivatives Association ("ISDA"), and both Markit Group Holdings Limited and its subsidiary Markit Group Ltd. (collectively, "Markit").

In less than two years following the appointment of lead counsel, the Class achieved a remarkable settlement. The Settlement will make a common fund of $1,864,650,000 available to Class members (the "Settlement Fund"), and require ISDA to take steps designed to increase transparency and competition in the CDS market. This Settlement was approved at the Fairness Hearing held on April 15, 2016. This Opinion further describes the basis for that approval, the rejection of the limited objections made to the Plan of Distribution that will govern the distribution of the Settlement Fund, the award of attorneys' fees and expenses, the rejection of the request for incentive awards for Class representatives, and the standard that may be applied to any request by class counsel for a bond pending an appeal by an objector.

BACKGROUND

The allegations in this litigation are described in detail in the Court's September 4, 2014 Opinion granting in part the defendants' motions to dismiss. See In re Credit Default Swaps Antitrust Litig., No. 13md2476 (DLC), 2014 WL 4379112 (S.D.N.Y. Sept. 4, 2014). In brief, a CDS is a derivative whose value depends on the value of an underlying debt instrument. Id. at *1. The buyer of the CDS purchases the seller's promise to pay on the occasion of a "credit event," such as a default on the debt instrument by a third party known as the "reference entity." Id. Market makers — also referred to as "dealers" — sell CDS to buyers, buy CDS from sellers, and hold CDS inventory until a match emerges. Id. A dealer offers a "bid" price at which the dealer will purchase and an "ask" price at which the dealer will sell. Id. By keeping their bid lower than their ask, dealers can capture the difference, known as the "bid/ask spread." Id.

The complaint alleges that, in and around 2008 to 2009, the defendants conspired to suppress price transparency and competition in the trading market for CDS and boycott the exchange trading of CDS, thereby maintaining supracompetitive bid/ask spreads. Among other things, the complaint alleges that the defendants conspired to block "CMDX," a proposed CDS electronic exchange platform that the Chicago Mercantile Exchange and Citadel Investment Group partnered to launch in the fall of 2008. Id. at *4-5.

The first complaint in this litigation was filed on May 3, 2013 in the Northern District of Illinois, and other related actions were filed in this district and elsewhere soon after. On October 22, the U.S. Judicial Panel on Multidistrict Litigation transferred all related class actions to this district. At a conference on December 5, 2013, the Court appointed Quinn Emanuel Urquhart & Sullivan, LLP ("Quinn Emanuel") as Lead Counsel. Shortly thereafter, Pearson, Simon & Warshaw, LLP ("Pearson Simon") was appointed Co-Lead Counsel (collectively, "Class Counsel"), and Salix Capital U.S., Inc. ("Salix") and the Los Angeles County Employee Retirement Association ("LACERA") were appointed Lead Plaintiffs for the Class. Plaintiffs filed the operative complaint on April 14, 2014, which brought claims under Sections 1 and 2 of the Sherman Antitrust Act ("Sherman Act"), 15 U.S.C. §§ 1-2, and under state unjust enrichment law.

At the time this case was filed, both the U.S. Department of Justice ("DOJ") and the European Commission ("EC") were conducting ongoing investigations of the defendants for collusion regarding the CDS market. The DOJ investigation is reported to have closed sometime in 2013; the EC investigation closed at least as to the Dealer Defendants in December 2015, just months after the Settlement was reached.

I. Discovery and Mediation

On September 4, 2014, the Court dismissed the complaint's claims under Section 2 of the Sherman Act, but allowed its other claims to proceed. Id. at *18. The parties proceeded to discovery immediately thereafter.

Class Counsel worked with lightning speed. Class Counsel obtained over fifty million pages of documents from the defendants and millions more from third parties. They developed and utilized research tools that allowed them to quickly locate critical documents, prepare a draft narrative of key events, and identify key witnesses. By July 29, 2015, they had taken twenty-seven of the forty-six depositions noticed by that date. During the initial discovery period, Class Counsel also obtained data for millions of CDS transactions from the Depository Trust & Clearing Corporation ("DTCC") and engaged experts to analyze this data and build a model capable of calculating damages for Class members.

In December 2014, and in parallel with ongoing discovery, the parties engaged the services of a renowned mediator, Daniel Weinstein. Under his supervision, the parties began mediation sessions on January 22, 2015. The mediator worked with the parties for nine months and invested over 400 hours of his own time in the mediation. Plaintiffs presented a detailed mediation brief and PowerPoint presentation on liability at the first mediation session and a preliminary damages model at the March 31 mediation session. At the urging of the mediator, plaintiffs produced a copy of their damages model, as well as the dataset used to calculate damages, to the defendants. The defendants presented a detailed critique of the plaintiffs' damages model during a June 8 mediation session.

Plaintiffs reached agreements in principle with all defendants to settle the case by mid-August, just prior to the August 31 deadline for the motion for class certification. Plaintiffs filed their motion for preliminary approval of the Settlement on October 16. The Court preliminarily approved the Settlement on October 29.

II. Terms of the Settlements

The separate Settlement agreements executed by each Dealer Defendant and Markit are virtually identical, except for the amount of money each has agreed to pay into the Settlement Fund. The total amount to be paid into that fund is $1,864,650,000. ISDA has also agreed to injunctive relief to bring greater transparency and competition to the CDS market. Among other things, ISDA agreed to create a new independent Licensing Sub-Committee consisting of equal buy- and sell-side members, and to make meetings of that Sub-Committee open to the public.

The Class agreed to the following release (the "Release"):

(Emphasis added.) The agreements define "CDS Transactions" as

The agreements define "Potential CDS Transactions" as "any CDS Transaction for which an offer or quote was obtained or sought, regardless of whether such transaction was actually entered into or executed with the party from which such offer was obtained or sought."

Excluded from the Release are claims by Class members not "domiciled or located in the United States at the relevant time"; claims based on transactions that "were not in or would not have been in United States commerce"; and claims based on transactions that "are or would have been subject only to foreign law."

III. Plan of Distribution

Because there have been four objections to the allocation of the Settlement Fund among Class members pursuant to the Plan of Distribution, the construction and structure of that Plan must be described in some detail. The Plan rests on the work Class Counsel performed with its experts to prepare a damages model.

A. Plan of Distribution Datasets

Class Counsel collected data on CDS transactions occurring from January 2008 to September 2015 in each of the major categories of CDS transactions: single-name, index, tranche index, structured credit, and CDS options. The data spans thousands of CDS contracts and millions of CDS transactions, with corresponding data on the bid/ask spreads quoted in these instruments each day or virtually every day. Class Counsel created its dataset by combining data from two sources: the DTCC's Trade Information Warehouse, which is a global repository for data concerning executed CDS transactions, and a Markit database that provides quote data.

DTCC provided over 159 million transaction records spanning nearly eight years, capturing over 90 percent of all CDS transactions. These include data on 3,500 distinct reference entities. For each transaction record, the DTCC data provides details about the transaction, including the trade date, the contract expiration date, and the key characteristics of the traded products.

While the DTCC data captures payment information, it does not capture the bid and ask prices quoted by the dealer pursuant to which the transaction was executed. To infer the bid/ask spreads incurred on a given CDS transaction, Class Counsel obtained data from Markit, which is a leading source of such data. Markit gathers information on the bid/ask spreads quoted by dealers during the course of a trading day by parsing electronic messages conveyed by dealers to market participants and extracting the relevant quote information from the messages. The Markit database produced in this litigation contained almost 3.2 billion records of bid/ask spreads.

B. The Mechanics of the Plan

The Plan determines the amount to be paid on each Class member's claim through three main steps: (1) identifying qualifying Covered Transactions; (2) estimating the amount of bid/ask spread inflation resulting from the Dealer Defendants' alleged conduct with respect to each Covered Transaction; and (3) calculating each claimant's recovery based on its pro rata share of the available Settlement Funds in relation to the recoveries to which all claimants who have submitted a valid claim are entitled.

Class Counsel and their consulting experts, led by Dr. Sanjay Unni of the Berkeley Research Group,3 used the DTCC dataset to identify Covered Transactions using the criteria specified in the Settlement agreements. The total notional volume of Covered Transactions from the DTCC dataset is approximately $69 trillion.

Under the Plan, the bid/ask spread paid on a given Covered Transaction is determined as the average spread quoted for the CDS contract actually involved in the transaction on the day of the transaction. Bid/ask spreads for a CDS can fluctuate during the day. While, in principle, the spread charged on a transaction should be measured as the spread prevailing at the time of the transaction's execution, the DTCC data only provides the day that a given transaction occurred, not the time within the day. As such, Class Counsel chose to associate each CDS transaction with the average daily bid/ask spread prevailing for that CDS on the day the transaction occurred.

In order to determine that average bid/ask spread, Class Counsel used the Markit dataset to identify the bid/ask spreads for the associated CDS during each hour of the date of the transaction.4 Class Counsel took the "inside spread," which was the smallest or tightest spread, during each trading hour to calculate an average spread for each day.5 Accordingly, the Plan measures the applicable bid/ask spread for an instrument on a given day as the average of the tightest bid/ask spreads prevailing in each hour of trading for that instrument. The average bid/ask spread is then reduced by half, as each CDS transaction is a buy or a sell transaction that only incurs half of the cost of the spread.

For one type of linked transaction, the Plan makes a further adjustment. This linked transaction is an "index roll." In March and September each year, index CDS are updated to reflect changes in the credit conditions of their constituent instruments. The updated index is assigned a new series number. When this happens, investors may "roll" over their exposure from the old position to the new series, thereby updating their risk exposure in the market segment covered by that index CDS. These investors do this through a two-legged transaction: selling one index series and buying the subsequent index series. Based on industry custom, the Plan exempts one leg of the roll from its spread calculations. To apply this adjustment conservatively, the Plan identifies the rolls as occurring when the trade and termination happen on the same day in different series of the same index.

Next, the Plan applies a spread compression percentage (the "Compression Rate") to reflect how the spread that historically prevailed in the CDS market would have tightened but for the defendants' actions. Based upon a review of empirical evidence on spread compression experienced in other markets, Class Counsel applied a Compression Rate of 20%.

IV. Notice to Class Members

Because of their access to trading records, Class Counsel were able to identify and reach most potential Class members. On January 11, 2016, Class Counsel mailed notice packets to each of 13,923 identified Class members. While some of the mailings were returned as undeliverable, reasonable efforts were made to locate every identified Class member, including identifying alternative mailing addresses. Ultimately, notice was successfully mailed to all but 548 of the Class members. In addition, given its magnitude, the Settlement received widespread publicity. See, e.g., Katy Burne, Big Banks Agree to Settle Swaps Lawsuit, Wall St. J. (Sept. 12, 2015); Jesse Druker, Wall Street Banks to Settle CDS Lawsuit for $1.87 Billion, Bloomberg (Sept. 11, 2015). The Summary Notice was also published on January 11 in several important business publications.

The Garden City Group (the "Claims Administrator") launched a website for the Settlement which posted the Settlement agreements, notices, court documents, and other information relevant to the Settlement. On January 11, a description of the Plan of Distribution was also posted on the website for Class members to review. Since January 28, each Class member has been able to log into a "Claimant Portal" on the Settlement website to review the Covered Transactions identified by Class Counsel as applicable to that Class member. Each Class member can review how the Plan of Distribution applies to each of its identified transactions. It may also challenge the accuracy of the information regarding posted Covered Transactions and submit additional transactions to the Claims Administrator for consideration as Covered Transactions.

Since going live, close to 10,000 distinct visitors have visited the website. In addition, the Claims Administrator received, as of April 15, approximately 700 claims. The last date for submission of claims is May 27, 2016.

Against this backdrop, only twenty-one requests for exclusion were timely submitted by February 29.6 Five entities are responsible for these twenty-one requests.7

There are effectively four objectors who submitted timely objections by February 29.8 The objectors are MF Global Capital LLC ("MF Global"); Silver Point Capital, L.P. ("Silver Point");9 Saba Capital Management ("Saba"); and Anchorage MTR Offshore Master Fund, L.P. ("Anchorage").10 Class Counsel have allowed the objectors to speak with their experts, and have had their experts conduct complex analyses of the damages dataset to analyze and respond to the objectors' critiques and proposals.

V. Fairness Hearing

The Fairness Hearing was held on April 15, 2016. Class Counsel and Michael Herrera, Senior Staff Counsel for LACERA, appeared at the hearing, as well as counsel for all defendants. Also present was counsel for a non-objecting Class member, BlueMountain Capital Management LLC.11 Of the objectors, only MF Global, Saba, and Silver Point were represented by counsel at the Fairness Hearing.12 The objectors were given an opportunity to be heard, but only Silver Point's counsel spoke.

DISCUSSION

I. Judicial Approval of Class Action Settlement

Pursuant to Rule 23(e), Fed. R. Civ. P., any settlement of a class action must be approved by the court. In determining whether to approve a class action settlement, the district court must "carefully scrutinize the settlement to ensure its fairness, adequacy and reasonableness, and that it was not a product of collusion." D'Amato v. Deutsche Bank, 236 F.3d 78, 85 (2d Cir. 2001) (citation omitted). In doing so, the court must "eschew any rubber stamp approval" yet simultaneously "stop short of the detailed and thorough investigation that it would undertake if it were actually trying the case." City of Detroit v. Grinnell Corp., 495 F.2d 448, 462 (2d Cir. 1974).

In making its determination, a district court should "review the negotiating process leading up to the settlement for procedural fairness." Charron v. Wiener, 731 F.3d 241, 247 (2d Cir. 2013). A court should assess whether the settlement resulted from "an arm's-length, good faith negotiation between experienced and skilled litigators," id., and whether plaintiffs' counsel engaged in discovery "necessary to the effective representation of the class's interests." D'Amato, 236 F.3d at 85.

The court must also evaluate the substantive fairness of a settlement by considering the nine factors set forth in Detroit v. Grinnell Corp.:

Charron, 731 F.3d at 247 (citation omitted).

Finally, the determination should recognize that there is a "strong judicial policy in favor of settlements, particularly in the class action context." Wal-Mart Stores, Inc. v. Visa U.S.A., Inc., 396 F.3d 96, 116 (2d Cir. 2005) (citation omitted). Similarly, "[t]he compromise of complex litigation is encouraged by the courts and favored by public policy." Id. at 117 (citation omitted).

A. Procedural Fairness

This Settlement was achieved after intense, lengthy negotiations among well-represented adversaries who were assisted by an able mediator. The existence and size of this Settlement is attributable in no small measure to the skill of Class Counsel and the litigation strategy it employed. Discovery was extensive and swiftly conducted. The opposing parties were able to assess quickly, in detail, and with care, the strength of the plaintiffs' theories of liability and damages.

The mediator has praised the work of Class Counsel as "one of the finest examples of efficient and effective lawyering by plaintiffs' counsel that" he has ever witnessed. In making this judgment, he took note of the complexity and size of the litigation, and the speed with which Class Counsel achieved a result of this magnitude. He also reports that the settlement negotiations were "conducted at arm's-length by sophisticated, knowledgeable, and fully-informed counsel who consulted directly with senior client representatives throughout the process."

B. Substantive Fairness

In addition, consideration of the Grinnell factors strongly favors approval of the Settlement.

1. Complexity, Expense, and Likely Duration of the Litigation

This is highly complex litigation. Antitrust cases are often challenging to investigate and litigate, and this litigation is no exception. It has also been extremely expensive to litigate. The Settlement was preceded by a period of intensive fact discovery, involving the production and examination of many millions of pages of documents. Twenty-seven depositions were conducted and more had been scheduled to occur. Only five months remained in the fact discovery period. The litigation, had it not been resolved through settlement, would have been very expensive to complete and may very well have required a trial of the plaintiffs' claims.

2. The Reaction of the Settlement Class

The Class has received effective and sufficient notice of the Settlement and the reaction of the Class has been overwhelmingly positive. While the reaction of a class to a settlement is always important, it is an especially telling here since the Class is composed of sophisticated parties who participate in buy-side trading of CDS. Out of almost 14,000 Class members, only twenty-one requests for exclusion were timely submitted. Only four objections have been pursued. This very low number of objections and requests for exclusion supports a finding that the Settlement is fair. See Grinnell, 495 F.2d at 462.

3. Stage of the Proceedings and Amount of Discovery

As noted above, the Settlement was achieved in the midst of the period assigned for fact discovery. The parties reached an agreement in principle on the eve of the date on which the plaintiffs' motion for class certification was due.

4. Risk Regarding Liability, Damages, and Class Certification Through Trial

The risk of establishing liability is somewhat difficult to assess in this case since the Settlement occurred before the filing of summary judgment motions or trial, and in the absence of the filing of any government charges arising from the alleged misconduct. The defendants intended to argue that they had not conspired with each other to violate our antitrust laws, that the CDMX would not have been a viable exchange platform, and that the plaintiffs' theory of damages was seriously flawed, among other things. Moreover, both the DOJ and EC investigations were closed without the filing of any charges against the Dealer Defendants. On the other hand, the size of the Settlement suggests that the plaintiffs' analysis of the document production and development of evidence through depositions of the defendants' witnesses held promise for the plaintiffs' success at trial and placed the defendants at risk of a substantial adverse verdict. Given the commonality of issues in the plaintiffs' theory of its case, it is likely that a class would have been certified and, if certified, maintained through the conclusion of the litigation.

The issues related to damages would have been hotly contested at each stage of the proceedings. Causation and the methodology for establishing damages would have been litigated extensively.

5. Ability of Defendants to Withstand Greater Judgment and the Range of Reasonableness of the Settlement Fund and Recovery

The defendants are generally large financial institutions and have the ability to withstand a greater judgment than the amount they each contributed to the Settlement. Nevertheless, the Settlement Fund, at nearly $1.9 billion, is a very substantial amount.

In fact, no one disputes that the Settlement is reasonable, both in light of the best possible recovery and in light of all the attendant risks of litigation. The mediator has praised the Settlement as "exceptional." In his opinion, the Settlement is not just fair and adequate, but "exceedingly favorable" to the Class, reflecting a recovery "well beyond" what he expected could be achieved.

To place the mediator's assessment in context, the plaintiffs' preliminary damages estimate forecast damages at roughly $8 to $12 billion. The recovery here, therefore, reflects 15 to 23% of the amount which plaintiffs may have sought at trial. Class Counsel estimate that over 1,300 Class members will each receive payments from the Settlement Fund exceeding $100,000, and over 230 of these will each receive more than $1,000,000. Given this significant recovery for the Class, it is unsurprising that no Class member has objected to the amount or fairness of the Settlement.

II. Objections by Class Members

Four sets of objection have been brought by Class members. The objections address the Plan of Distribution and the terms of the Release. None of the objections requires an alteration of the Plan or the Release.

A. Plan of Distribution

A district court "has broad supervisory powers with respect to the . . . allocation of settlement funds." In re Holocaust Victim Assets Litig., 424 F.3d 132, 146 (2d Cir. 2005) (citation omitted). The plan of allocation must "meet the standards by which the settlement [is] scrutinized — namely, it must be fair and adequate." Hart v. RCI Hosp. Holdings, Inc. No. 09cv3043 (PAE), 2015 WL 5577713, at *12 (S.D.N.Y. Sept. 22, 2015) (quoting In re WorldCom, Inc. Sec. Litig., 388 F.Supp.2d 319, 344 (S.D.N.Y. 2005)). A plan "need only have a reasonable, rational basis, particularly if recommended by experienced and competent class counsel." Id. (quoting In re WorldCom, Inc., 388 F. Supp. 2d at 344). A principal goal of a plan of distribution must be the equitable and timely distribution of a settlement fund without burdening the process in a way that will unduly waste the fund.

"[I]n the case of a large class action the apportionment of a settlement can never be tailored to the rights of each plaintiff with mathematical precision." In re PaineWebber Ltd. P'ships Litig., 171 F.R.D. 104, 133 (S.D.N.Y.), aff'd In re PaineWebber Inc. Ltd. P'ships Litig., 117 F.3d 721 (2d Cir. 1997). The challenge of precisely apportioning damages to victims is often magnified in antitrust cases, as "damage issues in [antitrust] cases are rarely susceptible of the kind of concrete, detailed proof of injury which is available in other contexts." J. Truett Payne Co., Inc. v. Chrysler Motors Corp., 451 U.S. 557, 565 (1981) (citation omitted)); See also In re Elec. Books Antitrust Litig., No. 11md2293 (DLC), 2014 WL 1282293, at *16 (S.D.N.Y. Mar. 28, 2014).

This sophisticated Class is in an excellent position to swiftly and competently assess whether the Plan, and the model upon which it is based, achieves a fair distribution of this very sizeable Settlement Fund. It has spoken. No Class member has objected that the Settlement Fund is inadequate. Many have already filed claims. Very few have opted out. Only four sets of objections to the Plan have been filed. This record is an overwhelming endorsement of the Plan and the fairness with which it will measure each member's entitlement to a distribution.

The Notice required any objections to the Settlement to be filed by February 29, 2106. Four sets of Class members objected to different elements of the Plan of Distribution. None of their objections provide a basis to alter the conclusion that the Plan is fair and entitled to adoption. Taken together, the four objectors make essentially three different types of arguments about the Plan. They complain that categories of linked or packaged trades are being over-compensated, that certain categories of investors will receive a disproportionate amount of the Settlement Fund, and that the 20% Compression Rate should not apply after December 31, 2012, when certain reforms in the Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. No. 111-203, 124 Stat. 1376 (2010) (the "Dodd-Frank Act"), that affect the CDS market took effect.

1. Overcompensation of Packaged Trades

MF Global, Silver Point, Saba, and Anchorage each contend that certain transactions were conducted as linked trades and enjoyed a zero or de minimis bid/ask spread on one leg of the trade. They object that, because the Plan does not treat the transactions as linked, it overcompensates Class members who engaged in the packaged trades.

MF Global lists types of packaged trades13 and proposes a methodology for more fairly calculating the spread for six types of packaged transactions. It describes these proposals as "non-controversial" adjustments. Silver Point contends that the full bid/ask spread is not charged on both legs of an index arbitrage. Saba contends that a "significant" number of trades undertaken by the Class were index arbitrage packages or correlation trade packages, and that the Dealer Defendants "significantly" mark down the spreads for such trades. It suggests that "off market trades" be included in any compilation of such trades and that additional information be obtained from the Dealer Defendants regarding these trades. Anchorage explains that a spread was generally paid on only one leg of a multi-leg CDS index related trade. It suggests three sets of adjustments.

As described above, the Plan identifies a type of linked trade — index rolls — and makes adjustments for that category of linked trade through the application of an objective, conservative test. The request to make adjustments to the Plan to identify more linked trades must be rejected.

As a practical matter, it is almost impossible to identify linked trades. To be a true package, linked trades have to be executed simultaneously. Only with simultaneous execution will the investor avoid the risk that the market will move against it in the interval between the trades. But, the data currently available to the Class does not permit the identification of simultaneous trades. The DTCC data does not identify the time of day when a CDS transaction occurs, only the trade date. No objector has identified a feasible way, much less a quick and inexpensive one, to obtain data for all CDS trades that identifies the precise time of the trades.

Beyond that problem, CDS are complex instruments and they are associated with many different complex trading strategies. Linkage of trades and arbitrage of a portfolio's investments can occur in many different ways, through the combination of many different instruments.14 Therefore, any attempt to construct a process to identify truly linked trades will necessarily be under-inclusive, and will almost certainly be over-inclusive as well.

Each of the potentially packaged trades identified by the objectors presents its own unique hurdles. One example will suffice. The packaged trades on which the objectors focus most intently is the index arbitrage. In an index arbitrage, an investor makes offsetting buys and sells of an index CDS and of its constituent single-name CDS. But, it is highly unlikely that an investor could execute an arbitrage of an index CDS with offsetting purchases or sales of all the index's constituents through a single dealer. As Dr. Unni explains, if a dealer's quotes for the index CDS are misaligned with the dealer's quotes for the many single-name CDS that form the constituents of the underlying index, and the investor seeks to execute simultaneous offsetting transactions with both the index and its constituents through this dealer, the dealer has an opportunity to move its quotes to tighten the spread or eliminate the arbitrage.15 Thus, the likelihood is that any investor who actually wishes to engage in an index arbitrage will try to execute the arbitrage through multiple dealers in order to mask its strategy. But, such a multi-dealer arbitrage strategy presents its own separate challenges, including how to calculate any appropriate discounted spread. The need to combine trades conducted through multiple dealers also makes it exceedingly difficult to apply an objective, neutral standard to identify a true arbitrage even when, as a theoretical matter, an opportunity for a reduced spread might have existed. Because such a multi-dealer arbitrage is complex and risky, it is also likely to be rare. Indeed, once an arbitrage is fragmented in this way, there is a real question as to whether it even qualifies as a packaged trade.

The objectors' varying proposals for identifying this type of packaged trade underscore this very problem. They use different tests to identify the packages to which the Plan should apply some yet-to-be-determined discounted spread. Saba opines that an index arbitrage should be identified as one in which there was the simultaneous trade of the index and all of the single-name entities making up that index. Anchorage asserts that the index arbitrage should be identified as one in which an index was traded on the same day as at least 75% of its constituents. MF Global contends that the index arbitrage should be identified as one in which the index was traded on the same day as at least 60% of its constituents.16 These competing and contradictory proposals themselves reflect the absence of any reliable, conservative, and fair standard for identifying an index arbitrage for purposes of distributing the Settlement Fund fairly to all members of the Class.

In anticipation of the April 15, 2016 Fairness Hearing, Class Counsel submitted its motion for final approval of the Settlement on April 1. This April 1 submission fully addressed each of the objections and explained in convincing fashion why the Plan should not be altered to account for more packaged trades. Then, in the days immediately preceding the Fairness Hearing, the objectors made additional submissions that included entirely new objections. To the extent that the objectors presented new objections in their April submissions, those objections are untimely and must be rejected on that basis alone.17 In any event, none of the eve-of-Hearing objections, whether new or renewed, provides a ground for altering the Plan.

On April 12, MF Global made a submission that added several new objections.18 That submission was supported by a declaration from a CDS trader adding personal observations based on his experience in the industry, but no analysis, study, or citation to research that would provide a basis to reject the detailed analysis presented by Class Counsel and its experts.

The MF Global submission also reflects a fundamental misunderstanding of the goal of any plan of distribution. A plan of distribution is not defective simply because it does not account for every individual trading strategy that may exist in a marketplace. As described above, a plan must fairly distribute the settlement funds across the entire class. No one denies that there are a variety of trading strategies that were used by many CDS market participants that are not accounted for in the Plan. But, unless there is reliable and fair way to both identify linked trades and adjust the spread associated with those trades, then that trading strategy should not and cannot be a component of a plan of distribution that seeks to treat all class members fairly. It is telling in this regard that none of the objectors has demonstrated that any unfairness will accrue to any specific group of investors if the Plan does not incorporate recognition of and adjustments for the particular kinds of linked trades on which they focus attention, much less that adoption of any of their proposals (assuming it were feasible to adopt any of them) would improve the fairness of the distribution to that group or to the Class generally. Finally, there has been no demonstration that the various idiosyncratic trading strategies discussed by the objectors account for any material portion of CDS trading.19

Silver Point also made a supplemental submission on April 12. While its timely objection made only a brief reference to the need to discount the spread for index arbitrage trading, its April 12 submission not only elaborates on that objection but also makes many new objections in a broad-based attack on the Plan.20 The Silver Point presentation does not come to grips with the detailed explanations of the Plan provided by Class Counsel and its experts. Nor does it provide any proposal for adjustments to the Plan that would make it more complete, reliable, or fair.

At the Fairness Hearing, Silver Point chose to make two points in oral argument. It asserted first that the DTCC dataset was populated with incorrect Trade Dates. Silver Point believes that a large number of Trade Date errors that it recently identified are associated with its assigning its rights in a CDS to another trader, although it did not believe that the errors affected the identification of its Covered Transactions or the calculation of its damages. It speculated that such error might overcompensate others in the Class by failing to identify a large number of index rolls. Second, Silver Point chose to emphasize that the Plan should use a wider spread for non-round tenors than the more liquid and therefore cheaper round tenors. Silver Point acknowledges that little or no data exists to identify the appropriate spread for non-round tenors and it has not offered a feasible way to do so. Neither of these points were made in any timely objection.

Since Silver Point had not provided Class Counsel with any data about incorrect Trade Dates, the Court invited Silver Point to do so. Silver Point submitted data to Class Counsel on April 18 and a suggestion that its experts evaluate using a Novation Date instead of the Trade Date. As explained in his April 22 submission, Dr. Unni determined that the transactions challenged by Silver Point consist almost entirely of assignments. Under the Plan, index rolls do not include assignment transactions. Even if the methodology for identifying index rolls were expanded to include assignment transactions as suggested by Silver Point, the pro rata share of the Settlement Fund attributable to each Class member would remain largely unchanged. Indeed, Silver Point's own spread inflation would decrease slightly. Silver Point's suggestions during the Fairness Hearing do not provide a reason to find that the Plan should be altered.

Saba made an additional submission on April 13. It acknowledges that Class Counsel provided Saba with the data from the model that was used to calculate Saba's potential recovery, but adds two new suggestions for altering the model in an effort to identify more packaged trades.21 These new suggestions would require a massive reworking of the entire Plan, would substantially delay any distribution, would cause an uproar from other Class members, and reflect a flawed understanding of the DTCC data.

Anchorage filed a brief letter on April 14 maintaining its objection to the Plan, but not addressing any of the analysis of that objection in the April 1 filings by Class Counsel.22 For the reasons explained in this Opinion, at the Fairness Hearing, and in Class Counsel's submissions, its single remaining objection does not require any change to the Plan.

There is one new request by an objector that deserves discussion, even though it is untimely. Silver Point now requests that it be given access to the entire database and an opportunity to work to try to improve the Plan. It has pointed to no legal authority to support this request. In January, Silver Point was given detailed information showing how the Plan's model applied to 6,400 of Silver Point's own transactions. It has not shown that additional access to its competitors' trading data is necessary for it to understand how the Plan works, how the Plan's implementation will impact it, or how the Plan's design might be improved.

There are several reasons to deny Silver Point's April 12 request. Given the access it has already had to the plaintiffs' experts and to the mechanics underlying the Plan, there is no reason to find that either more time or more data will permit Silver Point to develop for the first time a meritorious suggestion for improving the Plan. Moreover, giving Silver Point the access it requests risks substantial injury to other Class members if its competitors' data is used improperly. Finally, Silver Point's request would substantially delay the distribution of the Settlement Funds to the Class.

In sum, Class Counsel were responsive to each of the issues raised by the objectors and to questions posed by all Class members. Class Counsel spoke with the objectors frequently and let the objectors speak directly to the expert consultants retained by the Class. Class counsel also, at considerable expense, asked their experts to perform analyses of the dataset to respond to the objectors' proposals. None of that work, which is reported in detail in Dr. Unni's submissions of April 1, 14, and 22, suggests that there is a reliable way to correctly identify any of the proposed packaged trades to which one or more of the objectors contends a discounted spread should be applied, or that it would materially improve the fairness of the distribution to do so. As significantly, the objectors have not presented a model that would improve the Plan. Nor have they provided a reliable basis to find that the use of the Plan's current model treats any particular Class member or group of Class members unfairly.

2. Overcompensation of Categories of Class Members

MF Global and Silver Point complain that the Plan treats all transactions and therefore all traders equally when, in fact, the defendants treated categories of traders differently. While these two objectors contend that the Dealer Defendants offered certain classes of traders tighter spreads, they disagree as to whom the Dealer Defendants discriminated against.

MF Global contends that the Dealer Defendants "generally" treated Class members differently depending on whether they viewed the account as a "fast" money versus a "real" money account. According to MF Global, the Dealer Defendants offered wider bid/ask spreads to potential competitors, and active or speculative traders (that is, "fast" money), but tighter spreads to the remaining 75% of the Class members ("real" money). MF Global opines that "real" money accounts enjoyed an approximately 25% lower bid/ask spread "on average" and that this is correctly captured by the Plan's calculation of a trading day's average spread, which is built upon the narrowest observed spread each hour. It suggests that those Class members who can demonstrate that they were only offered the opportunity to trade with the Dealer Defendants "at consistently" wider bid/ask spreads "should be able to recover damages based on applying the bid/ask spread inflation to the actual spreads at which they entered into [a] Covered Transaction."

In contrast, Silver Point believes that the Dealer Defendants offered discounted bid/ask spreads to their "most active" clients, and discriminated against smaller traders. It does not make any proposal for how to identify the disadvantaged or advantaged group or for the size and system of applying any adjustment.

The objectors have not shown that adjustments should be made based on the identity of the buyer. As reflected in empirical studies, bid/ask spreads in the CDS market are driven by the nature of the particular instrument being traded and market conditions more generally. Specifically, it is driven by the types of product (e.g., whether an index or single-name CDS), the terms of the CDS contract at issue, the company or companies to which the product applies, and market conditions more generally. The Plan is so specific to each CDS contract and the prevailing spread for that product that the model properly accounts for each of the major forces that should be taken into account here.

It is noteworthy that in discussing discrimination against classes of traders, the objectors disagree as to who precisely was advantaged and disadvantaged in their negotiations with the Dealer Defendants. In addition, they have not pointed to empirical research supporting their premise that a buyer's identity had any effect on the spreads. Nor have they presented any fair and efficient process for identifying which traders belong within an advantaged or disadvantaged class. Despite these limitations, Class Counsel took the objection seriously. Dr. Unni did an analysis of some of MF Global's and Silver Point's most heavily traded CDS products during 2010 and 2011 and found no unfavorable bias against them or evidence of systematic bias in the market. These objections, which these two objectors have now essentially abandoned, provide no ground for rejecting or revising the Plan.

3. Uniform Compression Rate After Dodd-Frank Reforms

In its April 12 submission, Silver Point argues for the first time that the Plan's 20% Compression Rate should not be applied across the entire Class period. Silver Point argues that reforms in the Dodd-Frank Act led to greater transparency and dissemination of information in the CDS market. These reforms went into effect on December 31, 2012, and Silver Point contends that some unidentified but different compression rate should be applied after that date. Silver Point and Class Counsel presented oral argument on this issue at the Fairness Hearing.23

This objection does not require a change to the Plan of Distribution. As Class Counsel argued during the Fairness Hearing, the CDS spreads themselves are the most effective barometer of market efficiency. To the extent that spreads tightened generally after the implementation of the Dodd-Frank Act, then the 20% Compression Rate will be applied to that narrower set of spreads. Because the CDS market is so complex, with the multiple factors described above affecting the movement of bid/ask spreads, any attempt to tinker with the Compression Rate is an exercise in pure speculation. Applying different Compression Rates to two different periods would ultimately be arbitrary and less data-driven than the Plan's approach.

B. The Scope of the Release

Silver Point and MF Global both make objections related to the scope of the Release. MF Global argues that its claims arising from the defendants' efforts to prevent MF Global from launching its own clearing and market-making business may be barred by the scope of the Release. Silver Point objects to the release of any claims against the defendants "based on post-September 2015 trades."

Parties may "reach broad settlement agreements encompassing claims not presented in the complaint in order to achieve comprehensive settlement of class actions, particularly when a defendant's ability to limit his future liability is an important factor in his willingness to settle." In re Literary Works in Elec. Databases Copyright Litig., 654 F.3d 242, 247-48 (2d Cir. 2011). Accordingly, "class action releases may include claims not presented and even those which could not have been presented as long as the released conduct arises out of the `identical factual predicate' as the settled conduct." In re Am. Exp. Fin. Advisors Sec. Litig., 672 F.3d 113, 135 (2d Cir. 2011). The determination of whether a claim pleaded in a separate lawsuit is predicated on sufficiently similar facts as the class action claim to be barred by a class action settlement release "is inherently an individualized, fact-specific one." In re WorldCom, Inc., 388 F. Supp. 2d at 342 n.36.

The scope of a release is also limited by the adequacy of representation doctrine. "[A]dequate representation of a particular claim is determined by the alignment of interests of class members, not proof of vigorous pursuit of that claim." Wal-Mart Stores, Inc. v. Visa U.S.A., Inc., 396 F.3d 96, 113 (2d Cir. 2005). Because a settlement may bar future claims, "it is essential . . . that there be adequate notice of the effect of the release and compensation for released claims." In re WorldCom, Inc. Sec. Litig., No. 02cv3288 (DLC), 2004 WL 2591402, at *12 (S.D.N.Y. Nov. 12, 2004).

The Court has examined the Release with care. It is precise, reasonable, and appropriate to the circumstances of this case. Class members were given adequate notice of the terms of the Release on January 11, 2016, and were in a position to make an informed decision to opt out by February 29 if they were unhappy with the breadth or effect of the Release on any lawsuit they were contemplating.

Beyond these observations, it is premature to rule on whether any claim that might be brought in the future by some party would or would not be barred by the terms of the Release. If another lawsuit is brought and an application is made to this Court to enforce the Release, the application will be considered at that time.

C. Appealability of Claims Administrator Determinations

Several of the objectors had initially objected that the Settlement website failed to include some of their Covered Transactions. There is a process in place for Class members to bring additional trades to the attention of the Claims Administrator. Silver Point had complained that it has no appeal right should the Claims Administrator reject their proposed CDS transactions. Through an Order issued on April 18, it is now clear that any Class member has a right to appeal an adverse determination of the Claims Administrator to this Court. This includes a determination of the Claims Administrator regarding additional Covered Transactions.

III. Attorneys' Fees, Costs, and Incentive Awards

Class Counsel also sought approval for an award of $253,758,000 in attorneys' fees, $10,181,190.76 in expenses, and incentive awards of $200,000 and $193,700 for Class representatives LACERA and Salix, respectively. No Class member objected to those applications.

A. Attorneys' Fees

"Attorneys whose work created a common fund for the benefit of a group of plaintiffs" may receive "reasonable" attorneys' fees from the fund. Victor v. Argent Classic Convertible Arbitrage Fund L.P., 623 F.3d 82, 86 (2d Cir. 2010). Courts "may award attorneys' fees in common fund cases under either the `lodestar' method or the `percentage of the fund' method," although "the trend in this Circuit is toward the percentage method." McDaniel v. Cty. of Schenectady, 595 F.3d 411, 417 (2d Cir. 2010) (citation omitted). The six Goldberger factors "are applicable to the court's reasonableness determination whether a percentage-of-fund or lodestar approach is used." Id. at 423. They are:

In re World Trade Ctr. Disaster Site Litig., 754 F.3d 114, 126 (2d Cir. 2014) (quoting Goldberger v. Integrated Resources, Inc., 209 F.3d 43, 50 (2d Cir. 2000)).

Under the Private Securities Litigation Reform Act, there is a well-recognized rebuttable "presumption of correctness" given to the terms of an ex ante fee agreement between class counsel and lead plaintiff. See Flanagan, Lieberman, Hoffman & Swaim v. Ohio Pub. Employees Ret. Sys., 814 F.3d 652, 659 (2d Cir. 2016); see also In re Cendant Corp. Litig., 264 F.3d 201, 282 (3d Cir. 2001). But, the district court "must be mindful that it must act as a guardian of the rights of absent class members in assessing whether a presumption of correctness has been properly refuted and then, if indeed it has, determining on its own the appropriate fee allocation." Flanagan, 814 F.3d at 659 (citation omitted). There is no reason not to apply such a rebuttable presumption to the examination of an ex ante fee arrangement in a common fund antitrust case, at least where it has been negotiated with a sophisticated benefits fund with fiduciary obligations to its members and where that fund has a sizable stake in the litigation.

The requested attorneys' fees are calculated directly from the retainer agreement that Lead Plaintiff LACERA and Pearson Simon, its original counsel, negotiated in advance of LACERA joining this litigation. LACERA, with investment assets of over $48 billion, is one of the largest county retirement systems in the United States. At the Fairness Hearing, LACERA's Senior Staff Counsel, who was responsible for negotiating this agreement, obtaining board approval of it, and supervising the litigation, explained the process for arriving at the agreement. In response to LACERA's request for representation proposals, it received sixty-seven separate bids from counsel. LACERA evaluated those bids, considering both their terms and the quality of counsel. It then selected and negotiated a fee agreement with Pearson Simon. The agreement was reviewed and approved by LACERA's board.

The retainer agreement between LACERA and Pearson Simon provides for the following fee structure in the event the litigation is settled during the discovery period.24

The fee requested by Class Counsel is derived from this agreement. LACERA fully supports the fees requested by Class Counsel, and as noted, no Class member has objected.

The $253,758,000 in attorneys' fees which Class Counsel has sought is approximately 13.61% of the monetary value of the Settlement Fund. The loadstar calculation submitted by Class Counsel totals over $41 million as of April 1, reflecting over 93,000 hours of work by Class Counsel. This amount is equivalent to a loadstar multiple of just over 6.

While LACERA does not have the largest stake in the Settlement Fund, it has a very substantial one.25 This substantial stake gave LACERA a strong incentive to negotiate the retainer agreement with care when selecting counsel, as well as a strong incentive to examine the Settlement and the performance of Class Counsel with care.

The Goldberger factors weigh in favor of approval of Class Counsel's fee request. Although the requested fee is enormous, as just described, Class Counsel poured enormous resources into the litigation of this action, all on a contingency basis. It invested over 93,000 hours of time in this litigation, most of it over less than one year. The magnitude and complexity of this case have already been described, as has the risk of litigation. The quality of work performed on behalf of the Class by its counsel has been superb, as evidenced by Class Counsel's efficient and aggressive discovery work, the lack of objections to the large fee request,26 and the highly favorable outcome achieved for the Class in record-setting time. This success was obtained against a backdrop of government investigations that produced no charges against the Dealer Defendants.

The fee grid which LACERA negotiated with its counsel is generous.27 But, there is no reason to doubt that LACERA negotiated the best fee structure that it could given the difficulties it anticipated facing in this litigation and its desire to have excellent representation if it were to pursue a complex antitrust claim against many of the largest financial institutions in the nation. Indeed, the 13.61% in fees requested by Class Counsel is consistent with fees awarded in other large antitrust cases. See Brian T. Fitzpatrick, An Empirical Study of Class Action Settlements and Their Fee Awards, 7 J. Empirical Legal Stud. 811, 831 tbl. 7, 839 tbl. 11 (2010). In this context, then, the requested fee is reasonable in comparison to the size of the recovery for the Class.

Finally, there are significant public policy considerations that weigh in favor of approval. It is important to encourage top-tier litigators to pursue challenging antitrust cases such as this one. Our antitrust laws address issues that go to the heart of our economy. Our economic health, and indeed our stability as a nation, depend upon adherence to the rule of law and our citizenry's trust in the fairness and transparency of our marketplace. See F.T.C. v. Phoebe Putney Health Sys., Inc., 133 S.Ct. 1003, 1010 (2013) (noting the "fundamental national values of free enterprise and economic competition that are embodied in the federal antitrust laws").

B. Costs and Expenses

Class Counsel also sought reimbursement for over $10 million in expenses incurred. Most of these expenses were incurred in connection with retention of experts. The expert work was essential to the litigation and invaluable to the Class. There were no objections to this application and it was approved.

C. Incentive Fees

Class Counsel also sought an incentive award of $200,000 and $193,700 for Class representatives LACERA and Salix, respectively. The Salix incentive award request is brought on behalf of three individuals who have contributed significantly to Class Counsel's efforts in this litigation.28 These requests have been denied.

While class representatives should be compensated for out of pocket expenses and lost wages, incentive payments should not ordinarily be given. They "raise grave problems of collusion." Reed v. Continental Guest Servs. Corp., No. 10cv5642 (DLC), 2011 WL 1311886, at *4 (S.D.N.Y. Apr. 4, 2011). After all, representative plaintiffs "undertake to represent not only themselves, but all members of the class, in a fiduciary capacity, and are obligated to do so fairly and adequately, and with due regard for the rights of those class members not present to negotiate for themselves." Id. (citation omitted). When the settlement provides for incentive awards to the named plaintiffs not shared by the other class members, "a serious question arises as to whether the interests of the class have been relegated to the back seat." Id. (citation omitted). While there is no basis to find that Class representatives here have been tempted to receive high incentive awards in exchange for accepting suboptimal settlements for absent Class members, such an award would nonetheless inappropriately reward the representative Class members over absent ones.

IV. Rule 7 Bond Request

Class Counsel has stated that it will likely request that the Court require any objector who files an appeal from this Settlement to post a bond under Rule 7, Fed. R. App. P. Should such a request be made, the parties will be given an opportunity to address the following standard and the appropriateness of any bond.

Rule 7 provides that "the district court may require an appellant to file a bond or provide other security in any form and amount necessary to ensure payment of costs on appeal." A Rule 7 bond is prospective in its focus and "relates to the potential expenses of litigating an appeal." Adsani v. Miller, 139 F.3d 67, 70 n.2 (2d Cir. 1998) (citation omitted). The term "costs" in Rule 7 refers to "all costs properly awardable under the relevant substantive statute or other authority. In other words, all costs properly awardable in an action are to be considered within the scope of [the] Rule." Id. at 72 (citation omitted). The Adsani court explicitly rejected a definition of costs that would limit it to those costs enumerated in Rule 39, Fed. R. App. P. Id. at 74-75.

As explained in In re Gen. Elec. Co. Sec. Litig., 998 F.Supp.2d 145 (S.D.N.Y. 2014), Rule 38 of the Federal Rules of Appellate Procedure allows the Court of Appeals to award damages to appellees who are confronted with frivolous appeals. Id. at 151. An appeal is frivolous for the purpose of Rule 38 when it is "totally lacking in merit, framed with no relevant supporting law, conclusory in nature, and utterly unsupported by the evidence." Id. at 153 (citation omitted). Accordingly, "when an objector lodges a frivolous appeal to a class action settlement, a district court may impose a Rule 7 Bond in the amount of the additional administrative expenses that are reasonably anticipated from the pendency of the appeal." Id. A Rule 7 bond may also include attorneys' fees where the district court concludes that the court of appeals might award attorneys' fees as costs under Fed. R. App. P. 38 because the appeal is frivolous.29 Sckolnick v. Harlow, 820 F.2d 13, 15 (1st Cir. 1987); see also In re 60 E. 80th St. Equities, Inc., 218 F.3d 109, 118-19 (2d Cir. 2000) ("Rule 38 sanctions may include the granting of reasonable attorneys' fees to the party forced to defend the frivolous appeal.").

In setting the Rule 7 Bond, "a district court must not create an impermissible barrier to appeal." In re Gen. Elec. Co. Sec. Litig., 998 F. Supp. 2d at 151. As such, there are

Id. at 153 (citing Adsani, 139 F.3d at 76-79). As the Court of Appeals explained in Adsani, the "purpose of Rule 7 appears to be to protect the rights of appellees brought into appeals courts." Adsani, 139 F.3d at 75. In setting the amount of a Rule 7 Bond, a district court may "prejudge[ ]" the case's chances on appeal. Id. at 79. It is neither "bizarre [n]or anomalous for the amount of the bond to track the amount the appellee stands to have reimbursed." Id. at 75.

CONCLUSION

For the reasons stated herein and during the Fairness Hearing, Class Counsel's petition for approval of the Settlement and Plan of Distribution was granted, with the Court retaining jurisdiction to hear any disputes arising from the claims administration process. Class Counsel's application for attorneys' fees and expenses for the Settlement was also granted. Class Counsel's application for incentive awards for Class representatives LACERA and Salix was denied.

SO ORDERED.