In re Auction Houses Antitrust Litigation

197 F.R.D. 71 | S.D.N.Y. | 2000

OPINION

KAPLAN, District Judge.

Class action lawsuits protect plaintiffs’ rights and promote accountability by permitting dispersed, disorganized plaintiffs who may have suffered only small injuries to find redress by acting as a group where they would lack sufficient incentive to do so individually. At the same time, however, the relationship between a plaintiff class and its attorney may suffer from a structural flaw, a divergence of economic interests of the class and its counsel. The class action mechanism can redound more to the benefit of the attorney than to that of the class, as counsel has an incentive to act in its own best interest, rather than that of the class. Thus, the class action mechanism on occasion has proved to be Janus-faced.

This case has presented an occasion to seek to ease this tension and improve the class action as an instrument of justice. The Court, over the objection of some of plaintiffs’ counsel, employed an auction in selecting lead counsel. This opinion sets forth the basis for the Court’s decision to conduct an auction and the reasoning behind the manner in which it was conducted.

I

A. Background

Defendants Sotheby’s Holding, Inc. and its subsidiary Sotheby’s Inc. (collectively “Sotheby’s”) and Christie’s International PLC and its subsidiary Christie’s, Inc. (collectively “Christie’s”) are in the business of providing auction services of fine and applied arts, furniture, antiques, automobiles, collectibles and other items. The primary sources of revenues of-the defendant auction houses are so-called buyers’ premiums and sellers’ commissions. A buyer’s premium is, typically, a percentage of the price at which the buyer successfully bids on an item at auction that is added to the auction sales price and retained by the auction house. The seller’s commission is a percentage of the auction sales price deducted from the sale proceeds paid to the seller and retained by the auction house.

On December 24, 1999, Christie’s International’s former chief executive officer, Christopher Davidge, resigned abruptly. Subsequently, Christie’s reportedly provided evidence of price fixing with Sotheby’s to the Department of Justice and is said to have received conditional amnesty from criminal prosecution in exchange for providing evidence.

In late January and February 2000, following press reports of these events,¹ a large number of individual and class action complaints were filed in this District against Christie’s and Sotheby’s.² All were referred to the undersigned as related cases. The complaints allege that the auction house defendants, beginning at least as early as January 1, 1993, conspired to manipulate the prices at which they provided non-Internet auction services. The conspiracy allegedly began in 1993 with an agreement to employ a common rate schedule for the premiums charged to buyers. It allegedly was expanded in 1995, when they allegedly agreed to use substantially similar rates for sellers’ commissions. Further, plaintiffs maintain that the auction houses agreed in 1995 to terminate the previous practice of negotiating the amounts of sellers’ commissions with some of their customers.

The first status conference in this case was held on February 23, 2000. Dozens of plaintiffs’ attorneys attended, and a consortium of five law firms immediately proposed themselves as plaintiffs’ executive committee or co-lead counsel in the case. The group of five represented that it had been selected in an earlier meeting attended by all of the plaintiffs’ lawyers, that all possessed the highest credentials, and that the selection was unopposed.³ Nevertheless, a sixth firm then suggested to the Court that it be permitted to join the committee of five.⁴ And yet another objected to the proposed executive committee as too large and instead proposed an alternative executive committee consisting of itself and two other firms. The Court advised counsel that it had not decided how to select lead counsel for the class, if one were certified, but appointed interim lead counsel pending a decision on the class motion.⁵

On April 20, 2000, the Court certified the plaintiff class.⁶ In a separate order, the Court announced that it was considering the use of an auction to select lead counsel.⁷ The order set forth a tentative set of procedures governing the auction and solicited bids from interested counsel. The Court solicited also amicus briefs from a number of well-respected academic authorities in the field and invited counsel to submit briefs commenting on the merits of the proposed auction procedure.

B. First Proposed Fee Structure

The bids contemplated by the Court’s initial order were to contain three parts. First, each bid was to include information concerning the bidder’s qualifications and evidence that the bidder had evaluated fully the risks and potential rewards of the litigation. Second, each bid was to contain two figures, X and Y, on the basis of which the bidder was prepared to serve as lead counsel. The X and Y figures were to be determined based on the bidder’s evaluation of the case and the following fee structure: One hundred percent of any gross recovery obtained by the class or class members up to and including X would go entirely to the class or class members, free of attorney’s fees. One hundred percent of any gross recovery in excess of X, up to and including Y, would go to lead counsel. One fourth of any recovery in excess of Y would be paid to lead counsel as additional compensation and three fourths to the class. Third, each bidder was to submit a brief memorandum setting forth the basis for and supporting the bid. The briefs were to explain the bidders’ respective evaluations of the case, including their assumptions as to .possible and likely recoveries in the event liability were established, and the bases therefore.⁸ The order stated that, if the Court decided to use the bids in selecting lead counsel, lead counsel would be selected on the basis of both the economic terms of the bids and the qualifications of the bidder.⁹

On the appointed day, twenty law firms submitted bids for the position of lead counsel. Several included affidavits by economists supporting their bids. The Court received also three briefs amicus curiae¹⁰ and several submissions from bidders commenting on the merits of the auction procedure.

C. Second Proposed Fee Structure

After considering the comments of the amici and bidders, the Court issued a second order revising the fee structure and soliciting a new round of bids.¹¹ This second proposed fee structure included only one variable, X, rather than two. One hundred percent of any gross recovery up to and including X was to go to the class. And twenty-five percent of any recovery in excess of X would be paid to counsel, with the remainder going to the class. Each bid was to state the value of X pursuant to which the bidder was prepared to serve as lead counsel. As before, bidders were required to submit explanatory memoranda and sworn certifications. As with the previous round of bidding, the Court stated that it would select lead counsel based on its judgment as to which bidder was likely best to serve the interests of the class, taking into account the economic terms of the bids as well as the bidder’s qualifications.

All additional terms contained in the first proposed fee structure were included in the Court’s second proposal as well, including the provision that the attorney’s fee would be inclusive of all costs, disbursements and other charges incurred in connection with the litigation. The Court noted further that it did not intend to disclose any of the bids prior to the earlier of (a) final adjudication of the action, or (b) notice to the class of a proposed settlement, and it ordered that lead counsel thus selected not disclose the terms of its bid to defendants or anyone else without approval of the Court.

D. Disclosure of Interim Committee’s Expert Analysis

Prior to the submission of final bids, it became apparent that interim lead counsel had engaged in settlement discussions with defendants in the course of which they obtained information that their experts used to prepare studies of potential damages. Accordingly, the Court granted a motion by a prospective bidder and gave all counsel access to the damage studies solely for the purpose of preparing bids.¹²

E. In Camera Inspection of Twelve Key Documents

Immediately prior to the Court’s order revising the bidding structure, the Antitrust Division of the Department of Justice, which was and continues to be involved in a criminal investigation of the events here at issue, moved to stay discovery in this case with respect to twelve key documents furnished to it by Christie’s as part of its conditional amnesty agreement. In considering the merits of the motion, the Court ordered that the documents be made available for in camera inspection. The Department of Justice complied, and, after inspection, the Court issued a limited stay of discovery with respect to those documents.¹³

F. Selection of Lead Counsel for the Class

By May 25, 2000, the final day for submission of the bids, the Court had received twenty-one sealed bids for the position of lead counsel, of which seventeen complied with the Court’s proposed fee structure. After careful review, the Court selected David Boies and Richard B. Drubel of Boies, Schiller & Flexner, LLP as lead counsel in the case.

II

A. Problems of Choosing and Compensating Counsel

The modern class action device undoubtedly has proved an important innovation for plaintiffs’ rights. It provides a means of redress to dispersed and disorganized plaintiffs who may have suffered only small injuries and who, in its absence, likely would lack sufficient incentives to bring their own claims.¹⁴ By serving as a vehicle for these claims, the class action plays an important part in enforcement policy in many areas, including securities regulation and antitrust.¹⁵ Nonetheless, the class action mechanism is not free of problems, foremost among them for purposes of this ease difficulties in obtaining counsel who will manage the case efficiently and effectively on behalf of the class and the mismatch of economic incentives between the plaintiff class and its attorney.

When, as here, multiple related claims are filed by different plaintiffs’ attorneys, a case may threaten quickly to become unmanageable, as coordination and strategy problems arise. To remedy the problem of unmanageability, courts traditionally select lead counsel from among the attorneys representing the individual plaintiffs. Lead counsel typically is responsible for working with other counsel to develop positions on substantive and procedural issues in the case, presenting arguments to the court, initiating discovery requests and responses, employing expert witnesses, conducting depositions and insuring that schedules are met.¹⁶ By placing these responsibilities in the hands of one or a small group of counsel, the selection of lead counsel is meant to permit large numbers of cases in which common questions predominate to be prosecuted simultaneously as consolidated or class actions, thereby avoiding duplicated efforts, wasted resources and inconsistent or preclusive judgments. Nevertheless, problems of coordination and duplication of effort may exist.

Lead counsel generally litigates a class action case on behalf of dozens, hundreds or thousands of individual plaintiffs, all of whom seek to recover from defendants. Given the potential for massive plaintiffs’ recoveries in such cases, the lead counsel position may involve a potentially large attorney’s fee. The role therefore has become a coveted prize to be fought over or bargained for among competing plaintiffs attorneys. This process typically occurs in one of two ways, neither of which necessarily leads to an optimal outcome. Often, interested counsel jockey for the lead counsel position, leaving the court to choose one of the contenders, sometimes with little guidance. Counsel thus selected is not necessarily the most qualified or that who will best protect the interests of the class. Alternatively, the plaintiffs’ lawyers negotiate among themselves to select lead counsel or a team of lead counsel, and the choice is presented as a fait accompli for the court summarily to endorse. Here again, the choice is not necessarily in the plaintiffs’ best interests. These two scenarios threatened to replay themselves almost exactly in this case.

B. Compensation — Draivbacks of Commonly Utilized Fee Structures

Plaintiffs’ attorney is, of course, duty bound to act in the best interests of the class. However, because of the manner in which attorney’s fees in class actions frequently are calculated, the optimal recovery for the class often does not yield the highest attorney’s fee. Likewise, the result yielding the highest attorney’s fee is not necessarily in the class’ best interests. This tension can lead counsel to neglect the class’ interests in pursuit of a higher fee. These mismatched incentives predominate when the fee is determined by using either of the two most common fee structures used in common fund cases, the lodestar method and the percentage-of-recovery method.¹⁷

1. Lodestar Method,

The lodestar method essentially compensates plaintiffs’ counsel for the time expended in litigating the case, with the final result sometimes adjusted by application of a multiplier to reflect the risk assumed. It is determined by “multiplying the number of hours expended by each attorney involved in each type of work on the case by the hourly rate normally charged for similar work by attorneys of like skill in the area,” and “[o]nce this base or ‘lodestar’ rate [is] established,” calculating the final fee by then deciding whether to take into account “other less objective factors, such as the ‘risk of litigation,’ the complexity of the issues, and the skill of the attorneys.”¹⁸ Because this figure takes no account of the size of plaintiffs’ recovery, any of several perverse results can obtain.

First, the lodestar method may induce lead counsel to prolong the litigation beyond the optimal point from plaintiffs’ perspective simply in order to accrue more hours.¹⁹

Second, despite incentives to prolong the litigation to a certain point, counsel compensated by the lodestar method has also an incentive to settle the case before it reaches the trial stage, even if trial is in plaintiffs’ best interests.²⁰ This stems from the fact that, while these attorneys share with their clients the downside risk associated with trial (ie., a finding of no liability and therefore no attorney’s fee), they do not necessarily share the potential economic upside (ie., a substantial plaintiffs’ judgment), as trial usually requires few attorney hours relative to pretrial preparation.²¹

Third, the lodestar fee structure creates an incentive for the attorney to do unnecessary work such as filing motions with little merit, taking unnecessary depositions, or demanding production of huge volumes of documents, solely in order to accrue more hours. This risk is exacerbated where the class is represented by a committee of attorneys, rather than a single firm. The involvement of numerous counsel can create pressure to generate sufficient attorney hours to compensate all participating attorneys, and work may be allocated in order to further this objective, rather than in the most efficient and cost-effective manner. Appointment of a committee can lead also to administrative and cost problems, as coordination among committee members is time consuming and costly. All of these factors may result in a higher lodestar without commensurate benefit to the class.

Finally, the lodestar method can lead plaintiffs’ attorney to agree to a less-than-favorable settlement for the class while counsel collects a substantial fee. In the most egregious cases, such settlements have involved non-monetary consideration of virtually no value to all or part of the class while counsel -received substantial fees in cash.²² Although the Court is responsible for assessing the fairness of the settlement and fee application,²³ this task often is difficult. As soon as the parties to a class action lawsuit arrive at a settlement, all have an interest in seeing it approved by the court. This is particularly true in cases in which defendants face potentially enormous damages and therefore are inclined to settle quickly and comparatively cheaply, even if the fee application is unjustifiably high. Further, because attorney’s fees are taken from the common fund, rather than paid separately by defendants, defendants have little interest in contesting the amount of the fee. Instead, once a settlement is agreed upon, the adversary system typically abandons the judge, as plaintiffs’ lawyers and defendants band together to convince the court to approve the settlement and the fee award.²⁴ This creates substantial difficulties for the court in evaluating the fairness of both the settlement and the fee application.²⁵

Evaluation of the fee application can be complicated further where the class is represented by a committee, rather than a single firm. The process of reviewing retrospectively numerous time records and determining appropriate remuneration therefor is arduous, particularly when multiple firms are involved. More seriously, committees of counsel have been known to break down and submit separate contested fee applications to the Court, making accurate retrospective analysis almost impossible.²⁶

2. The Percentage-of-Recovery Method

The percentage-of-recovery method, in contrast, “is a simpler calculation of the fee award as some percentage of the fund created for the benefit of the class,”²⁷ frequently twenty to thirty percent. This method allows the attorney to share in both the upside and the downside risk of the litigation and thereby attempts to re-align the interests of plaintiffs’ and their attorney. Although eliminating incentives to prolong the litigation unnecessarily or accumulate needless hours, however, this method creates perverse incentives of its own. In particular, the percentage-of-recovery method might lead the plaintiffs’ attorney to settle the case prematurely as soon as counsel’s opportunity costs begin to mount.²⁸ Early settlement allows counsel to collect a large fee after investing relatively little time in the case, rather than continuing the litigation in order to maximize plaintiffs’ recovery but receiving a lower marginal rate of return on his or her work. Again, from the plaintiffs’ perspective, this outcome is suboptimal.

S. Collective Action Dilemma in Class Actions

These problems of mismatched incentives are present not only in class actions, but also in traditional attorney-client relationships where both the hourly rate fee structure and the contingency fee can motivate the attorney to pursue his or her own economic interest at the expense of the client. However, they often can be far more severe in the class action context, primarily because classes tend to be large, dispersed and disorganized and therefore suffer from a collective action dilemma not faced by individual litigants.²⁹ This collective action dilemma leads to significantly less monitoring of the attorney by the class and consequential higher agency costs. The danger of a suboptimal result for plaintiffs, therefore, is far more severe in the class action context than in traditional litigation.³⁰

J. Procedural Disadvantages for Class Action Plaintiffs

Plaintiffs are prohibited from exerting the same supervisory control over the litigation as exists in the non-class action context.³¹ They usually lack control even over the selection of counsel, giving rise to a situation in which a poorly qualified lawyer may be chosen to represent the class when few individuals in the class would have selected that lawyer in an open market. These problems further contribute to suboptimal outcomes in the class action context.

In consequence of these drawbacks, the class action mechanism cannot work wholly in the interests of the litigants. Under either of the most common fee structures, attorney/client agency costs are extraordinarily high. In some eases, they allow the class action device to serve the interests of the lawyers more than those of their clients. A few courts recently have begun to experiment with reform.

C. Use of Auctions to Select Lead Counsel

1. First Experiment with Lead Counsel Auction

Judge Vaughan Walker in the Northern District of California was the first to experiment with an auction to select and compensate lead counsel in a class action. In In re Oracle Securities Litigation,³² he declined to ratify the selection of two firms as co-lead counsel by a group fifteen plaintiffs’ lawyers involved in the case. Noting that although many among the fifteen were experienced antitrust attorneys, they displayed “cavalier indifference” to the spirit of the antitrust laws in their selection of counsel,³³ Judge Walker instead ordered the two selected firms and two others that jointly had contested the selection to submit budgets for the litigation, on the basis of which he proposed to choose lead counsel. On the appointed day, however, rather than submitting separate budgets, two of the four firms, one from each of the opposing camps, submitted a joint proposal to serve as lead counsel, thereby frustrating the court’s effort to inject competition into the process.

Judge Walker flatly rejected this proposal and instead ordered all interested counsel to submit bids, from among which the court would select lead counsel. The bids were required to state the bidder’s qualifications for the position and specify the percentage of any recovery the firm would charge as fees and costs.³⁴ The court prohibited the competing firms from submitting joint proposals and demanded that each bidder certify that its bid was prepared independently and that no part thereof had been revealed to any other bidder.

Following Judge Walker’s order, four firms submitted bids for the position of lead counsel. The bid selected proposed a sliding contingency fee arrangement with an early settlement discount and an expense cap of $325,000. Under the sliding fee arrangement, the percentage of plaintiffs’ recovery that would constitute the attorney’s fee was to decrease as the amount of recovery increased.³⁵ The court selected this bid because, in its view, (1) the declining percentage-of-recovery fee, unlike a flat or increasing percentage-of-recovery fee, would prevent a windfall recovery by lead counsel and instead would share counsel’s “economies of effort” with the class,³⁶ (2) the expense cap would prevent depletion of the common fund through inordinate litigation expenses,³⁷ (3) the early settlement discount would guard against cheap, collusive, early settlement,³⁸ (4) the bid was the most competitive in rates,³⁹ and (5) the successful bidder was at least as well qualified as the other bidders.⁴⁰

Some significant time after the selection of counsel, the parties in the Oracle case arrived at a settlement. Calculation of attorney’s fees based on the schedule proposed in the successful bid yielded an attorney’s fee of $4.8 million, or 19.2 percent of the settlement recovery.⁴¹ This compared favorably to what counsel would have been awarded using a standard 25 percent recovery method — $6.25 million. It is, of course, impossible to determine what counsel would have received using the lodestar method, as attorney time records are unavailable, and any multiplier the court would have used is unknown. Further, it is impossible to ascertain whether the compensation schedule had any effect, positive or negative, on the overall amount of the settlement.

2. Subsequent Experiments with Lead Counsel Auctions

Since Judge Walker first experimented with a lead counsel auction, several other courts have followed suit.⁴² A number have embraced fee structures with built-in incentives similar to those endorsed in Oracle, including the declining percentage-of-recovery fee and the expense cap.⁴³ Others have endorsed a cap on attorney’s fees, presumably in order to prevent windfall recovery by plaintiffs’ counsel.⁴⁴ Some have adopted Judge Walker’s notion of a fee discount for early settlement and requested bids that would adjust the fee based on the stage of the litigation at which the case is resolved, ranging from pleading to motions to dismiss, to summary judgment, to verdict after trial, to appeal.⁴⁵ This approach doubtless is based on the view that lead counsel should be encouraged to eschew cheap early settlement and rewarded for the risk attendant to continuing the litigation into a later stage. Still other courts have asked bidders to submit their evaluations of the case, including the probability of success, in order better to compare the competing proposals. Finally, some courts have given a right of first refusal to counsel for the lead plaintiff, allowing counsel to match the terms of the winning bid if it so chooses.⁴⁶ This undoubtedly reflects a presumption, ceteris paribus, in favor of counsel for the lead plaintiff.

The fee structures adopted in many of these cases attempt to address the high agency costs that pervade the traditional lodestar and percentage-of-recovery methods. Some of them, however, create perverse incentives of their own. The attorney’s fee cap, for example, addresses a major concern of the lodestar method — the investment of needless attorney hours in the case, including unnecessarily prolonging the litigation. However, the fee cap creates an incentive for lead counsel to settle the case exactly at the level at which the fee reaches its maximum, even if that level is suboptimal from plaintiffs’ perspective. If disclosed to defendants, the fee cap also can lead defendants to exploit the disjuncture of interests between plaintiffs’ and their counsel by making a firm settlement offer in the amount that would exactly maximize counsel’s fee, even if defense counsel otherwise would be prepared to go higher. Again, lead counsel would have an incentive to agree to settlement in this amount and not press for an award more favorable to plaintiffs.

The same problem arises with the use of a cap on expenses. Although doubtless reducing runaway litigation expenses, the expense cap encourages lead counsel to cease prosecuting the case as soon as expenses have reached the cap level.

The early settlement discount addresses a central risk of the traditional percentage-of-recovery method — early and cheap collusive settlements — by providing lead counsel with increasing marginal returns to effort over time. However, this method risks falling short, as it motivates counsel not to maximize the class’ recovery, but merely to extend the duration of the litigation, even if doing so is not in plaintiffs’ best interests. Therefore, although this arrangement might improve upon the flat percentage-of-recovery method, it does not align counsel’s interests fully with those of the class.

The declining percentage-of-recovery fee structure adopted in Oracle and other cases likewise addresses some of the concerns associated with the traditional flat percentage-of-recovery arrangement, yet contains its own problems. By adjusting downward the percentage of the recovery awarded to counsel as plaintiffs’ recovery increases, this arrangement arguably limits windfall attorney’s fee awards. However, this method may give rise to an attorney incentive problem by creating declining marginal returns to effort for counsel. If counsel’s opportunity costs begin to exceed the economic benefit to counsel of continuing to litigate, counsel may be more likely to settle the case and exit the litigation rather than prolonging the litigation and pushing for a higher recovery for the class, even if the added effort would be in plaintiffs’ best interest. Again, this method can create an incentive to settle quickly and cheaply, when the returns to effort are highest, rather than investing additional time and maximizing plaintiffs’ recovery.

An increasing percentage-of-recovery method likewise does not eliminate fully the disjuncture of interests between plaintiffs and lead counsel. As a rule, this method awards lead counsel a marginally greater percentage of plaintiffs’ recovery as the recovery incrementally increases, giving counsel an incentive to avoid premature settlement and push for a higher plaintiffs’ recovery. In theory, this approach can reduce agency costs by limiting the circumstances in which costs would outweigh the benefits of continuing to prosecute the litigation. Indeed, it gives counsel an incentive to push any settlement offer higher, as counsel’s marginal returns increase with plaintiffs’ recovery. However, this fee structure might have the effect of encouraging plaintiffs’ lawyers to eschew settlement in search of a very high recovery, even if this strategy is overly risky from plaintiffs’ perspective. Further, it is not clear a priori how to demarcate the increments of plaintiffs’ recovery according to which counsel’s fee correspondingly will increase. Setting the increments too low might eliminate the positive effect of the increasing pereentage-of-recovery method on counsel’s incentives because the opposing parties, after some discovery, will come to value the ease in the highest range, eliminating some of the upside benefit to lead counsel of a higher settlement. Conversely, if the increments are set too high, it might become apparent after some discovery that the case will be valued only in the lowest range. This in turn can make the litigation too costly for lead counsel, thereby encouraging premature cheap settlement in order to extricate counsel quickly from the case.

S. Possible Drawbacks of Lead Counsel Auctions

The use of auctions to select lead counsel in class actions has been the subject of much criticism. It has been argued that a simple auction that awards the lead counsel position to the bidder proposing the lowest fee carries substantial risks. Although this approach may keep attorney’s fees at a minimum, it limits the potential upside gain for counsel of a substantial award to plaintiff and consequently can encourage quick and cheap settlements.⁴⁷ Further, use of price as the sole criterion for selection does nothing to ensure that plaintiffs receive quality representation.⁴⁸

The lead counsel auction unwittingly may undermine also the efficacy of the class action device. Courts in certain cases have been known to award the lead counsel position to the attorney that files the first complaint in the case or to a group of which that attorney is a part.⁴⁹ The rationale behind this first-to-file rule is that it creates an incentive for attorneys to ferret out wrongs that may be difficult or impossible for individual plaintiffs ever to identify. By rewarding attorneys that incur these search costs, the award of the lead counsel position to the first attorney to file arguably makes the class action mechanism a more vital means of redress for injured plaintiffs. This, in turn, benefits society by creating a deterrent to wrongful behavior by others.

The routine selection of lead counsel by auction, in contrast, may discourage attorneys from searching out and identifying illegal activity, as the attorney who takes this initiative is not necessarily compensated for his or her effort. This casts doubt on the desirability of holding any auction at all, at least in cases in which attorney initiative played an important role in uncovering the alleged wrong.⁵⁰

Granting counsel to the lead plaintiff a right of first refusal conceivably might address this concern by promising the attorney that incurred the search costs, if willing to offer his or her services at a competitive price, a reward for this action. However, a right of first refusal takes control over the selection of lead counsel out of the court’s hands and thereby undermines the court’s ability to ensure that the class receives the highest quality representation.

Mindful of these considerations, the Court in this case undertook to establish a method of counsel selection and a fee structure that, in the context of this ease, would begin to address some of these concerns and seek to align counsel’s and plaintiffs’ interests more fully.

Ill

The Court was mindful of these considerations when considering the possibility of an auction for the position of lead counsel. It concluded that this case is singularly appropriate for the use of an auction for several reasons.

Unlike many class actions, no attorney initiative was required here to ferret out the alleged wrong committed by defendants. Rather, the alleged wrong came to light only after it was announced that the Department of Justice had begun to investigate defendants and that Christie’s had sought conditional amnesty from criminal prosecution. The attorney who filed the first complaint in this case therefore is not necessarily any more deserving of the lead counsel position than is any other attorney involved, and selection as lead counsel of someone other than' the first-to-file did not deprive an investigating attorney of his or her just reward or dissuade attorneys in other eases from searching out a wrong.

This case is well suited for a lead counsel auction also because several factors are present that permit an auction nearly to approximate an efficient market. First, this case has received extensive media attention and consequently attracted large numbers of able plaintiffs’ attorneys. Indeed, whereas most previous experiments with lead counsel auctions have involved bids from very few attorneys, the Court in this case received bids from upward of twenty firms in each of two rounds of bidding. As larger markets lead to more competition, and as competition leads to more efficient results, the number of prospective qualified bidders in this case undoubtedly contributed to the submission of many high quality bids from which to choose.

Second, the form of relief sought in this case is monetary damages, rather than equitable relief. This makes the case easier to evaluate, simplifies the bidding process and permits the Court more easily to compare the bids.

The circumstances in this case allowed the lead counsel auction to approach an efficient market for legal services for a third reason as well — the bidding attorneys had far more information with which to evaluate the case, both as to liability and damages, than typically is available. With respect to liability, this case differs from those in which plaintiffs simply make a claim that defendants deny, or even cases in which the government is undertaking a criminal investigation of defendants. Rather, Christie’s reportedly had sought to take advantage of the government’s amnesty program and allegedly has received conditional amnesty from prosecution. Although this alone certainly does not establish liability or speak to the scope or temporal duration of the alleged conspiracy, it appears to give plaintiffs a better prospect for success on the merits than is often the case.

With respect to damages, too, there are fewer unknowns here than often is the case. The essence of plaintiffs’ claim is that Christie’s and Sotheby’s acted as duopolists to rig prices in what is principally a two firm market. Significant information is available regarding the market shares of the two companies, and Sotheby’s is a publicly held company, the financial statements of which are available and informative. This information alone provided bidders with a strong base of information from which to calculate potential damages. Further, as the case developed, it became clear that there had been at least preliminary settlement negotiations in which defendants furnished financial information to Interim Lead Counsel, and they had ordered expert analysis of this information. The Court ordered that the expert analysis be made available to all bidders prior to the time the bids were due in order to equalize the information base and create the most competitive process possible.⁵¹ In consequence, there was an unusually substantial base of information from which bidders intelligently could evaluate the case.

A. Reasoning Behind the Court’s First Proposed Fee Structure

The Court’s first proposed fee structure was designed to avoid the agency pitfalls that characterize many of the fee structures discussed above. In order to create a disincentive to cheap, premature settlement, any recovery less than X was to go entirely to the class, depriving lead counsel of a fee. Although bidders presumably would choose a value for X below their expected value of the case, the pressure of competition would tend to drive X toward the expected recovery, appropriately discounted for the passage of time. Once the potential recovery surpassed X, however, counsel’s marginal returns to effort would increase dramatically, as all recovery between X and Y would go entirely to counsel. This was designed to motivate counsel to prosecute the ease as effectively as possible. As lead counsel’s returns to effort would be greatest if the case were resolved for exactly Y, bidders presumably would tend to choose a value for Y close to the expected value of the case. Finally, twenty-five percent of any recovery in excess of Y was to go to counsel, with the remainder going to the class. This flat percentage-of-recovery arrangement was designed to provide added motivation to counsel to continue to prosecute the case while avoiding the risk of over-prosecution that might result from an increasing percentage-of-recovery fee.

Two other features of the Court’s first proposal are worthy of note. The proposal provided that the successful bidder would be required to absorb all litigation expenses. This was intended to create an incentive to keep costs at a minimum and to avoid difficult problems in evaluating post hoc the propriety and utility of expenses. Further, the bids were to be kept confidential so as to prevent collusion by bidding attorneys.

B. Amicus Briefs

On the day the bids were due, the Court received also several amicus briefs and submissions from bidders commenting on the proposed fee structure. These submissions raised two principal issues with respect to the proposed auction structure.

First, one of the amici rightly pointed out that the initial proposed fee structure, that, by awarding one hundred percent of any recovery between X and Y to counsel, could create a stark conflict of interest between counsel and the class.⁵² If, for example, the winning bid placed X at $20 million and Y at $40 million, and if defendants were willing to settle only at $20 million, lead counsel would have an incentive to take the case to trial, even if the likelihood of verdict high of enough to bring added benefit the class (in excess of $40 million) were very small. Trial in such an instance clearly would not be in plaintiffs’ best interests. And indeed, were lead counsel to reject such an offer and take the case to trial, as the proposed fee structure implicitly encouraged it to do, counsel well might be in violation of counsel’s fiduciary duty to the class.

A second amicus pointed out that evaluation of the bids by the Court would be particularly complex by virtue of there being two variables, X and Y, rather than just one.⁵³ Without some relatively firm information on the distribution of possible recoveries, it would be difficult intelligently to compare two bids, one of which set slightly lower values on both X and Y than did the other.

C. Reasoning Behind the Court’s Second Proposed Fee Structure

In light of these comments, the Court revised the proposed fee structure to better align counsel’s and plaintiffs’ interests and facilitate ready comparison of the bids.⁵⁴ The use of a single variable, X, rather than two, as in the first proposal, was meant to eliminate the potential conflict of interest created by the first proposal. As counsel will receive no fee if plaintiffs’ recovery falls below X, counsel clearly is discouraged from settling prematurely and has an incentive to pursue a recovery higher than X. This effort will accrue to the benefit of both counsel and the class. As the value of the case surpasses X, counsel’s marginal returns to effort will increase steadily, as they will receive twenty-five percent of any amount in excess of X. Again, this creates an incentive for counsel to litigate the case aggressively. Insofar as the Y variable has been eliminated, so too has the conflict of interest.⁵⁵

The Court’s prohibition in the second proposal of disclosure of the terms of the successful bid was designed also to reduce perverse incentives that may have been created under the first proposal. Were defendants apprised of the amount of the bid, they might be inclined to formulate settlement offers in order best to take advantage of any perverse attorney incentives created by the fee structure.⁵⁶

D. Disclosure of Interim Committee’s Expert Analysis

The Court’s ruling that certain documents in possession of the Interim Committee be disclosed to all plaintiffs’ counsel also was intended to improve the quality of the auction process. This expert analysis contained damage assessments that materially would have assisted counsel in the formulation of bids that accurately took into account the value of the case. In consequence, these documents were ordered disclosed so as to even the playing field, facilitate bidders in assessing accurately the value of the case, and improve the overall quality of the bids submitted.

E. Selection of Lead Counsel

After careful review of the bids, the Court selected David Boies and Richard B. Drabel of Boies, Schiller & Flexner, LLP as lead counsel in the case. This choice does not reflect adversely on the capability or integrity of other bidders, many of whom are known to and respected by the Court. It merely reflects the Court’s judgment as to which bidder, in all the circumstances, likely would best serve the interests of the plaintiff class. In short, the Court sought to act as a fiduciary to the class in selecting counsel. In light of the pendency of the litigation, the Court is not prepared at this time to disclose the terms of the winning bid.

F. Potential Agency Costs of Second Proposed Fee Structure

At least one potential incentive problem with the attorney’s fee structure remains. Under the bid structure ultimately adopted, it theoretically might become apparent at some point that the case cannot be resolved in an amount greater than X, in which case counsel would receive no compensation. If that occurs, lead counsel will have an incentive to settle the case immediately and make a hasty exit. This may not be in plaintiffs’ interests and in any event, certainly raises the specter of an attorney-client conflict of interest. The potential conflict is exacerbated by the fact that lead counsel is required to pay all expenses out of the fee award, raising even further the opportunity costs for counsel of continued prosecution. Nonetheless, it appears that the unique circumstances of this case make this scenario unlikely for several reasons.

First, as in any class action, the Court is vested with authority to reject an inadequate settlement. The Court is fully prepared to do this were it apparent that counsel had failed to represent adequately the class.

Second, the Court in this case was in a uniquely advantageous position from which to evaluate the bids, helping to ensure that the bid selected was not unreasonably high. Following a motion by the government to stay discovery with respect to twelve key documents furnished by Christie’s to the Department of Justice, the Court ordered that these documents be made available for in camera inspection.⁵⁷ The information provided therein gave the Court with an additional tool with which to evaluate the bids, as did the plaintiffs’ damage analysis.

Third, the Court has required that notice to the class explain the manner in which lead counsel was selected and the risk for the class that may result from the manner in which lead counsel will be compensated.⁵⁸ This is designed to give class members sufficient information with which to evaluate the fee structure, allowing those who oppose it to opt out.

Finally, if the parties arrive at a proposed settlement, the Court will order notice to the class to disclose the fee arrangement. By revealing to the class the incentive structure under which counsel has been working, disclosure of the fee structure should permit class members adequately to evaluate any settlement and encourage any objectors to come forward if that proves appropriate.

IV

The benefits of any auction for lead counsel are difficult to assess. It is simple to compare post facto the fee awarded to counsel selected by auction to that which would have been awarded using a traditional percentage-of-recovery method. Likewise, ready comparison can be made with the fees that would have been awarded to other bidders, had their bids been selected. However, the relative value of the attorney’s fee does not adequately measure the success of the auction. Instead, the true value of the auction lies in its effect, if any, on the net recovery obtained by plaintiffs. In this respect, the jury on the lead counsel auction in this case is still out, but it is anticipated that the fee structure and the auction process will function as they were intended' — to align attorney-client interests more closely, reduce agency costs, and help ensure that the class action mechanism acts as an effective mechanism of justice.

SO ORDERED.

1 . Carol Vogel, Christie’s Says It Is Cooperating with Antitrust Inquiry in Art World, N.Y. TIMES, Jan. 29, 2000, at B4.

2 . Sir Anthony Tennant, A. Alfred Taubman, Christopher M. Davidge, and Diana D. Brooks later were added as defendants.

3 . Tr. (Feb. 23, 2000) at 17 (Docket Item "DI" 117).

4 . Id. at 18-19.

5 . Id. at 21.

6 . In re Auction Houses Antitrust Litigation, 193 F.R.D. 162 (S.D.N.Y.2000).

7 . Order, Apr. 20, 2000 (DI 119).

8 . This was proposed in a second order issued several days later. Order, Apr. 26, 2000 (DI 32).

9 . In addition to submitting the X and Y figures, each bidder was required to submit a sworn certification that the bidder had not, directly or indirectly, communicated with (1) any other bidder concerning the terms of the bid or its position with respect to whether the Court should adopt this method, (2) any defendant or prospective defendant following the issuance of the order concerning settlement or possible settlement of any or all of the actions, or (3) any other attorney or firm concerning its possible performance of legal or other services for the bidder in connection with this litigation in the event the bidder were selected as lead counsel. Order, Apr. 20, 2000 (DI 119).

On April 27, 2000, the Court denied a request by interim counsel that they be permitted to submit joint comments on the proposed bid structure. Memo-Endorsement on Apr. 26, 2000 letter from Interim Executive Committee. (DI 33).

The order provided also that any compensation awarded to the successful bidder was to be inclusive of all costs, disbursements and other charges incurred in connection with the litigation. Further, the Court reserved the right to compensate lead counsel on a different basis in the event the litigation were resolved in a manner that did not permit determination of a gross recovery by the class or if justice otherwise required. Finally, it noted that, in the event that lead counsel other than interim counsel were appointed and plaintiffs prevailed, it would accept a fee application on behalf of interim counsel for services performed on behalf of the class. Order, Apr. 26, 2000 (DI 32).

The bids were to be submitted sealed ex parte to the Court on or before May 12. The Court ordered also that it would receive on or before that date submissions from bidders, interim lead counsel and any class members or their counsel as to the advisability of employing this or a similar structure. Order, Apr. 20, 2000 (DI 119), at 3.

10 . The Court is grateful for the amicus submissions by Professors Jonathan R. Macey, John C. Coffee, Jr., and Randall S. Thomas and Robert G. Hansen, all of whom contributed measurably to the Court’s thinking in this case. See also John C. Coffee, Jr., "Auction Houses": Legal Ethics and the Class Action, N.Y.L.J., May 18, 2000, at 5.

11 . Order, May 17, 2000 (DI 61).

12 . Order, May 9, 2000 (DI 60).

13 . Order, May 17, 2000 (DI 59).

14 . See John C. Coffee, Jr., The Corruption of the Class Action: The New Technology of Collusion, 80 CORNELL L. REV. 851, 851 (1995); Jonathan R. Macey & Geoffrey P. Miller, The Plaintiffs’ Attorney’s Role in Class Action and Derivative Litigation: Economic Analysis and Recommendations for Reform, 58 U. CHI. L. REV. 1, 8 (1991).

15 . See generally 7B CHARLES ALAN WRIGHT, ARTHUR R. MILLER & MARY KAY KANE, FEDERAL PRACTICE AND PROCEDURE (hereinafter “WRIGHT & MILLER”) § 1781 (2d ed.1986).

16 . FEDERAL JUDICIAL CENTER, MANUAL FOR COMPLEX LITIGATION, THIRD § 20.221 (1995).

17 . Both of these methods have been endorsed by the Second Circuit for determination of attorney’s fees in common fund cases. See Goldberger v. Integrated Resources, Inc., 209 F.3d 43 (2d Cir.2000).

18 . City of Detroit v. Grinnell Corp., 560 F.2d 1093, 1098 (2d Cir.1977), abrogated on other grounds by Goldberger, 209 F.3d 43.

19 . See Janet Cooper Alexander, Do the Merits Matter? A Study of Settlements in Securities Class Actions, 43 STAN. L. REV. 497, 543 (1991) (hereinafter "Do the Merits Matter? ”).

20 . See Alexander, Do the Merits Matter?, 43 STAN. L. REV. at 543; John C. Coffee, Jr., Understanding the Plaintiff s Attorney: the Implications of Economic Theory for Private Enforcement of Law through Class and Derivative Actions, 86 COLUM. L. REV. 669, 717 (1986) no. 4 (hereinafter "Understanding the Plaintiff’s Attorney ”).

21 . Of course, in determining the lodestar post-trial, the court may apply a multiplier to reflect the added risk to counsel of proceeding to trial. This permits counsel to share in plaintiffs’ upside reward and might affect to some extent counsel's decision to go to trial or not. However, insofar as counsel does not know in advance how large a multiplier, if any, the Court will apply, the possibility of a multiplier does not eliminate the disjuncture of incentives between plaintiffs and counsel on the eve of trial.

22 . See, e.g., In re General Motors Corp. Pick-Up Truck Fuel Tank Prods. Liab., 55 F.3d 768 (3d Cir.1995) (proposing settlement in which class received rebate coupons on future truck purchases and attorneys received a $9.5 million fees); In re Ford Motor Co. Bronco II Prod. Liab. Litig., 981 F.Supp. 969 (E.D.La.1997) (rejecting proposed settlement involving "utility vehicle package” of safety materials, including safe driving instructional video, for class and $6 million in fees for attorneys); In re Domestic Air Transp. Antitrust Litig., 148 F.R.D. 297 (N.D.Ga.1993) (approving class action settlement which included discount air travel certificates for class and over $14 million in attorneys fees).

23 . FED. R. CIV. PROC. 23(e).

24 . See Coffee, Understanding the Plaintiff s Attorney, 86 COLUM. L. REV. at 714; see also In re Activision Sec. Litigation, 723 F.Supp. 1373, 1374 (N.D.Cal.1989).

25 . There is a long-standing principle in United States law in favor of adversarialness as a prerequisite to justiciability. This principle stems from Article Ill's case and controversy requirement and is manifest in numerous areas of the law. See, e.g., Lujan v. Defenders of Wildlife, 504 U.S. 555, 112 S.Ct. 2130, 119 L.Ed.2d351 (1992) (standing); North Carolina v. Rice, 404 U.S. 244, 246, 92 S.Ct. 402, 30 L.Ed.2d 413 (1971) (ban on advisory opinions); Flast v. Cohen, 392 U.S. 83, 88 S.Ct. 1942, 20 L.Ed.2d 947 (1968) (same); Powell v. McCormack, 395 U.S. 486, 496-97 & n. 7, 89 S.Ct. 1944, 23 L.Ed.2d 491 (1969) (mootness); Abbott Laboratories v. Gardner, 387 U.S. 136, 148-49, 87 S.Ct. 1507, 18 L.Ed.2d 681 (1967) (ripeness).

26 . See, e.g., In re Fine Paper Antitrust Litig., 751 F.2d 562 (3d Cir.1984).

27 . Savoie v. Merchants Bank, 166 F.3d 456, 460 (2d Cir.1999).

28 . See Coffee, Understanding the Plaintiff s Attorney, 86 COL. L. REV. at 687-90.

29 . MANCUR OLSON, THE LOGIC OF COLLECTIVE ACTION: PUBLIC GOODS AND THE THEORY OF GROUPS (1965).

30 . See Sanford I. Weisburst, Judicial Review of Settlements and Consent Decrees: An Economic Analysis, 28 J. LEGAL STUD. 55, 87 (1999).

31 . For instance, class members lack veto power over any proposed settlement, and they have no right to be kept informed of developments in the case. See Alexander, Do the Merits Matter?, 43 STAN. L. REV. at 535.

32 . 131 F.R.D. 688 (N.D.Cal.1990); 132 F.R.D. 538 (N.D.Cal.1990); 136 F.R.D. 639 (N.D.Cal. 1991).

33 . 131 F.R.D. at 690 n. 3.

34 . Id. at 697. The court noted also that payment of the fees and costs of any firm assisting the appointed firm in the prosecution of the actions would be the responsibility of the appointed firm. Id.

35 . 132 F.R.D. at 541.

36 . Id. at 543-44.

37 . Id. at 542.

38 . Id. at 545.

39 . Id. at 547.

40 . Id. at 542.

Subsequent to the appointment of lead counsel for the class, an additional defendant was added to the case, and the court solicited bids for counsel to prosecute claims against that defendant. After receiving bids from three firms, including the original successful bidder, the court selected the original successful bidder to prosecute the claims against the new defendant as well. Order, July 21, 1991 (cited in In re Oracle Securities Litig., 852 F.Supp. 1437, 1453 (N.D.Cal.1994)). The second successful bid proposed to aggregate the class’ recovery against all defendants for purposes of calculating the attorney’s fee and proposed the same decreasing percentage-of-recovery fee schedule set for the original claims, but did not include an early settlement discount. It capped additional litigation expenses (beyond those incurred in prosecution of the claims against the original defendants) at $500,000, for a total of $825,000 for the claims against all defendants. In re Oracle Securities Litig., 852 F.Supp. at 1453.

41 . Id. at 1457.

42 . See In re Amino Acid Lysine Antitrust Litig., 918 F.Supp. 1190 (N.D.Ill.1996) (Shadur, J.), No. 95 C 7679, 1996 WL 197671 (N.D.Ill.1996); In re Cendant Corp. Litig., 182 F.R.D. 144 (D.N.J.1998) (Williams, J.); In re Network Assoc. Inc., 76 F.Supp.2d 1017 (N.D.Cal.1999) (Alsup, J.); Sherleigh Assocs. LLC v. Windmere-Durable Holdings, Inc., 184 F.R.D. 688 (S.D.Fla.1999) (Lenard, J.); 186 F.R.D. 669 (S.D.Fla.1999); In re Lucent Technologies, Inc. Securities Litig., 194 F.R.D. 137 (D.N.J.2000) (Lechner, J.); In re Bank One Shareholders Class Actions, 96 F.Supp.2d 780 (N.D.Ill.2000) (Shadur, J.).

Judge Walker has used the auction method in several subsequent cases as well. See In re Wells Fargo Securities Litig., 157 F.R.D. 467 (N.D.Cal. 1994); In re California Micro Devices, 168 F.R.D. 257 (N.D.Cal.1996); Wenderhold v. Cylink, 188 F.R.D. 577 (N.D.Cal.1999); 191 F.R.D. 600 (N.D.Cal.2000).

43 . See, e.g., In re Wells Fargo Securities Litig., 157 F.R.D. 467; In re Amino Acid Lysine Antitrust Litig., 918 F.Supp. 1190; Wenderhold v. Cylink, 191 F.R.D. 600.

44 . See, e.g., In re Amino Acid Lysine Antitrust Litig., 918 F.Supp. 1190.

45 . See, e.g., In re Wells Fargo Securities Litig., 157 F.R.D. 467; Wenderhotd v. Cylink, 191 F.R.D. 600; In re Lucent Technologies, Inc. Securities Litigation, 194 F.R.D. 137.

46 . See, e.g., In re Cendant Corp. Litig., 182 F.R.D. 144.

47 . John C. Coffee, Jr., The Unfaithful Champion: The Plaintiff as Monitor in Shareholder Litigation, 48 LAW & CONTEMP. PROBS. 5, 77 (1985) (hereinafter "The Unfaithful Champion "),

48 . See John C. Coffee, Jr., The Regulation of Entrepreneurial Litigation: Balancing Fairness and Efficiency in the Large Class Action, 54 U. CHI. L. REV. 877, 933 (1987); Coffee, The Unfaithful Champion, 48 LAW & CONTEMP. PROBS. at 77.

49 . See James Bohn & Stephen Choi, Fraud in the New-Issues Market: Empirical Evidence on Securities Class Actions, 144 U. PA. L. REV. 903, 916-17 (1996); Jill E. Fisch, Class Action Reform: Lessons from Securities Litigation, 39 ARIZ. L. REV. 533, 549 (1997).

50 . See Coffee, Understanding the Plaintiff s Attorney, 86 COLUM. L. REV. at 691.

51 . Order, May 17, 2000 (DI 60).

52 . This point was raised by Professor Coffee as well as one of the plaintiffs’ counsel.

53 . This issue was raised by Professors Randall S. Thomas and Robert G. Hansen as well as one of plaintiffs' counsel.

54 . Order, May 17, 2000 (DI 61).

55 . Although generally critical of lead counsel auctions, Professor Coffee in his amicus submission to the Court voiced his approval of the use of an auction to select lead counsel in this case. As discussed above, he questioned the Court's first proposed bid structure on the ground that allocating one hundred percent of any recovery between X and Y to counsel would create an unnecessary conflict of interest. He suggested that the Court revise its proposal and employ instead an increasing percentage-of-recovery fee structure in which counsel would be awarded a marginally greater percentage of plaintiffs' recovery as the recovery increased. He proposed the increasing percentage-of-recovery method because, by giving counsel an increasingly large stake in a higher plaintiffs' recovery, such a fee structure would encourage plaintiffs’ attorney to "expend the additional effort, accept the additional risk, and wait out the greater delay to obtain such a recovery.” See also Coffee, "Auction Houses": Legal Ethics and the Class Action, N.Y.L.J., May 18, 2000, at 5. Professor Coffee’s position was well taken, and the Court's second proposed fee structure endeavors to do just that. By depriving counsel of a fee for any recovery below X and awarding counsel twenty-five percent of any recovery in excess of X, the Court's second proposed fee structure effectively institutes a two-tier, increasing percentage-of-recovery fee. By giving counsel no stake in any recoveiy below X and a substantial stake in any recovery in excess thereof, the second proposed fee structure encourages lead counsel to expend the effort, accept the risk and seek to obtain a recovery in excess of X.

56 . Professor Coffee’s amicus submission pointed out the danger of disclosing the terms of the winning bid to defendants.

57 . Order, May 17, 2000 (DI 59).

58 . Order, Jun. 20, 2000 (DI 87); Order, Jul. 19, 2000 (DI 94).