In Re Namenda Direct Purchaser Antitrust Litigation

Opinion

UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF NEW YORK

In re Namenda Direct Purchaser Antitrust Litigation No. 15 Civ. 7488 (CM)

OPINION AND ORDER REGARDING PLAINTIFFS’ MOTION FOR ATTORNEYS’ FEES, EXPENSES, AND INCENTIVE AWARDS

McMahon, C.J.: Following this Court’s preliminary approval of a settlement in this pharmaceutical direct purchaser antitrust class action for $750,000,000 (Dkt. No. 947), Class counsel made a timely motion for fees, expenses and incentive awards. (ECF Nos. 925-28.) Class counsel originally sought 27.5% of the common fund (or an award of X), reimbursement of expenses of $5,823,928.91, and incentive awards of $150,000 for each representative plaintiff. After negotiations with an objector, class counsel reduced the request to 21% of the common fund, or $157,500,000; the rest of the proposal remains the same. For the following reasons, Class counsel’s motion for a fee award is granted, but the amount awarded is lower than requested. FACTUAL BACKGROUND In order to prosecute this antitrust case based on the Defendants’ alleged abuse of the generic pharmaceutical approval process under the Hatch-Waxman Act, Class counsel was required to understand: (a) various complexities of patent law, in order to show that Mylan would have prevailed in showing that the ’703 patent was not infringed, and that Forest’s patent claims as well as the patent term extension were invalid, and to rebut Forest’s arguments to the contrary; (b) the biopharmaceutical aspects of NMDA receptor antagonism; (c) the relevant aspects of FDA and CMS drug regulation, including: (i) FDA regulations regarding approval of transfers of manufacturing technology (for Lexapro) from one site to another; and (ii) CMS regulations governing the Medicaid rebate liability consequences of selling an authorized generic

(Lexapro) in various ways. Class counsel then applied their knowledge of those regulations to a forensic examination of Forest’s deal valuation spreadsheets, developed a multi-input economic model to determine the earlier entry date a reverse-payment-free settlement between Forest and Mylan would have borne, determined the most likely market entry dates from ANDA approval and manufacturing capacity points of view for a host of generic companies and Forest’s “authorized generic” “but for” the reverse payment deal between Forest and Mylan, and developed economic modeling of the complicated interaction between the delay of generic Namenda IR entry from the reverse payment (on the one hand) and the hard switch product conversion enabled by that delay (on the other hand), which were interdependent sources of overcharges for direct purchasers. (See Dkt. No. 927, Gerstein Declaration, at ¶ 32.) Class

counsel also had to determine the quantum of damages based on various assumptions (see id. at ¶ 67), and combine all this evidence into a trial presentation that would be comprehensible to a jury (id. at ¶¶ 42-47). Of course, the law or the regulations were not new to class counsel; as they have advised the court repeatedly, they litigate “pay to delay” cases as their bread and butter and have filed dozens of them over the past two decades. The case settled on the eve of trial, October 28, 2019. On December 24, 2019, Class counsel filed a fully executed version of the Settlement Agreement with the Court (ECF No. 919-1), and a Motion for Preliminary Approval (ECF No. 917) requesting that the Court preliminarily approve the Settlement, approve the form and manner of notice to the Class, and set a schedule leading up to and including a Fairness Hearing. On January 6, 2020, this Court concluded that the Settlement between the Class and Forest was arrived at by arm’s-length negotiations by highly experienced counsel after years of

litigation and fell within the range of possibly approvable settlements, and preliminarily approved it. (ECF No. 920, at ¶ 6.) Concurrently, the Court appointed an escrow agent and claims administrator, approved the form and manner of notice to the Class, and set a schedule. (Id. at ¶¶ 7-18.) Thereafter, Forest deposited the settlement fund into an escrow account that is earning interest for the benefit of the Class. (ECF No. 927 at ¶ 56.) Pursuant to the Preliminary Approval Order, Class members had until March 30, 2020 to object to the Settlement and/or Class counsel’s fee request. (ECF No. 920 at ¶¶ 14, 15.) On March 17, 2020, Forest filed a response to the fee request, stating it took no position on the motion, but reemphasizing its position that Plaintiffs had not “succeeded in proving that Forest was liable for the alleged conduct or that Forest’s actions were actually found to be unlawful and

anti- competitive.” (ECF No. 929, at 1.) On March 30, 2020, the National Wholesalers filed an objection to Class counsel’s requested fee, sought more time to take discovery on Class counsel’s lodestar and submit an expert report and an additional brief on fees, and requested that the issue of attorneys’ fees be bifurcated from the issue of final review and approval of the Settlement. (ECF Nos. 932-33.) The National Wholesalers did not object to the Settlement nor dispute the requested expense reimbursement or incentive awards to the named Class representatives. (See ECF No. 932 at 1, 2 n.1, ¶ 5.) On April 1, 2020, the Court denied the National Wholesalers’ bifurcation request and stated that it was not inclined to permit expert testimony on the proposed fee or postpone the final fairness hearing scheduled for May 27, 2020. (ECF No. 935.) Thereafter, Class counsel and the National Wholesalers resolved the Objection. The National Wholesalers have agreed to

withdraw their Objection and support Class counsel’s requested fee of 21% of the gross Settlement amount.

DISCUSSION I. Class counsel’s Costs and Expenses Are Reasonable and Were Necessary to the Result.

There has been no objection to Class counsel’s request for reimbursement of costs and expenses of $5,823,928.91. These expenses were itemized by category for the Court’s convenience. (See ECF Nos. 926 at 24; 927 at ¶¶ 71-72.) Class counsel’s motion for an award of expenses in that amount is, therefore, GRANTED. II. A Reasonable Incentive Awards for the Class Representatives Is $75,000

There was no objection to the incentive awards of $150,000 for each of the two representative plaintiffs, and I cannot deny that like amounts have been awarded in in similar cases. But frankly, this was attorney-driven litigation. All the class representatives really did was sit for a deposition. (See ECF Nos. 926 at 24-25; 927 at ¶¶ 75-81.) As far as the court is concerned, they made only a minimal contribution to the prosecution of the case. An incentive award of $75,000 to each class representative is reasonable in this case. III. The Proposed Award of Attorneys’ Fees Is Not Fair and Reasonable. Class counsel’s request for an award of attorneys’ fees in the amount of $157,500,000 (plus proportionate accrued interest) -- i.e., 21% of the gross settlement amount -- is neither fair nor reasonable. I start from the proposition that class counsel have earned a “reasonable” attorney’s fee.

The word “reasonable” means “as much as is appropriate” or “fair.” A subsidiary definition of the word is “moderate.” By that definition, the requested fee award is, in the opinion of the Court, unreasonable. For it is anything but moderate. The fee award is requested on behalf of six law firms, who apparently operate in tandem on “pay for delay” cases, regardless of who is appointed lead counsel. The six firms, and the total billable hours for each firm, are as follows: (a) Berger & Montague, P.C. (12,470 hours); (b) Faruqi & Faruqi LLP (9,699.5 hours); (c) Garwin Gerstein & Fisher LLP (9,268.15 hours); (d) Heim Payne & Chorush LLP (4,243.8 hours); (e) Odom & Des Roches (10,135.6 hours); and (f) Smith Segura & Raphael LLP (6,294 hours). This makes up the “lodestar” amount of $34,769,008.35 in per hour billings.

Class counsel did not send time records along with their motion, instead submitting extremely general affidavits that summarized, in no level of detail whatever, the work done by each firm (not each lawyer from each firm) on the case. (Dkt. No. 927, Gerstein Decl., ¶ 70 & Exs. A-F.) Without having contemporaneous time records, the Court had no way to determine whether two or three lawyers at any firm were doing the work of one; or whether lawyers were billing time for monitoring aspects of the case for which they were not responsible or for attending meetings and depositions at which their presence was not required. Counsel did offer, in a throwaway line, to supply time records if the court wanted to see them. I cannot imagine why counsel would have thought that the court did not want to see them. It would have been irresponsible to decide the motion without having time records. Therefore, at the May 27, 2020 hearing finally approving the settlement, I asked for production of the time records, so that I could evaluate the reasonableness of the lodestar. Each

of the firms representing the Class then made their detailed billing records available to the Court on an in camera basis. Those records show that, as the court suspected, lawyers from more than one firm repeatedly engaged in duplicative work - or, in some cases triplicative or quadruplicative work – which inflated the lodestar. For example, three of the six firms prosecuting this case attended the deposition of Forest’s 30(b)(6) witness, David Solomon, with the four of the six billing more than $10,000 in preparing for it. The bills demonstrate the same sort of inefficiency in connection with the other depositions. Billable hours valued in the hundreds of thousands of dollars were devoted to litigating and relitigating an issue often raised in reverse-payment pharmaceutical antitrust cases: whether

brand-name drug producers like Forest can rely on the attorney-client privilege to shield discovery of their subjective beliefs regarding the strength of the underlying patent case. This basic issue had been settled in previous pay-for-delay cases involving the same law firms, so the learning curve was not steep. See, e.g., FTC v. Actavis, 570 U.S. 136, 133 S.Ct. 2223, 186 L.Ed.2d 343 (2013); In re Lidoderm Antitrust Litigation, No. 14-md-2521, 2016 WL 4191612 (N.D. Cal. Aug. 9, 2016)). Relying on those recent precedents, Magistrate Judge James Francis (now retired) made clear that any contention by Forest that it could submit its subjective beliefs into evidence without waiving its attorney-client privilege would be judged “based on the guidelines applied in Lidoderm and similar cases.” (Dkt. No. 249, at 15.) In Lidoderm, Judge William H. Orrick of the Northern District of California had ruled that defendants in a reverse- settlement antitrust case waive privilege over their attorneys’ views of the strength of the underlying patent case “if defendants choose to rely on a subjective belief.” Lidoderm, 2016 WL 4191612, at *6.

Yet, in spite of Judge Francis’s decision that the Lidoderm rule would apply in this case, figuring out how Forest’s decision not to waive attorney-client privilege would impact the production of documents led to an inordinate number of discovery disputes, which played out before Judge Francis and, on occasion, this court. Class counsel did not keep the cost of litigating these disputes down. For instance, on May 18, 2017, after fully briefing a motion to compel based on Forest’s purported privilege waiver, Class counsel sent three attorneys from Garwin Gerstein & Fisher, as well as one from Smith Segura, to an oral argument before Judge Francis on a discovery motion -- a gratuitous outlay of manpower that contributed approximately $10,000 to the lodestar. Moreover, despite that show of force, Plaintiffs were unsuccessful; the motion was denied on the grounds that it failed to identify specific documents over which the

Plaintiffs believed that Forest had waived privilege. (Dkt. No. 348, at 12.) In his opinion and order denying the motion, Judge Francis admonished the parties for their “failure to engage in a meaningful process of meeting and conferring prior to requesting court intervention . . . which confuses the issues and wastes judicial resources.” (Id. at 11.) After filing another motion to compel, one that narrowed the scope of Forest’s purported waiver, Plaintiffs obtained an order compelling Forest to produce….. five documents. (Dkt. No. 387.) But that was not the end of the waiver fight: Class counsel also filed a motion in limine that essentially called for relitigation of an issue already resolved by the Magistrate Judge, as to which there was no further need of briefing. (See Dkt. Nos. 724, 739.) That motion, too, was denied. (Dkt. No. 859 at 13.) It was a waste of time and resources, of both counsel and the court. Ideally, this Court could simply omit the inefficient hours from the lodestar when calculating the fee award. See Luciano v. Olsten Corp., 109 F.3d 111, 117 (2d Cir. 1997) (“[A]

district court can exclude excessive and unreasonable hours from its fee computation by making an across-the-board reduction in the amount of hours.”). However, the scope of that exercise is prohibitive for a court with a busy docket. Additionally, class counsel’s time sheets lack sufficient granularity to separate the productive hours from the wasted ones. For example, some attorneys have block billed entire days both to depositions that multiple firms attended and depositions that only one firm attended; that alone warrants a reduction of the lodestar. See Kortright Captial Partners L.P. v. Investcorp Investment Advisers Ltd., 392 F. Supp. 3d 382, 409 (S.D.N.Y. 2019) (reducing lodestar by 15% due to “inefficiencies and billing deficiencies”). Suffice it to say that this Court has uncovered enough inefficiency in the records provided to conclude that the $34 million lodestar could and should have been lower.

I note also that the underlying legal issues raised by this case were settled in other cases litigated over the past two decades. I do not mean to suggest the case was not complex – it was certainly complex for this court, which has not spent the past two decades immersed in pay-to- delay issues. But it was far less complex to lawyers who have litigated those issues repeatedly over that same period. One might even say that it was a routine case for these lawyers, albeit one with a few unusual twists. The principal such twist, an issue unique to this case, was the collateral estoppel impact of an early ruling by my predecessor, Judge Sweet, involving the so-called “hard stop.” I concede that this issue is not ordinarily presented in a pay-for-delay case. But it did not consume the lion’s share of $34 million in billings. Rather than undertake the arduous task of going time record by time record with the attorneys explaining what they did and why, courts may exercise their discretion “to deduct a

reasonable percentage of the number of hours as a practical means of trimming the fat from a fee application.” Kirsch v. Fleet Street, Ltd., 148 F.3d 149, 173 (2d Cir. 1998) (internal quotation marks omitted). Courts can and often do deal with inflated lodestars by reducing them by a fixed percentage, cutting the proposed multiplier, or both. See, e.g., Sea Spray Holdings, Ltd. V. Pali Fin. Group, Inc., 277 F. Supp. 2d 323, 326 (S.D.N.Y. 2003) (reducing lodestar by 15%); In re Petrobras Securities Litig., 317 F. Supp. 3d 858, 876 (S.D.N.Y. 2018) (reducing requested multiplier from 2.7 to 1.78 where class counsel stood to receive “very large amounts of money in absolute terms . . . under any analysis”). I have concluded that the best course is to leave the lodestar alone but cut the multiplier. The question is, by how much. Were I simply to award the lodestar plus expenses actually incurred (a verifiable number

and, given the number of experts in the case and its duration, not out of line), I would consider the fee award generous and proportionate to the work done (whether or not all that work was necessary – and some of it manifestly was not). Such an award qualifies as a “very large amount[] of money in absolute terms.” However, I do believe that some bump above lodestar is appropriate, given the size of the settlement. I fully understand why Forest settled this case for a large number. Its downside was huge; this was a “bet the company” case. Of course, the risk that the class would lose was also significant; the “hard stop” case was going to be an uphill climb, given that Judge Sweet had enjoined it; and Forest had arguments about the reasonableness of the Mylan settlement under Actavis that this Court considered quite viable, even though the case could not be resolved on a motion for summary judgment. The danger of an adverse verdict, however, was real, and Forest paid dearly to avoid that risk. The class has made out very well indeed. And when a class makes

out very well indeed, courts usually award a success fee, in the form of a bump over lodestar. This Court does not accept class counsel’s argument that an unprecedentedly large settlement warrants a commensurately unprecedentedly large fee award. (See, e.g., Dkt. No. 939, Reply Br., at 6-12.) Rather, in my view, the larger the settlement value, the less the need for a multiplier to nudge the plaintiffs’ bar to pursue large class actions like this one. Where class counsel reaches a settlement for hundreds of millions of dollars, a four-or-five times multiplier strikes me as both unseemly and unnecessary to protect the interests of the class. I am absolutely unmoved – indeed, I am offended -- by the argument that class counsel deserves a humongous fee award in this case because “the winning cases must help pay for the losing ones.” (Id. at 11 (quoting Union Carbide Corp. Consumer Prod. Bus. Sec. Litig., 724 F.

Supp. 160, 168 (S.D.N.Y. 1989).) The reasonableness of a fee award must be considered on a case by case basis, not in the context of lead counsel’s entire body of work. Each case must pay for itself, and if counsel bring meritless cases, neither Forest nor their class client in this case has any duty to mitigate their lawyers; past losses. To help guide this Court’s analysis of a reasonable multiplier, Class counsel have prepared a chart showing the large number of pay-to-delay cases in which courts have awarded fees ranging between 27.5% and 33.33% of the common fund. I take note of several things about the chart. First, it is heavily loaded with cases brought prior to 2013 (which is to say, pre-Actavis cases). Second, the settlements in the few more recently filed cases were substantially less than the settlement in this case, so the reasonableness of the fee award as a percentage of the common

fund is less questionable – again, in the opinion of this court, the larger the settlement, the lower the percentage of the common fund represents a reasonable fee award. Third, there is nothing in these charts that compares the fee award to the lodestar. I have found a number of recent pay to delay case fee awards (albeit in some very old cases) -- some involving these very law firms – which would counsel against a fee award in the range sought by Class Counsel here. For example, I am aware of a fee award in a pay-to-delay case that was 33.33% of the common fund but less than lodestar. See In re Prograf Antitrust Indirect Purchaser Litigation, No. 11-mdl-2242 (D. Mass. May 20, 2015) (Dkt. No. 677) (awarding $20 million, which was 33.33% of the common fund but just .772 of lodestar). The largest multiplier we located in a recent case was 2.3 times lodestar, and the fee amount was less than a quarter of

what plaintiffs are asking here – and that case, like this one, settled on the eve of trial, which means trial preparation had been completed. See Vista Healthplan, Inc., et al. v. Cephalon, Inc., et al., 2020 WL 1922029 at *26 (E.D. Pa, Apr. 21, 2020) (awarding $32 million, which was 33.33% of the common fund, but just 2.3 times lodestar, in connection with a settlement on the eve of trial); Especially notable is Lidoderm, No. 14-md-2421, 2018 WL 4620695 (N.D. Cal, Sept. 20, 2018) – a case that, like this one, took about four years to litigate, and in which many of the same issues were raised. There, Judge Orrick awarded counsel $34 million (equivalent to lodestar in this case), which was 33.33% of the common fund but just a 1.37 times multiplier over lodestar. As far as I can tell, the principal difference between our case and Lidoderm is that counsel did not have to prepare for trial in Lidoderm, though that is conjecture on my part. They did, however, have to litigate certain post-Actavis legal issues for the first time. The same cannot be said of this case. Given the sheer magnitude of the numbers, I am not prepared to award anything like the 4.5 times lodestar requested by class counsel — even if that number is only 21% of the common fund. Because I believe the lodestar to be inflated for the reasons stated, I have decided to award twice the lodestar as a “success fee,” which brings the counsel award to $69,538,016.70, representing approximately 9.3% of the common fund. It is still a handsome payday for counsel. CONCLUSION For the reasons set for above, the Court grants the motion for fees, expenses and incentive awards on part, approving awards in the following amounts: Class counsel shall be reimbursed for its expenses of $5,823,928.91. Each class representative shall receive an incentive award of $75,000. Class counsel shall collect an attorney’s fees award of $69,538,016.70 out of the common settlement fund. The Clerk of Court is directed to close the motion at Docket Number 925. This shall constitute the written opinion of the Court. Dated: June 15, 2020 New York, New York

ChiefJudge BY ECF TO ALL PARTIES